Herestraat 49, B-3000 Leuven Yves Kremer, CU Saint-Luc, Av

Home , Fascia iliaca block, Hyalobarrier

Editor | Prof. Dr. V. Bonhomme CO-Editors | Dr. Y. Kremer — Prof. Dr. M. Van de Velde ACTA ANAESTHESIOLOGICA JOURNAL OF THE BELGIAN SOCIETY OF ANESTHESIOLOGY, RESUSCITATION, PERIOPERATIVE MEDICINE AND PAIN MANAGEMENT (BeSARPP) BELGICA Indexed in EMBASE l EXCERPTA MEDICA ISSN: 2736-5239

Suppl. 1 202071

Master Theses

www.besarpp.be

Cover-1 -71/suppl.indd 1 12/01/2021 12:37 ACTA ANÆSTHESIOLOGICA BELGICA 2020 – 71 – Supplement 1

EDITORS

Editor-in-chief : Vincent Bonhomme, CHU Liège, av. de l’Hôpital 1, B-4000 Liège Co-Editors : Marc Van de Velde, KU Leuven, Herestraat 49, B-3000 Leuven Yves Kremer, CU Saint-Luc, av. Hippocrate, B-1200 Woluwe-Saint-Lambert Associate Editors : Margaretha Breebaart, UZA, Wilrijkstraat 10, B-2650 Edegem Christian Verborgh, UZ Brussel, Laarbeeklaan 101, B-1090 Jette Fernande Lois, CHU Liège, av. de l’Hôpital 1, B-4000 Liège Annelies Moerman, UZ Gent, C. Heymanslaan 10, B-9000 Gent Mona Monemi, CU Saint-Luc, av. Hippocrate, B-1200 Woluwe-Saint-Lambert Steffen Rex, KU Leuven, Herestraat 49, B-3000 Leuven Editorial assistant Carine Vauchel Dpt of Anesthesia & ICM, CHU Liège, B-4000 Liège Phone: 32-4 321 6470; Email: [email protected]

Administration secretaries MediCongress Charlotte Schaek and Astrid Dedrie Noorwegenstraat 49, B-9940 Evergem Phone : +32 9 218 85 85 ; Email : [email protected]

Subscription The annual subscription includes 4 issues and supplements (if any).

4 issues 1 issue (+supplements) Belgium 40€ 110€ Other Countries 50€ 150€

BeSARPP account number : BE97 0018 1614 5649 - Swift GEBABEBB

Publicity : Luc Foubert, treasurer, OLV Ziekenhuis Aalst, Moorselbaan 164, B-9300 Aalst, phone: +32 53 72 44 61 ; Email : [email protected]

Responsible Editor : Prof. Dr V. Bonhomme, Dpt of Anesthesia & ICM, CHU Liège av. de l’Hôpital 1, B-4000 Liège. Phone : +32 4 321 7117 ; Email : [email protected]

erratum2014(2).indd 2 4/08/14 14:04 Cover-2-Suppl.2020-nieuw kopie.indd 1 12/01/2021 12:38 (Acta Anaesth. Belg., 2020, 71, 1-6) Near-infrared spectroscopy for monitoring tissue oxygenation in breast reconstruction

E. Thys (*), J. De Puydt (*), F. Thiessen (*), T. Tondu (**), V. Saldien (*)

Abstract : Background : A breast reconstruction with a monly performed after mastectomies. The success deep inferior epigastric artery perforator flap (DIEP-flap) ratio of this type of surgery is high, more than 95 %. is commonly performed microvasculared procedure. It Nevertheless, a free flap failure can occur (1). The flap can is crucial to detect vascular complications as early as be jeopardized by vascular complications, which usually possible to maximize the chance on a successful flap occur in the first hours after surgery (2). The survival of revision. Near-infrared spectroscopy (NIRS) measures the free flap depends on the continuous arterial inflow the tissue oxygen saturation (StO2), by using infrared and the venous outflow through the new microvascular light and can be an interesting tool to monitor flap anastomoses. Thrombosis, vasospasm, kinking of the perfusion. vessels or external compression on blood vessels can Objective : The aim of this study is to determine whether cause ischemia of the flap. It is essential to recognize NIRS is suitable to assess tissue perfusion after DIEP- these complications as early as possible to maximize flap reconstruction. successful flap revision. Ischemia of the flap can become Methods : The trial was designed as a single cohort irreversible due to the no-reflow phenomenon (1). The interventional pilot study (approved by the Ethics key to enhance the flap success rate is early detection of Committee of the University Hospital of Antwerp - complications. Internal reference number : 18/12/173 ; Belgian regis- A DIEP-flap is commonly used to perform an tration number : B300201836129). autologous breast reconstruction. It uses tissue from All patients scheduled for uni- or bilateral breast the lower abdominal wall to reconstruct the breast. The reconstruction by means of a DIEP-flap were included, deep inferior epigastric artery and vein are responsible

if no exclusion criteria were present. The StO2 was for the regional perfusion of the skin and subcutaneous measured during the first 12 hours postoperatively. The fatty tissue below the umbilicus and above the pubic

StO2 measurements were performed using a sensor area. In the first phase, the plastic surgeon dissects the positioned on the free flap and one on the native tissue. flap. In a second phase, the vessels are microsurgically Results : Twelve women underwent a uni- or bilateral anastomosed to the internal mammary artery and vein. It DIEP-flap breast reconstruction. nI this study, the mean only uses skin and subcutaneous fatty tissue. The muscles

StO2 values of the flap and of the native tissue were 90.16 of the abdomen are left unaffected (3). ± 4.63 % (mean ± SD) and 91.52 ± 2.23 % (mean ± SD), Clinical examination is still the gold standard for

respectively. The mean difference in StO2 mea urements flap monitoring. Skin color, temperature and Doppler between the free flap and the native tissue was 1.36 ± signals are frequently examined (4). Clinical monitoring 3.46 % (mean ± SD). This difference was not statistically is subjective, intermittent, labor intensive and requires significant, according to the Wilcoxon signed-ranks test. experience. There is also a time delay between the

In one patient, the StO2 value of the free flap tissue had occurrence of vascular problems and the clinical signs fallen to 52%, due to a hematoma. The patient underwent (2). Because of these disadvantages, a more objective a successful re-exploration of the free flap. and continuous method is desirable. Conclusion : Our data showed that NIRS is an easy, tissue-specific, objective and sensitive way to monitor Ellen Thys, MD ; Joris De Puydt, MD ; Filip Thiessen, MD ; tissue oxygenation in a continuous manner, after free flap Thierry Tondu, MD ; Vera Saliden, MD, PhD breast reconstructions. (*) Department of Anesthesiology, University Hospital Ant- werp, 2650 Edegem, Belgium (**) Plastic Surgery Unit, University Hospital Antwerp, 2650 Keywords : Breast reconstruction ; oximetry ; Near Edegem, Belgium infrared spectroscopy ; non-invasive monitoring. Corresponding author : Department of Anesthesiology, Uni- versity Hospital Antwerp, Wilrijkstraat 10, 2650 Edegem, Belgium. E-mail : [email protected] Introduction Paper submitted on Mar 13, 2020 and accepted on May 26, Breast reconstructions with deep inferior epigastric 2020. artery and vein perforator flaps (DIEP-flaps) are com- Conflict of interest : none.

01-Thys.indd 1 12/01/2021 12:08 2 e. thys et al. Near-infrared spectroscopy (NIRS) is an easy-to- use, continuous, objective and noninvasive method. It provides real-time information about oxygen delivery. NIRS was first described in 1977 (5). NIRS uses infrared light waves to detect oxygen content of tissue. Near- Fig. 1. – Study design. a. NIRS in unilateral DIEP-flap infrared light waves are used to measure the level of breast reconstruction ; b. NIRS in bilateral DIEP-flap breast oxygen the hemoglobin of the tissue contains (= tissue reconstruction.

oxygen saturation, StO2). The tissue oxygen saturation

(StO2) measurements are based on the principles of received verbal and written information about this study optical spectrometry and the law of Lambert-Beer. during the preoperative consultation. If no exclusion Infrared light of five different wavelengths is emitted on criteria were present, the patients were included in our the flap. The flap reflects a part of the light waves on two study. The study design is illustrated in figure 1. ® ® detectors. These detectors measure the StO2. It reflects the In our study, the Foresight device and the Foresight microperfusion and the oxygen consumption in the free pediatric sensors (CASMED, Brandford, CT, USA) were flap. Changes in blood pressure and the supplementation used. The StO2 measurement depth is determined by the of oxygen to the patient influence the results (6). NIRS distance between the emitting and receiving source. The can detect ischemia and hypoxia. The continuous data Foresight® pediatric sensors infrared light penetrates up are a quantitative and objective parameter to evaluate to a depth of 0.75 to 2 cm in the tissue. The width of the tissue oxygenation. The method is currently used to measurement is 4 cm. The near infrared light consists of detect cerebral and regional tissue ischemia (7-12) during five different wavelengths (690, 730, 770, 810 and 870 multiple surgical procedures ; cardiac surgery, carotid nm). An algorithm analyzes the reflected light waves. endarterectomy, the detection of ischemia in shock, the Demographic patient details (age, body mass detection of compartment syndrome, pediatrics, etc. (10, index, diabetes mellitus, cardiovascular comorbidities, 13). substance abuse, ASA score, hemoglobin) were obtained During plastic surgery, NIRS has been used as an from the digital patient record. early predictor of vascular complications (14, 15). But Anesthesia was performed according to the there still is a lack of information about the predictability standard of care treatment and adapted to provide of data provided by Foresight® in monitoring of free optimal flap perfusion. The patients received analgesia, flaps. The technology is easy to use and inexpensive. It vasodilators and fluid titrated on an individual basis. does not harm the patient or the free flap. During surgery heart rate, non-invasive and invasive blood pressure, pulse oxygen saturation, temperature, Methodology respiratory rate, operation time and adverse events (need for blood transfusion or inotropic medication, pulse Our trial was designed as a single cohort oxygen desaturation <90 % or allergic/anaphylactic interventional pilot study. The study was conducted reaction) were recorded. with the approval of the Ethics Committee of the The self-adherent disposable Foresight® sensors University Hospital of Antwerp (Internal reference were placed in the center of the free flap on a surface number : 18/12/173 ; Belgian registration number : area of 16 cm², at the end of the surgery. Another sensor B300201836129). All patients scheduled for uni- or was placed on intact native abdominal tissue and served bilateral breast reconstruction by means of a DIEP-flap as a personalized reference value, as shown in figure 2. were included. The elective surgery was performed For practical reasons, such as sterility and surgical access by a plastic surgeon. The patients were recruited from to the work field, the StO2 measurements only started September 1, 2018 until November 1, 2019 in the directly after completing the surgery. University Hospital of Antwerp. The patients had to be After surgery, all patients were monitored on the older than 18 years. Exclusion criteria were a refusal of Post Anesthetic Care Unit (PACU) during 12 hours. informed consent, a redo-surgery and a known history Non-invasive and invasive blood pressure, pulse oxygen of clotting problems. Clotting problems can be the cause saturation, heart rate, diuresis, blood loss, and body of bleeding or thrombosis and may be detrimental to the temperature were assessed on the follow-up papers. outcome of the flap. The postoperative policy in patients In the first 12 hours postoperatively noninvasive with clotting problems should be individualized. In StO2 measurements were performed every 2 seconds and agreement with the plastic surgeon, we excluded these patients. We screened anamnestically for coagulation problems but did not screen with a blood test, due to the low incidence in the general population. After all, current guidelines do not recommend routine coagulation testing (16). All patients were invited in advance for a preoperative consultation. Based on anamnestic data and data from the clinical examination, it was determined whether the patient was ‘fit for surgery’. All patients Fig. 2. – NIRS setup.

01-Thys.indd 2 12/01/2021 12:08 near-infrared spectroscopy for monitoring tissue oxygenation 3 stored for analysis. It was checked whether the criterion (Tables 1 and 2). All patients had undergone a mastec-

of ‘∆StO2 decrease/h > 15 %’ used by M.H. Steele (17), tomy for preventive of curative breast cancer treatment. was achieved. The mean hemoglobin level at the start of the surgery Every hour the temperature, skin color and Doppler was 13.28 ± 0.78 g/dL (mean ± SD). Only one patient signals of the free flap were monitored. The temperature received blood transfusion during the surgery. On day 1 was subjectively determined by skin contact. In the event postoperatively, the mean hemoglobin was 9.98 ± 1.21

of abnormal StO2 measurements or abnormal clinical g/dL (mean ± SD). No non-Caucasian patients were signs, hemodynamic optimization was achieved based on included in the study. Four patients had been smoking individual needs and the plastic surgeon was contacted. in the past. None of the patient had diabetes. Only one Statistical analysis of the data was performed using patient had an ASA score of three, the other patients had the SPSS 22 software (SPSS, INC., Chicago, IL, USA). an ASA score of two or less. Anesthesia was adapted The Kolmogorov-Smirnov test and the Shapiro-Wilk to provide optimal flap perfusion. Ten out of twelve were used to check the data for normality. Because of patients were treated with a dobutamine infusion during the small group of patients, a Wilcoxon signed-ranks test the surgery, to achieve adequate perfusion pressures. All was used to determine the significance of the difference patients received a piracetam infusion postoperatively

between the StO2 values of the free flap and the native (Tables 3 and 4).

tissue. A probability value < 0.05 was considered The StO2 values are shown per 3 hours in table 5. significant. The mean StO2 value of the flap tissue measured 90.16 ± 4.63 % (mean ± SD). The mean StO2 value measured Results at the native tissue was higher, amounting 91.52 ± 2.23 % (mean ± SD). The Shapiro-Wilk test and the A total of twelve patients were included in the study. Kolmogorov-Smirnov test showed that our values are

The mean age was 47 ± 11 years (mean ± SD). The mean not normally distributed. The mean difference in StO2 body mass index was 25.8 ± 4.1 kg/m (mean ± SD). 50 measurements between the free flap and the native tissue % of the patients underwent an unilateral reconstruction was 1.36 ± 3.46 % (mean ± SD). The difference was not significant (p-value > 0.05) according to the Wilcoxon Table 1 signed-ranks test (Table 6). The Wilcoxon signed-ranks Demographics test showed a p-value of 0.308. This result is reassuring.

Number of women 12 If the mean difference in StO2 measurements between the Number of (ex) smokers 4 free flap and the native tissue was significant, we would Number of diabetics 0 be dealing with poorly oxygenated flaps. The flaps would Number of ASA 2 patients 10 not be viable.

Number of ASA 1 patients 1 In five patients, the StO2 values of the native tissue Number of ASA 3 patients 1 and the flap tissue showed consistently comparable Number of cardiovascular patients 1 values, throughout the entire measurement duration. Number of bilateral DIEP flap 6 In three patients, a very limited tissue oxygenation Number of unilateral DIEP flap 6 response was noted in the first hours after surgery. This Mean Age 47.1 ± 10.7 y response has already been described by Vranken et al. Mean Weight 68.44 ± 10.25 kg Mean BMI 25.8 ± 4.1 kg/cm2 (4). In the study of Vranken et al., the DIEP-flap showed Mean hemoglobin preoperatively 13.28 ± 0.78 g/dL a lower and more extended recovery phase. After this Mean hemoglobin postoperatively 9.98 ± 1.21 g/dL period, the StO2 values of the flap tissue remained stable.

Table 2 Demographics - extended Patient Age Weight BMI Smoker Radiation ASA Hemoglobin Cardio- Bilateral Hemoglobin Hemoglobin number (year) (kg) (kg/cm2) therapy score preoperatively vascular postoperatively drop (g/dL) (g/dL) problems (g/dL) 1 49 80.0 28.5 No present 2 13.5 absent no 11.6 1.9 2 52 85.0 30.1 No present 2 14.7 absent yes 9.8 4.9 3 66 82.4 28.5 No absent 2 13.6 present no 10.9 2.7 4 47 57.3 20.3 No present 2 12.8 absent no 8.3 4.5 5 46 69.9 26.0 No present 2 14.0 absent no 10.3 3.7 6 41 62.0 23.9 Yes present 3 14.1 absent yes 10.8 3.3 7 41 50.5 17.1 No present 2 12.7 absent no 10.3 2.4 8 42 60.2 26.4 Yes present 1 13.7 absent yes 10.4 3.3 9 33 74.0 33.0 No present 2 12.2 absent yes 10.5 1.7 10 60 61.0 24.8 Yes present 2 13.4 absent yes 7.0 6.4 11 60 67.0 23.7 No present 2 12.1 absent yes 9.1 3.0 12 28 72.0 27.4 Yes present 2 12.5 absent no 10.7 1.8

01-Thys.indd 3 12/01/2021 12:08 4 e. thys et al. Table 3 Peroperative course Bloodloss Oxygen Use of Allergic Use of Temperature Glycemia Use of saturation blood products reactions vasopression piracetam > 1000 ml < 90 % Present Present Present <= 35°C of <= 70 of >= Present >=38.5° C 180 mmol/l Number of patients 1 0 1 0 10a 0 0 12 Missing data 4 1 0 2 0 2 2 0 a All patients received a dobutamine infusion.

Table 4 Postoperative course Number of patients Heartrate < 60 bpm or > 110 bpm 0 - 12 h 1 Day 1 0 Day 2,3,4 1 Fig. 3. – Patient 2. Systolic <= 90 mmHg 0 - 12 h 5 bloodpressure Day 1 0 fallen to 52% (Table 5 and Fig. 3). The hemoglobulin Day 2,3,4 1 value had fallen from 14.7 g/dL preoperatively to 9.8 Oxygen < 90 % 0 - 12 h 0 saturation Day 1 0 g/dL postoperatively, mainly due to hemodilution and Day 2,3,4 0 peroperative blood loss. The hemoglobin value never Diuresis < 1000 ml/24h or < 500 0 - 12 h 2 met the transfusion criteria, namely a value of <7 g/dL. ml/12h Day 1 1 The color of the flap was reported as pale. The Doppler Day 2,3,4 - signal remained clearly audible. Since hemodynamic Bloodloss > 500 ml/24h or >250 0 - 12 h 0 optimization had already been achieved, the plastic ml/12h Day 1 1 surgeon was contacted and a large hematoma was drained. Day 2,3,4 - The patient was able to leave the hospital in good health Colour Pale or blue 0 - 12 h 1 after a few days. The criterion of ‘∆StO2 decrease/h > Day 1 2 15 %’ could only be achieved in this patient. The StO2 Day 2,3,4 2 values decreased before the appearance of abnormal Temperature <= 35°C or >= 38,5°C 0 - 12 h 0 clinical signs. We were therefore able to implement the Day 1 0 results of the NIRS in the decision-making regarding the Day 2,3,4 1 revision of the free flap. Doppler Absent 0 - 12 h 0 Day 1 0 In two other patients, the color of the flap was Day 2,3,4 0 reported as problematic and pale. However, the NIRS Skin irritation Present 0 - 12 h 0 values and the Doppler signals were preserved at all Day 1 0 times. So the arterial and venous blood flow was always Day 2,3,4 0 maintained and there was no need for a re-exploration. No patient suffered from skin irritation from the Vrancken et al. suggests that all patients experience such self-adherent sensors. The sensor can temporarily be a response, to a greater or lesser extent. The response is disconnected when necessary. Afterwards there is no difficult to detect in some patients due to small differences need for recalibration. NIRS monitoring is also not between the measurements of the native tissue and the influenced by patient movement in contrast to Doppler flap tissue. The clinical significance of this response is not measurements. yet clear. In the other nine patients, a tissue oxygenation

response period was not noted. The StO2 measurements Discussion and conclusion remained very stable during the postoperative period. Most patients underwent an uncomplicated post- The aim of this study is to determine whether NIRS operative course and were able to leave the hospital after is a suitable way for assessing tissue perfusion after five days. One patient underwent a re-exploration of the DIEP-flap breast reconstruction. Our data showed that flap. The 52-year-old woman underwent bilateral DIEP- NIRS is an easy, tissue-specific, objective and sensitive flap reconstruction. The flap perfusion was optimized way to monitor tissue oxygenation in a continuous by a dobutamine infusion during the initial surgery and manner, after free flap breast reconstructions. This a piracetam infusion postoperatively. No adverse events type of surgery is already common practice in many were reported during the surgery. During her stay at the hospitals. NIRS seems to be a promising clinical tool PACU, a constant increased difference between the StO2 in assessing the flap viability. It can identify early signs measurements of the native tissue and the flap tissue of vascular complications (2, 10, 18). Persistently low

was noted. The StO2 values of the free flap tissue had tissue oxygenation measurements predict poor outcomes

01-Thys.indd 4 12/01/2021 12:08 near-infrared spectroscopy for monitoring tissue oxygenation 5

Table 5

StO2 (%) versus time (h)

T0 T+3 T+6 T+9 T+12 T+15 T+18 1.1 90 91 89 92 92 1.0 92 93 93 93 93 2.1 80 52 53 52 2.2 91 92 93 93 2.0 86 86 88 87 3.1 93 92 93 88 92 92 89 3.0 93 93 94 94 94 94 94 4.1 92 91 89 92 92 91 4.0 92 92 92 93 93 92 5.1 93 93 93 93 93 93 92 5.0 93 92 93 93 93 93 92 6.1 93 93 93 93 93 93 6.2 92 93 93 93 93 92 6.0 92 93 93 93 93 93 7.1 93 93 93 93 93 93 93 7.0 93 93 93 92 93 93 93 8.1 83 89 92 92 89 91 8.2 82 93 93 93 92 93 8.0 90 92 92 92 92 92 9.1 92 89 91 92 90 9.2 88 92 92 92 92 9.0 86 87 92 92 86 10.1 92 88 91 89 89 10.2 93 93 93 92 92 10.0 92 91 87 87 86 11.1 89 92 92 92 92 11.2 81 93 94 94 94 11.0 91 93 93 93 93 12.1 87 90 90 90 92 94 12.0 92 93 92 93 93 94

Table 6 One of the limitations of this study is the small Wilcoxon Signed Ranks Test number of patients studied. Consequently further studies Wilcoxon Signed Ranks Test are necessary to evaluate the use of NIRS. The StO2 values differ among patients, flaps and StO2 free flap - StO2 native tissue even regions of the flap. Therefore, following a trend or a Z -1,020 drop ratio from the initial value seems more appropriate Asymp. Sig. (2-tailed) 0,308 than following absolute values. Further studies are needed to establish the correct criteria or cut-off values. However,

in previous studies (6). The results of our study are the criterion of ‘∆StO2 decrease/h > 15 %’ seems to be a comparable to the results of the literature (2, 4, 10, 14, reasonable standard (17). It is recommended to use NIRS 15, 17-19). as continuous monitoring (18). The drop ratio or trend of Some patients showed a tissue oxygenation the free flap and the native tissue must be compared with response, as previously described. While others patients each other. If both values show a falling trend or a sudden showed a stable tissue oxygenation. In this group of decrease, this indicates a generalized perfusion problem.

patients, the StO2 of the native tissue and the flap tissue Hemodynamic optimization with fluid administration, was comparable. The recovery of the tissue may have blood products or inotropics will be necessary. If only

already occurred before the StO2 measurements started. the values of the free flap show a falling trend or a large In total, one of twelve patients underwent a successful drop ratio, this indicates a flap oxygenation problem and re-exploration for a hematoma. a flap revisional surgery should be performed.

01-Thys.indd 5 12/01/2021 12:08 6 e. thys et al. NIRS monitoring does not inflict any discomfort. 7. Ruf B., Bonelli V., Balling G., Hörer J., Nagdyman N., The technique and study were considered safe and user Braun S.L. and Ewert P., et al. 2015. Intraoperative renal friendly by the patients and doctors. NIRS can even be near-infrared spectroscopy indicates developing acute kidney injury in infants undergoing cardiac surgery with used during the operation (18). It can provide added cardiopulmonary bypass : a case–control study. Crit Care information to compare the pre-, peri- and postoperative 19 : article number : 27. StO2 values. Flap failure usually occurs in the first 8. Shan X.F., Ouyang S.Y., Cai Z.G. and Zhang J. 2014. Evaluation of Foot Perfusion After Fibula Flap Surgery. J hours after surgery (2), but the ideal period of the StO2 measurements has yet to be determined. Although only Craniomaxillofac Surg 25(4) : 1346-1347. 9. De Santis V. and Singer M. 2015. Tissue oxygen tension Caucasian patients participated in the study, melanin monitoring of organ perfusion : rationale, methodologies, skin pigmentation does not interfere with the NIRS and literature review. Br J Anaesth. 115(3) : 357-365. measurements (10). 10. Steppan J. and Hogue C.W. 2014. Cerebral and Tissue The costs of the monitoring device and disposable Oximetry. Best Pract Res Clin Anaesthesiol. 28(4) : 429- sensors must be weighed against the profits in reducing 439. 11. Steffen R.J., Sale S., Anandamurthy B., Cruz V.B., Grady medical cost by eliminating the need for specialized P.M., Soltesz E.G. and Moazami N. 2014. Using Near- nursing and admission on the PACU. More studies are Infrared Spectroscopy to Monitor Lower Extremities needed to evaluate the economic impact. in Patients on Venoarterial Extracorporeal Membrane In some situations, NIRS cannot be used. For Oxygenation. Ann Thorac Surg. 98(5) : 1853-1854. example when the direct application of the NIRS sensor 12. Repez A., Oroszy D. and Arnez Z.M. 2006. Continuous postoperative monitoring of cutaneous free flaps using is not possible, as with heavy edema. Or when the StO2 near infrared spectroscopy. J Plast Reconstr Aesthet Surg. measurements may be incorrect, such as in case of 61(1) : 71-77. hypothermia or changes in metabolism in patients with 13. Murkin J.M. and Arango M. 2009.Near-infrared septic shock (6, 20). spectroscopy as an index of brain and tissue oxygenation. Br J Anaesth. 103 Suppl 1 :i3-13. 14. Colwell A.S., Wright L. and Karanas Y. 2008. Near-infrared spectroscopy measures tissue oxygenation in free flaps References for breast reconstruction. Plast. Reconstr. Surg. 121(5) : 344e-345e. 1. Chae M.P., Rozen W.M., Whitaker I.S., Chubb D., Grinsell 15. Keller A. 2007. Noninvasive tissue oximetry for flap D., Ashton M.W., Hunter-Smith D.J. and Lineaweaver W.C. monitoring : an initial study. J. Reconstr. Microsurg. 23(4) : 2015. Current evidence for postoperative monitoring of 189-197. microvascular free flaps. Ann Plast Surg. 74(5) : 621-632. 16. Developed by the National Guideline Centre, hosted by the 2. Lohman R.F., Langevin C.J., Bozkurt M., Kundu N. and Royal College of Physicians. 2016. Routine preoperative Djohan R. 2013. A Prospective Analysis of Free Flap tests for elective surgery. NICE guideline. Monitoring Techniques : Physical Examination, External 17. Steele M.H. 2015. Three-year experience using near Doppler, Implantable Doppler, and Tissue Oximetry. J infrared spectroscopy tissue oximetry monitoring of free Reconstr Microsurg. 29(1) : 51-56. tissue transfers. Ann. Plast. Surg. 66(5) : 540-545. 3. Futter C.M., Webster M.H., Hagen S. and Mitchell S.L. 18. Kagaya Y. and Miyamoto S. 2017. A systematic review 2000. A retrospective comparison of abdominal muscle of near-infrared spectroscopy in flap monitoring : Current strength following breast reconstruction with a free TRAM basic and clinical evidence and prospects. J Plast Reconstr or DIEP flap. Br J Plast Surg. 53(7) : 578-583. Aesthet Surg. 71(2) : 246-257. 4. Vranken N.P.A., Weerwind P.W., van Onna M.A., Boumanc 19. Lin S.J., Nguyen MD, Chen C., Colakoglu S., Curtis M.S., E.A.C., van der Hulsta R.R.W.J. 2017. Non-invasive tissue Tobias A.M. and Lee B.T. 2011. Tissue oximetry monitoring oximetry following unilateral DIEP-flap reconstruction : A in microsurgical breast reconstruction decreases flap loss pilot evaluation. JPRAS Open 12 : 59-65. and improves rate of flap salvage. Plast Reconstr Surg. 5. Jöbsis F.F. 1977. Noninvasive, infrared monitoring of 127(3) : 1080-1085. cerebral and myocardial oxygen sufficiency and circulatory 20. Guye M.L., Motamed C., Chemam S., Leymarie N., Suria parameters. Science. 23 ; 198(4323) : 1264-1267. S. and Weil G. 2016. Remote peripheral tissue oxygenation 6. Benedik P.S. 2014. Monitoring Tissue Blood Flow and does not predict postoperative free flap complications in Oxygenation : A Brief Review of Emerging Techniques. complex head and neck cancer surgery : A prospective Crit Care Nurs Clin N Am 26 : 345-356. cohort study. Anaesth Crit Care Pain Med. 36(1) : 27-31.

01-Thys.indd 6 12/01/2021 12:08 (Acta Anaesth. Belg., 2020, 71, 7-13) Regional anesthesia of the hip joint : review of regional anesthesia techniques

C. Coucke (*), M. Desmet (**), K. Vermeylen (***), M. Van de Velde (*)

Abstract : Introduction: As hip fractures are frequent nursing home and quality of life is heavily impaired. in the frail population, multiple regional anesthesia Several studies have shown that the use of regional approaches have been developed over the past years to anesthesia techniques improves pain control and achieve analgesia of the hip joint. Achieving optimal reduces complications (3). A knowledge of the analgesia with opioid reduction remains challenging neuro-anatomy of the hip joint is essential here. in this group with high morbidity and mortality. Many Sensory innervations of the hip capsule is provided techniques have changed over the years due to the by branches of the obturator, femoral, superior introduction of ultrasound and an optimisation of radiographic techniques such as MRI. There has been gluteal and sciatic nerves. The greater trochanter an upcoming interest in the fascia iliaca compartment is innervated by a branch of the femoral nerve (4, block (FICB), local infiltration analgesia (LIA) and 5). more recently the pericapsular nerve group block (PENG Since the anterior part of the capsule is most block). The aim of this review is to clarify the role of densely innervated, the lumbar plexus is the region these techniques in hip surgery. of interest to block. The lumbar plexus originates Methods : A systematic search was performed in Medline, from the ventral rami of lumbar spinal nerves L1- Embase and the Cochrane database, with the following 4, and divides in the ilioinguinal, iliohypogatric MeSH terms : conduction anesthesia, nerve block, local and genitofemoral nerves and more importantly anesthesia. in the femoral (FN), obturator (ON) and lateral Results : There is increasing evidence that a supra- femoral cutaneous nerves (LFCN). The fascia iliaca inguinal approach of the FICB is a valuable alternative covers the iliacus and psoas muscle, both muscles for hip analgesia, as there is more success in blocking the three targeted nerves (femoral, obturator and lateral inserting on the lesser trochanter of the femur. The femoral cutaneous nerve), compared to the classical fascia iliaca (FI) as well as both muscles border a infra-inguinal approach. Very recently, the pericapsular virtual compartment, the fascia iliaca compartment, nerve group block has seemed a promising option in extending from the lumbar vertebrae to the lesser providing analgesia of the hip joint, compared to LIA. trochanter. The FN, ON, LFCN all pass through this Conclusion : both the FICB as well as the PENG block compartment, and could theoretically be blocked have been shown to be safe and effective analgesic by one single injection. The aim of this paper is options in hip fracture. However, more research is needed to review different approaches of blocking these to determine the role of both techniques in different hip nerves and hence provide optimal analgesia and conditions. reduce postoperative complications.

Keywords : conduction anesthesia ; nerve block ; local anesthesia.

Introduction C. Coucke, MD; M. Desmet, MD, PhD; K. Vermeylen, MD, PhD; M. Van de Velde MD, PhD, EDRA (*) Department of Anesthesiology, UZ Leuven, Herestraat 49, There has been an increasing incidence of 3000 Leuven, Belgium hip fractures, a condition predominantly affecting (**) Department of Anesthesiology, AZ Groeninge, President the frail geriatric population, with poor outcomes Kennedylaan 4, 8500 Kortrijk, Belgium (***) Department of Anesthesiology, AZ Turnhout, Steenweg for many patients (1). The one year mortality is up op Merksplas, 44, 2300 Turnhout, Belgium to one third of these patients (2). This population Corresponding author: C. Coucke, Department of anesthesi- is at risk for postoperative complications due to ology, UZ Leuven, Herestraat 49, 3000 Leuven, Belgium. opioid consumption or poor pain control, leading E-mail: [email protected] to delirium and delayed return of functional status. Paper submitted on Apr 15, 2020 and accepted on Apr 24, 2020. Hence, many of these patients require admission at a Conflict of interest: none.

02-Coucke.indd 7 12/01/2021 12:09 8 c. coucke et al.

Methods findings were encouraging for more research of the FICB, and more progress was made with the A systematic search was performed in introduction of ultrasound. Medline, Embase and the Cochrane database, in order to review different techniques of blocking Ultrasound techniques the lumbar plexus. The following terms were used : fascia iliaca compartment block, hip analgesia, The first ultrasound guided FICB was described regional anesthesia, pericapsular nerve group block, by Dolan in 2008, by injecting LA just below the local infiltration analgesia. Randomised controlled FI. When compared with the blind technique, this trials were selected from the search results. For the resulted in more sensory loss and motor block. FICB, only results from the last ten years have been Adequate spread of LA is more reliable when seen reviewed, because of the abundance of results. on ultrasound than feeling two pops, because other fascial planes can result in ‘false positive pops’. Results There was almost a doubling in complete sensory loss of the thigh. Nevertheless, only 44% of patients Landmark techniques had a motor block of the ON (13). Given the highly variable cutaneous distribution of the ON, sensory The first anterior landmark technique to loss in the suspected cutaneous area of the ON block the lumbar plexus was described in 1973 could not be interpreted as successful ON block. by Winnie, as an inguinal paravascular injection Reliable ON block can only be tested by loss of technique, analogous to brachial plexus blocks. motor function, thus reduction in adduction strength Paresthesias of the FN were elicited, after inserting of the hip (14). the needle lateral of the femoral artery, and different The technique was modified in 2011 by volumes of local anesthetic (LA) were injected, Hebbard and was described as a supra-inguinal together with manual pressure distal of the needle, approach of the FI, keeping in mind that a more to promote cephalad spread to the lumbar plexus. Winnie concluded that 20 ml of LA was sufficient to block the FN, ON and LFCN, hence called 3-in- 1 block (9). However, this could not be clinically reproduced, neither with imaging techniques. An MRI-study by Swenson showed cephalad spread of LA to the level of the retroperitoneum, as well as medial and lateral spread at the level of the inguinal ligament. Despite this, LA did not reach the tissue plane in between the iliacus and psoas muscle. As a consequence, there was no blocking effect on the ON (10). However, when compared with a FICB, which will be discussed later, postoperative pain scores and opioid consumption were not significantly different (11). As the branches of the lumbar plexus are located posteriorly of the FI, Dalens et al. described a new landmark technique in 1989, the FICB. Drawing a line from the anterior superior iliac spine to the pubic tubercle, the needle was inserted half a centimetre caudal of the point between the lateral third and the medial 2 thirds. Two “pops” are felt, caused by the puncture of the fascia lata and the FI, followed by loss of resistance and the injection of a large volume of LA. Comparable to Winnie’s technique, the LA was massaged more proximally. Fig. 1. — Illustration of four different techniques of a FICB. Dalens compared this technique in children with a ASIS: anterior superior iliac spine ; PUB: pubic symphysis ; 3-in-1 block, and this resulted in a better block of US: ultrasound ; yellow line: femoral nerve ; red line: femoral artery ; blue line: femoral vein. Used with permission of K. the three targeted nerves (90% vs 15%) (12). These Vermeylen (7).

02-Coucke.indd 8 12/01/2021 12:09 regional anesthesia of the hip : review of options 9 proximal approach could lead to a better spread is that a proximal deposition of LA leads to a more under the FI and possibly spread towards the consistent block of the three targeted nerves, as they ON. In his procedure, the probe was placed in a are anatomically more closely related to each other longitudinal direction, a few centimeters below (16,17). the inguinal ligament. Successively the needle is advanced in a cranial direction under the iliac Infiltration techniques fascia up to the supra inguinal level, where LA was injected. However, Hebbard’s cadaver study did Local infiltration analgesia has a long history not investigate involvement of the ON, there was and seems beneficial in avoiding hyperalgesia nevertheless consistently good spread towards the and in providing postoperative pain control (18). FN and LFCN (15). The four previously mentioned However, there has been little evidence and lack techniques are illustrated in figure 1 (7). of randomized trials. Andersen et al. conducted This technique was slightly modified by a randomized, placebo-controlled trial in total hip Desmet, to a real supra-inguinal approach : placing arthroplasty patients, inspired by a non randomized the probe at the level of the anterior superior iliac case series of Kerr and Kohan that noticed earlier spine, then moving medially and caudally the probe mobilization and less opioid consumption using until the bow-tie sign is seen, which is formed LIA (19). The study group received a mixture by the sartorious muscle and the internal oblique of ropivacaine, ketorolac and epinephrine by muscle, the needle is inserted under the FI where peroperative infiltration. A catheter was placed in the LA is injected, as is illustrated in figure 2. It the hip joint, running down the capsule, muscles and then spreads as far as medial to the psoas muscle skin (20). They noticed lower pain scores on the first where the obturator nerve has its course. The idea postoperative day, less need for rescue analgesia and

Fig. 2. — ASIS: anterior superior iliac spine ; IM: iliac muscle ; arrows: iliac fascia ; SM: Sartorius muscle ; IOM: internal oblique muscle ; * bow-tie sign in A, needle in B, deep circumflex artery in C, LA in D. Part A shows relevant anatomy, part B shows a slightly more medial image with needle introduction, part C shows injection of LA and part D is an image after injection of LA. Used with permission of M. Desmet (8).

02-Coucke.indd 9 12/01/2021 12:09 10 c. coucke et al. less joint stiffness. However, these findings could have been reported, especially when the block not be reproduced in multiple randomized trials was performed postoperatively (30). The authors (21-24). A possible explanation is that in these trials hypothesized that surgical disruption of the normal only LA was used, whereas epinephrine and non- planes and anatomy leads to further spread of the steroidal anti-inflammatory drugs (NSAIDs) were LA, towards the FN. Other potential reasons of avoided. Epinephrine was not administered due to motor block are too superficial injection of LA, the risk of tissue necrosis (25) and NSAIDs were not injection of a few ml of LA in hydrolocalisation of used because of risks of renal impairment, bleeding, the needle tip, resulting in FICB or FN block. The cardiovascular adverse events in a general geriatric authors emphasized that the needle should be lateral frail population (26). of the psoas tendon, because a position medial of A new infiltration technique was then the psoas tendon would give good spread of LA, but developed, given the fact that the anterior hip also inadvertent spread towards the FN (28, 30). capsule is most densely innervated, by the ON, FN and accessory ON. The high articular branches Discussion of the latter two are located between the iliopubic eminence (IPE) and the anterior inferior iliac spine Patients needing hip surgery often have a (AIIS) (28). The block was named pericapsular frail, geriatric profile, resulting in high morbidity nerve group block (PENG block) (28-29). The block and mortality (1-3). As a consequence, an ideal was performed with a curvilinear ultrasound probe, treatment would consist of optimal analgesia transversally placed over the AIIS. The needle was requiring the least possible opioids and muscle in plane introduced from lateral to medial, under weakness, in order to mobilize as soon as possible the FN, with the needle tip in the plane between (2, 31). A larger number of regional anesthesia the psoas tendon and the pubic ramus. A total of 20 techniques are available in providing analgesia in ml was injected, with spread of the LA in the plane hip operations (31, 32). The diversity of techniques, under the psoas tendon and above the plane between made possible by anatomical studies, radiological the IPE and AIIS (Fig. 3). imaging studies and the introduction of ultrasound, Half an hour after the injection, the patients have made it difficult to assess the clinical relevance were asked to flex the hip ; they reported less pain of these techniques. It is challenging to point out the compared to baseline and had no clinical evidence optimal timing and dosage of every technique, and of quadriceps weakness (28). The authors believed to compare them. they could better reach the articular branches of the A Cochrane systematic review found only accessory ON and FN, because they are located moderate quality of evidence in favour of peripheral between the AIIS and IPE, whereas an FICB or nerve blocks compared to systemic analgesia, low 3-in-1 block are not likely to block these branches, quality of evidence in improving patients satisfaction as the spread of LA will not be cephalad enough. and very low quality of evidence in reducing length As the PENG block only blocks sensory branches, of stay or adverse events as pruritus and delirium motor block can be avoided, which is important in (33, 34). However, almost all peripheral nerve block early mobilization. However, cases of motor block techniques were included, and a detailed analysis for every technique might result in different findings. Many studies have described an opioid sparing effect of the FICB (16, 35-39), with even better results with a supra-inguinal approach (40- 42). These findings might be attributed to a better block of the ON with a supra-inguinal approach, which was found in electromyographical studies (43) as well as imaging studies. An MRI study by Vermeylen showed a more cephalad spread of LA, towards the lumbar plexus, with a supra- inguinal approach compared with an infra-inguinal technique on the contralateral side in a volunteer Fig. 3. — Ultrasound image after injection of LA. FA: femoral study, resulting in a significantly higher complete artery ; FV: femoral vein ; FN: femoral nerve ; PT: psoas sensory block of the anterior, lateral and medial tendon ; IPE: iliopubic eminence ; AIIS: anterior inferior iliac spine ; *: spread of LA ; arrow: needle trajectory, lateral to the thigh (44). A meta-analysis comparing FICB with FN. Used with permission of N. Black (38). no block showed reduction in opioid consumption,

02-Coucke.indd 10 12/01/2021 12:09 regional anesthesia of the hip : review of options 11 pain scores and postoperative nausea and vomiting preserved. However, the FICB has the advantage of in favour of the FICB. However, the weak points blocking all relevant cutaneous innervation of the in this meta-analysis were heterogenous FICB thigh. approaches, volumes and concentrations of LA and types of surgery (45). The most recent meta-analysis Conclusion comparing FICB with placebo was restricted to total hip arthroplasty, again with lower pain scores and It can be said that both the FICB as well as opioid consumption. Adverse events were similar in the PENG block are safe and effective techniques both groups and there was no increased fall risk in in providing analgesia of the hip joint. The supra- the FICB group, so the theoretical risk of muscle inguinal approach of the FICB seems to be most weakness proved clinically insignificant (46). effective, but ideal volume and concentration of One randomized trial found that an intra- LA still needs to be investigated. The PENG block articular injection of LA resulted in better pain relief definitely needs to be studied more in detail, and than an FICB in intracapsular hip fractures, although randomized trials are underway. Most likely, both the authors used an infra-inguinal FICB approach blocks are not interchangeable as they will each (47). A more recent trial demonstrated similar pain have their own benefit in different clinical situations, relief with peri-articular infiltration compared with which still remain to be determined. However, up supra-inguinal FICB, although muscle weakness to now, both techniques have been promising in was significantly more associated in the latter achieving lower pain scores and less need of opioids. technique. Nonetheless, the clinical impact of these These parameters are of paramount importance in findings remained unclear and few trials could the frail, geriatric population suffering from hip demonstrate any beneficial effect of LIA. conditions with high morbidity and mortality. With the PENG block being a relatively new kid on the block, it might be important to clinically References evaluate when this block could be beneficial compared to an FICB (27-28). Patients needing an 1. Nirav A., Jacob A., Travis F. and Hrayr B. 2017. Nerve Blocks in the Geriatric Patient With Hip Fracture : A Review elective total hip arthroplasty, most likely due to of the Current Literature and Relevant Neuroanatomy. hip osteoarthritis, are usually younger patients than Geriatric Orthopaedic Surgery & Rehabilitation. 8(4): 268- patients needing emergency surgery because of a 275. 2. Dyer S., Crotty M., Fairhall N., Magaziner J., Beaupre L. hip fracture. Patients with hip fracture might benefit and Cameron I., et Al. 2016. A critical review of the long- more from a PENG block, whereas patients needing term disability outcomes following hip fracture (Fragility total hip arthroplasty are most likely better off with Fracture Network Rehabilitation Research Special Interest an FICB, depending on the surgical incision site. If Group). BMC Geriatrics. 16: 158-175. 3. Freeman N. and Clarke J. 2016. Perioperative pain a posterior approach is used, the patient is not likely management for hip fracture patients. Orthopaedics and to benefit from an FICB. So far, no large randomized Trauma. 30 (2): 145-152. clinical trials have been published concerning the 4. Genth B., Von During M., Von Enghelhardt L., Ludwig J., Teske W. and Von Schulze-pellengahr C. 2012. Analysis PENG block, all available studies are very small in of the Sensory Innervations of the Greater Trochanter for patient number and rather descriptive (48, 49). First Improving the Treatment of Greater Trochanteric Pain of all, comparison of the PENG block with placebo Syndrome. Clinical Anatomy. 25: 1080-1086. 5. Birnbaum K., Prescher A., Hessler S. and Heller D. 1997. is necessary, followed by trials comparing this new The sensory innervation of the hip joint - An anatomical technique to other regional anesthesia techniques study. Surg Radiol Anat. 19: 371-375. in hip surgery as a supra-inguinal approach of the 6. Capdevila X., Coimbra C. and Choquet O. 2005. Approaches FICB. to the Lumbar Plexus: Success, Risks, and Outcome. Reg Anesth Pain Med. 30 (2): 150-162. Very recently, several of these trials have 7. Vermeylen K., Van Herreweghe I., Leunen I., Soetens F. already started. One small study investigated pain and Van de Velde M. 2019. History of the fascia iliaca scores, onset of action and duration of block in compartment block. Acta Anaesth. Belg. 70: 131-140. 8. Desmet M., Balocco A. and Van Belleghem V. 2019. patients with fracture of the femoral neck. Half Fascia iliaca compartment blocks: Different techniques and of them received a FICB, whereas the other half review of the literature. Best Practice & Research Clinical received a PENG block. The PENG block had a Anaesthesiology. 33: 57-66. 9. Winnie A., Ramamurthy S. and Duranni Z. 1973. The quicker onset of action, compared with an FICB, Inguinal Paravascular Technique of Lumbar Plexus but overall pain control and duration of the block Anesthesia: The “3 - in - 1 Block”. Anaesth Analg. 5: 1-9. were similar in both groups (50). Compared with 10. Swenson J., Davis J. and Stream J. 2015. Local anesthetic FICB, the advantage of the PENG block, being a injection deep to the fascia iliaca at the level of the inguinal ligament: the pattern of distribution and effects on the true capsular block, is that motor function will be obturator nerve. J Clin Anesth. 27: 652-657.

02-Coucke.indd 11 12/01/2021 12:09 12 c. coucke et al. 11. Deniz S., Atim A., Kurklu M., Cayci T. and Kurt E. 2014. 28. Black N. and Chin KJ. 2019. Pericapsular nerve group Comparison of the postoperative analgesic efficacy of an (PENG) block: Comments and practical considerations. ultrasound-guided fascia iliaca compartment block versus 3 Journal of Clinical Anesthesia. 56: 143-144. in 1 block in hip prosthesis surgery. AĞRI. 26(4):151-157.. 29. Delgado A. and Portugalyan M. 2019. Pericapsular nerve 12. Dalens B., Vanneuville G. and Tanguy A. 1989. Comparison group block (PENG) for postoperative analgesia following of the fascia iliaca compartment block with the 3-in-1 block femoral neck fracture surgery (dynamic hip screw). Reg in children. Anaest Analg. 69: 705-713. Anesth Pain Med. 44: A221-A222. 13. Dolan J., Williams A., Murney E., Smith M. and Kenny G. 30. Yu H., Moser J. and Chu A. 2019. Inadvertent quadriceps 2008. Ultrasound Guided Fascia Iliaca Block: A Comparison weakness following the pericapsular nerve group (PENG) With the Loss of Resistance Technique. Anesth Pain Med. block. Reg Anesth Pain Med. 44: 611-613. 33: 526-531. 31. Macfarlane A., Prasad G., Chan V. and Brull R. 2009. 14. Bouaziz H., Vial F., Jochum D., Macalou D., Heck M. Does regional anaesthesia improve outcome after total hip and Meuret P. et Al. 2002. An Evaluation of the Cutaneous arthroplasty? A systematic review. BJA. 103 (3): 335-45. Distribution After Obturator Nerve Block. Anesth Analg. 32. Newman B., McCarthy L., Thomas P., May P., Layzell M. 94: 445-449. and Horn K. 2013. A 1comparison of pre-operative nerve 15. Hebbard P., Ivanusic J. and Sha S. 2011. Ultrasound-guided stimulator-guided femoral nerve block and fascia iliaca supra-inguinal fascia iliaca block: a cadaveric evaluation of compartment block in patients with a femoral neck fracture. a novel approach. Anaesthesia. 66: 300-305. Anaesthesia. 68: 899-903. 16. Desmet M., Vermeylen K., Van Herreweghe I., Carlier 33. Atchabahian A., Schwartz G., Hall C., Lajam C. and L., Soetens F. and Croes K., et Al. 2017. A Longitudinal Andreae M. 2015. Regional analgesia for improvement Supra-Inguinal Fascia Iliaca Compartment Block Reduces of long-term functional outcome after elective large joint Morphine Consumption After Total Hip Arthroplasty. Reg replacement. Coch Database System Rev. 8: CD010278. Anesth Pain Med. 42: 327-333. 34. Guay J., Johnson R. and Kopp S. 2017. Nerve blocks or no 17. Singh H. and Jones D. 2013. Hourglass-Pattern Recognition nerve blocks for pain control after elective hip replacement Simplifies Fascia Iliaca Compartment Block. Reg Anesth (arthroplasty) surgery in adults. Coch Database System Pain Med. 38: 467-468. Rev. 10: CD011608. 18. Dahl J., Moiniche S. and Kehlet H. 1994. Wound infiltration 35. Gao Y., Tan H., Sun R. and Zhu J. 2019. Fascia iliaca with local anaesthetics for postoperative pain relief. Acta compartment block reduces pain and opioid consumption Anaesthesiol Scand. 38: 7-14. after total hip arthroplasty: A systematic review and meta- 19. Kerr D. and Kohan L. 2008. Local infiltration analgesia: analysis. Int J Surg. 65: 70-79. a technique for the control of acute postoperative pain 36. Thompson J., Long M., Rogers E., Pesso R., Galos D., following knee and hip surgery: A case study of 325 Dengenis R. and Ruotolo C. 2020. Fascia Iliaca Block patients. Acta Orthopaedica. 79 (2): 174-183. Decreases Hip Fracture Postoperative Opioid Consumption: 20. Andersen L., Poulsen T., Krogh B. and Nielsen T. 2007. A Prospective Randomized Controlled Trial. J Orthop Postoperative analgesia in total hip arthroplasty: A Trauma. 34 (1): 49-54. randomized double-blinded, placebo-controlled study on 37. Hong H. and Ma Y. 2019. The efficacy of fascia iliaca peroperative and postoperative ropivacaine, ketorolac, and compartment block for pain control after hip fracture: A adrenaline wound infiltration. Acta Orthopaedica. 78 (2): meta-analysis. Medicine. 98:28 (e16157). 187-192. 38. Stevens M., Harrison G. and McGrail M. 2007. A modified 21. Tan N., Gotmaker R. and Barrington M. 2019. Impact of fascia iliaca compartment block has significant morphine- Local Infiltration Analgesia on the Quality of Recovery sparing effect after total hip arthroplasty. Anaesth Intensive After Anterior Total Hip Arthroplasty: A Randomized, Care. 35: 949-952. Triple-Blind, Placebo-Controlled Trial. Anesth Analg. 129: 39. Bin Y., Miao H., Guang-Yu C., Tian-Xiao Z. and Na Z. 1715-1722. 2016. Ultrasound-guided continuous femoral nerve block 22. Bech R., Ovesen O., Lauritsen J., Emmelut C., Lindholm P. vs continuous fascia iliaca compartment block for hip and Overgaard S. 2018. Local Anesthetic Wound Infiltration replacement in the elderly. Medicine. 95:42 (e5056). after Osteosynthesis of Extracapsular Hip Fracture Does 40. Perry C., Fahs A., Kurdziel M., Koueiter D., Fayne R. and Not Reduce Pain or Opioid Requirements: A Randomized, Verner J. 2018. Intraoperative Psoas Compartment Block Placebo-Controlled, Double-Blind Clinical Trial in 49 vs Preoperative Fascia Iliaca Block for Pain Control After Patients. Pain Res Manag. Article ID 6398424. Direct Anterior Total Hip Arthroplasty: A Randomized 23. Apratoa A., Audisioa A., Santorob A., Grossoa E., Controlled Trial. The Journal of Arthroplasty. 33: 1770- Devivoa S., Berardinob M. and Massèa M. 2018. Fascia- 1774. iliaca compartment block vs intra-articular hip injection 41. Zhou Y., Zhang W., Chong H., Xi Y., Zheng S., Wang G. for preoperative pain management in intracapsular hip and Wu X. 2019. A Prospective Study to Compare Analgesia fractures: A blind, randomized, controlled trial. Injury, Int. from Femoral Obturator Nerve Block with Fascia Iliaca J. Care Injured. 49: 2203-2208. Compartment Block for Acute Preoperative Pain in Elderly 24. Kuchálik J., Magnuson A., Lundin A. and Gupta A. 2017. Patients with Hip Fracture. Med Sci Monit. 25: 8562-8570. Local infiltration analgesia: a 2‑year follow‑up of patients 42. Kumar K., Kumar R., Pandey A., Bhall A., Kashyap L. undergoing total hip arthroplasty. J Anesth. 31: 837-845. and Garg R. et Al. 2015. Comparison of conventional 25. Toftdahl K., Nikolajsen L., Haraldsted V., Madsen F., infrainguinal versus modified proximal suprainguinal Tonnesen E. and Soballe K. 2007. Comparison of peri- and approach of Fascia Iliaca Compartment Block for post- intra-articular analgesia with femoral nerve block after operative analgesia in Total Hip Arthroplasty. A prospective total knee arthroplasty: a randomized clinical trial. Acta randomized study. Acta Anaesth. Belg. 66: 95-100 Orthopaedica. 78: 172-179. 43. Yuying Q., Juanjuan G., Qunying Z., Xiaofeng A., 26. Souter A., Fredman B. and White P. 1994. Controversies Huansheng H., Fengqig Z. and Xianghe W. 2020. Electro- in the perioperative use of nonsterodial anti-inflammatory myographic Comparison of the Efficacy of Ultra-sound- drugs. Anesth Analg. 79: 1178-1190. guided Suprainguinal and Infrainguinal Fascia Iliaca 27. Girón-Arang L., Peng P., Chin KJ., Brull R. and Perlas A. Compartment Block for Blockade of the Obturator Nerve 2018. Pericapsular Nerve Group (PENG) Block for Hip in Total Knee Arthroplasty. The Clinical Journal of Pain. Fracture. Reg Anesth Pain Med. 43: 859-86. 36 (4): 260-266.

02-Coucke.indd 12 12/01/2021 12:09 regional anesthesia of the hip : review of options 13 44. Vermeylen K., Desmet M., Leunen I., Soetens F., Neyrinck 47. Gasanova I., Alexander J. and Estrera K. 2019. Ultrasound- A. and Carens D. et Al. 2019. Supra-inguinal injection for guided suprainguinal fascia iliaca compartment block fascia iliaca compartment block results in more consistent versus periarticular infiltration for pain management after spread towards the lumbar plexus than an infra-inguinal total hip arthroplasty: a randomized controlled trial. Reg injection: a volunteer study. Reg Anesth Pain Med. 44: Anesth Pain Med. 44: 206-211. 483-491. 48. Ueshima H. and Otake H. 2018. Clinical experiences of 45. Yang L., Li M., Chen C., Shen J. and Bu X. 2017. Fascia pericapsular nerve group (PENG) block for hip surgery. iliaca compartment block versus no block for pain control Journal of Clinical Anesthesia. 51: 60-61. after lower limb surgery: a meta-analysis. Journal of Pain 49. Fusco P., Di Carlo S. and Paladini G. 2019. Could the Research. 10: 2833-2841. combination of PENG block and LIA be a useful analgesic 46. Zhang X. and Ma J. 2019. The efficacy of fascia iliaca strategy in the treatment of postoperative pain for hip compartment block for pain control after total hip replacement surgery? Reg Anesth Pain Med. 44: 531-532. arthroplasty: a meta-analysis. Journal of Orthopaedic 50. Bhattacharya A., Bhatti T. and Haldar M. 2019. Pericapsular Surgery and Research. 14: 33-39. nerve group block – is it better than the rest for pain relief in fracture neck of femur? Reg Anesth Pain Med. 44: A116.

02-Coucke.indd 13 12/01/2021 12:09

(Acta Anaesth. Belg., 2020, 71, 15-20) Sevoflurane consumption with the Flow-i ventilator in two versions of automatic gas control algorithms and two settings of manually controlled anesthesia : an economic and ecological assessment

F. De Rydt (*,**), A. Kalmar (*), M. Van de Velde (**)

Abstract : Volatile anesthetics have long-lived increased interest to reduce the ecological footprint greenhouse gas properties. These molecules undergo resulting from anesthesia practice (1,4). In this virtually no biotransformation in vivo, so in a single respect, it is important to realize that volatile patient they can be recirculated in the closed-circuit anesthetics undergo virtually no transformation in system or expelled into the atmosphere. In this respect, humans (1). The consumption of a volatile agent is manufacturers of modern-day anesthesia machines have type-specific (MAC-specific), time-dependent and developed automated target control low-flow tools for proficient volatile delivery. Volatile agent consumption largely determined by the FGFs anesthesiologists at the Maria Middelares Hospital, Ghent, within four use to saturate the ventilator and patient with defined flow control groups was observed retrospectively narcotics (1,5-7). Pursuing lower FGFs and agent by chronologically assigning the first 25 encountered consumption to deliver sustainable anesthesia can patients (who met inclusion criteria) to their respective result in major financial and environmental benefits group. In these groups, either (1) a Minimal (300 mL (1, 6, 8). -1 min ) or (2) a 2-liter FGF (2L FGF) ; (3) AGCV4.04 or Historically, the absence of reliable gas moni- the newer (4) AGCV4.07 was used to provide sustainable toring, trustworthy vaporizers and early recognition anesthesia. Digitalized data was extracted and visualized alarms were an important limitation to this practice for each case. Average (standard deviation ; SD) (9-12). Minimal flow anesthesia lowers remains to consumption was calculated at 5 minutes intervals. After be considered as labor-intensive and risks being 45 minutes, the average (SD) consumption of sevoflurane performed suboptimal at moments where other was 14,4 (1,3) ; 5 (1,2) ; 7 (1,4) ; and 6,3 (1,3) mL in Ò aspects of anesthesia care require focused attention the 2L FGF; Minimal ; AGC V4.04 and AGCV4.07 group, respectively. Compared to Minimal, AGCÒ groups (8, 12, 13). In this respect, there are opposing stakes consumed significantly more sevoflurane at the early between ecological/economic concerns and stability stage of anesthesia, but then decreased to a comparable of anesthetic delivery (8, 10, 12-14). rate. Our results indicate that manually regulated minimal Minimal flow anesthesia requires frequent flow is more economic and ecological, compared to manual adjustment of the ventilator settings in order AGCÒ. Conventional “low-flow anesthesia” practice to avoid accidental hypoxia or the inappropriate (e.g. 2L-FGF) should be aborted as AGCÒ offers a vital dosage of anesthetic agents (due to various solution to the opposing stakes of ecological/economic reasons, such as circuit leakage or distraction of concerns and stability/ease of anesthetic delivery, but the anesthesiologist) (12, 15, 16). Except for rare the algorithms should be furtherly optimized. Timely situations, such as carbon monoxide poisoning upgrade towards most recent software versions can or decompensated diabetic states, there are no deliver significant cost savings and a reduction of the ecological footprint.

Keywords : Volatile anesthetic ; closed-circuit anes- Frederic De Rydt, MD ; Alain Kalmar MD, PhD ; Marc Van de Velde, MD, PhD thesia ; vaporizers ; greenhouse gas ; health care econo- (*) Department of Anesthesiology, University Hospitals of the mics and organizations. KU Leuven, Herestraat 49, 3000 Leuven, Belgium (**) Department of Anesthesia and Critical Care Medicine, Maria Middelares Hospital, Buitenring Sint-Denijs 30, 9000 Introduction Ghent, Belgium Corresponding author: Frederic De Rydt, MD. Dpt of Anesthesia and Critical Care Medicine, Maria Middelares Volatile anesthetics have long-lived greenhouse Hospital, Buitenring Sint-Denijs 30, 9000 Ghent, Belgium. gas properties and evaporate to the atmosphere E-mail: [email protected] after clinical application (1-3). With growing Paper submitted on Apr 19, 2020 and accepted on Apr 24, acknowledgement of the disastrous consequences 2020. of a derailed greenhouse effect, there has been an Conflict of interest: none.

03-De Rydt.indd 15 12/01/2021 12:10 16 f. de rydt et al. contraindications to perform low flow anesthesia can be used, depending on the clinician’s personal (10-12). In this respect, manufacturers of modern- preference, experience and clinical circumstances. day anesthesia machines developed automated Ventilatory data and the sequential consumption target-control low-flow tools for volatile delivery of sevoflurane is being reported into ICCA at a 15 (16-20). seconds interval by the Flow-i with a precision of The Flow-i anesthesia machine (Getinge, 0.1ml (26). Goteborg, Sweden), for instance, can be supplied Case selection and inclusion with AGCÒ (16,19). In AGCÒ , the anesthesiologist chooses the appropriate speed 1 (slow) to 8 (fast) to Assuming a normal distribution of the reach desired end-tidal concentrations of anesthetics consumption data, we considered a mean difference (16,19,21). The AGCÒ algorithm gradually lowers of 1 ml after 45 minutes between AGC® and minimal the FGF to a rate of 300 ml min-1 (16,19). Several flow to be significant (estimated SD of 1.1 ml, based studies have established that low flow anesthesia on pilot data). To detect this difference with an α of and automated gas control software can significantly 0.05 and a power of 0.95, a total of 25 records are reduce consumption of anesthetic agent (19, 22, needed (27). 23). The use of automated software eliminates the The consumption patterns of sevoflurane within requirement of frequent manual adjustment of the four defined flow control groups were observed ventilator settings during minimal flow anesthesia retrospectively by chronologically assigning the and in addition optimizes the stability of the first 25 encountered patients (who met inclusion administered anesthetics and oxygen (16, 17, 19, criteria) to each group. 24). In the first two groups, records were included This study aimed to compare the sevoflurane where AGCÒ was continuously employed at consumption in a) conventional low flow anesthesia standard speed 6 during the whole procedure, with

(2L FGF), b) manually adjusted Minimal FGF and a set FiO2 (inspiratory oxygen fraction) of 80%, ® AGC software versions c) AGC4.04 and d) AGC4.07 and a target EtSevo of 2%. In the first (AGC4.04) and Ò at different stages of anesthesia. An additional second (AGC4.07) group, AGC software V4.04.01 analysis carried out to describe whether differences and V4.07.00 was used, respectively. In the third in consumption vary in time after the start of volatile group (2L FGF), a continuous FGF of 2 L min-1 was agent delivery. A secondary outcome variable was used. In the fourth group (Minimal), the FGF was

the stability of the end-tidal sevoflurane (Etsevo) manually titrated to attain and maintain an Etsevo of concentrations during steady-state anesthesia. 2% with the lowest possible flow of 300 ml min-1 on the Flow-i ventilator ; at the start of sevoflurane Methods administration and intermittently during the procedure, a higher flow could be applied for a few Hardware and data registration seconds. Data extraction and analysis After institutional ethical approval (MMS. 2019.069), digital anesthesia reports between 01 All anesthetic data were extracted and October 2019 and 31 March 2020 were extracted subsequently imported into Microsoft Excel 2010® and analyzed on the used mode of volatile anesthetic (Microsoft, Redmond, USA) for analysis. The administration, ventilator setting and demographic records were synchronized (T0 was defined as a

data. All patients had given routine written consent Fisevo > 0.2%). The evolution of the values for all for retrospective data research. This manuscript study variables were analyzed for 45 minutes,

adheres to the applicable STROBE guidelines. starting from the moment Fisevo exceeded 0.2%. For Intellispace Critical Care and Anesthesia (ICCA) visual assessment, the evolution of the individual (Philips, Amsterdam, Netherlands) is used for values is depicted in figure 1. For comprehensive recordkeeping of intraoperative data and anesthetic comparison of the different groups, the average interventions such as provided airway and values were shown in figure 2.

administered drugs (25). The time until Etsevo 1.8% was reached in the Ò The Flow-i ventilators are equipped with AGC averaged Etsevo curve was counted as the time to

version V4.04.01 or V4.07. All anesthesiologists reach the desired Etsevo. Ò are aware of the AGC feature, but there is no The average and standard deviation (SD) of Etsevo departmental protocol on preferential settings. As between 25 and 35 minutes after start was counted such, AGCÒ or manually performed flow modes as the stability of sevoflurane administration.

03-De Rydt.indd 16 12/01/2021 12:10 sevoflurane consumption in manual vs automatic modes 17 The rate of sevoflurane consumption at a certain average injection rate of sevoflurane for each group -1 minute (Rm, expressed as ml h ) was calculated as is shown in figure 3. Patient characteristics were

the increase in cumulative consumption (Cm) over comparable between groups.

the coming minute : Rm = (C(m+1)-Cm) *60 (figure 3). The cumulative consumption of sevoflurane Normality was tested with the Kolmogorov- (ml) at different time points for each group, as well Smirnov test. Continuous data are expressed as as the consumption in the 15-45 minutes interval mean (SD). significantly differs between groups (Table 1, Fig. 2). The average (SD) values of the analyzed After 45 minutes, an average (SD) of 14,4 (1,2) ml, variables were determined at 5 minutes intervals. In 7,0 (1,4) ml, 6,3 (1,3) ml and 5,0 (1,2) ml sevoflurane

addition, the cumulative sevoflurane consumption liquid is consumed in 2L FGF, AGC4.04, AGC4.07 and between 15 and 45 minutes was determined. Results Minimal, respectively. The total consumption at 45 were subject to the general linear model repeated minutes is significantly different between all groups measures ANOVA with subsequent unpaired t-tests. (table 1), with Minimal being significantly the most Significance level was set at 5% (p<0.05). proficient, although the difference with both AGC® settings is mainly accrued in the first 10-15 minutes Results after start (Fig. 1).

The standard deviation of the Etsevo was The patient characteristics and main outcome significantly lower in both ACG settings than in data are summarized in table 1. The evolution of minimal flow, indicating more stable sevoflurane the individual cumulative sevoflurane consumption delivery in AGC®. The time to reach the 1.8%

and the individual Etsevo, as well as the average value sevoflurane-target was significantly different be- for each setting is shown in figure 1. For a clear tween groups (Table 1). comparison, the average curves of each of these The anesthetic management was adequate in settings is shown in figure 2. The time course of the all groups and no critical events occurred.

Table 1

Demographics, cumulative sevoflurane consumption, average end-tidal sevoflurane (Etsevo) concentrations during steady-state and time to reach the target value in the different settings

Group Statistical analysis

2L FGF AGC4.04 AGC4.07 Minimal Differences between groups Patient demographics Subjects 25 25 25 25 Male/female 11/14 14/11 9/16 9/16 Weight (kg) 72 (12) 82 (21) 75 (22) 74 (18) p = 0,27 Age 59(15) 55(18) 52(15) 58(14) p = 0,32 Cumulative sevoflurane liquid consumption (ml)

Time (min)

5’ 2,2 (0,3) 1,7 (0,4) 1,7 (0,3) 1,1 (0,4) AGC4.04 and AGC4.07 do not differ (p=1) 10’ 4,0 (0,5) 2,9 (0,7) 2,6 (0,3) 1,7 (0,5) AGC4.04 and AGC4.07 do not differ (p=0,06) 15’ 5,7 (0,6) 3,7 (0,9) 3,3 (0,4) 2,3 (0,5) All differ (p<0,01) 20’ 7,2 (0,7) 4,4 (1,0) 3,8 (0,5) 2,9 (0,6) All differ (p<0,01) 25’ 8,7 (0,8) 5 (1,1) 4,3 (0,7) 3,4 (0,7) All differ (p<0,01) 30’ 10,2 (0,8) 5,5 (1,2) 4,8 (0,8) 3,8 (0,8) All differ (p<0,01) 35’ 11,6 (0,9) 6,0 (1,3) 5,3 (1) 4,3 (1,0) All differ (p<0,01) 40’ 13,1 (1,0) 6,5 (1,4) 5,8 (1,1) 4,6 (1,1) All differ (p<0,01) 45’ 14,4 (1,2) 7,0 (1,4) 6,3 (1,3) 5,0 (1,2) All differ (p<0,01)

AGC4.04 and AGC4.07 do not differ (p=0,22) 15-45’ 8,7 (0,9) 3,3 (0,8) 3,0 (1,0) 2,7 (1,0) AGC4.07 and Minimal do not differ (p=0,22) AGC4.04 and Minimal differ (p=0,012)

Etsevo during steady-state anaesthesia (= between 25 and 35 minutes after start) 25-35’ 2,07 (0,21) % 1,99 (0,09) % 2,03 (0,08) % 2,16 (0,29) %

Time to reach Etsevo of 1,8% (min:sec) 04:00 08:00 05:15 12:15 Results are presented as average values (SD).

03-De Rydt.indd 17 12/01/2021 12:10 18 f. de rydt et al.

Discussion

This study shows that AGC® reduces sevoflurane waste compared to conventional low flow anesthesia (e.g. 2L FGF), while the consumption in Minimal demonstrates room for improvement of the current AGC® algorithms. Reducing waste has financial and ecological benefits. While largely ignored by the medical and political community because of many undue reasons, volatile anesthetics are very potent greenhouse gases (1, 4). As the urgency of decisive climate action grows, so too has interest curbing emissions of

greenhouse gases other than CO2. Following initial clinical administration and after passage through the carbon dioxide absorber, volatile anesthetics can be recirculated. Apart from rare conditions such as CO intoxication, there is no reason not to aim for a minimal gas flow (10). Figure 2 shows a reduction of volatile anesthetic consumption with AGC ® or by swiftly applying minimal FGFs, compared to the 2L FGF anesthesia. Fig. 1. — Individual and average cumulative sevoflurane Cumulative sevoflurane usage after 45 minutes is consumption (ml) (a) and corresponding end-tidal sevoflurane lowest in Minimal throughout all clinical anesthesia concentrations (%) (b) in all groups. Thin black lines represent individual cases, thick red lines are the calculated mean values. procedures in comparison to the AGC4.04 and 2L FGF, 29% and 65%, respectively. In the AGC® cases, the cumulative sevoflurane consumption

is still even 9% lower in AGC4.07 compared to the

AGC4.04. The sevoflurane consumption remains highest in all stages of the anesthesia procedure in the 2L FGF group, while there is a significant decline of consumption in time in all other groups. Sevoflurane usage in the wash-in period is kept low in Minimal, while in AGC® there is a higher consumption at the start. This disparity becomes smaller in due course (Fig. 3).

EtSevo 2% was reached fastest in 2L FGF,

followed by AGC4.07, AGC4.04 and Minimal. This Fig. 2. — Average cumulative liquid sevoflurane consumption (ml) (a) and corresponding end-tidal sevoflurane concentrations partly explains the higher consumption rates of the

(Etsevo) (b) in all four groups. 2L FGF group, but as our results indicate, in time, all other groups were more proficient in terms of volatile delivery once the target was reached and consumed less volatile agent at any given moment. Importantly, in this study, the speed setting in both AGC® modes

was 6 in all records. Probably, the requested EtSevo can be reached faster with a higher speed setting. Conversely, a lower speed setting should result in lower consumption, which would be closer to the consumption figures we see in Minimal. As observed by Carette et al., the consumption in the Flow-i after 45 minutes with AGC® version V4.0.0 in speed 2 was 25% lower than in speed 6 (19). If this ratio would be extrapolated to version V4.07.0, it would Fig. 3. — Average sevoflurane injection rate (ml h-1) in all four groups. correspond to a consumption of 4,7ml in AGC4.07 in

03-De Rydt.indd 18 12/01/2021 12:10 sevoflurane consumption in manual vs automatic modes 19 speed 2 after 45 minutes – which is less than the 5 ml delivery. Carette et al. reported volatile agent we observed in Minimal. This assumption should be consumption to be 8,4 ml with AGC® V4.0.0 after confirmed in further research. To our opinion, there 45 minutes (19). We observed a volatile agen

is no physically inevitable obstacle to explain the consumption of 7,0 ml and 6,3 ml in AGC4.04 and

abundant consumption at higher speeds compared AGC4.07, respectively. to lower speeds. Eventually, the essential amount of There are several limitations to this study. sevoflurane liquid is the amount needed to saturate First of all, as this is a retrospective analysis of breathing system and patient to acquire the desired data records, possible confounding factors are target. As such, we believe that, certainly at higher not controlled. Secondly, the number of extra speeds, there is clear room for improvement of the interventions from the vigilant anesthesiologist algorithms. could not be deducted from our data, and as such, Although the cost of inhalational drugs remains the workload for the additional interventions could a small part of the total operation costs, cutting not be quantified. Also, clinical parameters of the down any form of wastage has environmental and patients were not compared and while the patients financial benefits which should not be ignored (4, 6). did not differ significantly in age, there was no Resulting from the complexities of the atmospheric adjustments for age on MAC. Finally, an additional physics and chemistry, which is described elsewhere, consideration is the increased consumption of sevoflurane has a global heating effect which is soda lime when low flow settings are used. In a

349 times worse than CO2, while desflurane is 3714 comprehensive analysis, this should be included times worse (2). Because volatile anesthetics are in future research. Nevertheless, at a conventional widely and often continuously used in operating 2-2,5 L FGF, an 8-hour effective anesthesia would theatres (assuming an effective anesthesia time of result in a sevoflurane consumption of 160ml per day, a ventilator of 8 hours per day for 45 weeks per resulting in a cost of around 64€ and greenhouse gas

year), the aggregate consumption of sevoflurane emissions of 56 kg CO2-equivalents. Any increased in conventional flow settings of 2-2,5 L/min FGF consumption of soda lime resulting from decreased may easily amount to 40 liters of sevoflurane or FGF would clearly be comparatively negligible. 100 liters of desflurane or more per ventilator per year (1,2). Assuming a market price of 0,4 € per Conclusion ml, this represents an annual cost of 16.000 € and a greenhouse gas equivalent of 14 or 371 metric tons This study indicates that manually regulated

of CO2 for sevoflurane or desflurane, respectively minimal flow is more economic and ecological, (3). As such, a reduction in anesthetic waste of well compared to AGCÒ, at a cost of lower precision

over 50% would result in significant savings - easily of the Etsevo. Compared to all other groups, 2L covering the additional cost of modern equipment FGF consumes at least twice as much sevoflurane - and less atmospheric pollution. As a reference, after only 45 minutes. Conventional “low-flow one roundtrip intercontinental flight Brussels-New anesthesia” (e.g. 2L FGF) should be abandoned as York in economy class results in 2 metric tons of AGCÒ offers a vital solution to the opposing stakes

CO2 emissions. Since use of automatic gas control of ecological/economic concerns and stability/ease technology permits safe and convenient reduction of anesthetic delivery. Still, there are significant of anesthetic waste of easily 50%, implementation gains to be made in improving the algorithms, of this technology would result in a financial saving especially in the first 10 minutes. The cost savings well over 8000 € per year and an equivalent of 7 or through implementation of automated low-flow

185 tons of CO2 emission when used with sevo- tools more than cover the investment costs in flurane or desflurane, respectively (19). Of note, modern machines. Timely upgrade towards most although the relative savings are vastly higher recent software versions can deliver additional cost with desflurane, compared to sevoflurane, from reductions. an ecological point of view, we would strongly discourage the use of desflurane wherever sevo- References flurane were considered a reasonable clinical alter- 1. Vollmer M., Rhee T., Rigby M., Hofstetter D., Hill native (1). M., Schoenenberger F and Reimann S. 2015. Modern After comparison of the subsequent AGC® inhalation anesthetics : Potent greenhouse gases in the software versions, it becomes clear that the upgrades global atmosphere. Geophys Res Lett. 42: 1606-1611. ® 2. Ryan S. and Nielsen C. 2010. Global warming potential of AGC software versions have led to significant of inhaled anesthetics: application to clinical use. Anesth improvements of the efficiency of volatile agent Analg. 111: 92-98.

03-De Rydt.indd 19 12/01/2021 12:10 20 f. de rydt et al. 3. Sulbaeck-Andersen M., Sander S., Nielsen O., Wagner D., 18. Spaeth J., Schweizer T., Schmutz A., Buerkle H and Sanford T and Wallington T. 2010. Inhalation anaesthetics Schumann S. 2017. Comparative usability of modern and climate change. Br J Anaesth. 105: 760-766. anaesthesia ventilators: a human factors study. Br J Anaesth. 4. McGain F., Bishop J., Elliot-Jones L., Story D and Imberger 119: 1000-1008. G. 2019. A survey of the choice of general anaesthetic 19. Carette R., De Wolf A., Hendrickx J. 2016. Automated gas agents in Australia and New Zealand. Anaesth Intens Care. control with the Maquet FLOW-i. J Clin Monit Comput. 247: 235-341. 30: 341-346. 5. Hendrickx J and De Wolf A. 2008. Special aspects of 20. Struys M., Kalmar A., De Baerdemaeker L., Mortier E., pharmacokinetics of inhalation anesthesia. Handb Exp Rolly G and Manigel J., et al. 2005. Time course of inhaled Pharmacol. 182:159-186. anaesthetic drug delivery using a new multifunctional 6. Griend B. and Kennedy R. 2019. Time to stop the go slow closed-circuit anaesthesia ventilator. In vitro comparison on the low flow. Pediatr Anesth. 29: 300-301. with a classical anaesthesia machine. Br J Anaesth. 94: 306- 7. Nair B., Peterson G., Neradilek M., Newman S., Huang 317. E and Schwid H. 2013. Reducing wastage of inhalation 21. Lucangelo U., Ajčević M., Accardo A., Borelli M., anesthetics using real-time decision support to notify of Peratoner A and Comuzzi L., et al. 2017. FLOW-i ventilator excessive fresh gas flow. Anesthesiology. 118: 874-884. performance in the presence of a circle system leak. J Clin 8. Suttner S. and Boldt J. 2000. Low-flow anaesthesia. Does Monit Comput. 31: 273-280. it have potential pharmacoeconomic consequences? Phar- 22. Moran P., Barr D and Holmes C. 2019. Saving Sevoflurane: macoeconomics. 17: 585-590. Automated gas control can reduce consumption of 9. Baum J. and Aitkenhead A. 1995. Low-flow anaesthesia. anesthetic vapor by one third in pediatric anesthesia. Pediatr Anaesthesia. 50: 37-44. Anesth. 29: 310-314. 10. Upadya M. and Saneesh P. 2018. Low-flow anaesthesia - 23. De Rydt F., Goossens J., Allaert S. and Kalmar A. 2017. underused mode towards “sustainable anaesthesia”. Indian Sevoflurane consumption using automatic gas control versus J. Anaesth. 62: 166-172. manual gas control with Flow-I: comparative consumption 11. Nunn G. 2008. Low-flow anaesthesia. BJA Educ. 8: 1-4. in 200 patients. 2017. Sevoflurane consumption using 12. Brattwall M., Warrén-Stomberg M., Hesselvik F and automatic gas control versus manual gas control with Jakobsson J. 2012. Brief review: theory and practice of Flow-I : comparative consumption in 200 patients. Eur J minimal fresh gas flow anesthesia. Can J Anaesth. 59: 785- Anaesthesiol. 43: 35-35 797. 24. Skalec T., Górecka-Dolny A., Zieliński S., Gibek 13. Kennedy R. and French R. 2014. A ten-year audit of fresh M., Stróżecki Ł and Kübler A. 2017. Comparison of gas flows in a New Zealand hospital: the influence of the anaesthetic gas consumption and stability of anaesthesia introduction of automated agent delivery and comparisons using automatic and manual control over the course of with other hospitals. Anaesth Intensive Care. 42: 65-72. anaesthesia. Anaesthesiol Intensive Ther. 49: 34-39. 14. Tay S., Weinberg L., Story D and Briedis J. 2013. Financial 25. Roumeliotis N., Parisien G., Charette S., Arpin E., and Environmental Costs of Manual versus Automated Brunet F and Juvet P. 2018. Reorganizing Care With the Control of End-Tidal Gas Concentrations. Anaesth Inten- Implementation of Electronic Medical Records: A Time- sive Care. 41: 95-101. Motion Study in the PICU. Pediatr Crit Care Med. 19: 172- 15. Ghijselings I., De Cooman S., Carette R., Peyton P., De 179. Wolf A and Hendrickx J. 2016. Performance of an active 26. Dehouwer A., Carette R., De Ridder S., De Wolf A and inspired hypoxic guard. J Clin Monit Comput. 30: 63-68. Hendrickx J. 2016. Accuracy of inhaled agent usage 16. Lucangelo U., Garufi G., Marras E., Ferluga M., Turchet displays of automated target control anesthesia machines. F and Bernabè F., et al. 2013. End-tidal versus manually- J Clin Monit Comput. 30: 539-543. controlled low-flow anaesthesia. J Clin Monit Comput. 28: 27. Faul F., Erdfelder E., Buchner A and Lang A. 2009. 117-121. Statistical power analyses using G*Power 3.1: Tests for 17. Singaravelu S. and Barclay P. 2013. Automated control of correlation and regression analyses. Behav Res Methods. end-tidal inhalation anaesthetic concentration using the GE 41: 1149-1160. Aisys CarestationTM. Br J Anaesth. 110: 561-566.

03-De Rydt.indd 20 12/01/2021 12:10 (Acta Anaesth. Belg., 2020, 71, 21-27) Case report and narrative review on regional anesthesia for closed shoulder reduction

J. Kennes (*), S. Coppens (*), J. Deprest (**), M. Van de Velde (*)

Abstract : Objective : obstruction and hemodynamic vascular injury. Radiographic imaging confirmed instability. a shoulder dislocation and excluded a humeral Methods : A literature search in Cochrane central, fracture. General anesthesia and sedation were MEDLINE and Embase was performed on 23-03-2020 avoided for the risk of airway obstruction, gastric using following search strings : “shoulder reduction”, content aspiration, hemodynamic instability and its “shoulder dislocation”, “locoregional anesthesia”, repercussions on the growing fetus. After consulting “supraclavicular block” and “interscalene block”. In total gynecology, traumatology and anesthesiology, the 134 articles were found after duplicate removal. Nine articles were included for full text screening. Case series patient was transferred to the operation room for were excluded for analysis. Three articles were selected shoulder reduction with locoregional anesthesia. A and a fourth one was added from the references of the single shot interscalene brachial plexus block was analyzed articles. performed under ultrasound guidance and nerve Results : Length of stay at the emergency department is stimulation. Eighteen milliliters of ropivacaine significantly shorter when using locoregional anesthesia 0.5% was injected with observation of good compared to general anesthesia (1-4) (table 1). The spread of the local anesthetic. Shoulder reduction results for pain scores and patient satisfaction are more was successfully performed using the Kocher variable. No complications of locoregional anesthesia method and the patient experienced no pain during were observed in these trials. Studies comparing the maneuver. Fetal monitoring was performed different locoregional anesthetic techniques for shoulder throughout the procedure and was uneventful. dislocation were not found. The patient was admitted at the delivery room for Discussion and Conclusion : Current literature supports the feasibility of locoregional anesthesia for shoulder monitoring and was discharged the same day. reduction. Complications due to locoregional anesthesia were absent in the investigated trials though the current Introduction trials are too small to draw conclusions. The use of locoregional anesthesia can avoid complications due to The overall incidence of shoulder dislocation procedural sedation and general anesthesia. Locoregional in the United Kingdom over a 20 year period is anesthesia offers the advantage of shorter length of stay 40.4/100000 person-years in men and 15.5/100000 in the emergency department. Results of pain scores and person-years in woman (1). The incidence is patient satisfaction differ amongst the analyzed trials. highest in young men. The incidence in woman Further research is needed to confirm these results and to rises with advanced age. The etiology of shoulder investigate the most appropriate locoregional anesthetic dislocations is predominantly traumatic (fall and technique.

Key words : shoulder dislocation ; regional anesthesia ; J. Kennes, MD, S. Coppens, MD, J. Deprest, MD, PhD, M. nerve block. Van de Velde, MD, PhD (*) Dpt of Anesthesiology, University Hospital Leuven, Herestraat 49, 3000 Leuven, Belgium. (**) Dpt of Gynecology, University Hospital Leuven, Herestraat Case report 49, 3000 Leuven, Belgium. Corresponding author: J. Kennes, MD. Dpt of Anesthesiology, A 33 year old woman presented at the University Hospital Leuven, Herestraat 49, 3000 Leuven, emergency department with a right sided shoulder Belgium E-mail: [email protected] dislocation after a traumatic fall. No other injuries were observed. She was 40 weeks + 4 Paper submitted on May 25, 2020 and accepted on May 26, days pregnant. Her medical history included a 2020. Conflict of interest: none. Bankart shoulder repair 5 years earlier. Clinical This master thesis has been presented at the SARB Graduation examination excluded concomitant nerve and Day on 23/05/2020.

04-Kennes.indd 21 12/01/2021 12:11 22 j. kennes et al.

high energy injury during sports or recreation) (2). Results Patients presenting with a shoulder dislocation at the emergency department often complain of We analyzed the results of studies comparing discomfort and persisting pain until the moment regional anesthesia with general anesthesia or of reduction. Most patients need analgesia and procedural sedation for the outcome parameters of sedation for shoulder manipulation. Multiple length of stay, pain scores and patient satisfaction reasons make procedural sedation and general (Table 1). Three articles compared the effect of anesthesia unattractive (non-fasted state, cardiac or an interscalene block to general anesthesia for respiratory co-morbidities, pregnancy, …). In this shoulder reduction. Blaivas et al. (4) compared review we will analyze the existing literature on the an interscalene block to procedural sedation with use of regional anesthesia for shoulder reduction. etomidate in a prospective randomized controlled trial with 21 patients in each group. The randomized controlled trial of Doost et al. (5) compared an Methods interscalene block to procedural sedation with propofol and fentanyl and included 30 patients in A literature search in Cochrane central, each group. Kreutziger et al. (6) did a retrospective MEDLINE and Embase was performed on 23-03- data analysis of patients with shoulder dislocation. 2020 using following search strings : “shoulder They compared patients receiving an interscalene reduction”, “shoulder dislocation”, “locoregional block, general anesthesia and intravenous analgesia, anesthesia”, “supraclavicular block” and “inter- respectively 31, 18 and 14 patients. All studies scalene block”. In total 134 articles were found after showed successful use of the interscalene block for duplicate removal. Nine articles were included for shoulder reduction though the outcome parameters full text screening. Case series were excluded for vary amongst the different studies. Tezel et al. (7) analysis. Six articles were excluded for the absence performed a prospective randomized controlled trial of a control group. Three articles were selected and comparing a suprascapular block to sedation with a fourth one was added from the references of the ketamine for shoulder reduction in respectively 21 analyzed articles (Fig. 1). and 20 patients.

Fig. 1.

04-Kennes.indd 22 12/01/2021 12:11 regional anesthesia for closed shoulder reduction 23

Table 1 Studies comparing a regional technique with procedural sedation/general anesthesia

Author Year Study type n LOS in minutes Pain score Patient satisfaction

Blaivas et al. (4) 2011 prospective RCT 42 100,3 (±28,3 SD) vs. 177,3 0 (0-3 range) vs. 0 (0-4) 8.3 (±1.1 SD) vs. 8.2 (±1.3) US ISB (21) vs. GA (21) (±37,9) p<0,0001 p=0,7990 p=0.9275 Doost et al. (5) 2017 RCT 60 80.2 (±25.2 SD) vs 108.6 3,43 (±1,6 SD) vs. 0,38 3,0(±0,6 SD) vs. 3,6(±0,4) US ISB (30) vs. GA (30) (±42.1) p=0,005 (±0,5) p<0,001 p<0,001 Kreutziger et 2019 retrospective data analysis 63 67.5 (48.8-93.5 IQR) vs. 0 (0-0 IQR) vs. 2 (0- ISB 12(9-15) vs. GA 17(12-24) al. (6) US ISB(31) vs. GA(18) 102 (100-110) p=0.003 vs. 2,5) p=0,04 vs. 2(0-2) p=0.03 vs. IVA(14) 125 (102.5-147.5 p=0,0007 p=0,002 ISB 12(9-15) vs. IVA 11(8-13) p=0.22 ˠ Tezel et al. (7) 2014 prospective RCT 41 25 (21-36 range) vs 125 NA NS US SNB (21) vs GA (20) (120-138) p<0,001 GA general anesthesia, IVA intravenous analgesia, US ISB ultrasound guided interscalene block, US SNB ultrasound guided suprascapular nerve block, LOS length of stay, RCT randomized controlled trial, SD standard deviation, IQR interquartile range, NA not applicable, NS not significant.

Discussion Table 2 Reduction techniques (11-13) Shoulder dislocation and reduction techniques Reduction techniques for anterior dislocations Kocher method The overall incidence of shoulder dislocation Spaso method External rotation maneuver in the United Kingdom over a 20 year period is Milch technique 40.4/100000 person-years in men and 15.5/100000 Chair method person-years in woman (1). Comparable results were Traction counter-traction observed in other nations (2, 8, 9). The incidence Scapular manipulation is highest in young men. The incidence in woman Stimson technique rises with advanced age. The etiology of shoulder FARES technique (FAst, REliable and Safe) dislocations is predominantly traumatic (fall or Hippocratic method Eskimo technique high energy injury during sports or recreation) Boss-Holzach-Matter method (2). Shoulder dislocation is the most frequent joint The snowbird method dislocation in the human body (10). Three types Best of Both Maneuvers of shoulder dislocation have been described : Legg Reduction Maneuver anterior, posterior and inferior. Anterior dislocation Bokor-Billmann technique is most frequent followed by posterior and inferior Cunningham technique Davos method/Aronen technique dislocation. Glenohumeral reduction can be Elbow technique obtained by one of many techniques described in GONAIS method the literature (Table 2) (11-13). Some techniques Janecki’s forward elevation maneuver advocate to be less traumatic and painful, others Nicola method yield serious discomfort. Ancient techniques like Zahiri technique the Hippocratic method have become obsolete Reduction techniques for posterior dislocations Wilson’s technique because of the increased risk of fractures and DePalma’s ‘‘lever’’ method neurovascular injury (12). A systematic review of Caudal traction closed shoulder reduction techniques of Alkaduhimi Reduction techniques for inferior dislocations et al. (11) proposed that the scapular manipulation Traction/countertraction technique is the fastest and most successful and the Two-step maneuver FARES (FAst REliable and Safe) technique offers a good alternative. However, there is no “one fits most institutes intravenous analgesia, sedation or all” reduction technique and clinicians need to general anesthesia are used. Shoulder dislocations master alternative techniques in order to adapt to the are mostly seen in trauma and these patients are clinical situation. Alkaduhimi et al. also reassessed seldomly sober which increases the risk for gastric pain scores during reduction ranging from 1.38 to content aspiration. Data on the incidence of gastric 6.9 on a 10-point VAS (Visual Analog Score) scale. content aspiration in procedural sedation in the Therefore most patients need a form of analgesia emergency department are limited. Application of or sedation to obtain successful shoulder reduction. the same fasting guidelines as for general anesthesia The two major elements needed for shoulder are questioned (14). Postponing a shoulder reduction reduction are analgesia and muscle relaxation. In for the risk of aspiration can reduce the success

04-Kennes.indd 23 12/01/2021 12:11 24 j. kennes et al. rate of the procedure (15). This dilemma between adherence to fasting guidelines and fast shoulder reduction can be solved by regional anesthesia.

Innervation of the shoulder

A thorough knowledge of anatomy and innervation is essential to facilitate shoulder reduction with regional anesthesia. The brachial plexus (Fig. 2) is formed by the anterior rami of the nerve roots of C5 to T1 with variable contributions of C4 and T2. Two terminal nerves arise at the level of the roots : the dorsal scapular and long thoracic nerve that will innervate respectively the rhomboid muscles and the serratus anterior muscle. The superior (C5-C6), middle (C7) and inferior trunk (C8-T1) are formed by fusion of the nerve roots at the level of the interscalene groove. The suprascapular Fig. 2. — Brachial plexus. Image reprinted by permission and subclavian nerve arise from the superior trunk from NYSORA e-Learning System (https://nysoralms.com/) and innervate the supra- and infraspinatus muscles (31). and the subclavian muscle. Each trunk divides in an anterior and posterior division at the level of the anteriorly (22). These branches originate from the clavicle without formation of terminal nerves. The C5 and C6 nerve roots. Tensile stress on the shoulder anterior divisions of the superior and middle trunk capsule provokes major pain during dislocation. merge and create the lateral chord. The medial chord Muscular spasm develops progressively in shoulder is created by the anterior divisions of the middle and dislocation and impedes simple reduction. The inferior trunk. The posterior divisions of all trunks target of a regional technique should comprise the form the posterior chord. Multiple nerves originate at aforementioned nerves to obtain pain relief and the cords before formation of the terminal branches. muscular relaxation of the shoulder girdle. The lateral pectoral nerve originates from the lateral chord and innervates the pectoralis major muscle. Regional techniques The medial pectoral, medial brachial cutaneous Brachial plexus blockade for shoulder reduc- and medial antebrachial cutaneous nerve arise from tion is described in literature using interscalene, the medial chord. The thoracodorsal nerve and the supraclavicular and suprascapular nerve block. Most medial and lateral subscapular nerve emerge at the articles used an interscalene block for this indication. posterior chord. More distally, the chords form Since the first description of the interscalene block the terminal nerve branches : musculocutaneous, in 1970 by Winnie et al. (23) it has been successfully axillary, radial, median and ulnar nerve (16-19). utilized for shoulder and upper arm surgery. Several Many variations to this classic presentation early reports confirmed the feasibility of regional of the brachial plexus exist. The contribution of anesthesia for shoulder reduction (24, 25) though it the C4 nerve root to the brachial plexus is greater has never become a standard of care. Literature on in a prefixed plexus (C4-C8) with implications for the use of regional anesthesia for shoulder reduction the innervation of the shoulder girdle. A postfixed is scarce. plexus (C6-T2) shows greater contribution of the T2 nerve root and less contribution of the C4 and Interscalene block (ISB) C5 roots (18). Additionally the C5 and C6 roots are often located inside or anterior to the anterior The interscalene block targets the brachial scalene muscle as established in cadaver studies plexus at the level of the trunks at the interscalene (20,21). The presence of a cervical rib can deform groove between the anterior and middle scalene the normal anatomy of the brachial plexus (18). muscle. This is an optimal technique for shoulder The shoulder capsule is predominantly inner- and upper arm anesthesia but unreliable for hand vated by the suprascapular nerve and axillary and lower arm anesthesia because the C8-T1 nerve on the posterior side and branches of the roots are frequently spared (17, 19). The majority subscapular, axillary and lateral pectoral nerve of the included studies used an interscalene block

04-Kennes.indd 24 12/01/2021 12:11 regional anesthesia for closed shoulder reduction 25 for shoulder reduction because of the complete The use of a fast and short acting local anesthetic is coverage of the C5 and C6 nerve roots. therefore promoted.

Supraclavicular block (SCB) Pain scores

The supraclavicular block targets the brachial The results for pain scores are more variable. plexus at the level of the divisions between the The study of Blaivas et al. (4) could not show a clavicle and the first rib. The brachial plexus lies significant difference in self-reported pain scores in close proximity with the subclavian artery and during reduction between the ISB group and injection in the corner pocket between the artery procedural sedation. Doost et al. (5) even found and the nerves is most successful (26, 27). This significant lower pain scores during reduction in technique is optimal for complete upper limb the procedural sedation group. These results need anesthesia with one injection. Rostral spread of the to be interpreted with caution. Patients receiving local anesthetic offers good conditions for shoulder procedural sedation are unconscious at the moment anesthesia. Stone et al. (28) described the use of the of reduction. Humans need consciousness to be able supraclavicular block for upper arm emergencies to experience pain and this element could bias the including fractures, dislocations and abscesses. observed result. Moreover, the affective component They concluded that a supraclavicular block was of pain could be increased by the snapping sound feasible for upper arm emergencies and led to a of the shoulder, the view of the reduction and effort shorter length of stay. Two patients in their study of the physician to obtain a successful reduction. population suffered from shoulder dislocation but More reliable would be the pain score at different they received procedural sedation (control group). moments in time as done in the retrospective study Minville et al. (29) performed supraclavicular by Kreutziger et al. (6). They observed pain scores brachial plexus block for shoulder reduction in 55 on six different moments in time (pre-hospital, out of 97 patients receiving regional anesthesia. on admission, after pain drugs, during reduction, Seventeen patients needed conversion to general on discharge, first evening). Their observations anesthesia, which seems unusually high. This study show a significant reduction in pain intensity on did not report on pain scores, length of hospital stay discharge and on the first evening compared to or patient satisfaction and was excluded for further general anesthesia and intravenous analgesia. analysis. Their numbers confirm non-significant pain scores between regional anesthesia and general anesthesia Suprascapular block (SNB) at the moment of shoulder reduction.

The suprascapular nerve block limits its action Patient satisfaction to one terminal nerve of the brachial plexus. Sparing Patient satisfaction varies amongst the ana- of the other nerve branches can result in suboptimal lyzed studies. Kreutziger et al. (6) showed a signi- reduction conditions but offers a better safety profile ficantly increased patient satisfaction with regional compared to an interscalene or supraclavicular anesthesia while Doost et al. (5) observed the block. opposite effect. The other trials could not obtain a significant result for the outcome of patient Length of hospital stay satisfaction (4,7). Patients are overall satisfied with whatever technique has been used and they would All studies uniformly show a significant choose the same analgesic technique again (6). reduction in length of stay (LOS) in the regional anesthesia group regardless of different definitions Complications of LOS in the emergency department. This effect is explained by the increased monitoring time after Complications of shoulder reduction itself procedural sedation or general anesthesia compared include nerve, vascular and musculoskeletal injury. to regional anesthesia. Blaivas et al. (4) confirmed The incidence of complications differs amongst the a significantly increased one-on-one health care reduction techniques (13). Moreover, a pre-existing provider time for procedural sedation. Kreutziger humeral fracture increases the risk for vascular and et al. (6) showed a prolonged time to shoulder nerve injury during reduction. reduction in the ISB group compared to the general Anesthetic complications depend on the anesthesia group explained by the onset of the block. technique used. General anesthesia or procedural

04-Kennes.indd 25 12/01/2021 12:11 26 j. kennes et al. sedation increases the risk of airway obstruction References with desaturation, gastric content aspiration and hemodynamic instability. The studies analyzed in 1. Shah A., Judge A., Delmestri A., Edwards K., Arden NK. this review all used a different sedative (propofol, and Prieto-Alhambra D. et al. 2017. Incidence of shoulder etomidate, ketamine with or without the use of dislocations in the UK, 1995–2015: a population-based an opioid) which makes it difficult to compare. cohort study. BMJ Open. 7(11):e016112. 2. Zacchilli MA. and Owens BD. 2010. Epidemiology of Every sedative has its own side effect profile. Shoulder Dislocations Presenting to Emergency Depart- The retrospective study of Minville et al. (29) ments in the United States. J Bone Joint Surg Am Actions. analyzed the results of 224 shoulder dislocations 92(3):542-549. and encountered a 8% complication rate, all linked 3. Moher D., Liberati A., Tetzlaff J., Altman DG., The PRISMA Group. 2009. Preferred Reporting Items for to general anesthesia. Tezel et al. (7) observed Systematic Reviews and Meta-Analyses: The PRISMA a 15% incidence of nausea and vomiting, 10% Statement. PLoS Med 6(7): e1000097. supplemental oxygen need and 15% agitation in 4. Blaivas M., Adhikari S. and Lander L. 2011. A prospective comparison of procedural sedation and ultrasound-guided their ketamine sedation group. interscalene nerve block for shoulder reduction in the Complications of regional anesthesia are emergency department. Acad Emerg Med. 18 : 922-927. numerous (bleeding, infection, nerve injury, intra- 5. Raeyat Doost E., Heiran MM., Movahedi M. and Mirafzal A. 2017. Ultrasound-guided interscalene nerve block vs vascular injection with seizures, epidural injection, procedural sedation by propofol and fentanyl for anterior spinal cord injury, pneumothorax, local anesthetic shoulder dislocations. Am J Emerg Med. 35 : 1435-1439. systemic toxicity, allergic reactions, …). A physical 6. Kreutziger J., Hirschi D., Fischer S., Herzog RF., Zbinden examination to detect prior neurological deficits and S., and Honigmann P. 2019. Comparison of interscalene block, general anesthesia, and intravenous analgesia for radiographic evidence of the absence of pre-existing out-patient shoulder reduction. J Anesth. 33 : 279-286. fractures should be performed before shoulder 7. Tezel O., Kaldirim U., Bilgic S., Deniz S., Eyi YE. reduction. Other side effects such as Horner syn- and Ozyurek S. et al. 2014. A comparison of suprascapular nerve block and procedural sedation analgesia in shoulder drome, recurrent laryngeal nerve paralysis and dislocation reduction. Am J Emerg Med.32 : 549-552. hemidiaphragmatic paralysis occur in a high number 8. Liavaag S., Svenningsen S., Reikerås O., Enger M., of patients and can be detrimental in patients with Fjalestad T. and Pripp AH. et al. 2011. The epidemiology of shoulder dislocations in Oslo. Scand J Med Sci Sports. co-morbidities (30). The absence of any of these 21 : e334-e340. side effects in the analyzed trials is remarkable and 9. Leroux T., Wasserstein D., Veillette C., Khoshbin A., Henry could be explained by underreporting of side effects P. and Chahal J. et al. 2014. Epidemiology of Primary and complications. These side effects can guide the Anterior Shoulder Dislocation Requiring Closed Reduction in Ontario, Canada. A m J Sports Med. 42 : 442-50. clinician to choose a different regional technique. 10. Cutts S., Prempeh M. and Drew S. 2009. Ante-rior shoulder The clinician needs to have a profound training and dislocation. Ann R Coll Surg Engl. 91 : 2-7. sufficient experience with regional anesthesia (4, 11. Alkaduhimi H., van der Linde JA., Willigenburg NW., van Deurzen DFP. and van den Bekerom MPJ. 2017. A 5). There is no scientific ground to avoid regional systematic comparison of the closed shoulder reduction techniques for the risk of complications in these techniques. Arch Orthop Trauma Surg. 137 : 589-599. patients at this moment though the trials are too 12. Gottlieb M. 2020. Shoulder Dislocations in the Emergency Department: A Comprehensive Review of Reduction small to draw conclusions. Techniques. J Emerg Med. pii: S0736-4679 : 31090-X. 13. Dannenbaum J., Krueger CA. and Johnson A. 2012. A Conclusion review of reduction techniques for anterior glenohumeral joint dislocations. J Spec Oper Med. 12 : 83-92. 14. Green SM. and Krauss B. 2002. Pulmonary aspiration risk Current literature supports the feasibility of during emergency department procedural sedation--an regional anesthesia for shoulder reduction. Com- examination of the role of fasting and sedation depth. Acad plications due to regional anesthesia were absent Emerg Med. 9 : 35-42. 15. Kanji A., Atkinson P., Fraser J., Lewis D. and Benjamin in the investigated trials though the current trials S. 2016. Delays to initial reduction attempt are associated are too small to draw conclusions. The use of with higher failure rates in anterior shoulder dislocation: a regional anesthesia can avoid complications due retrospective analysis of factors affecting reduction failure. Emerg Med J. 33(2):130-3. to procedural sedation and general anesthesia. 16. Leinberry CF. and Wehbé MA. 2004. Brachial plexus Regional anesthesia offers the advantage of shorter anatomy. Hand Clin. 20(1):1-5. length of stay in the emergency department. Results 17. Mian A., Chaudhry I., Huang R., Rizk E., Tubbs RS. of pain scores and patient satisfaction differ amongst and Loukas M. 2014. Brachial plexus anesthesia: A review of the relevant anatomy, complications, and anatomical the analyzed trials. Further research is needed to variations. Clin Anat.27 : 210-221. confirm these results and to investigate the most 18. Johnson EO., Vekris M., Demesticha T. and Soucacos PN. appropriate regional anesthetic technique. 2010. Neuroanatomy of the brachial plexus: normal and variant anatomy of its formation. Surg Radiol Anat.32 : 291-297.

04-Kennes.indd 26 12/01/2021 12:11 regional anesthesia for closed shoulder reduction 27 19. Orebaugh SL. and Williams BA. 2009. Brachial plexus 26. Soares LG., Brull R., Lai J. and Chan VW. 2007. Eight ball, anatomy: normal and variant. ScientificWorldJournal. 28 : corner pocket: the optimal needle position for ultrasound- 300-312. guided supraclavicular block. Reg Anesth Pain Med 32 : 20. Harry W.G., Bennett J.D.C., and Guha S.C. 1997. Scalene 94-95. muscles and the brachial plexus. Clin. Anat. 10 : 250–252. 27. Brull R. and Chan VW. 2011. The corner pocket revisited. 21. Natsis K., Totlis T., Tsikaras P., Anastasopoulos N., Skan- Reg Anesth Pain Med. 36(3):308. dalakis P., and Koebke J. 2006. Variations of the course 28. Stone MB., Wang R. and Price DD. 2008. Ultrasound- of the upper trunk of the brachial plexus and their clinical guided supraclavicular brachial plexus nerve block vs significance for the thoracic outlet syndrome: a study on 93 procedural sedation for the treatment of upper extremity cadavers. Am. Surg. 72 : 188-192. emergencies. Am J Emerg Med. 26 :706-710. 22. Aszmann OC., Dellon AL., Birely BT. and McFarland 29. Minville V., Plante T., Eychenne B., Colombani A. EG. 1996. Innervation of the human shoulder joint and its and Fourcade O. 2009. Retrospective study on six years implications for surgery. Clin Orthop Relat Res. 330 : 202- of anaesthesia for reduction of shoulder dislocation in 207. emergency. Ann Fr Anesth Reanim. 28 : 949-953. 23. Winnie AP. 1970. Interscalene brachial plexus block. 30. Urmey WF., Talts KH. and Sharrock NE. 1991. One Anesth Analg. 49(3):455-66. hundred percent incidence of hemidiaphragmatic paresis 24. Underhill TJ., Wan A. and Morrice M. 1989. Interscalene associated with interscalene brachial plexus anesthesia as brachial plexus blocks in the management of shoulder diagnosed by ultrasonography. Anesth Analg. 72 : 498-503. dislocations. Arch Emerg Med. 6 : 199-204. 31. Carrera A., Lopez AM., Sala-Blanch X., Kapur E., 25. Gleeson AP., Graham CA., Jones I., Beggs I. and Nutton Hasanbegovic I. and Hadzic A. 2020. NYSORA. Functional RW. 1997. Comparison of intra-articular lignocaine and regional anesthesia anatomy. (Online). Retrieved from https: a suprascapular nerve block for acute anterior shoulder //www.nysora.com/foundations-of-regional-anesthesia/ dislocation. Injury. 28 :141-142. anatomy/functional-regional-anesthesia-anatomy/

04-Kennes.indd 27 12/01/2021 12:11

(Acta Anaesth. Belg., 2020, 71, 29-35) The added value of simulation-based medical education : a quality improvement study

R. De Meirsman, E. Van Gerven

Abstract : Objective : Current literature is unsure of Innovations in flight simulators, computers and the added value of simulation-based medical education. audio-visual technology were the driving forces The aim of this study was to make a basic assessment to incorporate simulation in medical education. of simulation based medical education within the Slow integration was due to skepticism, the lack university hospitals of Leuven, questioning the necessity of validated studies and the lack of good primary and the proposition for a simulation-based curriculum, outcomes (1). Simulation-based medical education, estimating the clinical competence of anesthesia trainees as described by Gaba, is a technique to replace real during perioperative emergency events and assessing the application of a crisis resource management (CRM) life experiences with simulated experiences that protocol. evoke or replicate substantial aspects of the real Background : Simulation based medical education world in a fully interactive manner (2). There is a evolved from aviation industry, where it was used lot of diversity in simulated-based medical training, to practice technical skills and apply crisis resource e.g. skill training, scenario training, emergency management to enhance safety, technical and non- training, etcetera... Skill training is a type of training technical skills. where you can practice and perfect a specific skill ; Methods : This manuscript adheres to the SQUIRE Scenario training is a type of training which uses guidelines for quality improvement studies. We designed a medical case to set up a scenario. It consists of a a questionnaire comprising 25 questions to test the added briefing, a scenario and a debriefing and is usually value of simulation in the university hospitals of Leuven. done in a team setting, which allows for technical Results : Sixty-one percent (61%) (102 of 167) responded and non-technical skills, e.g. Crisis Resource to the questionnaire ; 83% (85 of 102) of participants describe an added value attributed to simulation ; 93% Management (CRM), to be practiced. Technology (93 of 102) wants to integrate simulation into their in simulation can range from high fidelity to low education ; barriers towards integration of simulation, fidelity. Fidelity is a way to describe the realism according to the participants, are time-related in 78% (79 behind a simulation. Recent advances in computer of 102) of the answers ; perioperative emergency events sciences and virtual reality are making simulations happened to 74% (75 of 102) of the participants, on a six- more life-like than ever before. An in depth analysis monthly basis. Applicable knowledge of CRM protocol of all the different variables regarding simulation- was described by 11% (12 of 102) of the participants. based medical education is not the scope of this Discussion : The majority of participants (83%) reported manuscript. The great diversity and heterogeneity an added value of simulation based education and of simulation based medical training could explain there was a positive attitude towards integration of a the paucity of strong evidence based research (2). simulation-based curriculum. Majors barriers towards Medical errors due to human related factors are the implementation are time and resource/money related. The incidence of perioperative emergency events remains common (3). Despite the lack of high-quality studies unclear and underreported in the current literature. The there are recent articles that try to demonstrate the application of Crisis Resource Management (CRM) added value of simulation-based medical education is currently lacking, and more effort should be made in terms of performance, medical errors and cost- towards a systematic implementation. R. De Meirsman , MD ; E. Van Gerven , MD Keywords : Simulation training ; quality Improvement ; Department of Anesthesiology, University Hospitals of Leuven, Herestraat 49, 3000 Leuven, Belgium education ; medical ; anesthesia Corresponding author : De Meirsman R., Department of Anesthesiology, University Hospitals of Leuven, Herestraat 49, 3000 Leuven, Belgium. Introduction E-mail: [email protected]

Evidence based literature is unsure of the Paper submitted on May 23, 2020 and accepted on May 26,2020. added value of simulation-based medical education. Conflict of interest: none.

05-De Meirsman.indd 29 12/01/2021 12:11 30 r. de meirsman, e. van gerven

effectiveness. This could lead to optimization of Methods procedures enhancing safety, quality and efficiency (4). Evans et al demonstrated that simulation-based This manuscript adheres to the SQUIRE medical education has been shown to be effective guidelines for quality improvement studies. The in enhancing the performance of central venous committee for medical simulation at University catheter (CVC) insertion in comparison with hospitals of Leuven wanted to discuss the current traditional training (5), resulting in a better state state of their simulation-based medical education, of care and reduced medical errors (6). Cohen et al questioning the necessity and the proposition for a also demonstrated that simulation-based medical simulation-based curriculum. To assess the need for education was proven to be cost-effective in the simulation-based medical education and to evaluate prevention of catheter-related bloodstream infections its potential added value a detailed questionnaire was (CRBI) on the intensive care unit. After one year developed for the department of anesthesiology’s of simulating, data showed there was a reduction senior residents and junior residents. The research of approximately 10 CRBI with a cost reduction was conducted from November 2017 to March 2018. of almost 800,000 dollars. In comparison the costs The participants were recruited from a database of the simulation program were calculated to be containing anesthesia residents currently active 100,000 dollars giving them a return on investment in the university hospitals of Leuven. Informed of 7 to 1 (7). The costs of simulation-based medical consents were obtained from all participants prior education are dependent upon the variables used to to their inclusion in the research. No remarks were set up a simulation. A low fidelity simulation will made by the Ethics committee of the University cost substantially less than a high fidelity simulation Hospitals of Leuven. The questionnaire consisted (2). The aim of this manuscript was to make a basic of 25 questions (see table 1 for details) and was assessment of simulation based education within created using google documents. A subdivision was the university hospitals of Leuven, questioning the made based on demographics, simulation- based necessity and the proposition for a simulation-based medical education, skill training and scenario- curriculum, estimating the clinical competence of training. The questions were scored using a Likert- anesthesia trainees during perioperative emergency scale, ranging from 1 up to 10. Data collection was events and assessing the application of a crisis carried out anonymously by means of an electronic resource management protocol. link, which led to the google docs, questions were

Table 1 Questionnaire 1. “Did you ever participate in simulation-based medical training?” 2. ‘’How many simulation-based medical educations have you taken part in?’’ 3. “Do you want to integrate simulation-based medical training into your education?” 4. “How often do you want simulation-based medical training?” 5. “Can you describe the added value of simulation-based medical education on a Likert scale?” 6. “Which topics would you like to practice during simulation-based medical training?” 7. “Are you willing to pay for simulation-based medical training as part of your educational program?” 8. “How much are you willing to spend on one training session?” 9. “When would you like to practice simulation-based medical education?” 10. “On a Likert scale of 1 to 10 do you think there is an added value of simulation-based medical education towards maintaining accreditation after your graduation as an anesthesiologist?” 11. “What do you think the barriers are towards practical integration of simulation-based medical education?” 12. “Have you ever encountered a peri-operative emergency event during your education, which would have been better trained beforehand during a simulation-based medical training session?” 13. “How often did you encounter such a peri-operative emergency event?” 14. ‘’Give a short description of such an event” 15. “Have you ever participated in skill training?” 16. “How are these skills taught?” 17. “Are these skills taught on a standardized manner?’’ 18. “Have you ever participated in a scenario training?” 19. “On a Likert scale of 1 to 10 did you find scenario training to be a good learning experience?” 20. “Do you feel more prepared for a peri-operative emergency event after completing a scenario training? 21. “Would you like to do a scenario based simulation-based training before a specific rotation, e.g. weaning of cardiac bypass” 22. “Are you familiar with Crisis Resource Management (CRM)?” 23. “I do not know the core principles of crisis resource management, but I have heard of the concept CRM” 24. “I do not know crisis resource management, and I have not heard of the concept CRM”. 25. “Do you think your knowledge of CRM is sufficiently large/ applicable in a real emergency situation?”

05-De Meirsman.indd 30 12/01/2021 12:11 the added value of simulation-based medical education 31 asked about demographic data, training year and previous experience in simulation education. During statistical analysis no editing of data was performed, except when input errors were detected. Datasets were compiled in a Microsoft Excel spreadsheet, coded and stored on a cloud based google server. Statistical analysis was extracted from data entered into the google documents and was descriptive by nature. Nominal variables were compared according to training years, dividing senior and junior residents.

Results

We contacted 167 people, hereafter referred to as the participants, of which 102 people completed Fig. 1. — Total participants. the questionnaire. Accounting for a response rate of 61% (Fig. 1). To the question : “Did you ever participate in simulation-based medical training?” 70% (71 of 102) answered : “yes”. When asked how many simulation-based medical educations the participants had undergone 58% (59 of 102) answered 1-3 times, 5% (5 of 102) answered 3-5 times, 1% answered (1 of 102) answered 5-10 times, 7% (7 of 102) answered more than 10 times. To the question : “Do you want to integrate simulation- based medical training into your education? “ 93% (93 of 102) answered “Yes”. Subsequently, to the question : “How often do you want simulation- based medical training?” 30% (30 of 102) answered once every month, 36% (35 of 102) answered once every three months, 22% (22 of 102) answered once every six months, 10% (10 of 102) answered once every year, 2% (2 of 102) answered not to be interested in simulation-based medical training (Fig. 2). To the question : “Can you describe the added value of simulation-based medical education Fig. 2. — Total participants that wants integration of on a Likert scale?” 83% (78 of 93) answered an 8 simulation. or more, 16% (14 of 93) answered between 6-7 and 1% (1 of 93) answered less than 5 (Fig. 3). To the question : “Which topics would you like to practice during simulation-based medical training?”. 87% (88 of 102) answered : “anesthetic complications and their management”, 83% (84 of 102) answered : “scenario training”, 68% (69 of 102) answered : “advanced life support”, 50% answered : “crisis resource management”, 48% (49 of 102) answered : “practice of technical skills”, 2% (2 of 102), answered : “I do not want to practice simulation- Fig. 3. — Added value of simulation on a scale 1-10. based medical training”. To the question : “Are you willing to pay for simulation-based medical training as part of your educational program?”, 40% (41 of sequently, to the question : “How much are you 102) answered : “Yes”, 36% (37 of 102) answered : willing to spend on one training session”, 60% (60 “Maybe”, 23% (24 of 102) answered : “No”. Sub- of 102) answered : “less than 100 euro”, 16% (17 of

05-De Meirsman.indd 31 12/01/2021 12:11 32 r. de meirsman, e. van gerven 102) answered : “Between 100 and 200 euro”, 1% (1 of 102) answered : “more than 300 euro”, 23% (24 of 102) answered : “I do not want to pay for simulation- based medical education”. To the question : “When would you like to practice simulation-based medical education?”, 87% (88 of 102) answered : “during working hours”, 43% (44 of 102) answered : “after hours”, 12% (12 of 102) answered : “in weekends”, 2% (2 of 102) answered : “I do not want simulation-based medical training”. To the question : “On a Likert scale of 1 to 10 do you think there is an added value of simulation-based medical Fig. 4. — Prevalence of perioperative emergency events. education towards maintaining accreditation after your graduation as an anesthesiologist?”, 48% (49 (9 of 51) of cases, “Problems outside the OR”, in of 102) answered with a score of 8 or higher, 20% 8% (4 of 51) of cases, “Difficult airway” in 8% (4 answered (19 of 102) a score of 7, 17% answered of 51) of cases, “a non-specific scenario”. in 25% (16 of 102) answered 5 or 6, 18% (18 of 102) (13 of 51) of cases, Senior assistants reported : answered a score of 4 or lower. To the question : “Difficult airway” in 30% of cases (7 of 23), “What do you think the barriers are towards “Problems outside the OR” in 30% of cases (7 practical integration of simulation-based medical of 23), “Anaphylaxis” in 13% (3 of 23) of cases, education?”, 78% (79 of 102) answered : “time”, “Reanimation” in 13% (3 of 23) of cases, “A non- 71% (72 of 102) answered : “resources”, 47% (48 of specific scenario” in 13% of cases (3 of 23) (Fig. 102) answered : “management/ event organizing”, 5). To the question : “Have you ever participated in 21% (22 of 102) answered : “peer judgement, skill training?”, 69% (70 of 102) answered : “Yes”, stressful environment”. To the question : “Have 31% (32 of 102) answered : “No”. To the question : you ever encountered a perioperative emergency “How are these skills taught?”, 67% (65 of 102) event during your education, which would have answered : “Simulation-based medical training”, been better trained beforehand during a simulation- 49% (47 of 102) answered : “textbook”, 47% (46 of based medical training session?”, 73% (75 of 102) 102) answered : “On a patient under supervision”, answered : “Yes”, 27% (27 of 102) answered : “No” 35% (34 of 102) answered : “ex cathedra”, 31% (Fig. 4). Subsequently, to the question : “How often (30 of 102) answered : “live demonstration”, 11% did you encounter such a perioperative emergency (10 of 105) answered : “On a patient without event?”, 51% (52 of 102) reported as often as once supervision”. To the question : “Are these skills per semester, 20% (18 of 102) reported once per taught on a standardized manner?”, 40% (40 of year, 29% (32 of 102) reported to not have had any 102) answered : “Yes”, 38% (38 of 102) answered : perioperative emergency events, which they could “Some skills are, other are not”, 22% (24 of 102) not manage during their education. Subsequently, to answered : “No”. To the question : “Have you ever the question : “Give a short description of such an participated in a scenario training?”, 76% (78 of event”, Junior assistants reported : “Anaphylaxis” 102) answered : “Yes”, 24% (24 of 102) answered : in 23% of cases (12 of 51), “Bronchospasm” in “No”. Subsequently to the question : “On a Likert 19% of cases (10 of 51), “Reanimation” in 17% scale of 1 to 10 did you find scenario training to

Fig. 5. — Type of peri-operative emergency events. A : input junior residents, B: input senior residents.

05-De Meirsman.indd 32 12/01/2021 12:11 the added value of simulation-based medical education 33 Potential barriers towards the integration of simulation-based education are time and resource/ money related (78%). The costs of simulation based education are proportional to the amount of fidelity (2), more fidelity doesn’t always mean better (8), Friedman et al showed that a simple model was as useful for learning how to place an epidural as an expensive simulator was (9). Nagendran et al. noted that virtual reality training appears to decrease operating time and operating performance Fig. 6. — Correlation between prior experience with CRM and of trainees learning laparoscopic procedures, im- its applicability. plying there could be a reduced cost associated with simulated based training (10). Cohen et al be a good learning experience?”, 75% (58 of 78) showed that simulation-based education was cost- answered with a score of 8 or higher, 14% (12 of 78) effective by preventing CBRI (7). Ongoing research answered with a score of 7, 11% (8 of 78) answered is focusing on simple, but effective simulated with a score of 6 or lower. To the question : “Do you based training tools to further evaluate the cost- feel more prepared for a perioperative emergency effectiveness of simulation based education (11). event after completing a scenario training?”, 75% Participants of the questionnaire were asked if (58 of 78) answered : “ Yes”, 20% (15 of 78) they would want to make a contribution, no clear answered : “Maybe”, 5% (4 of 78) answered : “No”. results were found. Examining the prevalence and To the question : “Would you like to do a scenario the incidence of perioperative emergency events based simulation-based training before a specific amongst anesthesia trainees gave surprising results. rotation, e.g. weaning of cardiac bypass”, 78% (80 The majority of trainees (75%) have encoun-tered of 102) answered : “Yes”, 7% (7 of 102) answered : perioperative emergency events which they had “No”, 15% (15 of 102) answered : “No idea”. To the rather trained for in a simulated setting. Encoun- question : “Are you familiar with Crisis Resource tering a perioperative emergency event during Management (CRM)?”, 30% (31 of 102) answered : anesthesia is highly probable, the most frequent “Yes”, 34% (32 of 102) answered : “I do not know experienced perioperative emergency events were : the core principles of crisis resource management, anaphylaxis, bronchospasm, difficult airway, life but I have heard of the concept CRM”, 36% (37 support. The incidence of perioperative emergency of 102) answered : “I do not know crisis resource events is not well documented and frequencies vary management, and I have not heard of the concept considerably between studies (12). An approximate CRM”. Subsequently to the question : “Do you frequency of IgE-mediated anaphylaxis is estimated think your knowledge of CRM is sufficiently large/ on 1 per 5,200 general anesthesia (12). Severe applicable in a real emergency situation?”, 18% (12 bradycardia (heart rate <50 beats per minute) and of 63) answered : “Yes”, 48% (30 of 63) answered : severe hypotension (mean arterial pressure <55 “No”, 34% (19 of 63) answered : “I don’t know” mm Hg), within a 10-minute window after in- (Fig. 6). duction of anesthesia has an incidence of 5.6% (13). Bronchospasm during ENT surgery has Discussion an incidence of 0.4% and laryngospasm has an incidence of 5% (14). Knowing the most reported To assess the added value of simulation-based perioperative emergency events and their relative training and to improve the quality of education incidence could help with the development of a in the university hospitals of Leuven’s curriculum tailored based curriculum for anesthesia trainees for anesthesiology residents a questionnaire was and can be an important factor during training and developed and administered. A response rate of could lead to a reduction in medical errors. However 61% was noted. The majority of participants (83%) more validated studies with a good study design reported a score of 8 or higher on a Likert scale must be performed to check for potential benefits towards the added value of simulation and there was of simulation-based training. The results of the a positive attitude with a high proposition towards questionnaire showed that only 30% participants integration of a simulation-based curriculum. know the basic principles of CRM and 18% of the These findings had an important effect towards participants feels comfortable in applying CRM into the integration of a simulation-based curriculum. daily practice. Proper usage of the CRM protocol

05-De Meirsman.indd 33 12/01/2021 12:11 34 r. de meirsman, e. van gerven should be imbedded in to the educational program, unclear and underreported in the current literature. and results should be researched. No clear evidence We were able to identify the most frequent adverse was found regarding CRM (15, 16). events from the questionnaire, which allows the possibility for focused simulation based training. Limitations Although the application of non-technical skills during emergency events is an essential part of Some questions of the survey have a different its management our questionnaire suggests basic response rate. This is due to some responders knowledge of Crisis Resource Management (CRM) not having participated in simulation, resulting is currently lacking, and more effort should be in different denominators and hence different made towards a systematic implementation, which percentages throughout the results section. A could be trained during simulation. The results of questionnaire based study design brings along this study should help set up future research on the bias. The heterogeneity of participants, i.e. starting topic and more well-conducted studies are needed trainees to graduated anesthesia residents, accom- to demonstrate a proven added value regarding panies the possibility for exposure bias. Also the simulation based medical education. attitude towards simulation of participants and nonparticipants should be taken into account. Experience gained outside of simulated based References education is different for every participant, as is 1. Rosen KR. 2008. The history of medical simulation. the retention of knowledge before and after training J Crit Care 23:157-166. https://doi.org/10.1016/j.jcrc. for every participant. This heterogeneity makes 2007.12.004 for a large bias in data analysis and validation of 2. Gaba DM. 2004. The future vision of simulation in health care. Qual Saf Heal Care 13:2–10. https://doi.org/10.1136/ the study. The study is limited only to university qshc.2004.009878 hospitals of Leuven, the results found are thereby 3. Kohn LT, Corrigan JM, Molla S. 1999. To Err Is Human. a sample size of currently practicing residents and Medicine (Baltimore) 126:312. https://doi.org/10.1017/ S095026880100509X should not be considered as representative for the 4. Ndividuals COREI. 2007. H Ealth C Are R Eform 2007 - I whole community. Ssues O Verview and P Roposal C Omparison. 43: 5. Evans L V., Dodge KL, Shah TD, et al. 2010. Simulation training in central venous catheter insertion: Improved onclusion C performance in clinical practice. Acad Med 85:1462–1469. https://doi.org/10.1097/ACM.0b013e3181eac9a3 Recent technological advances, increasing 6. Al McGaghie WC, Issenburgh SC CE. e. 2011. Does numbers of procedures and increasing numbers of Simulation based medical education yield better results than traditional clinical education? A meta-analytic compartive young anesthesia trainees, were the reasons to assess review of the evidence. Acadm Meded 86:706–711. https:// the added value of simulation-based training. This doi.org/10.1097/ACM.0b013e318217e119.Does has led to the development of a simulation-based 7. Cohen ER, Feinglass J, Barsuk JH, et al. 2010. Cost savings from reduced catheter-related bloodstream infection after medical curriculum, which allows for training of simulation-based education for residents in a medical common perioperative emergency/adverse events intensive care unit. Simul Healthc 5:98–102. https://doi. in a low risk environment. Simulation-based medi- org/10.1097/SIH.0b013e3181bc8304 8. Diederich E, Mahnken JD, Rigler SK, et al. 2015. The cal education makes it possible to teach theoretical effect of model fidelity on learning outcomes of a simu- and practical aspects of anesthesia and also lation-based education program for central venous allows for non-technical skill to be practiced, catheter insertion. Simul Healthc 10:360–367. https://doi. e.g. communication. The majority of participants org/10.1097/SIH.0000000000000117 9. Friedman Z, Siddiqui N, Katznelson R, et al. 2009. Clinical reported an added value towards simulation based Impact of Epidural Anesthesia Simulation on Short- and medical education and towards the integration of Long-term Learning Curve. Reg Anesth Pain Med 34:229– a simulation-based curriculum. Majors barriers 232. https://doi.org/10.1097/AAP.0b013e3181a34345 10. Nagendran M, Gurusamy KS, Aggarwal R, et al. 2013. towards the implementation are time and resource/ Virtual reality training for surgical trainees in laparoscopic money related. The current evidence on simulation surgery. Cochrane Database Syst Rev. https://doi. based education remains poorly substantiated, partly org/10.1002/14651858.CD006575.pub3 11. 2018. Innovations. Acad Emerg Med 25:S285–S298. due to the great heterogeneity in the type of training. https://doi.org/10.1111/acem.13433 More clarity is needed in this regard to help facilitate 12. Harboe M.D. T, Guttormsen M.D., Ph.D. AB, Irgens M.Sc. the integration of simulation based medical education A, et al. 2005. Anaphylaxis during Anesthesia in Norway: into routine practice. Perioperative emergencies A 6-Year Single-center Follow-up Study. Anesthesiol J Am Soc Anesthesiol 102:897-903 are to be expected during anesthesiology training, 13. Solomon SC, Saxena RC, Neradilek MB, et al. 2020. the incidence of these emergency events remains Forecasting a Crisis: Machine-Learning Models Predict

05-De Meirsman.indd 34 12/01/2021 12:11 the added value of simulation-based medical education 35 Occurrence of Intraoperative Bradycardia Associated With 15. Salas E, Wilson KA, Burke CS, Wightman DC. 2006. Hypotension. Anesth Analg 130: Does Crew Resource Management Training Work? An 14. von Ungern-Sternberg BS, Sommerfield D, Slevin L, et Update, an Extension, and Some Critical Needs. Hum al. 2019 Effect of Albuterol Premedication vs Placebo on Factors J Hum Factors Ergon Soc 48:392-412. https://doi. the Occurrence of Respiratory Adverse Events in Children org/10.1518/001872006777724444 Undergoing Tonsillectomies: The REACT Randomized 16. McGaghie WC. 2008. Research opportunities in simulation- Clinical Trial. JAMA Pediatr 173:527-533. https://doi. based medical education using deliberate practice. Acad org/10.1001/jamapediatrics.2019.0788 Emerg Med 15:995–1001. https://doi.org/10.1111/j.1553- 2712.2008.00246.x

05-De Meirsman.indd 35 12/01/2021 12:11

(Acta Anaesth. Belg., 2020, 71, 37-44) Comparison of the analgesic effects of perineural and intravenous dexamethasone on peripheral nerve block : a systematic review

I. Piepers (*), M. Desmet (**), M. Van de Velde (*)

Abstract : Objective : The objective of this systematic patients do not have access to potent intravenous review is to compare the duration of analgesia between analgesics (7, 8). Adjuvants and nerve catheters perineural and intravenous dexamethasone as an with continuous delivery of local anesthetics adjuvant to a peripheral nerve block with long-acting have tried to solve this problem. Catheters are local anesthetics. costly and practical problems of the postoperative Background : Perineural dexamethasone has proved to follow-up limit the use of catheter nerve blocks prolong the postoperative analgesia of local anesthetic in the ambulatory setting (9-11). Dexamethasone peripheral nerve blocks. However, a similar increase in postoperative analgesia duration is seen with the use of is a high-potent, long-acting glucocorticoid with intravenous dexamethasone. little mineralocorticoid effect and frequently used Methods : A systematic search in Medline and Scopus in anesthesia to prevent postoperative nausea and was executed up to April 2020 using the following MeSH vomiting (12). It has shown to prolong the duration terms : dexamethasone, nerve block, brachial plexus of analgesia as an adjuvant to a peripheral nerve block, perineural. block (1, 12-17). Intravenous dexamethasone Results : The analgesic duration of perineural versus demonstrates a similar prolongation of analgesia intravenous administration as an adjuvant to a long- in addition to a peripheral nerve block(18), even acting local anesthetic peripheral nerve block, was in without a nerve block is a substantial analgesic seven trials with perineural dexamethasone significantly effect noticeable (13, 19). The mechanism of prolonged of which two studies were evaluated as high- action of dexamethasone is still unclear. There risk bias. However, in five trials, no difference between are arguments for a local influence containing perineural and intravenous dexamethasone was observed. Discussion and Conclusion : No clinically significant vasodilatation and inhibition of nociceptive C-fibres difference in the analgesic duration of perineural versus (20-25). However, a systemic effect through the intravenous administration as an adjuvant to a long- known anti-inflammatory characteristics is another acting local anesthetic peripheral nerve block could be explanation of the analgesic prolongation (19, withheld. The intravenous route is the preferred method 26, 27). From this Cochrane systematic review of administration as no clinically significant analgetic (28), we learn with a moderate grade of evidence advantages are seen with the perineural route, and the that the duration of sensory block with perineural remaining questions exist of possible neurotoxicity with dexamethasone is significantly longer (three hours) perineural dexamethasone use. than with the intravenous counterpart. However, this does not translate in clinically significant lower Keywords : Dexamethasone ; nerve block ; brachial pain scores at 12 and 24 hours postoperatively and plexus block ; perineural. no difference at all at 48 hours.

Introduction I. Piepers, MD ; M. Desmet, MD, PhD ; M. Van de Velde, MD, PhD, EDRA (*) Department of Anesthesiology, UZ Leuven, 3000 Leuven, Regional anesthesia is an excellent addition Belgium and alternative for general anesthesia and proved to (**) Department of Anesthesiology, AZ Groeninge, 8500 reduce the postoperative use of opioid analgesics (1- Kortrijk, Belgium Corresponding author : I. Piepers, MD. Department of 5). Painful procedures, such as shoulder surgery, are Anesthesiology, UZ Leuven, Herestraat 49, 3000 Leuven, now commonly performed in an ambulatory setting Belgium. (6). Despite the use of peripheral nerve blocks, which E-mail : [email protected] facilitated the transition to ambulatory surgery for Paper submitted on May 23, 2020 and accepted on May these painful procedures, is postoperative pain 26,2020. the most limiting factor in this setting in which Conflict of interest: none.

06-Piepers.indd 37 12/01/2021 12:12 38 j. piepers et al. The perineural use of dexamethasone is off- the prolongation of a ropivacaine peripheral sensory label as a result of possible neurotoxic effects(29). nerve block when intravenous dexamethasone An increased ropivacaine-neurotoxicity was is administrated in volunteers. In this volunteer- seen on isolated nerve cells when large doses model, no surgery-induced inflammation could be of dexamethasone was added to a ropivacaine simulated. These studies observed no benefit of administration(30). Solutions free of neurotoxic intravenous dexamethasone relative to ropivacaine preservatives, solvents, and particulates should be alone. Therefore, as it seems that the perioperative used (31-33). As long as the possible neurotoxic local or systemic inflammation is involved, effect is not yet clarified, a preference for the only randomized controlled trials with patients intravenous route of administration can become undergoing surgery were withheld from the first evident. Definitely, if perineural administration can pool of studies. not demonstrate a prolonged duration of analgesia relative to intravenous dexamethasone. Intervention and comparator The objective of this systematic review is to compare the analgesic duration of perineural Participants receiving perineural dexametha- versus intravenous administration as an adjuvant sone, in addition to a peripheral nerve block with to a long-acting local anesthetic peripheral nerve a long-acting local anesthetic, is the investigated block. We chose to evaluate this comparison in intervention. The comparison are those receiving randomized controlled trials, including an adult a similar peripheral nerve block with intravenous patient population. dexamethasone administration.

Methods Outcomes

This systematic review reports according to The duration of postoperative analgesia is Preferred Reporting Items for Systematic Reviews the outcome measure. Trials can use different and Meta-Analyses (PRISMA) recommendations measuring techniques : time until first pain at the for systematic reviews(34). A protocol for this study surgical site, time until first analgesic request, time was not registered. According to the Education until first breakthrough analgesic medication. The Guidance Committee for the Ethical Approval of degree of pain was estimated using different rating Master’s Theses of the Biomedical Sciences Group scales and cut-off values, for example, the Visual at KU Leuven, no application for ethical approval is Analogue Scale (VAS), a validated pain intensity needed for a systematic review. measurement (37-39).

Literature search and inclusion criteria Trial characteristics

Published reports of randomized controlled Following trial characteristics were extracted trials evaluating the duration of analgesia were for evaluation : the number of patients in the explored using the National Library of Medicine’s study, the age of the population, the used long- PubMed database (Medline database) and Scopus acting local anesthetic, the concentration of local with the last search on the 20th April 2020. No anesthetic, the volume of local anesthetic, the type language and year limits were placed on the of peripheral nerve block, the type of surgery, the search. Medline database, for example, was dose of perineural and intravenous dexamethasone, electronically investigated using “dexamethasone” the duration of analgesia outcome parameter, the AND “nerve block” AND “perineural”. A second postoperative analgesia duration, if there was series of keywords existed of “dexamethasone” funding, and conflicts of interest. AND “brachial plexus block” AND “perineural”. Both keyword-series were also used in the Scopus Rating of the studies database as well. We used the Cochrane Collaboration’s tool Population for assessing risk of bias in randomized controlled trials (40). The following topics are included in The investigated population consists of the evaluation : selection bias, performance bias, adult patients undergoing surgery. The studies by detection bias, attrition bias, reporting bias, and Marhofer et al. (35) and Short et al. (36) investigated other bias.

06-Piepers.indd 38 12/01/2021 12:12 review perineural versus intravenous dexamethasone 39 Study selection and data extraction Two studies did not explicitly mention an adult population in their article. Morales-Munoz et al. The first collection of trials from the literature (41) investigated a population receiving knee- search was further eliminated independently by the replacement surgery with an average 70 years of investigator through removing trial duplicates and age, and Vetriselvan et al.(42) studied a population using the inclusion principles (randomized controlled of women with infertility averaging 28 years of trial, population, intervention, comparator, and age. Various long-acting local anesthetics are used, outcome criteria). For this elimination, the study mainly ropivacaine and bupivacaine. Two trials title and abstract were used. If insufficient details used a mixture of bupivacaine and lidocaine (43, were available, a full-text article was evaluated 44), and one study worked with levobupivacaine in a second step. Of the remaining items, the trial (42). A significant difference in the applied volume characteristics were extracted. of long-acting anesthetic is seen, especially for interscalene nerve blocks (12 mL versus 30 mL) Results (45, 46). The largest share of surgery consisted of arthroscopic shoulder interventions, followed by the Study selection remainder of the upper limb, foot and ankle surgery, knee replacement, and laparoscopic gynecologic In total, 250 studies have been screened (102 operations. The majority of trials evaluated equal items from Pubmed, 148 from Scopus) for inclusion doses of perineural and intravenous dexamethasone, and exclusion criteria using the title and abstract, most frequently eight milligrams. Marty et al.(47) figure 1. After removing the duplicates, 18 full- used four milligrams of dexamethasone perineurally text articles were assessed for eligibility. Another and ten milligrams intravenously. A variety of six studies had to be excluded based on different outcomes are used across the studies in an attempt outcome parameters, volunteer-population, and an to contain the most clinically relevant postoperative equivalence trial instead of a randomized controlled time window of analgesia. The time window (superiority) trial. always starts from the moment the nerve block is administered. The endpoint is in five studies the first Study characteristics pain in the surgical area regardless of the intensity. In four trials is the use of a rescue (opioid) analgetic Table 1 contains an overview of the study the marker. In the remaining three studies is the characteristics. All 12 studies evaluated an adult intake of any analgetic the endpoint. No randomized population ranging from 39 to 280 patients. control trials reported any conflict of interest. Four studies published a source of funding (46-49).

Bias

Figure 2 contains an overview of the assessment of risk of bias. Six studies were considered low-risk bias by the reviewer. Kawanishi et al.(50) did not mention the method of random sequence generation and, therefore, was deemed to be unclear risk bias. The method of concealment was not clearly mentioned in three trials resulting in an unclear risk bias (45, 47, 51). Blinding of participants and personnel, the third item of the Cochrane Collaboration’s tool for assessing risk of bias in randomized trials, was not fully described in three of the investigated studies regarding the participants, study employees, and all concerned health care providers (45, 50, 52). Morales-Munoz et al. (41) described that the operating room anesthesiologist was aware of Fig. 1. — Flow diagram of articles for our systematic review. the patient’s group allocation. This insufficient n, number of studies. blinding was considered a high-risk bias. Only two

06-Piepers.indd 39 12/01/2021 12:12 40 j. piepers et al.

Table 1 Trial characteristics. TAP, transversus abdominis plane. PN, perineural. IV, intravenous

06-Piepers.indd 40 12/01/2021 12:12 Table 2 Overview of trial bias assessment review perineural versus intravenous dexamethasone 41 studies did not specify the blinding of outcome sufficient to be clinically significant (46). A second assessment resulting in an unclear risk bias (47, trial’s analgesia duration contained another order of 50). All randomized controlled trials used complete effect, an increase of almost a thousand minutes of outcome data, or could explain the excluded patients analgesia for the perineural dexamethasone group with a balanced drop-out between groups. Chun et compared to the two other intervention groups al. (45) used a different primary outcome parameter (intravenous dexamethasone and control group) in the registered protocol and the published article (41). This trial had a high risk of bias concerning conclusion. A high-risk bias was given for this operating room anesthesiologists not being blinded. adjustment. In the study of Kawanishi et al.(50) The remaining five studies could not demonstrate there is an outlier in the intravenous dexamethasone any significant difference between perineural and group regarding the median time to first request intravenous dexamethasone administration as an for analgesia. This patient was not included in the adjunct to a long-acting peripheral nerve block (47- analysis resulting in a high-risk bias for selective 50, 52). reporting. No other type of bias was noted. Table 2 contains the bias data. Summary of findings This reviews primary endpoint, the analgesia with perineural versus intravenous dexamethasone, The analgesic duration of perineural versus is significantly longer in seven out of twelve intravenous administration as an adjuvant to a randomized controlled trials. Of these seven trials, long-acting local anesthetic peripheral nerve block, one trial analgesia time prolongation was not was in seven trials with perineural dexamethasone significantly prolonged of which two studies were evaluated as high-risk bias. However, in five trials, no difference between perineural and intravenous dexamethasone was observed.

Discussion

Studies involving dexamethasone and nerve block frequently exclude specific populations ; people with diabetes mellitus, peptic ulcer, high BMI, American Society of Anesthesiologists (ASA) IV classification, pregnancy, severe respiratory disease, cognitive or psychiatric history, and chronic pain. As a result, systematic reviews and meta-analysis regarding this subject cannot make statements for these populations, and caution is advised regarding extrapolating the resulting findings to these specific patients. In the pediatric population is pain management an essential item in the perioperative phase. According to the Cochrane study, no evidence in children exist, and future research should include the pediatric population (28). The included studies used a variety of long- acting local anesthetics, mainly ropivacaine and bupivacaine. A more pronounced increase in the duration of analgesia with one of the two mentioned long-acting local anesthetic is shown in trials(53, 54). Additional research could clarify which local anesthetic is preferred in combination with dexamethasone. Potential neurotoxicity should also be evaluated in future trials when specific dexamethasone augmented local anesthetic Fig. 2. — Risk of bias assessment. neurotoxicity could be confirmed (30). The volume

06-Piepers.indd 41 12/01/2021 12:12 42 j. piepers et al.

Table 2 Overview of trial bias assessment

and concentration of both the local anesthetic of ropivacaine 0.5%. The control and intravenous and dexamethasone are essential factors in such dexamethasone group reached an average duration research. In this systematic review, multiple of femoral nerve block analgesia of around locations of peripheral nerve blocks are performed 200 minutes. On the other hand, the perineural to anesthetize different surgery areas. The site dexamethasone group averaged an estimated 1160 of the nerve block has an impact on the tissue minutes of femoral block analgesia (using figure 2 blood perfusion, the resulting systemic absorption of their article). rates, the depot function for both local anesthetic Further studies investigating the mechanism of and dexamethasone, and the anatomic content action, the possible toxic effects of dexamethasone, of the nerve (ratio fascicles versus connective and their potential systemic implications on tissue) all potential influencing the duration of the infection rate and wound healing are needed. postoperative analgesia. The majority of included The different outcome parameters make the studies contained a population receiving upper limb drafting of a meta-analysis more difficult. The same surgery. More studies concerning the lower limb applies to the use of various statistical units. Desmet are needed to be able to answer the question of et al. (52) presented their results using the median perineural versus intravenous dexamethasone fully. time, with an interquartile range. The median is very Marty et al. (47) used a different dose of similar to the mean when the data exhibits a normal dexamethasone in the perineural and intravenous distribution. However, medians are often reported groups to analyze the superiority of the perineural because the data is skewed (55). route. They argued that in previous studies, an equivalence between routes was seen when Conclusion high doses of dexamethasone was used, and an advantage of perineural administration with low No clinically significant difference in the doses. Nevertheless, no significant difference was analgesic duration of perineural versus intravenous reported in their trial. administration as an adjuvant to a long-acting local No risk of bias at the outcome level was anesthetic peripheral nerve block could be withheld. carried out, only at the study level. The results from The inconsistency in the results of the included Morales-Munoz et al. (41) could have called some trials in this systematic descriptive review and the questions. Different results are observed regarding risk of bias are essential factors. The intravenous the duration of femoral block analgesia with 20 mL route is the preferred method of administration as

06-Piepers.indd 42 12/01/2021 12:12 review perineural versus intravenous dexamethasone 43 no clinically significant analgetic advantages are intercostal blockade (T9 to T11) in healthy volunteers. seen with the perineural route, and the remaining Anesth Analg. 96: 576-582. 15. Holte K., Werner MU., Lacouture PG., Kehlet H. questions exist of possible neurotoxicity with 2002. Dexamethasone prolongs local analgesia after perineural dexamethasone use. subcutaneous infiltration of bupivacaine microcapsules in human volunteers. Anesthesiology. 96: 1331-1335. 16. Albrecht E., Kern C., Kirkham KR. 2015. A systematic Acknowledgments review and meta-analysis of perineural dexamethasone for peripheral nerve blocks. Anaesthesia. 70: 71-83. The authors have no conflict of interest 17. Huynh TM., Marret E., Bonnet F. 2015. Combination of dexamethasone and local anaesthetic solution in peripheral to report. There is no source of funding to be nerve blocks: A meta-analysis of randomised controlled mentioned. trialsEuropean Journal of Anaesthesiology. 32: 751-758. 18. Cummings KC., Napierkowski DE., Parra-Sanchez I., Kurz A., Dalton JE., Brems JJ., Sessler DI. 2011. Effect of dexamethasone on the duration of interscalene nerve blocks References with ropivacaine or bupivacaine. Br J Anaesth. 107: 446- 453. 1. Choi S., Rodseth R., McCartney CJL. 2014. Effects of 19. De Oliveira GS., Almeida MD., Benzon HT., McCarthy RJ. dexamethasone as a local anaesthetic adjuvant for brachial 2011. Perioperative single dose systemic dexamethasone plexus block : A systematic review and meta-analysis of for postoperative pain: A meta-analysis of randomized randomized trials. Br J Anaesth. 112: 427-439. controlled trials. Anesthesiology. 115: 575-588. 2. Fredrickson MJ., Krishnan S., Chen CY. 2010. Postoperative 20. Wang PH., Tsai CL., Lee JS., Wu KC., Cheng KI., Jou IM. analgesia for shoulder surgery: A critical appraisal and 2011. Effects of topical corticosteroids on the sciatic nerve: review of current techniques. Anaesthesia. 65: 608-624. An experimental study to adduce the safety in treating 3. Axelsson K., Gupta A. 2009. Local anaesthetic adjuvants: carpal tunnel syndrome. J Hand Surg Eur Vol. 36: 236-243. Neuraxial versus peripheral nerve block. Curr Opin 21. Shishido H., Kikuchi S., Heckman H., Myers RR. 2002. Anaesthesiol. 22: 649-654. Dexamethasone decreases blood flow in normal nerves and 4. Brummett CM., Williams BA. 2011. Additives to local dorsal root ganglia. Spine (Phila Pa 1976). 27: 581-586. anesthetics for peripheral nerve blockade. International 22. Castillo J., Curley J., Hotz J., Uezono M., Tigner J., Anesthesiology Clinics. 49: 104-116. Chasin M. et al. 1996. Glucocorticoids prolong rat sciatic 5. Borgeat A., Tewes E., Biasca N., Gerber C. 1998. Patient- nerve blockade in vivo from bupivacaine microspheres. controlled interscalene analgesia with ropivacaine after Anesthesiology. 85: 1157-1166. major shoulder surgery: PCIA vs PCA. Br J Anaesth. 81: 23. Johansson A., Hao J., Sjölund B. 1990. Local corticosteroid 603-605. application blocks transmission in normal nociceptive 6. Ardon AE., Prasad A., McClain RL., Melton MS., Nielsen C-fibres. Acta Anaesthesiol Scand. 34: 335-338. KC., Greengrass R. 2019. Regional Anesthesia for 24. Attardi B., Takimoto K., Gealy R., Severns C., Levitan ES. Ambulatory Anesthesiologists. Anesthesiol Clin. 37: 265- 1993. Glucocorticoid induced up-regulation of a pituitary 287. K+ channel mRNA in vitro and in vivo. Recept Channels. 7. Rancourt MPM., Albert NT., Côté M., Létourneau DR., 1: 287-293. Bernard PM. 2012. Posterior tibial nerve sensory blockade 25. Eker HE., Cok OY., Aribogan A., Arslan G. 2012. duration prolonged by adding dexmedetomidine to Management of Neuropathic Pain with Methylprednisolone ropivacaine. Anesth Analg. 115: 958-62. at the Site of Nerve Injury. Pain Med. 13: 443-451. 8. Marhofer D., Kettner SC., Marhofer P., Pils S., Weber M., 26. Barnes PJ. 1998. Anti-inflammatory actions of gluco- Zeitlinger M. 2013. Dexmedetomidine as an adjuvant to corticoids: Molecular mechanisms. Clin Sci. 94: 557-572. ropivacaine prolongs peripheral nerve block: A volunteer 27. Movafegh A., Razazian M., Hajimaohamadi F., Meysamie study. Br J Anaesth. 110: 438-442. A. 2006. Dexamethasone added to lidocaine prolongs 9. Ilfeld BM. 2011. Continuous peripheral nerve blocks: A axillary brachial plexus blockade. Anesth Analg. 102: 263- review of the published evidence. Anesth Analg. 113: 904- 267. 925. 28. Pehora C., Pearson AME., Kaushal A., Crawford MW., 10. Yanovski B., Gaitini L., Volodarski D., Ben-David B. 2012. Johnston B. 2017. Dexamethasone as an adjuvant to Catastrophic complication of an interscalene catheter for peripheral nerve block. Cochrane Database Syst Rev. 11. continuous peripheral nerve block analgesia. Anaesthesia. 29. Sakae TM., Marchioro P., Schuelter-Trevisol F., Trevisol 67: 1166-1169. DJ. 2017. Dexamethasone as a ropivacaine adjuvant for 11. Choquet O., Capdevila X. 2013. Three-dimensional ultrasound-guided interscalene brachial plexus block: A high-resolution ultrasound-guided nerve blocks: A new randomized, double-blinded clinical trial. J Clin Anesth. panoramic vision of local anesthetic spread and perineural 38: 133-136. catheter tip location. Anesth Analg. 116: 1176-1181. 30. Williams BA., Hough KA., Tsui BYK., Ibinson JW., Gold 12. De Oliveira GS., Castro-Alves LJS., Ahmad S., Kendall MS., Gebhart GF. 2011. Neurotoxicity of adjuvants used MC., McCarthy RJ. 2013. Dexamethasone to prevent in perineural anesthesia and analgesia in comparison with postoperative nausea and vomiting: An updated meta- ropivacaine. Reg Anesth Pain Med. 36: 225-230. analysis of randomized controlled trials. Anesth Analg. 31. Neal JM., Rathmell JP., Rowlingson JC. 2009. Editorial: 116: 58-74. Publishing studies that involve “off-label” use of drugs 13. Waldron NH., Jones CA., Gan TJ., Allen TK., Habib formalizing regional anesthesia and pain medicine’s policy. AS. 2013. Impact of perioperative dexamethasone on Reg Anesth Pain Med. 34: 391-392. postoperative analgesia and side-effects: Systematic review 32. Benzon HT., Gissen AJ., Strichartz GR., Avram MJ., and meta-analysis. Br J Anaesth. 110: 191-200. Covino BG. 1987. The effect of polyethylene glycol on 14. Kopacz DJ., Lacouture PG., Wu D., Nandy P., Swanton mammalian nerve impulses. Anesth Analg. 66: 553-559. R., Landau C. 2003. The dose response and effects 33. Rosenfeld DM., Ivancic MG., Hattrup SJ., Renfree KJ., of dexamethasone on bupivacaine microcapsules for Watkins AR., Hentz JG. et al. 2016. Perineural versus

06-Piepers.indd 43 12/01/2021 12:12 44 j. piepers et al. intravenous dexamethasone as adjuncts to local anaesthetic 45. Chun EH., Kim YJ., Woo JH. 2016. Which is your choice brachial plexus block for shoulder surgery. Anaesthesia. 71: for prolonging the analgesic duration of single-shot 380-388. interscalene brachial blocks for arthroscopic shoulder 34. Liberati A., Altman DG., Tetzlaff J., Mulrow C., Gøtzsche surgery? intravenous dexamethasone 5mg vs. perineural PC., Ioannidis JPA. et al. 2009. The PRISMA statement for dexamethasone 5mg randomized, controlled, clinical trial. reporting systematic reviews and meta-analyses of studies Medicine (Baltimore). 95. that evaluate healthcare interventions: explanation and 46. Holland D., Amadeo RJJ., Wolfe S., Girling L., Funk F., elaboration. BMJ. 339: 1-34. Collister M. et al. 2018. Effect of dexamethasone dose and 35. Marhofer P., Columb M., Hopkins PM., Greher M., Marhofer route on the duration of interscalene brachial plexus block D., Levi Bienzle MR., Zeitlinger M. 2019. Dexamethasone for outpatient arthroscopic shoulder surgery: a randomized as an adjuvant for peripheral nerve blockade: a randomised, controlled trial. Can J Anesth. 65: 34-45. triple-blinded crossover study in volunteers. Br J Anaesth. 47. Marty P., Rontes O., Chassery C., Vuillaume C., Basset 122: 525-531. B., Merouani M. et al. 2018. Perineural Versus Systemic 36. Short A., El-Boghdadly K., Clarke H., Komaba T., Dexamethasone in Front-Foot Surgery under Ankle Block: Jin R., Chin KJ., Chan V. 2020. Effect of intravenous A Randomized Double-Blind Study. Reg Anesth Pain Med. dexamethasone on the anaesthetic characteristics of 43: 732-737. peripheral nerve block: a double-blind, randomised 48. Abdallah FW., Johnson J., Chan V., Murgatroyd H., Ghafari controlled, dose-response volunteer study. Br J Anaesth. M., Ami N. et al. 2015. Intravenous dexamethasone and 124: 92-100. perineural dexamethasone similarly prolong the duration 37. Thong ISK., Jensen MP., Miró J., Tan G. 2018. The validity of analgesia after supraclavicular brachial plexus block: A of pain intensity measures: What do the NRS, VAS, VRS, randomized, triple-arm, double-blind, placebo-controlled and FPS-R measure? Scand J Pain. 18: 99-107. trial. Reg Anesth Pain Med. 40: 125-132. 38. Ferreira-Valente MA., Pais-Ribeiro JL., Jensen MP. 2011. 49. Rahangdale R., Kendall MC., McCarthy RJ., Tureanu L., Validity of four pain intensity rating scales. Pain. 152: Doty R., Weingart A., De Oliveira GS. 2014. The effects 2399-2404. of perineural versus intravenous dexamethasone on sciatic 39. Alghadir AH., Anwer S., Iqbal A., Iqbal ZA. 2018. Test- nerve blockade outcomes: A randomized, double-blind, retest reliability, validity, and minimum detectable change placebo-controlled study. Anesth Analg. 118: 1113-1119. of visual analog, numerical rating, and verbal rating scales 50. Kawanishi R., Yamamoto K., Tobetto Y., Nomura K., Kato for measurement of osteoarthritic knee pain. J Pain Res. 11: M., Go R. et al. 2014. Perineural but not systemic low-dose 851-856. dexamethasone prolongs the duration of interscalene block 40. Higgins JPT., Altman DG., Gøtzsche PC., Jüni P., Moher with ropivacaine: a prospective randomized trial. Local D., Oxman AD. et al. 2011. The Cochrane Collaboration’s Reg Anesth. 7: 5-9. tool for assessing risk of bias in randomised trials. BMJ. 51. Godbole M., Karhade S., Parihar P. 2019. A prospective 343. study of comparison of analgesic efficacy of dexamethasone 41. Morales-Muñoz C., Sánchez-Ramos J-L., Díaz-Lara MD., as an adjuvant in supraclavicular block with intravenous González-González J., Gallego-Alonso I., Hernández- dexamethasone after supraclavicular block in patients Del-Castillo MS. 2017. Analgesic effect of a single- undergoing forearm surgeries. Anesth Essays Res. 13: 31- dose of perineural dexamethasone on ultrasound-guided 35. femoral nerve block after total knee replacement. Rev Esp 52. Desmet M., Braems H., Reynvoet M., Plasschaert S., Van Anestesiol Reanim. 64: 19-26. Cauwelaert J., Pottel H. et al. 2013. I.V. and perineural 42. Vetriselvan P., Mandal B., Bhatia N., Jain V. 2019. Effect dexamethasone are equivalent in increasing the analgesic of dexamethasone on analgesic efficacy of transverse duration of a single-shot interscalene block with ropivacaine abdominis plane block in laparoscopic gynecological for shoulder surgery: A prospective, randomized, placebo- procedures: A prospective randomized clinical study. J controlled study. Br J Anaesth. 111: 445-452. Anaesthesiol Clin Pharmacol. 35: 165-169. 53. Baeriswyl M., Kirkham KR., Jacot-Guillarmod A., 43. Aliste J., Leurcharusmee P., Engsusophon P., Gordon A., Albrecht E. 2017. Efficacy of perineural vs systemic Michelagnoli G., Sriparkdee C. et al. 2017. Comparaison dexamethasone to prolong analgesia after peripheral nerve randomisée entre la dexaméthasone intraveineuse et block: A systematic review and meta-analysis. Br J Anaesth périneurale pour réaliser un bloc axillaire échoguidé. Can 119: 183-191. J Anesth. 64: 29-36. 54. Bindal D., Narang N., Mahindra R., Gupta H., Kubre J., 44. Leurcharusmee P., Aliste J., Van Zundert TCRV., Saxena A. 2018. Effect of dexamethasone on characteristics Engsusophon P., Arnuntasupakul V., Tiyaprasertkul W. et of supraclavicular nerve block with bupivacaine and al. 2016. A multicenter randomized comparison between ropivacaine: A prospective, double-blind, randomized intravenous and perineural dexamethasone for ultrasound- control trial. Anesth Essays Res. 12: 234-239. guided infraclavicular block. Reg Anesth Pain Med. 41: 55. Higgins JPT., Green S. 2011. Cochrane Handbook for 328-333. Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration.

06-Piepers.indd 44 12/01/2021 12:12 (Acta Anaesth. Belg., 2020, 71, 45-56) Pain management after open inguinal hernia repair : an updated systematic review and procedure-specific postoperative pain management (PROSPECT/ESRA) recommendations

S. Coppens1 (*), J. Gidts1 (*), P. Huynen (*), M. Van de Velde (*), G. P. Joshi (*), on behalf of the PROSPECT Working Group collaboration : G. P. Joshi, E. Pogatzki-Zahn, M. Van de Velde, S. Schug, H. Kehlet, F. Bonnet, N. Rawal, A. Delbos, P. Lavand’homme, H. Beloeil, J. Raeder, A. Sauter, E. Albrecht, P. Lirk, D. Lobo, S. Freys

Abstract : Background : Open inguinal hernia repair role of novel regional analgesic techniques such as can be associated with moderate-to-severe postoperative erector spinae blocks and to confirm the influence of the pain, which can delay return to activities of daily living. recommended analgesic regimen on postoperative pain The aim of this systematic review was to update the relief in an enhanced recovery setting. available literature and develop recommendations for optimal pain management after open inguinal hernia Key words : Open inguinal hernia repair ; pain ; analge- repair. A systematic review utilizing PROcedure sia ; systematic review ; evidence-based medicine. SPECific Postoperative Pain ManagemenT (PROSPECT) methodology was undertaken. Methods : Randomised controlled trials published in Recommendations the English language between January 1st 2009 and August 31st 2019, evaluating the effects of analgesic, 1. Systemic analgesia should include paracetamol anesthetic, and surgical interventions were retrieved from and a non-steroidal anti-inflammatory drug or cyclo- MEDLINE, EMBASE and Cochrane Databases. Of 203 oxygenase-2 selective inhibitor administered pre- eligible studies identified, 37 studies met the inclusion operatively or intra-operatively and continued post- criteria. operatively. Results : Interventions that improved postoperative 2. Local anesthetic infiltration and/or regional analgesia pain relief included paracetamol and nonsteroidal anti- technique (ilio-hypogastric/ilio-inguinal nerve blocks or inflammatory drugs or cyclooxygenase-2 selective 1 1 inhibitors, as well as local anesthetic infiltration and S. Coppens , MD ; J. Gidts , MD ; P. Huynen, MD ; M. Van de Velde, MD, PhD ; G. P. Joshi, MD, PhD ; on behalf of the regional analgesia techniques such as ilio-hypogastric/ PROSPECT Working Group collaboration. ilio-inguinal nerve blocks and transversus abdominis 1 S. Coppens and J. Gidts equally share first authorship plane blocks. Although effective, epidural analgesia or (*) Department of Cardiovascular Sciences, Section paravertebral blocks are considered invasive and harmful, Anesthesiology, KU Leuven and University Hospital and thus not recommended. Insufficient evidence was Leuven, Belgium (**) Department of Anesthesiology and Pain Management, found for psoas block, extended release local anesthetics, University of Texas Southwestern Medical Center, Dallas, wound infiltration using non-steroidal anti-inflammatory Texas, USA drugs, clonidine or opioids, topical conventional non- Corresponding author : Marc Van de Velde, Department steroidal anti-inflammatory drugs, systemic clonidine, of Cardiovascular Sciences, Section Anesthesiology, KU corticosteroids and ketamine, intravenous lidocaine Leuven and University Hospital Leuven, Belgium E-mail : [email protected] infusion, cryoanalgesia techniques, and nerve section. Inconsistent evidence was found for the use of Paper submitted on May 25, 2020 and accepted on May 26, gabapentinoids. 2020. Conclusion : The analgesic regimen for open inguinal Conflict of interest : PROSPECT is supported by an unrestricted hernia repair should include paracetamol and nonsteroidal grant from the European Society of Regional Anesthesia and Pain Therapy (ESRA). In the past, PROSPECT has anti-inflammatory drug or cyclooxygenase-2 selective received unrestricted grants from Pfizer Inc. New York, NY, inhibitor administered pre-operatively or intra- USA and Grunenthal, Aachen, Germany. GPJ has received operatively and continued post-operatively. In addition, honoraria from Baxter and Pacira Pharmaceuticals. MVdV local anesthetic infiltration and/or a regional analgesia has received honoraria from Sintetica, Grunenthal, Vifor technique (ilio-inguinal nerve blocks or transversus Pharma, MSD, Nordic Pharma, Janssen Pharmaceuticals, Heron Therapeutics and Aquettant. EP-Z has received abdominis plane blocks), with opioids used as rescue honoraria from Mundipharma, Grunenthal, MSD, Janssen- analgesics. Further studies are required to assess the Cilag GmbH, Fresenius Kabi and AcelRx.

07-Coppens.indd 45 12/01/2021 12:13 46 s. coppens et al. transversus abdominis plane blocks) is recommended. trations and use of paracetamol and non-steroidal 3. A single dose of intravenous dexamethasone is anti-inflammatory drugs (NSAIDs) or cyclo-oxy- recommended for its ability to increase the analgesic genase (COX)-2 selective inhibitors (1). Therefore, duration of the block, decrease analgesic use, and an updated systematic review on analgesic antiemetic effects. interventions for pain management in IHR was 4. Opioids should be reserved as rescue analgesics in the performed. In addition, it was deemed necessary to post-operative period. reassess the recommendations to align them with Why was this guideline developed? the updated PROSPECT approach that considers current clinical relevance and clinical effectiveness Open inguinal hernia repair is associated with moderate- by balancing the invasiveness of the analgesic to-severe postoperative pain which may influence interventions and the degree of pain after surgery, as recovery. The aim of this guideline is to provide clinicians well as balancing efficacy and adverse effects (6,7). with an updated evidence for optimal pain management after open inguinal hernia repair. The aim of this update was to systematically review the available literature assessing the anal- What other guidelines are available on this topic? gesic, anesthetic, and surgical interventions on pain after open IHR. Postoperative pain scores were The PROSPECT recommendations for pain management the primary outcome measures. Other recovery after open inguinal hernia repair have been published outcomes assessed included cumulative opioid previously, however, an update assessing analgesic interventions was necessary. requirements and adverse effects, when reported, and the limitations of the data were reviewed. The How does this guideline differ from other guidelines? ultimate aim was to develop recommendations for pain management after IHR. The updated systematic review further confirms the previous recommendations. Also, an updated Methodology PROSPECT approach was used to develop the current recommendations such that the available evidence is critically assessed for current clinical relevance The methods of this review adhered to the and the use of simple, non-opioid analgesic such as PROSPECT methodology as previously reported paracetamol and non-steroidal anti-inflammatory drugs (7). Specific to this study, the EMBASE, MEDLINE, as baseline analgesics. This approach reports true PubMed and Cochrane Databases (Cochrane Central clinical effectiveness by balancing the invasiveness of Register of Controlled Trials, Cochrane Database of the analgesic interventions and the degree of pain after Abstracts or Reviews of Effects, Cochrane Data- surgery, as well as balancing efficacy and adverse effects. base of Systematic Reviews) were searched for randomized controlled trials (RCTs) published between January 1st 2009 and August 31st 2019. Introduction The search terms used was (hernia OR inguinal OR inguinal hernia repair OR herniorrhaphy OR Open Inguinal Hernia Repair (IHR) is associated herniorraphy) AND (pain OR postoperative pain with moderate-to-severe postoperative pain, which OR analgesia OR anesthesia OR anesthetic) AND may delay recovery and return to activities of daily (anesthetics neuraxial OR intrathecal OR spinal living (1). In addition, inadequate pain control can OR epidural analgesia OR paravertebral blocks OR increase unplanned admission rate and readmission peripheral nerve OR peripheral block OR regional after discharge home (1). Furthermore, inadequate nerve OR transversus abdominis plane block OR post-operative pain relief may lead to hyperalgesia infiltration OR instillation OR NSAID OR COX-2 and persistent postoperative pain (2). OR paracetamol OR acetaminophen OR gabapentin Based on a systematic review performed OR pregabalin OR clonidine OR opioid OR in 2009, the PROSPECT (PROcedure SPECific ketamine OR corticosteroid OR dexamethasone OR Postoperative Pain ManagemenT) Working Group heavyweight mesh OR polypropylene mesh). (3,4), which is a collaboration of surgeons and Quality assessment, data extraction and data anesthetists, previously provided recommendations analysis adhered to the PROSPECT methodology for pain management in patients undergoing IHR [7]. Pain intensity scores were used as the primary (5). A recent International guideline for groin hernia outcome measure. In this study, we defined a change surgery provided non-specific statements regarding of more than 10 mm on the visual analogue scale pain management recommending perioperative (VAS) or numerical rating score (NRS) as clinically field blocks and/or subfascial/subcutaneous infil- relevant (8). The effectiveness of each intervention

07-Coppens.indd 46 12/01/2021 12:13 pain management after open inguinal hernia repair 47

Fig. 1. — PRISMA flow diagram of studies.

for each outcome was evaluated qualitatively by shown in Table S2 and Table S3. assessing the number of studies showing a signi- Systemic non-opioid analgesics ficant difference between treatment arms (p< 0.05 as reported in the study publication). A meta- One placebo-controlled RCT (n=50) reported analysis was not performed due to heterogeneity in gabapentin (400 mg) prior to surgery reduced pain study design and result reporting, restricting pooled scores and postoperative morphine requirements analysis. in the immediate postoperative period. No obvious Recommendations were made according side-effects were reported (9). Another placebo- to PROSPECT methodology (7). In brief, this controlled RCT (n=60) reported that preoperative involved a grading of A–D according to the overall gabapentin (1200 mg) significantly reduced pain level of evidence, as determined by the quality of scores (at rest and on sitting) and reduced the total studies included, consistency of evidence and study tramadol consumption at 8, 12, 16, 20 and 24 hours design. The proposed recommendations were sent after surgery (10). Pain scores at 1, 3 and 6 months to the PROSPECT Working Group for review and after surgery were lower in the gabapentin group than comments and a modified Delphi approach was in the placebo group (10). A large placebo-controlled utilised as previously described. Once a consensus RCT (n=425) found that preoperative pregabalin was achieved the lead authors drafted the final (50, 150, or 300 mg/day) as adjuvant analgesic did document, which was ultimately approved by the not influence the intensity of postoperative pain at Working Group. 24 hours (11). Of note, there was a wide variability in the anesthetic and analgesic techniques. Results One placebo-controlled RCT (n=77) found no statistical significant difference in pain during the PRISMA flow chart demonstrating the search first 24 hours, postoperatively with a TNF-α inhibitor, data are presented in Figure 1. The methodological etanercept 50 mg, administered subcutaneously 90 quality assessments of the 38 RCTs included for the min before incision (12). final qualitative analysis are summarized in Table A placebo-controlled RCT (n=55) evaluated S1. The characteristics of the included studies are the effects of rofecoxib 50 mg (a COX-2 selective

07-Coppens.indd 47 12/01/2021 12:13 48 s. coppens et al. Table S1 Quality assessment and level of evidence assigned to the randomised trials included in the review for analgesia after inguinal hernia repair surgery

Allocation Statistical analyses and Study Jadad Score Level of Evidence concealment: A-D patient follow-up Mahoori A et al 2014 (9) A 3 no 1 Sen H et al 2009 (10) A 3 no 1 Singla NK et al 2014 (11) A 4 no 1 Cohen SP et al 2013 (12) A 5 yes 1 Schurr M et al 2009 (13) A 3 no 1 Somri M et al 2017 (14) A 5 yes 1 Kang H et al., 2011 (15) A 4 no 1 Chaparro LE et al 2012 (16) A 5 yes 1 Mentes O et al 2009 (17) A 4 no 1 Viscusi E et al 2019 (18) A 5 yes 1 Velanovich V et al 2019 (19) A 5 yes 1 Neisioonpour SH et al 2013 (20) A 3 no 1 Santos SC et al., 2011 (21) C 3 no 2 Saeed M et al., 2015 (22) C 4 no 2 Baerentzen F et al., 2012 (23) A 4 no 1 Vizcaino-Martinez L et al 2014 (24) A 3 yes 1 Aveline C et al 2011 (25) A 5 no 1 Okur O et al 2017 (26) A 3 yes 1 Hosalli V et al 2019 (27) A 5 yes 1 Theodoraki K et al 2019 (28) A 5 yes 1 Wegner R et al 2017 (29) A 5 no 1 Razavizadeh MR et al 2017 (30) B 4 no 1 Akyol BC et al 2018 (31) A 3 yes 2 Heil JW et al 2014 (32) A 4 no 1 Bhattacharya P et al 2010 (33) A 4 no 1 Magnusson J et al 2016 (34) A 3 yes 1 Kingsnorth A et al 2012 (35) A 3 no 1 Manyilirah W et al 2012 (36) A 4 yes 1 Gundre NP et al 2012 (37) A 2 no 1 Torcivia A et al 2011 (38) A 1 no 1 Canonico S et al 2013 (39) B 3 no 1 Fan JKM et al 2017 (40) A 5 yes 1 Carro JLP et al 2017 (41) B 3 yes 1 Bracale V et al 2014 (42) A 2 no 1 Shen YM et al 2012 (43) A 2 no 1 Bona S et al 2018 (44) A 3 yes 2 Hoyuela C et al 2016 (45) A 5 yes 1

inhibitor that has been withdrawn from the market preoperatively (14). Pain scores at rest and on straight due to potential cardiac complications) given one leg raise were significantly lower in the etoricoxib hour pre-operatively followed by once daily for 4 group at 16 and 24 hours, and on discharge. days after surgery (13). Rofecoxib reduced pain A placebo-controlled RCT (n=64) assessed the scores on the first postoperative day, but the opioid effects of intra-operative intravenous bolus injection requirements remained similar to the placebo of 1.5 mg/kg lidocaine followed by a continuous group. A placebo-controlled RCT (n=60) evaluated infusion (2 mg/kg/h) (15). Fentanyl consumption the effects of oral etoricoxib 120 mg given 1 hour was significantly lower in the lidocaine group until

07-Coppens.indd 48 12/01/2021 12:13 pain management after open inguinal hernia repair 49 Table S2 Summary of key results from studies evaluating systemic analgesics, analgesics adjuncts, regional anesthesia, and surgical procedures used to support the recommended interventions in patients after open inguinal hernia repair

Study Study design Pain Scores Cumulative opioid dose Baseline Analgesia Pre-operative Systemic non-opioid analgesics Schurr M et al. 2009 (13) Rofecoxib 50 mg or placebo 1 hour Significantly lower ratings for No difference in milligrams Patients were given prior to surgery and once daily on least pain in patients with COX-2 hydrocodone bitartrate hydrocodone bitartrate as needed postoperative day 1-4. inhibitor intake on the operative consumed. postoperative. day. Somri M et al. 2017 (14) Etoricoxib 120 mg or placebo 1 hour Significantly lower pain scores in Amount of rescue analgesic When VAS≥4 at PACU 1-3mg preoperative. patients with COX-2 inhibitor at doses of morphine boluses of morphine were used 16, 24 h and on discharge, in rest (mg) within 24 h were until a response of 1 on a verbal and with active straight leg raise. significantly lower in COX-2 point descriptive scale of 5 was inhibitor group. achieved. Intra-operative Topical and surgical site infiltration with extended-release local anesthetics Nesioonpour SH et al. Wound infiltration prior to incision Mean pain score at all time The amount of opioid use for If NRS pain score > 4 intravenous 2013 (20) infiltrated by the surgeon. One group points (every 4 h during 24 h the bupivacaine group was admission of 15 mg pethidine was received 10 ml of bupivacaine 0.5% postoperative) were lower in the 80% less in the first 24 h. injected. and the control group placebo (10 ml treatment group. normal saline). Ilio-inguinal and ilio-hypogastric nerve block Santos SC et al. 2011 (21) Ilioinguinal and Iliohypogastric Visual analogue scale at rest was Not mentioned. VAS > 4 received intravenous nerve block with 10 mL of 0.75% significant lower 3 h after surgery. dipyrone (2.0 g). In case ropivacaine, and wound infiltration No difference at 6 and 12 h after ineffective, intravenous ketorolac with 10 mL of 0.75% ropivacaine surgery and with movement. (30 mg) and, whenever necessary, versus no block nor wound infiltration intravenous nalbuphine (3 mg), were added. Saeed M et al. 2015 (22) Iliohypogastric and ilioinguinal block Median VAS was lower in the Not mentioned, only time for VAS>3, on demand paracetamol (12ml 0.75% ropivacaine) and incision intervention group. demand first dose analgesia 1000mg and if necessary 75 mg infiltration prior to incision. diclofenac sodium. Baerentzen F et al. 2012 Ilio-hypogastric and ilio-inguinal Significant lower pain scores at No significant difference / (23) block with 20 mL bupivacaine 0.5% mobilization and rest upon arrival between the two groups. or a placebo block with 20 ml saline. in PACU and after 30 minutes.

TAP-block Aveline C et al. 2011 (25) Ultrasound-guided TAP block (levo- Vas pain scores at rest were lower No significant difference. Paracetamol 1000 mg and keto- bupivacaine 0.5%, 1.5mg/kg) versus in the US-guided TAP group at profen 100 mg i.v. during surgery. blind IH block (levobupivacaine 0.5%, 4h, 12h and 24h. At PACU 3 mg i.v. morphine 1.5mg/kg) before surgery. boluses at 5 min intervals, until VAS ≤3. Okur O et al. 2017 (26) TAP block versus ilio-hypogastric and Pain scores were significantly No significant difference Acetaminophen 10–15 mg/ ilio-inguinal nerve block versus no lower at all time points of the in additional analgesic kg once every 6 h and salvage block. Blocks were done with 20 ml study groups versus placebo. Only requirement was found analgesia of intravenous tramadol of 0.25% isobaric bupivacaine. at 24 h there was a significantly between study groups. hydrochloride 50 mg. difference between TAP and IHIN in favor of TAP.

Hosalli V et al. 2019 (27) Dual TAP-block and ilioinguinal/ VAS at 12, 24 and 48 h were A higher number of patients Fentanyl 0.5 μg/kg and, after 2 iliohypogastric block versus significantly lower in Dual TAP- required rescue analgesics in h if the pain persists (VAS score ilioinguinal/ iliohypogastric block. group compared to IL/IH group IL/IH group. of ≥4 at rest) even after giving alone. fentanyl intravenous tramadol 50 mg added. Theodoraki K et al. 2019 TAP-block (ropivacaine 0.75% 20ml) Significant different pain scores Lower median morphine con- Paracetamol 1g and parecoxib (28) versus placebo (saline 20ml) in favour of TAP-block in rest at 6 sumption via the PCA device 40mg intravenously 30min and at 24 h. For movement at 3, 6 for the first 24 h in comparison approximately before the end of and 24 h postoperatively. to the placebo group. surgery. Razavizadeh MR et al. Addition of dexamethasone 8 mg to Onset/duration of anesthesia Not mentioned. Not mentioned. 2017 (30) epidural block into epidural space was significantly more rapid and versus placebo (saline 2ml). longer in the dexamethasone group. Post-operative TAP-block Akyol BC et al. 2018 (31) Intravenous analgesic regimen Pain scores significantly lower in NRS-score > 4 received an During close-up of the surgery, (group I) versus subcostal TAP- group II compared with group I: additional dose of 1 mg/kg patients in both groups were block postoperatively in addition to postoperative 15th-minute; 1st-, tramadol IV. given 100mg tramadol and an intravenous analgesic regimen 6th-, 12th-, and 24th-hour; 15th- 1000mg paracetamol IV. (group II) day and 1st-month NRS values. NS = no significant difference between groups, POD = Postoperative Day.

07-Coppens.indd 49 12/01/2021 12:13 50 s. coppens et al. Table S3 Summary of key results from studies evaluating systemic analgesics, analgesics adjuncts, regional anesthesia, and surgical procedures used to support the NOT recommended interventions in patients open inguinal hernia repair surgery

Study Study design Pain Scores Cumulative opioid dose Baseline analgesia Pre-operative Systemic non-opioid analgesics Gabapentinoids Mahoori A et al. Gabapentin 400 mg versus placebo 2 h Reduced pain scores postoperative at Reduced postoperative mor- Morphine 0.05 mg/kg intra- 2014 (9) prior to herniorraphy 2, 4, 12 h and 24 h phine requirements venously on demand Sen H et al. 2009 Gabapentin 1200 mg or placebo 1 hour Gabapentin group had signifi-cantly Tramadol consumption was PCA device postoperative with (10) prior to surgery lower VAS scores at time 1, 4, 8, lower at time 8, 12, 16, 20 bolus of tramadol 20 mg (lockout 12, 16, 20 and 24 h post-operative and 24 h postoperative. interval 15 min and a 4 hour and also at 1,3 and 6 months maximal dose of 150 mg) VAS>4 postoperative were given 75mg diclofenac intramuscular Singla NK et al. 2014 Pregabalin 50, 150 or 300 mg/day or No statistically significant dif- Decreased by 41% and by 500 mg naproxen (bid) for 3 days. (11) placebo ference in pain scores between 300 59% in patients receiving 50 mg tramadol and 500–650 12 and 2 h prior to surgery and once mg/day pregabalin and placebo 150 and 300 mg/day pre- mg acetaminophen every 4 h, as daily 1 week postoperative gabalin, respectively, when needed. if still inadequate, 5 mg compared with placebo oxycodone every 4 h, as needed. TNF-alpha inhibitor Cohen SP et al. 2013 Etanercept 50 mg subcutaneous or The NRS pain scores were lower Etanercept group required Morphine as needed post-ope- (12) placebo 90 minutes prior to surgery. in the etanercept group though less opioids (morphine) rative. these results were not statistically though not statistically significant. significant. Intra-operative Systemic non-opioid analgesics Lidocaine Kang H et al. 2011 Intravenous bolus injection of 1.5 mg/ VAS pain scores in the lidocaine Total fentanyl consumption PCA infusion of 0.1 μg/kg/h (15) kg lidocaine followed by a continuous group were significantly lower at 2, (PCA and rescue admi-nistration fentanyl and on-demand lidocaine infusion of 2 mg/kg/h. 4, 8 and 12 h after surgery. was significantly lower in the release of a 0.1 μg/kg bolus. lidocaine group. When VAS>3 50 μg of fentanyl was administered. Dipyrone Chaparro LE et al. Dipyrone 15mg vs Dipyrone 40mg intra- D15 group, the D40 group showed There was more morphine Not mentioned. 2012 (16) venously after induction a lower incidence of moderate to consumption in the D15 group. severe pain only in the first 30 minutes. Topical and surgical site infiltration with extended-release local anesthetics HTX-011 Viscusi E et al. 2019 Near completion surgery local infil- HTX-011 significantly reduced Opioid consumption over 72 As needed: oral oxycodone (18) tration with HTX-011 300 mg/9 mg mean pain intensity by 23% versus h was reduced by 38% versus (10mg max/4h), and/or IV (bupivacaine/meloxicam) or bupivacaine placebo. placebo and 25% versus bupi- morphine as needed(10mg HCl 0.25%, 75 mg or saline placebo. A significant reduction of 21% for vacaine HCl. max/2h), and/or oral aceta- pain intensity over 72 h of HTX-011 minophen (1000mg max/6h). compared to bupivacaine. INL-001 Velanovich V et al. Comparison of INL-001 (implant Patients who received INL-001 Less opioid analgesic usage in Postoperative parenteral 2019 (19) designed to provide extended delivery reported lower pain intensity through the INL-001 group compared morphine as needed. of bupivacaine to the area around the 72 h. throughout the first 24 h in both surgical wound) with placebo implant. studies Ilio-inguinal and ilio-hypogastric block izcaino-Martinez L General anesthesia with ilioinguinal nerve Only significant 2 h after ad- Not mentioned. Paracetamol (1 g/8 h), meta- et al. 2014 (24) block versus spinal anesthesia alone mission and before discharge. mizol (2 g/8 h), dexketoprofen Analgesic drug significantly bigger (50 mg/8 h) and fentanyl in spinal anesthesia group. (bolus of 25 μg/15 min i.v.) until VAS<3. TAP-block Wegner R et al. 2017 Dexamethasone 8 mg added to ropivacaine The dexamethasone had a greater Not mentioned. Not mentioned. (29) TAP-block. change in pain score though no statistical significant difference was noted. Heil JW et al. 2014 Continuous TAP-block (catheter 2 days Not significant Not significant. Intravenous fentanyl was (32) postoperative) with administered at PACU at Ropivacaine 0.5% 20ml bolus initially need. versus saline injection. Anesthetic technique Bhattacharya P et al. Unilateral lumbar PVB versus con- No statistical significant dif- Only the time to first post- VAS scores of 4 or more 2010 (33) ventional spinal anesthesia ference. operative analgesic require-ment were treated with rescue was longer in the PVB group. analgesic tramadol in boluses of 50mg i.v.

07-Coppens.indd 50 12/01/2021 12:13 pain management after open inguinal hernia repair 51

Surgical technique Magnusson J et al. UltraPro Hernia System, versus Pro- NS. Not mentioned. Not mentioned. 2016 (34) lene Hernia System versus Lich-tenstein hernia repair. Kingsnorth A et al. Sutureless Parietex™ ProGrip™ mesh Pain score was lower in the Not mentioned. Not mentioned. 2012 (35) repair versus traditional Lichtenstein parietex™ group at discharge and at repair with lightweight polypropylene 7 days postoperative. mesh secured with sutures. Manyilirah W et al. Comparison of non-mesh (Desarda) and There was no significant difference Not mentioned. Not mentioned. 2012 (36) mesh (Lichtenstein) technique for open 1-2 h postoperative and not after 3 inguinal hernia repair. days. Gundre NP et al. Polyethylene vs polypropylene mesh The mean average post-operative Not mentioned. Not mentioned. 2012 (37) pain was similar between groups Torcivia A et al. 2011 Glucamesh® vs Polypropylene® On the evening of the intervention Not mentioned Not mentioned. (38) the pain was less intense in the Glucamesh® group. Canonico S et al. Lightweight, large-pore mesh vs heavy Patients with lightweight, large- Not mentioned. Not mentioned. 2013 (39) weight mesh with small pores pore mesh reported less pain postoperative. Fan JKM et al. 2017 Self gripping semi reabsorbable mesh No significant pain differences at Not mentioned Not mentioned. (40) (Progrip) vs polypropylene mesh. any time point. Carro JLP et al. 2017 Low density vs high density Mesh Only on the 1st postoperative day Not mentioned. Not mentioned. (41) and the 7th postoperative day). Bracale V et al. 2014 Fibrin glue vs Lichtenstein fixation of Postoperative pain was lower in the Patients requiring analgesics Not mentioned. (42) Mesh fibrin glue group at 1 week and 1 in the fibrin glue group were month. significantly less. Shen YM et al. 2012 NBCA (n-butyl-2-cyanoacrylate) medical 24 h after surgery VAS was Not mentioned. Not mentioned. (43) adhesive vs traditional suture significant lower in the NBCA- group when compared with the control group. Bona S et al. 2018 Lightweight mesh (Ultrapro®) vs A statistically significant difference NS. NS. (44) heavyweight mesh (Prolene®) in patients in favour of lightweight mesh for with primary inguinal hernia repair pain at 1 week and 6 months after (Lichtenstein technique) surgery Hoyuela C et al. Glue (Histoacryl®) or non-absorbable Postoperative pain at 8 h, 24 h, Not mentioned. All patients received para- 2016 (45) polypropylene sutures for fixation of 7 days and 30 days was less when cetamol 1000 mg eyery 8h lightweight polypropylene mesh. glue was used instead of sutures for and dexketoprofen 25 mg all measures. every 8h for 7 days. Post-operative Systemic non-opioid analgesics Mentes O et al. 2009 Lornoxicam 8 mg at the end of surgery and 12 No statistical significant Patients were not receiving Study protocol. (17) h after versus tramadol 1 mg/kg at the end of difference in the effect of pain rescue analgesia protocol. surgery and every 6 h up to 24 h postoperative. 12 h and 24 h after surgery. NS = no significant difference between groups, POD = Postoperative Day.

12 hours after surgery after which the difference use of HTX-011 (a novel, extended release, fixed- became insignificant. Another RCT (n=162) dose combination local anesthetic comprising evaluated the effects of intra-operative dipyrone 15 bupivacaine and low-dose meloxicam, incorporated mg/kg or 40 mg/kg IV during the operation (16). in a proprietary Biochronomer® polymer) (18). It A statistically significant difference was only found showed a 23% reduction in mean pain intensity over during the first 30 minutes in the PACU. One RCT 72 hours compared to placebo. At all timepoints (n=160) reported no difference in pain scores 12 through 72 hours, the mean pain scores were lower and 24 hours after surgery, between lornoxicam (an in the HTX-011 group. A reduction of 21% for pain NSAID administered at the end of the operation intensity over 72 hours was observed when HTX- and 12 hours post-operatively) and tramadol (17). 011 was compared to plain bupivacaine. Of note, However, the tramadol group experienced more basic analgesics – paracetamol and/or NSAIDs nausea. were not used. One study, which consisted of two RCTs Local/Regional Analgesia Techniques (n=417), compared the effect of INL-001 (an implant designed to provide extended delivery of Topical and surgical site infiltration with extended- bupivacaine to the area around the surgical wound) release local anesthetics with placebo implant (19). Patients who received INL-001 reported lower pain intensity through 72 h One placebo-controlled RCT (n=418) for the two pooled studies. One placebo-controlled showed a significant reduction in pain with the RCT (n=60) in patients receiving spinal anesthesia

07-Coppens.indd 51 12/01/2021 12:13 52 s. coppens et al. examined the effects of wound infiltration with US-guided TAP block with ropivacaine vs an US- bupivacaine 0.05% 10 mL prior to incision (20). guided TAP block with saline and concluded that Early postoperative pain was significantly lower at pain was reduced with a ropivacaine TAP block all timepoints in favour of the bupivacaine group. (28). Statistical significant differences were noted at rest, at 6 and 24 hours and with movement at 3, 6 Ilio-inguinal and ilio-hypogastric block and 24 hours, in favour of TAP block (28). A placebo-controlled RCT (n=82) found no A placebo-controlled RCT (n=34) investigated significant difference between the adjuvant effect the possible benefits of an additional ilio-inguinal, of dexamethasone to ropivacaine in TAP blocks ilio-hypogastric block (10mL 0.75% ropivacaine) (29). Another placebo-controlled RCT (n=44) in- and surgical wound infiltration (10mL 0.75% vestigated the effect of adding dexamethasone ropivacaine) (21). Pain scores were lower in the 8 mg to a lumbar epidural anesthesia block. The control group 3 hours post-surgery. There was no onset of epidural anesthesia, the primary outcome difference in pain scores at 6 and 12 hours post- of this study, was significantly more rapid in the surgery. Another placebo controlled RCT (n=60) dexamethasone group than in the control group. The found that an ilio-inguinal and ilio-hypogastric block duration of analgesia was markedly prolonged in (12mL 0.75% ropivacaine) and wound infiltration the dexamethasone group than in the control group prior to incision provided significantly lower (30). median VAS scores compared with placebo (22). A Another RCT (n=100) compared the effect of placebo-controlled RCT (n=60) also investigated the a post-operative subcostal TAP block to a standard effects of ultrasound-guided ilioinguinal and ilio- analgesic regimen. At different timepoints after hypogastric nerve blocks. A significant reduction in surgery (15 min, 1 h, 6 h, 12 h, 24 h ; 15 days and pain scores at mobilization and rest was recorded one month) there was a significantly difference in in the bupivacaine group vs placebo upon arrival pain scores between the two groups in favour of the in the post anesthesia care unit (PACU) and again TAP block group (31). after 30 minutes. However, opioid consumption was One randomized triple-masked, placebo- not significantly different between groups (23). Yet controlled study (n=20) investigated the effects of another RCT (n=32) investigated general anesthesia single injection TAP block vs a continuous TAP combined with ilio-inguinal nerve block vs spinal block through a perineural catheter for 2 days post- anesthesia alone. Except for the first 2 h after operative (32). There were no differences between a surgery, there was no difference in pain scores for single injection and continuous infusion at 6, 24, 48 the 24 h study period (24). and 72 hours after surgery. However, this study was underpowered. Transversus abdominis plane (TAP) blocks Anesthetic technique A RCT (n=273) compared ultrasound-guided TAP block (levobupivacaine 0.5%, 1.5mg/kg) One placebo-controlled RCT (n=58) study with blind ilio-hypogastric and ilio-inguinal block found no difference in pain scores at 2, 12 and 24 (levobupivacaine 0.5%, 1.5mg/kg) performed be- hours between spinal anesthesia and paravertebral fore surgery. Pain at rest was lower during the first block (PVB). Of note, these techniques had a 7% 24h after ultrasound-guided TAP block, but this failure rate (33). difference was not observed during movement (25). Okur et al. (n=90) found that pain scores with TAP Surgical technique block were significantly lower at 24 h compared with ilio-inguinal-ilio-hypogastric blocks in a post- One placebo-controlled RCT (n=309) com- hoc analysis (26). A prospective RCT (n=197) pared Lichtenstein technique with Prolene Hernia compared the effects of ultrasound-guided TAP System (PHS) and UltraPro Hernia System (UHS). block combined with Ilio-inguinal/Ilio-hypogastric There were no differences between groups regarding nerve blocks vs Ilio-inguinal/Ilio-hypogastric nerve perioperative course, operating time, postoperative blocks alone. Pain scores at 12, 24 and 48 hours pain or rehabilitation (34). Pain was assessed daily were significantly lower in the TAP block and ilio- until 14 days postoperatively by a research nurse inguinal/ilio-hypogastric block group compared to using a telephone interview. Need time points Ilio-inguinal/Ilio-hypogastric nerve block alone (27). for pain assessments. A placebo-controlled RCT Theodoraki et al. compared the effects of an (n=302) used Sutureless Parietex™ ProGrip™

07-Coppens.indd 52 12/01/2021 12:13 pain management after open inguinal hernia repair 53 mesh repair vs traditional Lichtenstein repair with Table 1 lightweight polypropylene mesh secured with Overall recommendations for pain management in patients sutures. The pain scores were significantly lower undergoing open inguinal hernia repair in the sutureless mesh group at discharge and at 7 Pre-operative and intra-operative days, but not at 1 month after surgery (35). One RCT ● Paracetamol (Grade D) (n=101) evaluated the effects of non-mesh (Desarda) ● Cyclo-oxygenase-2 selective inhibitor (Grade D) vs mesh (Lichtenstein) methods for inguinal hernia ● Intravenous dexamethasone (Grade B) repair. No differences in terms of postoperative pain ● Local anesthetic infiltration and/or regional analgesia (ilio-inguinal/ilio-hypogastric nerve block or transversus abdominis plane between the two techniques were noted (36). There block) was no significant difference in pain scores at 1-2 Postoperative hours postoperative and after 3 days. ● Paracetamol (Grade D) Another placebo-controlled RCT (n=70) found ● Cyclo-oxygenase-2 selective inhibitor or non-steroidal anti- no significant differences in early postoperative inflammatory drugs (Grade D) pain, infection, seroma, recurrence and other ● Opioid for rescue (Grade D) complications between different types of mesh Anesthetic technique ● Field block (e.g., ilio-inguinal/ilio-hypogastric block) with or (37). However, this study might be underpowered. without wound infiltration as a sole anesthetic or as an adjunct to Another placebo-controlled RCT (n=50) comparing general anesthesia (Grade A) Glucamesh® vs Polypropylene® mesh found that Surgical technique the pain scores (measured twice daily for 7 post- ● Mesh techniques in preference to non-mesh techniques (Grade A) operative days) were significantly lower before discharge in the glucamesh group (38). this systematic review were determined to be of A placebo-controlled study (n=80) evaluated high quality and all but three showed statistically the effect of sutureless fixation with fibrin glue of significant results. The updated literature strengthens lightweight mesh in open inguinal hernia repair. the previous PROSPECT recommendations for pain Patients in the lightweight group reported less management in patients undergoing open IHR (5). pain during the first month after surgery (39). The updated PROSPECT methodology further Another placebo-controlled RCT (n=45) found no strengthens the recommendations, because it goes differences in pain scores between self-gripping beyond assessment of the available evidence based mesh vs polypropylene mesh (40). In a comparison solely on the simple statistical analysis (7). between low density vs high density mesh, there Combination of basic analgesics such as para- were differences in pain scores only on the 1st and cetamol and NSAID or COX-2 selective inhibitor the 7th postoperative days in favour of low density along with a local/regional analgesic techniques mesh (41). One randomized controlled trial (n=102) (e.g., local infiltration with or without a field blocks compared the effect of fibrin glue vs Lichtenstein such as ilio-inguinal/ilio-hypogastric nerve block mesh repair (42). The fibrin group had lower or interfascial plane blocks such as TAP block) is pain scores and opioid requirements at 1 week. recommended pre-operatively and/or intra-ope- A randomized controlled trial (n=110) compared ratively. Paracetamol and NSAID have been shown fibrin glue as medical adhesive NBCA (n-butyl- to provide excellent analgesia and reduce opioid 2-cyanoacrylate) vs suture for patch fixation in requirements, and should be continued in the Lichtenstein inguinal herniorrhaphy (43). The pain postoperative period, with opioids used as rescue scores in the fibrin group were lower at 24 hours after analgesics (Table 1). Although wound infiltration surgery. One multicenter RCT (n=808) compared is recommended, the role of extended release the effects of lightweight mesh (Ultrapro®) vs local anesthetics, topical NSAIDs, clonidine, heavyweight mesh (Prolene®), and reported lower dexamethasone, or opioids remain controversial. postoperative pain scores in favour of lightweight With regards to the choice of anesthetic mesh (44). One multicenter RCT (n=370) compared technique, the current systematic review strengthens the effects of glue (Histoacryl®) or non-absorbable our previous recommendation of using a field block polypropylene sutures for fixation of lightweight (e.g., ilio-inguinal/ilio-hypogastric block) with or polypropylene mesh (45). Postoperative pain was without wound infiltration as a sole anesthetic or less in the group where glue was used. as an adjunct to general anesthesia. In fact, even the recent International guidelines recommend Discussion and conclusion the use of local anesthesia provided the surgeon is experienced in this technique (1). Obviously, patient The majority of the studies included in selection and acceptance to a sole regional/local

07-Coppens.indd 53 12/01/2021 12:13 54 s. coppens et al. Table 2 Analgesic interventions that are not recommended for pain management in patients undergoing open inguinal hernia repair

Intervention Reason for not recommending

Gabapentinoids Conflicting procedure-specific evidence

Clonidine Insufficient procedure-specific evidence Ketamine Insufficient procedure-specific evidence Pre-operative Etanercept Insufficient procedure-specific evidence Although effective these techniques are relatively invasive Epidural analgesia with potential for complications Although effective these techniques are relatively invasive Paravertebral blocks with potential for complications Psoas block Insufficient procedure-specific evidence Wound infiltration with extended release bupivacaine Insufficient procedure-specific evidence Wound infiltration using conventional non-steroidal anti- Insufficient procedure-specific evidence Intra-operative inflammatory drugs , clonidine or opioids Topical non-steroidal anti-inflammatory drugs Insufficient procedure-specific evidence Intravenous lidocaine infusion Insufficient procedure-specific evidence Topical extended-release local anesthetic Limited procedure-specific evidence Transcutaneous Electrical Nerve Stimulation Limited procedure-specific evidence Postoperative Compressive cryotherapy or ice wrapping Lack of procedure-specific evidence Zolpidem as a sleep aid Limited procedure-specific evidence Nerve section Limited of procedure-specific evidence Surgical technique Cryoanalgesia techniques Insufficient procedure-specific evidence TENS Inconclusive procedure-specific evidence

anesthetic is imperative. A recent meta-analysis of data, or lack of evidence (Table 2). These include RCTs comparing local anesthesia vs other forms of analgesic adjuncts such as gabapentinoids, clonidine, anesthesia (including general anesthesia) concluded corticosteroids, ketamine, intravenous lidocaine that local anesthesia allows shorter operating infusion, and etarnacept. Neuraxial analgesic room times and is associated with lower incidence techniques (epidural or paravertebral analgesia) are of urinary retention (compared with neuraxial not recommended despite the fact that they provide anesthesia) (46). Also, patient satisfaction with local excellent pain relief because they are invasive and anesthesia was similar to that with other anesthetic have potential complications. Psoas-block is not techniques (46). Although in comparison with recommended at the moment due to insufficient general anesthesia, neuraxial block is associated with evidence. However, newer interfascial plane blocks reduced pain scores and decreased post-operative such as erector spinae blocks may provide pain analgesic requirement, their use is associated with relief and negate the concerns of neuraxial blocks side effects (e.g., postural hypotension) that might (47). However, evidence currently is lacking for delay recovery and discharge home in a day care open inguinal hernia repair. setting. The limitations in this review are related to It is recommended that the surgical technique those of the included studies. There was considerable should be based on surgeon’s expertise, hernia- heterogeneity between studies with regards to related characteristics, and availability of local dosing regimens and route of administration as well resources (1). Nevertheless, as far as postoperative as timing of pain assessments. The small size of pain is concerned, a mesh technique is in preference most studies has the potential for estimation effect to non-mesh techniques. This recommendation is and do not provide safety profile of the analgesic also validated by the International guidelines for interventions. In a majority of the studies the groin hernia management (1). analgesic intervention was not evaluated against an Several analgesic interventions are not recom- optimized multimodal analgesic regimen. mended due to either conflicting data, insufficient In summary, this review has identified an

07-Coppens.indd 54 12/01/2021 12:13 pain management after open inguinal hernia repair 55 analgesic regimen that for optimal pain management randomized, comparative trial of a COX-2 selective after open inguinal hernia repair as well as nonsteroidal anti-inflammatory drug versus placebo in inguinal herniorrhaphy patients. Hernia. 13: 491-497. interventions that are not recommended for pain 14. Somri M., Hawash N., Hadjitoffi C., Ghantous-Toukan M., management in patients undergoing IHR. Future tome R., Yodashkin M. et al. 2017. Protective multimodal adequately powered studies should assess the effects analgesia with etoricoxib and spinal anesthesia in inguinal hernia repair: a randomized controlled trial. J Anaesth. 31: of analgesic interventions not only on pain, opioid 645–650. consumption, opioid-related adverse events and 15. Kang H. and Kim BG. 2011. Intravenous lidocaine complications associated with the intervention, but for effective pain relief after inguinal herniorrhaphy: a prospective, randomized, double-blind, placebo-controlled also outcome measures such as time to ambulation, study. J Inter Med Res. 39: 435-445. length of hospital stay, and the occurrence of chronic 16. Chaparro LE., Lezcano W., Alvarez HD. And Joaquin W. pain or chronic opioid consumption. Furthermore, 2012. Analgesic Effectiveness of Dipyrone (Metamizol) the influence of analgesic intervention on patient- for Postoperative Pain after Herniorrhaphy: A Randomized, Double-Blind, Dose-Response Study. Pain Pract. 12: 142- specific factors such as chronic pain and chronic 147. opioid therapy need to be assessed. 17. Mentes O. and Bagci M. 2009. Postoperative pain management after inguinal hernia repair: Lornoxicam versus tramadol. Hernia. 13: 427-430. References 18. Viscusi E., Minkowitz H., Winkle P., Ramamoorthy S., Hu J. and Singla N. 2019. HTX-011 reduced pain intensity 1. Simons MP., Smietanski M., Bonjer HJ., Bittner R., Miserez and opioid consumption versus bupivacaine HCl in M. and Aufenacke TJ. et al. 2018. International guidelines herniorrhaphy: results from the phase 3 EPOCH 2 study. for groin hernia management. Hernia. 22: 1-165. Hernia. 23: 1071-1080. 2. Wheeler DW., Bhatia A., Mani V., Kinna S., Bell A. and 19. Velanovich V., Rider P., Deck K., Minkowitz S., Leiman D. Boyle Y. et al. 2019. Evaluation of postsurgical hyperalgesia and Jones N. et al. 2019. Safety and Efficacy of Bupivacaine and sensitization after open inguinal hernia repair: a useful HCl Collagen-Matrix Implant (INL-001) in Open Inguinal model for neuropathic pain? J Pain. 20: 1199-1208. Hernia Repair: Results from Two Randomized Controlled 3. Joshi GP., Schug SA. and Kehlet H. 2014. Procedure- Trials. Adv Ther. 36: 200-216. specific pain management and outcome strategies. Best 20. Nesioonpour Sh., Akhondzadeh R., Pipelzadeh MR., Pract Res Clin Anesth. 28: 191-201. Rezaee S., Nazaree E. and Soleymani M. 2013. The effect 4. Lee B., Schug SA., Joshi GP. and Kehlet H. 2018. of preemptive analgesia with bupivacaine on postoperative Procedure-Specific Pain Management (PROSPECT) - An pain of inguinal hernia repair under spinal anesthesia: a update. Best Pract Res Clin Anesth. 32: 101-111. randomized clinical trial. Hernia. 17: 465-470. 5. Joshi GP., Rawal N., Kehlet H., Bonnet F., Camu F. and 21. Santos GC., Braga GM., Queiroz FL., Navarro TP. And Fischer HBJ. et al. 2012. Evidence-based management Gomez RS. 2011. Assessment of postoperative pain and of postoperative pain in adults undergoing open inguinal hospital discharge after inguinal and iliohypogastric nerve hernia surgery. Br J Surg. 99: 168-185. block for inguinal hernia repair under spinal anesthesia: a 6. Joshi GP. and Kehlet H. 2017. Guidelines for perioperative prospective study. J Braz Med Assoc. 57: 545-549. pain management: need for re-evaluation. Br J Anaesth. 22. Saeed M., Andrabi WI., Rabbani S., Zahur S., Kahmood 119: 703-706. K. and Andrabi SI. et al. 2012. The impact of preemptive 7. Joshi GP., Van de Velde M. and Kehlet H. 2019. ropivacaine in inguinal hernioplasty a randomized Development of evidence-based recommendations controlled trial. Int J Surg. 13: 76-79. for procedure-specific pain management: PROcedure- 23. Bærentzen F., Maschmann C., Jensen K., Belhaege B., SPECific Pain ManagemenT (PROSPECT) methodology. Hensler M. and Børglum J. et al. 2012. Ultrasound-guided Anesthesia. 74: 1298-1304. nerve block for inguinal hernia repair: a randomized, 8. Myles PS., Myles DB., Galagher W., Boyd D., Chew C. and controlled, double-blind study. Reg Anesth Pain Med. 37: MacDonald N. et al. 2017. Measuring acute postoperative 502-507. pain using the visual analog scale: the minimal clinically 24. Vizcaíno-Martínez L., Gómez-Ríos MÁ. and López- important difference and patient acceptable symptom state. Calviño B. 2014. General anesthesia plus ilioinguinal nerve Br J Anesth. 118: 424-429. block versus spinal anesthesia for ambulatory inguinal 9. Mahoori A., Noroozinia H., Hasani E. and Hosainzadeh herniorrhapy. Saudi J Anesth. 8: 523-528. S. 2014. The effect of pre-operative administration of 25. Aveline C., Le Hetet H., Le Roux A., Vautier P., Cognet F. gabapentin on post-operative pain relief after herniorrhaphy. and Tison C. et al. 2011. Comparison between ultrasound- Saudi J Anesth. 8: 220-223. guided transversus abdominis plane and conventional 10. Sen H., Sizlan A., Yanarates O., Senol MG., Inangil G. ilioinguinal/iliohypogastric nerve blocks for day-case open and Sücüllü I. et al. 2009. The effects of gabapentin on inguinal hernia repair. Br J Anaesth. 106: 380-386. acute and chronic pain after inguinal herniorrhaphy. Eur J 26. Okur O., Tekgul ZT. and Erkan N. 2017. Comparison Anaesth. 26: 772-776. of eficacy of transversus abdominis plane block and 11. Singla NK., Chelly JE., Lionberger DR., Gimbel J., Sanin iliohypogastric/ilioinguinal nerve block for postoperative L., Sporn J. et al. 2014. Pregabalin for the treatment of pain management in patients undergoing inguinal her- postoperative pain: Results from three controlled trials niorrhaphy with spinal anesthesia: a prospective randomized using different surgical models. J Pain Res. 8: 9-20. controlled open label study. J Anaesth. 31: 678-685. 12. Cohen SP., Galvagno SM., Plunkett A., Harris D., Kurihara 27. Hosalli V., Ayyanagouda B., Hiremath P., Ambi U. and C., Ali T. et al. 2013. A multicenter, randomized, controlled Hulkund SY. 2019. Comparative efficacy of postoperative study evaluating preventive etanercept on postoperative analgesia between ultrasound-guided dual transversus pain after inguinal hernia repair. Anesth Analg. 116: 455- abdominis plane and Ilioinguinal/Iliohypogastric nerve 462. blocks for open inguinal hernia repair: An open label 13. Schurr MJ. and Faucher LD. 2009. A prospective, prospective randomised comparative clinical trial. Ind J

07-Coppens.indd 55 12/01/2021 12:13 56 s. coppens et al. Anaesth. 63: 450-455. inguinal hernia meshplasty as a safe and cost-effective 28. Theodoraki K., Papacharalampous P., Tsaroucha A., alternative to polypropylene mesh. Updates Surg. 64: 37- Verzakis A. and Arygyra E. 2019. The effect of transversus 42. abdominis plane block on acute and chronic pain after 38. Torcivia A., Vons C., Barrat C., Dufour F. and Champault inguinal hernia repair. A randomized controlled trial. Int J G. 2011. Influence of mesh type on the quality of early Surg. 63: 63-70. outcomes after inguinal hernia repair in ambulatory 29. Wegner R., Akwar D., Guzman-Reyes S., Pednekar G., setting controlled study: Glucamesh® vs Polypropylene®. Caudhry R. and Grewal N. et al. 2017. Evaluating the Langenbeck’s archives of surgery. 396: 173-178. Adjuvant Effect of Dexamethasone to Ropivacaine in 39. Canonico S., Benevento R., Perna G., Guerniero R., Transversus Abdominis Plane Block for Inguinal Hernia Sciaudone G. and Pellino G. et al. 2013. Sutureless fixation Repair and Spermatocelectomy: A Randomized Controlled with fibrin glue of lightweight mesh in open inguinal Trial. Pain Phys. 20: 413-418. hernia repair: Effect on postoperative pain: A double- 30. Razavizadeh MR., Fazel MR., Heydarian N. and Atoof F. blind, randomized trial versus standard heavyweight mesh. 2017. Epidural Dexamethasone for Postoperative Analgesia Surgery. 153: 126-130. in Patients Undergoing Unilateral Inguinal Herniorrhaphy: 40. Fan JKM., Yip J., Foo DCC., Lo OSH. and Law WL. A Comparative Study. Pain Res Manag. Epub 28 Feb 2017. 2017. Randomized trial comparing self gripping semi re- 31. Akyol Beyoğlu Ç., Özdilek A., Erbabacan E., Özmumcu absorbable mesh (PROGRIP) with polypropylene mesh in EA, Ekici B. and Köksal G. et al. 2018. Evaluation of the open inguinal hernioplasty: the 6 years result. Hernia . 21: effects of subcostal transversus abdominis plane block on 9-16. acute and subacute pain development following inguinal 41. Carro JLP., Riu SV., Lojo BR., Latorre L., Teresa M. and herniography: Randomized clinical study. The journal of Garcia A. et al. 2017. Randomized Clinical Trial Comparing the Turkish Society of Algology. 30: 123-129. Low Density versus High Density Meshes in Patients with 32. Heil JW., Nakanote KA., Madison SJ., Loland VJ., Mariano Bilateral Inguinal Hernia. Am Surg. 83: 1352-1356. ER. and Sandhu NS. et al. 2014. Continuous transversus 42. Bracale U., Rovani M., Picardo A., Merola G., Pignata G. abdominis plane (TAP) blocks for postoperative pain and Sodo M. et al. 2014. Beneficial effects of fibrin glue control after hernia surgery: a randomized, triple masked, (Quixil) versus Lichtenstein conventional technique in placebo-controlled. Pain Med. 15: 1957-1964. inguinal hernia repair: a randomized clinical trial. Hernia. 33. Bhattacharya P., Mandal MC., Mukhopadhayay S., Das S., 18: 185-192. PAl PP. and Basu SR. 2010 Unilateral paravertebral block: 43. Shen YM., Sun WB., Chen J., Liu S. and Wang M. 2012 an alternative to conventional spinal anesthesia for inguinal NBCA medical adhesive (n-butyl-2-cyanoacrylate) hernia repair. Acta Anesth Scand. 54: 246-251. versus suture for patch fixation in Lichtenstein inguinal 34. Magnusson J., Nygren J., Gustafsson UO., and Thorell herniorrhaphy: A randomized controlled trial. Surgery. 151: A. 2016. UltraPro Hernia System, Prolene Hernia System 550-555. and Lichtenstein for primary inguinal hernia repair: 3-year 44. Bona S., Rosati R., Opocher E., Fiore B. and Montorsi M. outcomes of a prospective randomized controlled trial. 2018 Pain and quality of life after inguinal hernia surgery: Hernia. 20: 641-648. a multicenter randomized controlled trial comparing 35. Kingsnorth A., Gingell-Littlejohn M., Nienhuijs S., Schüle lightweight vs heavyweight mesh (Supermesh Study). S., Appel P. and Ziprin P. et al. 2012. Randomized controlled Updates Surg. 70: 77-83. multicenter international clinical trial of self-gripping 45. Hoyuela C., Juvany M., Carvajal F., Veres A., Troyano D. ParietexTM ProGripTM polyester mesh versus lightweight and Trias M. et al. 2017. Randomized clinical trial of mesh polypropylene mesh in open inguinal hernia repair: interim fixation with glue or sutures for Lichtenstein hernia repair. results at 3 months. Hernia. 16: 287-294. Br J Surg. 104: 688-694. 36. Manyilirah W., Kijjambu S., Upoki A. and Kiryabwire 46. Argo M., Favela J., Phung T. and Huerta S. 2019. Local VS. J. 2012. Comparison of non-mesh (Desarda) and mesh other forms of anesthesia for open inguinal hernia repair: A (Lichtenstein) methods for inguinal hernia repair among meta-analysis of randomized controlled trials. Am J Surg. black African patients: a short-term double-blind RCT. 218: 1008-1015. Hernia. 16: 133-144. 47. Abu Elyazed MM., Mostafa SF., Abdelghany MS. and Eid 37. Gundre NP., Iyer SP. and Subramaniyan P. 2012 Prospective GM. 2019. Ultrasound-guided erector spinae plane block randomized controlled study using polyethylene mesh for in patients undergoing open epigastric hernia repair: a prospective randomized controlled study. Anesth Analg. 129: 235-240.

07-Coppens.indd 56 12/01/2021 12:13 (Acta Anaesth. Belg., 2020, 71, 57-65) The effect of dexamethasone on post-tonsillectomy pain, nausea and vomiting in children : a review of the evidence

L. Beckers, V. Saldien, S. Maes

Abstract : Background : pediatric tonsillectomy ing, both of which can substantially affect child is associated with significant postoperative pain, health status and quality of life. When compared nausea and vomiting. Postoperative vomiting with healthy children, those with recurrent throat occurs more frequently after tonsillectomy than any infections have a greater perception of pain and other commonly performed pediatric operation. It have poorer general health and physical wellbeing. is also the commonest cause of morbidity and re- Similarly, sleep-disordered breathing is associated admission following tonsillectomy. Furthermore, with cognitive and behavioral im-pairment. Fur- inadequate pain management results in considerable thermore, tonsillectomy has been shown to have a increase in morbidity. clear benefit on quality of life (1). Purpose of the review : a number of authors have Despite advances in anesthetic and surgical explored the effects of systemic corticosteroids in techniques, post-tonsillectomy morbidity remains the reduction of morbidity after tonsillectomy. At a significant clinical problem not only for the the Antwerp University Hospital, dexamethasone is patient and the family, but also for medical care not currently used as part of the standard care for providers. The most important complications this procedure. However, our surgical colleagues are pain and PONV (postoperative nausea and have shown interest in this adjunctive treatment for vomiting). In order to address this issue, a number tonsillectomy. This literature review was therefore of authors have explored the effects of systemic conducted to examine the efficacy of dexamethasone corticosteroids in reducing postoperative morbidity, in the prophylaxis of postoperative pain, nausea and but no consensus has yet been reached about their vomiting, following tonsillectomy in the pediatric routine use. At the Antwerp University Hospital, population. DXM (dexamethasone) is not currently included in Design : a retrospective analysis of English-speaking routine standard care for elective tonsillectomy. Our literature from January 2008 until December 2019. surgical colleagues, however, have shown interest Conclusion : preoperative intravenous dexametha- in implementing DXM treatment, either alone or sone clearly decreases post-tonsillectomy morbidity, as part of a multimodal regimen, for prophylaxis of but as yet, no consensus has been reached regarding PONV and postoperative pain. the optimum recommended dose. In order to examine the efficacy of DXM for these indications, we conducted an electronic Keywords : Dexamethasone ; Pediatric tonsillec-tomy ; literature review focusing on English-speaking Pain ; Nausea ; Vomiting. literature published between 1st January 2008 and 31st December 2019.

Introduction

L. Beckers, M.D. ; V. Saldien, M.D. PhD ; S. Maes, M.D. Tonsillectomy is one of the most commonly Department of Anesthesiology, Antwerp University Hospital, performed surgical procedures in children. It is 2650 Edegem, Belgium defined as a surgical procedure (performed with or Corresponding author : Beckers L., M.D., assistant without adenoidectomy) that completely removes Anesthesiology, Department of Anesthesiology, Antwerp University Hospital, Wilrijkstraat 10, 2650 Edegem, the palatopharyngeal tonsillar tissue with its capsule. Belgium. This involves dissection of the peritonsillar space E-mail : [email protected] between the tonsil capsule and the muscular wall. Paper submitted on May 26, 2020 and accepted on May 26, Indications for surgery include recurrent 2020. pharyngeal infections and sleep-disordered breath- Conflict of interest : none.

08-Beckers.indd 57 12/01/2021 12:14 58 l. beckers et al. PONV following tonsillectomy 3. Patient factors

PONV occurs more frequently after The likelihood of PONV is age-related. The tonsillectomy than any other commonly performed incidence is low in infancy, but increases to a peak pediatric operation. It occurs independent of the in late childhood (between 6-16 years) and later surgical technique and has been reported to occur declines in adulthood. This pattern is observed in more than 70% of children who do not receive regardless of the type of operation performed. prophylactic antiemetics (1). Tonsillectomy is most frequently performed in the PONV is the most frequent reason for re- late childhood age group, which consistently shows admission following day-case ear, nose, throat a higher incidence of PONV. surgery (2). It leads to decreased oral intake, dehydration and metabolic disturbances and is distressing Pain following tonsillectomy for both the child as well as the care providers. Not only does it contribute to an overall lower level Tonsillectomy is one of the most painful of satisfaction with the procedure, it may place procedures performed in an outpatient setting the child at increased risk for bronchopulmonary (4). When compared with urology, orthopedics, aspiration and potential pneumonitis. There is and general surgery, otolaryngology procedures an increased chance of immediate postoperative have been shown consistently to be associated hemorrhage and there is a substantially prolonged with increased postoperative pain. In one study of time in the recovery room, with associated increases outpatient surgery, only three factors significantly in healthcare costs. correlated with high postoperative pain scores : The rate of vomiting following pediatric female gender, duration of anesthesia and oto- tonsillectomy can be influenced by several factors laryngology procedures (4). As tonsillectomy is one categorized into surgical, anesthetic and patient of the most commonly performed otolaryngology factors (2) : procedures, the literature concerning the assessment and management of post-tonsillectomy pain is 1. Surgical factors highly relevant. Severe pain following tonsillectomy results The vomiting reflex is controlled and coor- in delayed oral intake and an increased risk of dinated by the vomiting center in the medulla dehydration and weight loss. Inadequate analgesia oblongata. Afferent inputs are located in the is related to increased morbidity. In addition, posterior oropharynx and so surgery in this area pain has a major impact on the psychological and is frequently associated with a higher incidence of social aspects of a patient’s life, and should not be vomiting. Furthermore, blood from the operative underestimated (5). If pain is not treated effectively, site is easily ingested by the patient during and after it can cause avoidance behaviors with regard tonsillectomy. This blood causes gastrointestinal to future health care. Adequate assessment and irritation and increases the likelihood of post- management is indicated in this high-risk group. operative vomiting. Therefore, it is important to Currently, the optimal treatment regimen has not yet minimize the volume of blood ingested during been identified. surgery. This can be achieved by packing the throat Post-tonsillectomy pain is of multi-factorial before removing the tonsils (3), emptying the origin. There are two distinct pain mechanisms that stomach with an aspiration tube before extubation result from the stimulus of surgical trauma. Firstly, and strict attention for hemostasis. there is inflammatory pain due to the local effect of surgical trauma. Secondly, there is a physiological 2. Anesthetic factors or functional pain resulting from stimulation of The emetic effect of opioid analgesics is well- the central nervous system (6). Post-tonsillectomy known. They stimulate the chemoreceptor trigger pain probably results, therefore from muscle zone, which send afferents to the vomiting center. spasm caused by inflammation and irritation of When compared with total intravenous anesthesia the pharyngeal musculature (7). It is more intense using propofol, volatile anesthetic agents increase within the first three post-operative days, but can the risk of PONV. Although nitrous oxide is com- persist until day 10. Although the introduction monly used to accelerate gas induction and reduce of electrocautery during surgery has almost the dose of volatile anesthetic required, it is asso- eliminated immediate postoperative hemorrhage, ciated with a 28% increased risk of vomiting (2). it may cause increased pain and discomfort due to

08-Beckers.indd 58 12/01/2021 12:14 pain and ponv in pediatric tonsillectomy 59 local inflammation, nerve irritation and laryngeal thereby inhibiting synthesis of potent mediators spasm (6). Hanasono et al studied the effect of of inflammation such as prostaglandins and DXM in patients undergoing tonsillectomy by leukotrienes (5, 6). Decreased prostaglandin levels hot (electrocautery) and cold (sharp) dissection might further contribute to analgesia by inhibiting techniques and found the combination of intravenous the synthesis of COX-2 (cyclo-oxygenase isoform steroid with cold dissection to be the more effective 2) in peripheral tissues and in the CNS (central in reducing postoperative pain (8). nervous system) (6). Glucocorticoids also inhibit leucocyte infiltration at the site of inflammation Mechanisms of action of intravenous dexamethasone and inhibit mediators of the inflammatory response such as TNF-α (tumor necrosis factor-α), IL-ß

Dexamethasone (C22H29FO5) has a molecular (interleukin-1ß), and IL-6 (interleukin-6) (6). weight of 392.5 and functions as a glucocorticoid that The degree of clinical effect usually relates induces negative inhibition of cortisol secretion in to the dose administered. Likewise, the numerous individuals with a functional hypothalamopituitary adverse effects related to corticosteroid use usually axis. In vivo, these steroid molecules exhibit a depend on both the dose and the duration of therapy multitude of effects not only on carbohydrate, fat and (7). protein catabolism, but also immunosuppression, DXM has an onset of action of 4-8 hours. The modulation of vascular tone and anti-inflammatory biological half-life is prolonged at about 36-72 properties (4). Glucocorticoids are also associated hours which explains the continued postoperative with multiple side effects including peptic ulce- beneficial effects lasting about three days (5, 11). ration, osteoporosis, hypertension, diabetes mellitus and fluid retention (7). Methodology The most serious potential risk with glucocorticoids is immune system suppression. Although We performed an electronic search of the the immunosuppressive and anti-inflammatory English-speaking literature in PubMed from 1st effects of DXM are approximately 30 times greater January 2008 until 31st December 2019. The following than with cortisol, the incidence of harmful effects search terms were used : ‘dexamethasone’, ‘pediatric occurs only rarely. Moreover, a single dose of DXM, tonsillectomy’, ‘tonsillectomy’, ‘children’, ‘pain’, even in high doses, is considered to be safe (9). ‘nausea’ and ‘vomiting’. The search strategy was as follows : ‘dexamethasone’ AND ‘pediatric Antiemetic effect of dexamethasone tonsillectomy’ AND ‘pain’ AND ‘nausea’ AND ‘vomiting’ ; ‘dexamethasone’ AND ‘tonsillectomy’ DXM has antiemetic properties in the surgical AND ‘children’ AND ‘pain’ AND ‘nausea’ AND setting (10), but the mechanism of its antiemetic ‘vomiting’ ; ‘dexamethasone’ AND ‘tonsillectomy’ action is not well understood. A commonly held AND ‘children’ AND ‘pain’ ; ‘dexamethasone’ AND theory is that corticosteroids exert their antiemetic ‘tonsillectomy’ AND ‘children’ AND ‘nausea’ AND activity via prostaglandin antagonism. Others have ‘vomiting’. We screened the titles and abstracts of suggested that the release of endorphins, resulting in all identified articles to determine their eligibility. mood elevation, a sense of well-being, and appetite Duplicate articles were manually deleted and we stimulation, plays an important role (10). checked the reference lists of the selected articles in order to identify further relevant studies. No meta- Analgesic effect of dexamethasone analysis was performed. The mechanism by which DXM exerts an analgesic effect is not fully understood, but Results believed to be related to its anti-inflammatory properties. Indeed, as mentioned previously, The results of the search strategy are DXM has strong anti-inflammatory effects, which summarized in Table 1. decreases postoperative tissue injury, edema and The study includes papers retrieved by pain. Glucocorticoids bind to specific receptors electronic searches from January 2008 until in the cytosol of cells and suppress tissue levels December 2019. In total, 212 papers were identified, of bradykinin and the release of neuropeptides of which 15 were retrieved for full-text review and from the nerve endings. This results in diminished were discussed in the final analysis. nociception in inflamed tissue (5). In addition, The characteristics with main findings of the DXM is able to inhibit the arachidonic acid cascade, included studies are summarized in Table 2.

08-Beckers.indd 59 12/01/2021 12:14 60 l. beckers et al. Table 1 The effect of dexamethasone on PONV Results of PubMed search strategy Publication breakdown Number of articles A RCT (randomized controlled trial) Prospective randomized controlled trials 5 conducted among 215 children undergoing elective Prospective audits 1 tonsillectomy pointed out that there was a beneficial dose-dependent effect of DXM on PONV (10). Reviews and meta-analyses 5 This conclusion was confirmed by the study Retrospective studies 2 of Karaman et al. They looked at the incidence Quasi-experimental studies 1 of PONV within the first 24 hours after (adeno) Guidelines 1 tonsillectomy in three groups receiving respectively Total 15 0.2 mg kg-1 intraoperative DXM (group 1), 0.7 mg

Table 2 Characteristics of the included studies Czarnetzki 2008

Methods Randomized, placebo-controlled

Participants [age (yr)] 215 patients (2-17) undergoing tonsillectomy

Surgical technique One of 3 surgical techniques: - A cold technique: dissection with cold steel instruments, hemostasis with gauze compression - A hot technique: dissection and hemostasis with electric bipolar forceps - A combination of both Interventions (DXM mg kg-1) DXM 0.05 IV or DXM 0.15 IV or DXM 0.5 IV (max dose 20 mg) or placebo

Outcomes Incidence of PONV at 24 hours The need for ibuprofen at 24 hours Evaluation of adverse effects Main findings Beneficial dose-dependent effect of DXM on PONV DXM - decreased the use of NSAID’s at 24 hours, dose-independent - no effect on the need of morphine in the PACU - no decrease in the proportion of children with inadequate pain relief at 24 hours - a dose-dependent increased risk of postoperative bleeding Bennett AM 2008

Methods Prospective audit

Participants (age (yr)) 107 patients (< 16)

Interventions (DXM mg kg-1) DXM 0.1

Outcomes Incidence of PONV

Main findings A statistically significant reduction in vomiting (p < 0.05)

Karaman 2009

Methods Randomized, single-blind, placebo-controlled

Participants (age (yr)) 150 patients (3–11) undergoing (adeno)tonsillectomy

Interventions (DXM mg kg-1) DXM 0.2 IV (group 1) or DXM 0.7 IV (group 2) (max dose 25 mg) or placebo

Outcomes Incidence of PONV within first 24 hours Incidence of patient’s tolerability to take oral semisolid/solid foods within first 24 hours Main findings Intravenous DXM 0.7 effectively decreases postoperative morbidity, in the absence of measurable side effects

Steward 2011

Methods Meta-analysis: randomized, double-blind, placebo-controlled trials

Participants (age (yr)) 1756 patients (1–18) undergoing (adeno)tonsillectomy

Interventions (DXM mg kg-1) DXM 0.15 – 1 IV

Outcomes Incidence of PONV within first 24 hours The number of children returning to a soft/solid diet by postoperative day 1 Incidence of significant pain at 24 hours

08-Beckers.indd 60 12/01/2021 12:14 pain and ponv in pediatric tonsillectomy 61

Main findings Decreased incidence of PONV at 24 hours Increased tolerability to a soft/solid diet on postoperative day 1 Improved postoperative pain Baugh 2011

Methods Clinical practice guideline: systematic reviews, guidelines, meta-analyses, randomized controlled trials

Participants (age (yr)) Patients (1-18) under consideration for tonsillectomy

Interventions Evidence based recommendations on the preoperative, intraoperative and postoperative care and management

Outcomes Primary purpose: to provide clinicians with evidence-based guidance in identifying children who are the best candidates for tonsillectomy Secondary objectives: to optimize the perioperative management of children undergoing tonsillectomy Main findings Strong recommendation that clinicians should administer a single, intraoperative dose of intravenous DXM to children undergoing tonsillectomy Windfuhr 2011

Methods Retrospective cohort study

Participants (age (yr)) 272 patients (2-15) undergoing tonsillectomy

Surgical technique A cold technique + bipolar cautery when necessary

Interventions (DXM mg kg-1) DXM 4 mg IV or DXM 8 mg IV (< 0.15 or > 0.15) or placebo

Outcomes Incidence of PONV

Incidence of post-operative bleeding Main findings DXM does not appear to have an effect on PONV or bleeding following tonsillectomy

Hermans 2012

Methods Randomized, double-blind, placebo-controlled

Participants (age (yr)) 134 patients (2-8) undergoing (adeno)tonsillectomy

Surgical technique A dissection technique combined with electrocoagulation

Interventions (DXM mg kg-1) DXM 0.15 IV (max dose 8 mg) or DXM 0.5 IV (max dose 15 mg) or placebo

Outcomes Incidence 0for additional analgesia

Main findings Decreased incidence of PONV The use of rescue anti-emetics was lower in the DXM groups No dose-dependent effect of DXM on PONV Reduced pain only on postoperative day 2 No effect on the need of morphine in the PACU or tramadol on the ward Hashmi 2012

Methods Randomized, double-blind, placebo- controlled

Participants (age (yr)) 100 patients (4–12) undergoing (adeno)tonsillectomy

Surgical technique Electrocautery

Interventions (DXM mg kg-1) DXM 0.5 IV (max dose 8 mg) or placebo

Outcomes Incidence of PONV Mean time of first oral intake Incidence of severe pain Main findings Decreased incidence of PONV Mean time to first oral intake: earlier in the DXM group Lower pain scores + less painful swallowing in patients receiving DXM Buland 2012

Methods Quasi-experimental study

Participants (age (yr)) 60 patients (10–40) undergoing tonsillectomy

Interventions (DXM mg kg-1) DXM 0.15 IV or placebo

Outcomes Postoperative pain scores

Main findings Lower pain scores at 6 and 12 hours postoperatively

08-Beckers.indd 61 12/01/2021 12:14 62 l. beckers et al.

Keller 2013

Methods Overview of the literature

Participants

Interventions Intraoperative DXM IV

Outcomes Safety of intraoperative DXM

Risk of post-tonsillectomy hemorrhage Mahant 2015

Methods Retrospective cohort study

Participants (age (yr)) 111 813 patients (1–18) undergoing (adeno)tonsillectomy

Interventions Evidence-based processes suggested by the guidelines: preoperative DXM, no antibiotic use and outcomes (30 day tonsillectomy complication-related revisits) Outcomes DXM and antibiotic use Ratio of revisits to the hospital in the first 30 days after surgery Main findings Some improvement in evidence-based processes of care but no improvement in complication-related outcomes in the first 2 years after publication of the AAOHNS guideline High incidence of hospital revisits due to pain Carpenter 2017

Methods Review – tonsillectomy perioperative care and outcomes

Participants

Interventions Assessing the literature regarding postoperative management after tonsillectomy in children

Outcomes

Main findings DXM improves analgesia, decreases PONV and does not appear to increase the risk of post-tonsillectomy hemorrhage Batistaki 2017

Methods Review of RCT’s and prospective consecutive studies

Participants

Interventions

Outcomes Efficacy of DXM on postoperative pain/analgesic consumption/time for first rescue analgesic administration

Main findings Beneficial impact of DXM on reducing post-tonsillectomy analgesic needs

Hanasono 2004

Methods Randomized, double-blind, placebo- controlled

Participants (age (yr)) 219 patients (1–12) undergoing tonsillectomy

Interventions (DXM mg kg-1) DXM 1 IV or placebo

Outcomes Pain scores on postoperative day 1 Oral intake on postoperative day 1 Emesis on postoperative day 1 Main findings Reduced postoperative pain with the combination of intravenous steroid and cold dissection surgical technique

Athanassoglou 2017

Methods Review: any publications of any type

Participants

Interventions

Outcomes Identify a potential benefit of throat pack use in anesthetic practice

Main findings There is no evidence supporting the use of throat packs insertion routinely after intubation; most evidence concerns complications of throat pack use

08-Beckers.indd 62 12/01/2021 12:14 pain and ponv in pediatric tonsillectomy 63 kg-1 intraoperative DXM (with a maximum dose of hours, independent of the dose DXM administered 25 mg) (group 2) and no intravenous DXM (control (p= 0.01). DXM did not decrease the need for group). There was a significantly higher incidence morphine in the PACU (postanesthetic care unit) of PONV in the control group when compared (p= 0.42) nor did it decrease the proportion of with group 1 (p : 0.001 ; p < 0.01) and group 2 children with inadequate pain relief during the first (p : 0.001 ; p < 0.01). Furthermore, tolerance of 24 hours (p= 0.37) (10). food intake was better in group 2 when compared In their randomized double-blind trial, with group 1, although there was no statistically Hermans et al found that a single intravenous dose significant difference in PONV ratio between group of DXM before surgical incision reduced pain 1 and group 2 (p : 0.400 ; p > 0.05). Karaman et al only on the second postoperative day (p < 0.001). therefore concluded that intravenous DXM (0.7 mg The authors studied 134 children undergoing kg-1) effectively decreases postoperative morbidity, tonsillectomy who received DXM 0.15 mg kg-1, when compared with 0.2 mg kg-1, and in the absence 0.5 mg kg-1, or placebo. Like Czarnetzki et al, they of measurable side effects (11). also pointed out that administration of DXM during Hermans et al also demonstrated that at in- induction of anesthesia did not reduce the number of duction of general anesthesia, a single intravenous patients receiving morphine in the PACU (p= 0.321) injection of DXM was effective in reducing the or tramadol on the ward (p= 0.331). However, in the incidence of PONV, both during hospitalization and placebo group, time to first dose of tramadol on the on postoperative day 2. The use of rescue antiemetics ward was shorter than in the DXM groups (12). was lower in the DXM groups compared with The same year, Hashmi et al showed in a placebo. However, they were unable to identify a double-blind RCT that pain scores were significantly dose-dependent effect of DXM on PONV (0.15 mg lower and swallowing was less painful in patients kg-1 appeared to be as effective as 0.5 mg kg-1, p= after 4 (p= 0.001), 8 (p= 0.001), 12 (p= 0.003) and 0.118) (12). 24 hours (p= 0.001) in DXM group. Pain scores Hashmi et al evaluated in a RCT the effect of were on average 0.8 – 1.2 times lower in the DXM a single preoperative dose of DXM (0.5 mg kg-1) group than with placebo in the first 24 hours using on PONV in children undergoing electrocautery a VAS (visual analog score) of 1 – 10 (6). The tonsillectomy. The study showed significantly less trial evaluated the effect of a single pre-operative postoperative vomiting as compared to placebo dose of DXM (0.5 mg kg-1) with the frequency (p=0.027 for early vomiting (within 4 hours of postoperative PONV, the severity of throat pain postextubation) ; p=0.046 for late vomiting (4-24 and painful swallowing in children undergoing hours postextubation). Furthermore, the mean time electrocautery tonsillectomy (6). to first oral intake was earlier in the DXM group (6). Finally, Buland et al reported the effect of A prospective audit of pediatric post-tonsil- intravenous DXM (0.15 mg kg-1) after induction lectomy vomiting in 107 patients showed a statis- of general anesthesia compared with placebo. In tically significant reduction in vomiting due to use this study, which included adult patients as well of intra-operative DXM (0.1 mg kg-1) (p < 0.05) (2). as children, patients in the DXM group reported In contrast, Windfuhr et al were unable to detect significantly lower pain scores at 6 (p= 0.028) and any beneficial effect on the incidence of PONV 12 hours (p= 0.020) postoperatively. There was no with the use of DXM post-tonsillectomy. This significant difference in pain scores at 2 hours (p= study was limited by its retrospective observational 0.093) (7). design. The major shortcomings of this type of Discussion and conclusion study, acknowledged by the authors, were the lack of standardization and control of confounding Tonsillectomy is one of the most commonly variables e.g. anesthetic technique, use of narcotics, performed surgical procedures in children and its volume of ingested blood. Thus, the quality of importance relates to the documented benefit on evidence provided by this negative study may be child health status and quality of life. However, post- insufficient to provide definitive answers (13). tonsillectomy morbidity, including postoperative pain and PONV, remains a significant clinical The effect of dexamethasone on pain problem for the patient, family and care providers. This review aimed to determine the efficacy of In the RCT of Czarnetzki et al, DXM DXM on postoperative pain and PONV in children decreased the use of nonsteroidal anti-inflammatory undergoing tonsillectomy. Fifteen studies were medications for rescue analgesia during the first 24 included.

08-Beckers.indd 63 12/01/2021 12:14 64 l. beckers et al. Key findings creases PONV, and does not appear to increase the risk of post-tonsillectomy hemorrhage (4). Despite the fact that the exact mechanisms We need to remember that this guideline does are not fully understood, the literature supports not apply to tonsillotomy, intracapsular surgery, or use of a single dose of peroperative intravenous other partial removal techniques, because of the DXM to reduce post-tonsillectomy morbidity. relatively sparse high-quality published evidence In their Clinical Practice Guideline Report, the on these techniques and limited long-term follow- AAOHNS (American Academy of Otolaryngology, up. Similarly, the guideline does not apply to Head and Neck Surgery) panel made a ‘strong populations of children excluded from most recommendation’ that clinicians should administer tonsillectomy research studies, including those a single, intraoperative dose of intravenous DXM to with diabetes mellitus, cardiopulmonary disease, children undergoing tonsillectomy (1). craniofacial disorders, congenital anomalies of the Not all studies are in agreement, however. head and neck region, sickle cell disease, and other Windfuhr et al reported no effect of DXM on coagulopathies or immunodeficiency disorders. PONV (13). Mahant et al concluded from their In conclusion, given the frequency with retrospective cohort study that the 2011 AAOHNS which tonsillectomy is performed and the benefit tonsillectomy guideline was associated with some this procedure has on child quality of life, there is improvement in evidence-based processes of sufficient evidence to promote the use of a single care, but no improvement in complication-related peroperative dose of intravenous DXM. This safe, outcomes in the first 2 years after its publication. In low cost treatment provides benefits with regard to spite of their study limitations, a major concern was postoperative analgesia, PONV and return to a soft/ the high incidence of hospital revisits due to pain solid diet. (9%, p < 0.001). The results were almost certainly influenced by confounding factors such as surgical Limitations of the review technique, specific discharge practices, analgesic prescription and outpatient follow-up (14). Substantial variability exists with respect to Summarizing the available evidence, a study design as well as surgical, anesthetic and Cochrane meta-analysis of RCT’s (with a total hemostatic techniques used in the different articles of 1756 participants) demonstrated that children included in this analysis. Conclusions are therefore receiving a single intraoperative dose of DXM difficult to draw. (dose range 0.15-1 mg kg-1) were half as likely to There is also great variability in the numbers vomit in the first 24 hours compared to children of study participants included ; some trials were receiving placebo. Children receiving DXM were limited by their small sample size. The manner also more likely to advance to a soft/solid diet in which pain was assessed in these studies also on post-tonsillectomy day 1 than those receiving differed. This is a difficult task, especially in placebo. On average, about 4 children would need children. Finally, preoperative fasting influences the to receive intravenous DXM to result in 1 fewer incidence of PONV. None of the studies, however, patient experiencing post-tonsillectomy emesis report the fasting guidelines they applied, and this (number needed to treat = 4) (9). The Cochrane may in turn have affected the results. study also demonstrated that postoperative pain was improved in children receiving DXM as measured Future research opportunities by a VAS 0 to 10 which correlates clinically to a reduction in pain score from 4.72 to 3.65. They The literature provides no consensus as to concluded that a single dose of DXM is ‘virtually the optimum dosage of intravenous DXM. Most without harmful effect’ ; no adverse events were published studies have used a DXM dose of 0.5 mg noted in the included studies (9). kg-1, although the Cochrane systematic review of In their review Batistaki et al highlighted RCT’s showed that doses have ranged from 0.15 to a beneficial impact of DXM on reducing post- 1 mg kg-1, with maximum doses ranging from 8 to tonsillectomy analgesic needs. They concluded that 25 mg intravenously (9). DXM administered at a dose of 8 mg before surgical Studies aimed at determining the effective incision may be beneficial in tonsillectomies in analgesic dose of DXM are, to date, inconclusive children (5). (5). In 2017, a review by Carpenter et al concluded Further research regarding the optimum dose that perioperative DXM improves analgesia, de- of DXM for tonsillectomy is warranted.

08-Beckers.indd 64 12/01/2021 12:14 pain and ponv in pediatric tonsillectomy 65 Acknowledgement steroid (dexamethasone) for postoperative pain relief in tonsillectomy patients. J Coll Physicians Surg Pak. 22: 349-352. We would like to thank Dr. Morrison S., M.D., 8. Hanasono MM., Lalakea ML., Mikulec AA., Shepard KG., of the Department of Anesthesiology, Antwerp Wellis V. and Messner AH. 2004. Perioperative steroids University Hospital, for his contribution to this in tonsillectomy using electrocautery and sharp dissection techniques. Arch Otolaryngol Head Neck Surg. 130: 917- master thesis. 921. 9. Steward DL., Grisel J. and Meinzen-Derr J. 2011. Steroids for improving recovery following tonsillectomy in children. References Cochrane Database Syst Rev. 8: CD003997. 10. Czarnetzki C., Elia N., Lysakowski C., Dumont L., Landis 1. Baugh RF., Archer SM., Mitchell RB., Rosenfeld RM., BN. and Giger R., et al. 2008. Dexamethasone and risk Amin R. and Burns JJ., et al. 2011. Clinical practice of nausea and vomiting and postoperative bleeding after guideline : tonsillectomy in children. Otolaryngol Head tonsillectomy in children: a randomized trial. JAMA. 300: Neck Surg. 11: S1-S30. 2621-2630. 2. Bennett AM. and Emery PJ. 2008. A significant reduction 11. Karaman M., Ilhan AE., Dereci G. and Tek A. 2009. in paediatric post-tonsillectomy vomiting through audit. Determination of optimum dosage of intraoperative Ann R Coll Surg Engl. 90: 226-230. single dose dexamethasone in pediatric tonsillectomy and 3. Athanassoglou V., Patel A., McGuire B., Higgs A., Dover adenotonsillectomy. Int J Pediatr Otorhinolaryngol. 73: M.S. and Brennan P.A. et al. 2017. Systematic review of 1513-1515. benefits of harms of routine anaesthetist-inserted throat 12. Hermans V., De Pooter F., De Groote F., De Hert S. and packs in adults: practice recommendations for inserting and Van der Linden P. 2012. Effect of dexamethasone on counting throat packs. Anaesthesia. 73: 612-618. nausea, vomiting, and pain in paediatric tonsillectomy. Br J 4. Carpenter P., Hall D. and Meier JD. 2017. Postoperative Anaesth. 109: 427-431. care after tonsillectomy: what’s the evidence? Curr Opin 13. Windfuhr JP., Chen YS., Propst EJ. and Güldner C. 2011. Otolaryngol Head Neck Surg. 25: 498-505. The effect of dexamethasone on post-tonsillectomy nausea, 5. Batistaki C., Kaminiotis E., Papadimos T. and Kosto- vomiting and bleeding. Braz J Otorhinolaryngol. 77: 373- panagiotou G. 2017. A narrative review of the evidence on 379. the efficacy of dexamethasone on postoperative analgesic 14. Mahant S., Hall M., Ishman SL., Morse R., Mittal V. consumption. Clin J Pain. 33: 1037-1046. and Mussman GM., et al. 2015. Association of national 6. Hashmi MA., Ahmed A., Aslam S. and Mubeen M. 2012. guidelines with tonsillectomy perioperative care and Post-tonsillectomy pain and vomiting: role of pre-operative outcomes. Pediatrics. 136: 53-60. steroids. J Coll Physicians Surg Pak. 22: 505-509. 15. Keller M. and Brigger MT. 2013. The steroid controversy: 7. Buland K., Zahoor MU., Asghar A., Khan S. and Zaid AY. where are we? ORL J Otorhinolaryngol Relat Spec. 75: 2012. Efficacy of single dose perioperative intravenous 155-164.

08-Beckers.indd 65 12/01/2021 12:14

(Acta Anaesth. Belg., 2020, 71, 67-72) Best access for interscalene plexus block : a prospective observational study

P. Waelkens (*), D.S. Dirzu (**), S. Coppens (*)

Abstract : Background : Interscalene brachial Introduction plexus block has been widely implemented for anesthesia and analgesia in shoulder, clavicle and Locoregional anesthesia has a key role in humerus surgery. However, complications can perioperative pain management. It significantly occur in a complex and densely organized nerve decreases postoperative pain, reduces supplemental area. Through the years many techniques have analgesia requirements and probably enhances been described to access this region in the safest recovery (1). The interscalene brachial plexus block possible way. We aimed to describe one superior (ISB) is worldwide approved for shoulder, clavicle technique for needle introduction, with minimal and humerus surgery (2). Since its introduction complications, using sonoanatomy to detect ideal in 1970 by Winnie (3), it is commonly used. He pathways. proposed an anterior approach to the interscalene Methods : After obtaining approval from the ethics groove. The puncture site was the top of the triangle committee and written informed consent, patients formed by the anterior and middle scalene muscle. were included in this single centre, prospective The needle was introduced perpendicular to the clinical observational trial. We initiated systematic skin in all planes and advanced in a mesiad, dorsal ultrasound scanning and identified structures and slightly caudal direction until paresthesia was before block performance. Anatomical differences elicited. The block is easy to perform and provides were compared. Secondly, on these images, a excellent postoperative analgesia. It has fewer side perfect needle pathway was drawn. We compared effects and has a superior patient satisfaction and anatomical variance to calculate safest practice. The patient outcomes compared to general anesthesia (4). study was conducted according to the STROBE The first descriptions of the block were anatomical guidelines for reporting observational trials. based techniques (5-8). Pippa et al. (9) described an Results : From May 2019 to May 2020, 32 alternative to the Winnie method. They presented participants were enrolled in the study. Due to the a landmark based, posterior approach. The needle small sample size, statistical significance could was inserted posteriorly three cm lateral to a point not be met for our endpoints. Mean depth of the midway between the projection of the spinal process brachial plexus is 9.70 mm (SD 1.56 mm). Ideal of the 6th and 7th cervical vertebrae. The needle was needle angle position is 27.11° (SD 4.15°). There then advanced in an anterolateral direction until a is no clear correlation between the needle angle and loss of resistance to air was felt (10). Both Winnie and the subpopulations. Pippa are comparable regarding clinical efficacy for Conclusion : Interscalene brachial plexus blocks anesthesia of the shoulder and upper arm (11). After should be placed with neurostimulation and ultrasound guidance. We promote a lateral insertion technique with an in-plane ultrasound view. The Piet Waelkens MD, Dan Sebastian Dîrzu MD, Ph.D ; Steve needle should enter the skin next to the linear Coppens, MD transducer with the plexus brachialis in the middle (*) Department of Anesthesiology, University Hospitals of the screen. An angle of 27.11° (SD 4.15°) should Leuven, Belgium (**) Department of Anesthesia and Intensive Care, Emergency be applied to the sagittal plane. The external jugular County Hospital Cluj Napoca, Romania vein, the dorsal scapular nerve and the long thoracic Corresponding author : Piet Waelkens, Anesthesie en reani- nerve are avoided in this way in the majority of the matie, UZ Leuven, Herestraat 49, 3000 Leuven, Belgium. E-mail : [email protected] cases. Attention should be paid on identifying these structures on ultrasound in every patient before Paper submitted on May 27, 2020 and accepted on May performing the block. 28,2020. Conflict of interest : none.

09-Waelkens.indd 67 12/01/2021 12:14 68 p. waelkens et al. the introduction of ultrasound, multiple approaches Table 1 to the interscalene groove have been reported. There Inclusion and exclusion criteria isn’t one approach superior to another and many Inclusion criteria Exclusion criteria approaches are used nowadays without a consensus American Society of Patient refusal of an optimal technique. However, even with Anesthesiologist (ASA) Infection at the insertion point neurostimulation and ultrasound guidance, the risk physical status class I-III of nerve injury still remains. The anterior approach Age 18 – 60 years old Active bleeding in the cervical region has the risk of causing phrenic nerve (PN) damage, Body mass index < 35 Anticoagulants intake (except for aspirin) Severe pulmonary diseases while the lateral approach encounters the dorsal Contralateral paresis of the diaphragm scapular nerve (DSN) and the long thoracic nerve Pre-existing neurological deficit of the (LTN) (12, 13). In this trial, we tried to find the ideal limb approach to the interscalene groove to avoid nerve injury or other complications. C6. This is usually the puncture place. The neck was dynamically scanned from cranial to caudal and back Methods to identify several structures such as nerve roots C5- C6-C7, dorsoscapular nerve, phrenic nerve, long This study is a single centre, prospective, thoracic nerve, carotid artery, internal and external observational clinical trial. Approval by the local jugular veins and the vertebral artery. These ethics committee (Medical Ethics Committee of structures were mapped on a template (Appendix the University Hospitals and Catholic University 2). A sonographic picture was made of theoretical Leuven – registration number B322201940509) puncture place. On this picture, an ideal pathway for was obtained and registered (CTG S62468) in ISB was drawn. The angle of needle entrance was accordance to the International Conference on then measured. The primary endpoint was mapping Harmonization Guidelines on good Clinical Practice the inter-individual differences in anatomy of the (ICH-GCP) and the latest Declaration of Helsinki. interscalene groove and the cervical region. Based This observational trial is conducted according to on that information, we tried to describe an ideal the STROBE (STrengthening the Reporting of pathway for needle entrance as secondary endpoint. OBservational studies in Epidemiology) guidelines The goal was to obtain less nerve damage such as (14). the dorsal scapular nerve, long thoracic nerve or A total of 32 patients undergoing an ISB for phrenic nerve. elective shoulder, clavicle or scapula surgery were A statistical analysis was made in order to included in the trial. Patients were identified and define relationships between patient subgroups approached during their presurgical evaluation in the and measured variables on ultrasound. Values were anesthesia preparation room. They were informed submitted to a two-tailed student t-test. Results were about the study and if they agreed to participate, considered to be statistically significant at P < 0.05. written informed consent was obtained before A small literary review was performed to provide a placement of the block. Patient characteristics were brief update of the current practice guidelines. registered such as age, gender, height and length. Inclusion and exclusion criteria are shown in table Results 1. The ISB procedure was standardized for From May 2019 to May 2020, 32 participants all patients. All procedures were performed in a were enrolled in the study. Among them 12 female locoregional block room near the operating theatre. and 20 male patients. Patient characteristics and All patients received an intravenous line and full results are shown in table 2. ASA monitoring (ECG, non invasive blood pressure Mapping of the different structures around and pulse oxygenation). All patients were positioned the interscalene groove was done according to the with a rotation of the head to the contralateral protocol and ultrasound images were saved (Fig. side of the block. The angle between the bed and 1A). The internal jugular vein, the external jugular the head was 30°, accomplished by a preformed vein, the carotid artery, the cervical nerve roots C5 wedge (Appendix 1). Ultrasound transmission gel and C6 could be visualized in all cases. Nerve root C7 (Aquasonic 100®) was applied on the cervical area. was seen in 62.5% of the cases. Identification of the Then a high-frequency (15 MHz) linear ultrasound DSN, the PN and the LTN was 62.50%, 93.75% and transducer (BK Medical Ultrasound ®) was placed 9.38% respectively. No large anatomical differences in the transverse plane on the neck at the level of could be seen in the population. The position of

09-Waelkens.indd 68 12/01/2021 12:14 from winnie to pippa : best access for an interscalene plexus block 69

Fig. 3. — Box plot : comparison of the distribution of the measured angles of needle entrance in all patients.

Fig. 1. — A. Ultrasound image of a right-side ISB. Ideal needle pathway is drawn and the angle (26.3°) measured just next to the transducer. Abbreviations : ASM ; anterior scalene muscle, MSM ; middle scalene muscle, DSN ; dorsal scapular nerve, LTN ; long thoracic nerve. B. Position of the external jugular vein above the interscalene groove at the right side. 1 : anterior scalene muscle, 2 : middle scalene muscle, 3 : external jugular vein, 4 : C5, 5 : C6. An ideal pathway is drawn and the angle is measured (avoiding the external jugular vein, the DSN, the LTN and away from the PN).

Fig. 2. — Scatter diagrams demonstrating the distribution of the Fig.4. — A. Interscalene brachial plexus block puncture at the needle angle in degrees with increasing age, length, weight and left side. The needle (*) is piercing the middle scalene muscle BMI. Every dot represents the measurement of a single patient. in the middle. Our ideal pathway, with an angle of 25.3° and insertion just next to the transducer, is going more superficial in the middle scalene muscle. Nerve damage is less likely. the external jugular vein is latero-superior to the Abbreviations : ASM ; anterior scalene muscle, MSM ; middle interscalene groove (Fig. 1B). Eleven patients had scalene muscle, C5 ; cervical root 5, SCM ; sternocleidomastoid surgery in the day care centre, 21 patients in the muscle. B. Relationship between the plexus and the vertebral artery. When the tip is not well visualized, puncture of the operating theatre and had to stay overnight. This vertebral artery is possible when inserting the needle too depends on the type of surgery. For analysis of the deep. SM : anterior scalene muscle, VA : vertebral artery, IJV : continuous variables age, length, weight and BMI, internal jugular vein, SCM : sternocleidomastoid muscle, *: we plotted the needle angle in scatter diagrams (Fig. needle tip. 2). When drawing the ideal pathway to the deviation (SD) of 4.15 degrees. A comparison brachial plexus at the interscalene groove, the mean between discrete variables is schematically shown needle angle was 27.11 degrees with a standard by means of a box plot (Fig. 3). We did not find

09-Waelkens.indd 69 12/01/2021 12:14 70 p. waelkens et al. Table 2 peripheral nerve injury following ISB is 3 to 9% Pertinent Patient Data and results of the mapping, (15, 16, 17). There is an increased incidence with angle and depth measurements neurostimulation guidance compared to ultrasound Study population guidance (18, 19). Symptoms resolved in the Gender (m/f) 20/12 majority of the cases. Less than 0.51% of peripheral Age (years) 44.84 ± 10.50 neurologic symptoms persisted beyond 1 year for Body weight (kg) 83.03 ± 16.65 both groups, which is still alarmingly high. The DSN Body height (cm) 175.2 ± 9.86 Side (right/left) 20/12 and the LTN are more encountered in the lateral to Type of surgery medial approach. The LTN arises from the anterior – Arthroscopy 17 branches of the 6th cervical nerve root, the DSN – Total shoulder arthroplasty 5 from the 5th cervical nerve root. In most cases, both – Latarjet 4 are piercing the middle scalene muscle. In a study of – Humerus fracture 2 Kim et al. (20), 70 patients were observed during in- – Clavicle fracture 3 – Scapula fracture 1 plane lateral ISB placement with neurostimulation Visualization of: and ultrasound guidance. The DSN was encountered – Dorsal scapular nerve 62.50% in 62,8%, the LTN in 21,4% and both in 14.3% of – Phrenic nerve 93.75% cases. Neurostimulation is a great tool to detect – Long thoracic nerve 9.38% – Carotid artery 100% DSN (levator scapulae muscle and rhomboids) – Internal jugular vein 100% and LTN (serratus anterior muscle) involvement. – External jugular vein 100% If this happens, the needle should be withdrawn – Cervical nerve root C5 100% – Cervical nerve root C6 100% and another path must be chosen. DSN injuries can – Cervical nerve root C7 62.50% cause a chronic pain syndrome, often called the Needle angle 27.11° ± 4.15° DSN syndrome. LTN injuries can cause shoulder Depth of the brachial plexus 9.70 mm ± 1.56 mm weakness and serratus anterior muscle palsy. ISB Values are expressed as mean ± standard deviation ; f : female ; m : is associated with high rates of hemidiaphragmatic male. paralysis due to inadvertent blockade of the phrenic nerve. Incidence has been reported to occur up to a statistically significant difference among gender 100% (21). The phrenic nerve originates from the (p = 0.59) and side of block placement (p = 0.41). cervical spinal roots C3, C4 and C5, arises at the The mean depth of the interscalene brachial plexus, lateral border of the anterior scalene muscle and which was measured from the top of C5, was 9.70 then passes over it at the anterior surface. The PN mm (SD 1.56 mm). Obese patient had a trend of a is thought to be more involved in the out-of-plane deeper (p = 0.13) lying brachial plexus and a steeper anterior approach, described by Winnie in 1970. But needle angle (p = 0.21), without reaching statistical Bergmann and colleagues refute this assumption significant difference. Due to the relative small (22). In their randomized controlled trial, there was sample population, statistical significance could not no statistically significant difference in the incidence be met for any endpoint. of phrenic nerve blocks between the anterior and When we looked at the effective needle path- lateral approach. Numerous modifications have way retrospectively used for block placement, we been proposed to avoid phrenic nerve involvement found that it was much more parallel to the skin including avoidance of catheter infusions, low compared to the ideal pathway. The insert point volume and concentration of local anaesthetics and was not just next to the probe, but mostly one cm ultrasound guidance (23). Recently, new phrenic- more lateral. Angle of needle insertion is difficult sparing nerve blocks have been developed with to measure because of the curvature of the neck promising results like the superior trunk block from and compression with the transducer. The risk of Kim et al. (24) or suprascapular and axillary nerve creating a parallel path to the skin is piercing a blocks. bigger portion of the middle scalene muscle. The Historically, needle placement has been DSN and the LTN can be damaged in this way (Fig. described as perpendicular to the skin in every 4A). plane (25). Wong et al. evaluated the ideal angle in patients undergoing MRI of the cervical region Discussion (26). They found that a needle path of 60° to the sagittal plane was ideal for block placement. When ISB is the most implicated block when we drew our ideal pathway, the mean needle angle reviewing literature on nerve damage. The rate of was 27.11° in the sagittal plane. The difference can

09-Waelkens.indd 70 12/01/2021 12:14 from winnie to pippa : best access for an interscalene plexus block 71 be explained by compression with the ultrasound block placement. Secondly, the ultrasound scan of probe. Nevertheless, if this angle is applied, the the interscalene region is operator dependent which DSN and LTN were avoided and the brachial plexus could give some bias. Finally, due to the small easily reached at a mean depth of 9.70 mm. Plexus sample no definite conclusions could be drawn. brachialis lies deeper in patients with a higher In conclusion, an ISB should be placed body mass index (BMI), and thus the needle angle with neurostimulation and ultrasound guidance. is steeper. This trending difference was observed, Stimulation of the DSN provokes contraction of the however was not statistically significant. rhomboid muscles and the levator scapulae muscle In our findings, we support the use of a lateral and LTN stimulation provokes serratus anterior to medial approach, with in-plane ultrasound view muscle contractions. We promote a lateral insertion and nerve stimulation (27). This support is based technique with an in-plane ultrasound view. The on the knowledge that out of plain approach (OOP) needle should enter the skin right next to the linear grants less needle visibility. OOP approach to the transducer with the plexus brachialis in the middle medial side of the plexus is very close to the phrenic of the screen. An angle of 27.11° (SD 4.15°) should nerve, and also has a chance of hitting the vertebral be applied. The external jugular vein, the DSN and artery when depth is not adequately visualized. the LTN are avoided in this way in the majority of Using the lateral border brings you also dangerously the cases. Attention should be paid on identifying close to the initial branching of the DSN. Out of these structures on ultrasound in every patient plane approach without proper needle visualization before performing the block. at that same lateral border could also jeopardize the LTN when going too deep. Hydro dissection when Acknowledgement using OOP can help needle visualization, however can anesthetize phrenic nerve superficial cervical PW and SC included and scanned the patients, plexus, unless physiological serum is used. together with Aline Thouktarian, Caroline Gosh and Founded on our observations of the PN, the André Theron, three locoregional fellows who were DSN, the LTN and the external jugular vein, the active at the University Hospitals of Leuven. PW needle should be inserted lateral to the linear probe wrote the manuscript. The manuscript was reviewed in an angle of 27.11° in the sagittal plane. In this by SC and DSD. manner, the needle pathway is superior to the DSN and LTN and injection is not near the PN. Although References this seems to be the ideal pathway, PN involvement cannot be excluded. Anatomy is patient dependent 1. Hughes MS, Matava MJ, Wright R, Brophy RH and Smith MV. 2013. Interscalene brachial plexus block for and every case should be evaluated separately. arthroscopic shoulder surgery. J Bone Joint Surg Am. 95: Attention should be also paid on avoiding 1318-24. puncture of the external jugular vein at the needle 2. Toma O, Persoons B, Pogatzki-Zahn E, Van de Velde M and Joshi GP. 2019. PROSPECT guideline for rotator insertion point. The vein lies mostly above and cuff repair surgery: systematic review and procedure- lateral from the brachial plexus. Puncture of the specific postoperative pain management recommendations. vertebral artery (Fig. 4B), the carotid artery and Anaesthesia. 74(10):1320-1331. the internal jugular vein are described but are rare. 3. Winnie. Interscalene brachial plexus block. 1970. Anaest Analg. 49:455-466. Finally, case reports describe epidural spread and 4. Hadzić A, Vloka JD, Kuroda MM, Koorn R and Birnbach total spine after ISB (28, 29). This is rare but severe DJ. 1998. The practice of peripheral nerve blocks in the complication. Full ASA monitoring during and after United States: a national survey. Reg Anesth Pain Med. 3:241-6. block placement, is still the key point. 5. Boezaart AP, Koorn R and Rosenquist RW. 2003. There are limitations to this prospective Paravertebral approach to the brachial plexus: an anatomic observational study. It was not possible to identify improvement in technique. Reg Anesth Pain Med. 28(3):241-244. all the different nerve structures in every patient, 6. Pham-Dang C, Gunst JP, Gouin F, et al. 1997. A novel although literature suggests that the DSN and the supraclavicular approach to brachial plexus block. Anesth LTN can be identified in the majority of patients Analg. 85:111-6. (30). This could be due to the fact that identifying 7. Borgeat A, Dullenkopf A, Ekatodramis G and Nagy L. 2003. Evaluation of the lateral modified approach for continuous these structures is not standard in our hospital interscalene block after shoulder surgery. Anesthesiology. or in international guidelines like the NYSORA 99:436-42. guidelines for interscalene block placement. From 8. Nguyen HC, Fath E, Wirtz S and Bey Tareg. 2007. Transscalene brachial plexus block: a new posterolateral now on, identification of the DSN, PN and LTN approach for brachial plexus block. Anesth Analg. 105:872- should be part of the anatomic evaluation before 5.

09-Waelkens.indd 71 12/01/2021 12:14 72 p. waelkens et al. 9. Pippa P, Cominelli E, Marinelli C and Aito S. 1990. during ultrasound-guided interscalene brachial plexus Brachial plexus block using the posterior approach. Eur J block with nerve stimulator. Korean J Pain. 29(3):179-84. Anaesthesiol. 7:411-420. 21. Urmey WF, Talts KH and Sharrock NE. 1991. One hundred 10. Rucci FS, Pippa P, Barbagli R and Doni L. 1993. How many percent incidence of hemidiaphragmatic paresis associated interscalenic blocks are there? A comparison between the with interscalene brachial plexus anesthesia as diagnosed lateral and posterior approach. Eur J Anaesthesiol. 10:303- by ultrasonography. Anest Analg. 72:498-503. 307. 22. Bergmann L, Martini S, Kesselmeier M, Armbruster W, 11. Rettig HC, Gielen MJM, Jack NTM, Boersma E and Klein Notheisen T, Adamzik M and Eichholz R. 2016. Phrenic J. 2006. A comparison of the lateral and posterior approach nerve block caused by interscalene brachial plexus block: for brachial plexus block. Reg Anesth Pain Med. 31:119- breathing effects of different sites of injection. BMC 126. Anesthesiol. 16(1): 45. doi: 10.1186/s12871-016-0218-x. 12. Kessler J, Schafhalter-Zoppoth I and Gray AT. 2008. An 23. Hogan QH. 2013. Phrenic nerve function after interscalene ultrasound study of the phrenic nerve in the posterior block revisited. Anesthesiol. 119(2): 250-252. cervical triangle: implications for the interscalene brachial 24. Kim HD, Lin Y, Beathe JC, et al. 2019. Superior trunk block. plexus block. Reg Anesth Pain Med. 33:545-50. A phrenic-sparing alternative to the interscalene block: a 13. Kim HJ, Park SH, Shin HY and Choi YS. 2014. Brachial randomized controlled trial. Anesthesiology. 131:521-33. plexus injury as a complication after nerve block or vessel 25. Long TR, Wass CT and Burkle CM. 2002. Perioperative puncture. Korean J Pain. 27:210-8. interscalene blockade: an overview of its history and 14. Cuschieri S. 2019. The STROBE guidelines. Saudi J current clinical use. J Clin Anesth. 14:546-556. Anaesth. 13:S31-S34. 26. Wong GY, Brown DL, Miller GM and Cahill DR. 1998. 15. Holbrook HS and parker BR. 2017. Peripheral nerve injury Defining the cross-sectional anatomy importance to following interscalene blocks: a systematic review to guide interscalene brachial plexus block with magnetic resonance orthopedic surgeons. Orthopedics. 41(5):e598-e606. imaging. Reg Anesth Pain Med. 23:77-80. 16. Sharma BS and Chung C. 2020. Incidence and risk factors 27. Abrahams MS, Aziz MF, Fu RF and Horn JL. 2009. for neurological injury after interscalene peripheral nerve Ultrasound guidance compared with electrical neurosti- block at an academic institution. J Clin Anaesth. 61: (in mulation for peripheral nerve block: a systematic review press). and meta-analysis of randomized controlled trials. Br J 17. Kim HJ, Park SH, Shin HY and Choi YS. 2014. Brachial Anaesth. 102:408-17. plexus injury as a complication after nerve block or vessel 28. Stundner O, Meissnitzer M, Brummett CM, et al. puncture. Korean J Pain. 27(3):210-218. 2016. Comparison of tissue distribution, phrenic nerve 18. Schnabel A, Meyer-Friesem CH, Zahn PK and Pogatzki- involvement, and epidural spread in standard vs low- Zahn EM. 2013. Ultrasound compared with nerve volume ultrasound-guided interscalene plexus block using stimulation guidance for peripheral nerve catheter place- contrast magnetic resonance imaging: a randomized, ment : a meta-analysis of randomized controlled trials. Br J controlled trial. Br J Anaesth. 116(3):405-12. Anaesth. 111(4):564-72. 29. Grefkens J Mand Bürger K. 2006. Total spinal anesthesia 19. Kapral S, Greher M, Huber G, Willschke H, Kettner after an attempted brachial plexus block using the posterior S, Kdolsky R and Marhofer P. 2008. Ultrasonographic approach. Anaesthesia. 61:1105-108. guidance improves the success rate of interscalene brachial 30. Saporito A. 2013. Dorsal scapular nerve injury: a plexus blockade. Reg Anesth Pain Med. 33: 253-258. complication of ultrasound-guided interscalene block. Br J 20. Kim YD, Yu JY, Shim J, Heo HJ and Kim H. 2016. Risk Anaesth. 111(5)840-9. of encountering dorsal scapular and long thoracic nerves

09-Waelkens.indd 72 12/01/2021 12:14 (Acta Anaesth. Belg., 2020, 71, 73-77) Virtual reality hypnosis for postoperative pain after total knee arthroplasty

L. Depauw (*), A. Bosteels (**), S. Maes (*), L. Sermeus (***), V. Saldien (*)

Abstract : Background and aims : Postoperative pain on quality of life and implies an important cost to frequently occurs after total knee arthroplasty. This society in terms of subsequent healthcare costs and is a major cause of distress after surgery and often social support systems (3). Therefore, adequate causes complications such as persistent pain. Adequate postoperative pain control is mandatory. management of postoperative pain is therefore mandatory. Opioids are often used in postoperative Virtual Reality (VR) is a non-pharmacological way to pain management protocols, as they are effective decrease pain in a number of different situations, with the analgesics and part of the well-known analgesic pain extended benefit of no opioid related side effects. VR can also be used as a method to induce therapeutic hypnosis, ladder for relief of cancer pain of the World Health which enables the patients to be in control of their pain Organisation. Unfortunately, opioids are known to using (self-)induced hypnosis. have major potential side effects and complications, Methods : The VR4POPKA (Virtual Reality for Post- such as toxicity, dependency and abuse (4). This Operative Pain management after Knee Arthroplasty) knowledge prompted a lot of research in the field trial was set up as an interventional, randomized, of alternative, non-opioid and non-pharmacological controlled, single centre, double blind, parallel group ways to treat pain. superiority trial. Eligible patients are assigned to Virtual Reality (VR) is a non-pharmacological standard treatment group with a sham VR program or way to decrease pain. It has been around for a while, standard treatment with VR hypnosis. They will receive but has only become a popular medical treatment VR treatment on the day of surgery (DOS) and DOS+1, since the introduction of accessible, affordable and +2 and +3 during physiotherapy. Outcome variables are wearable VR devices such as the Oculus Rift HMD measured by Numeric Rating Scale (NRS) pain scores, State-Trait Anxiety Inventory (STAI) Y questionnaires, (head-mounted display). VR has been proven to piritramide consumption and degree of passive flexion reduce pain in a great number of different situations, during physiotherapy. from experimental pain in healthy subjects to Results : For various reasons, few data were available for dressing changes in burn patients and as an analgesic analysis. Hence no conclusions can be drawn from these adjunct during physiotherapy (5, 6, 7). However, results. Mean piritramide consumption was lower in the literature regarding the use of VR for postoperative VR Hypnosis group compared to the VR sham group, but pain is still scarce. this difference did not reach statistical significance. Conclusion : This study could not find statistical significant differences between the VR hypnosis and the L. Depauw, MD ; A. Bosteels, MD ; S. Maes, MD ; L. Sermeus, MD PhD ; V. Saldien, MD, PhD VR sham group. It is however, a unique and interesting (*) Department of Anesthesia, University Hospital of Antwerp, study design that comprises three heavily studied fields Wilrijkstraat 10, 2650 Edegem, Belgium and yields possibilities for future research. (**) Department of Anesthesia, Clinique Saint-Jean, 1000 Brussels, Belgium Keywords : Virtual reality ; hypnosis ; total knee (***) Department of Anesthesia, Cliniques Universitaires Saint-Luc, 1200 Brussels, Belgium arthroplasty ; analgesia. Corresponding author : L. Depauw, Department of Anesthesiology, University Hospital of Antwerp, Wilrijkstraat 10, 2650 Edegem, Belgium. E-mail : [email protected] Introduction Paper submitted on May 27, 2020 and accepted on May 28, 2020. Postoperative pain frequently occurs after total Conflict of interest : none. knee arthroplasty (TKA) (1,2). Persistent surgical, Abstract was accepted for a poster presentation at the 10th or chronic surgical pain, defined as pain persisting World Congress of the World Institute of Pain (WIP), originally scheduled to take place at 20-23 May, 2020, for more than 3 months after surgery, poses a Rome, Italy, but postponed due to the outbreak of SARS- major health care issue. It has a significant impact CoV-2.

10-Depauw.indd 73 12/01/2021 12:15 74 l. depauw et al. In medical VR different types of environments – History of seizures, drug abuse or psychiatric exist. Some of these act as simple distractions (games, disease natural scenery), others have included elements – Patients with impaired hearing, impaired vision of medical hypnosis (8). Medical hypnosis has or afraid of submarine worlds been shown to reduce intensity and unpleasantness – Chronic pain patients of noxious stimuli (9) and to reduce anxiety and – Claustrophobia pain during postoperative recovery (10, 11). Since – Lacking decent knowledge of Dutch or French medical hypnosis is a one-on-one intervention, language. requiring a skilled, highly trained individual (8), it is difficult to standardize this technique for research. Interventions Applying medical hypnosis to great numbers of patients is also deemed a time consuming, costly Eligible patients were approached on the practice which is unfit for daily use. Using VR as a day before surgery or on the day of surgery from means of delivering a hypnotic intervention, solves September 2019 until February 2020. After a lot of the aforementioned problems. VRH (Virtual verifying the absence of any exclusion criteria and Reality Hypnosis) has already been established as presenting the study, the written, informed consent a means to reduce acute postoperative and chronic was obtained. Patients were randomly assigned to neuropathic pain (8, 12). one of two groups : one including virtual reality Therefore, we used a VR environment to hypnosis (VRH) complementary to standard treat- reduce pain after total knee arthroplasty, the envi- ment (VR Hypnosis group) and one including a ronment including a therapeutic hypnotic script. VR sham program complementary to standard We wanted to assess the analgesic effect of treatment (VR Sham group). Both environments medical VRH combined with standard treatment were designed by Oncomfort ® SA. The VR in the postoperative period, during sessions of phy- module, AQUA, comprised an animated underwater siotherapy. experience guided by a whale. A concomitant hypnotic script brought the patient in a state of auto- Materials and methods hypnosis. The Sham module consisted of the same animated scene depicting an underwater scene and Study design and setting a whale, similar to the AQUA module, but without any hypnotic suggestions in the animation. Patients This study was set up as a randomized con- and physiotherapist were blinded to which of the trolled, single centre, parallel group, double blind modules the patients received. The VR device used superiority trial to evaluate virtual reality hypnosis in this trial is a VR Gear from Samsung Oculus, as a component of a multimodal postoperative paint completed by a Samsung smartphone type Galaxy treatment. The protocol was approved by the ethics S7 and a pair of headphones. committee of the Antwerp University Hospital All patients received the VR device on the (UZA) on 24/09/2018. The study took place in a day of surgery during preoperative placement of tertiary hospital where more than 150 TKA are the femoral nerve block (15-minute session) and performed each year. Informed, written consent on postoperative days 1 through 3 (POD1, 2 and was obtained prior to participation. Randomization 3) during passive mobilisation with the continuous was done using a web-based block randomization passive motion (CPM) device : Kinetec (45-minute program in a 1:1 allocation. session). Placement of the femoral nerve block and perioperative analgesia was executed as standardized Patient selection per protocol. All patients received a patient con- Inclusion criteria : trolled intravenous analgesia device (PCIA) with piritramide (Dipidolor®), titrated to corrected – Scheduled for TKA (total knee arthroplasty) body weight (CBW). Before surgery and after each – Receiving general anesthesia and femoral nerve postoperative session, patients were asked to fill out block a STAI-Y questionnaire concerning state and trait – Adults (>18 years) anxiety factors. Degree of passive flexion was noted – Elective procedures by the physiotherapist, as were the NRS pain scores – Granted informed consent and nausea scores after each physiotherapy session. Exclusion criteria : Total piritramide consumption and length of stay – Refusal of consent were recorded as well.

10-Depauw.indd 74 12/01/2021 12:15 vr hypnosis for postoperative pain after tka 75 Outcomes

The comparison of piritramide consumption between VR Hypnosis group and VR Sham group was the primary outcome. This was recorded after each physiotherapy session and at POD 3. Secondary outcome measures were anxiety, pain scores, length of stay, nausea and vomiting, and overall rehabilitation.

Results

Figure 1 depicts the flow of enrolment of patients during the course of the study. It shows the reasons why the majority of approached patients did not get enrolled and the reasons for drop out during the study. For instance, when patients would be scheduled first on the day of surgery, the regional nerve blocks would be performed in the operating Fig.1. — Flowchart of enrolment of patients.

Table 1 Patient demographics and baseline values

VR Hypnosis VR Sham p-value* gender (male, %) 0 75 age (years, mean ± SD) 71.5 ± 6.4 57.3 ± 14.1 0.26 BMI (kg.m-², mean ± SD) 26.5 ± 5.0 33.8 ± 8.7 0.57 NRS D0 8 0 9 0 0 2 0.77 STAIY D0 33 62 42 32 36 36 0.31

BMI : body mass index ; SD: standard deviation ; NRS : numeric rating scale ; STAIY : state and trait anxiety index form Y. * t-test

Table 2 Primary and secondary endpoints

VR Hypnosis VR Sham Pitiramide consumption (ml, mean ± SD) 34.8 ± 28 79.6 ± 62.2 p = 0.46 NRS D1 5 4 0 0 0 7 NRS D2 1 1 / 5 0 5 NRS D3 0 2 / 2 / 6 nausea D1 2 1 0 0 0 3 nausea D2 0 2 / 0 0 3 nausea D3 0 1 / 0 / 0 STAIY D1 30 56 32 33 35 32 STAIY D2 30 60 / 42 35 31 STAIY D3 27 61 / 34 / 30 degree of flexion D1 70 115 75 100 95 80 degree of flexion D2 85 115 / 85 100 95 degree of flexion D3 85 115 / 85 / 110 length of stay (days) 5 5 6 5 8 6

NRS : Numeric Rating Scale ; 0 : no nausea ; 3 : >=2 episodes of nausea and vomiting. STAIY : State and Trait Anxiety Inventory form Y.

10-Depauw.indd 75 12/01/2021 12:15 76 l. depauw et al. room instead of the preparatory rooms, where VR and trained researchers can be guaranteed and more treatment could not take place due to lack of time. patients can be included in and followed through the The number of patients who dropped out because course of the study. of being uncomfortable with the study, was evenly Following the month of February, enrolment distributed among the VR Sham group and the VR had come to an unexpected halt for two reasons. First Hypnosis group. of all, our hospital changed the standard analgesia Table 1 shows the general demographic cha- regimen in TKA patients to include the iPACK block racteristics and baseline values of patients enrolled (infiltration between Popliteal Artery and Capsule in the study. Mean age of patients was 62 ±13.5 of the Knee) (13). Following this change, our study years and mean BMI was 32.33 ± 7,4 kg/m2. Patient protocol did not correlate to our hospital’s standard characteristics and baseline values did not differ of care anymore and enrolment could not continue significantly between the two groups. at that point. Secondly, the emerging COVID-19 Table 2 shows the results of the actual primary pandemic caused a cessation of all non-essential and secondary endpoints of the study. Secondary surgery, including TKA procedures. This measure endpoints are represented as raw data to point out the was taken by the Belgian federal government to individual trends of these values. Mean piritramide ensure hospital capacity would allow care for consumption was lower in the VR Hypnosis group all incoming COVID-19 patients in both regular compared to the VR Sham group, but the difference hospital wards and intensive care wards. did not reach statistical significance. Nausea was Furthermore, to allow more patients to be attributed to the piritramide by all patients, not to the analyzed in future research, measures will have to VR intervention. All but one patient reported lower be taken to ensure fewer patients dropping out. The pain scores on the third postoperative day (D3). All inclusion of more research personnel was already patients’ STAI-Y scores were lower on D3. Degree cited earlier. Drop out also occurred because of of flexion improved over the three day observation patients’ discomfort in the study, stating that they period in half of the observed patients. did not feel comfortable (anymore) participating in Additionally, two patients reported explicitly a medical study. Two patients dropped out stating to have enjoyed the VR module, one in each they perceived the VR program repetitions as intervention group. One of them, who had boring. This could be counteracted by expanding experienced regional anesthesia in the past, stated the number of available programs in the VR device, spontaneously that the procedure was much more so patients can choose from a selection of different enjoyable with VR Hypnosis than without. programs. As the number of analyzed patients is low, it is Discussion difficult to draw decisive conclusions from our data. Also, this magnitude may partially explain why Interpretation of results no statistical difference was found in piritramide consumption between the two groups. Another Initially, this study was set up as a randomized element to consider is the choice of intervention. controlled superiority trial. Due to various reasons, Since it is not possible to blind patients for an the number of patients analyzed was too low to intervention requiring a pair of VR goggles, this generate sufficient power to prove the hypothesis. research included no control arm in the strict sense. Therefore, the results acquired were mostly pre- Instead, the sham module provided a means to find sented as raw data and possibilities to yield higher a difference in analgesic effect between a simple numbers of patients in future research were explored. VR distraction and an included hypnotic script As Figure 1 shows, a high number of patients delivered by a VR device. Seeing as both groups had to be considered before one completed the received a VR treatment, this could attribute to the study. This was partly due to patients meeting fact that the difference found was too small to reach an exclusion criterion, but also due to logistical statistical significance. reasons. Also, enrolment took place at the day of surgery or the day before. If study information was Strengths and limitations of the study given sufficient time beforehand, more patients may be inclined to participate in the study. These The major strength of this study is the unique organizational problems might be solved by study design that uses virtual reality hypnosis as an assigning more researchers and VR devices to the analgesic in the setting of total knee arthroplasty, study protocol. This way availability of materials which is known to be a painful procedure with a

10-Depauw.indd 76 12/01/2021 12:15 vr hypnosis for postoperative pain after tka 77 high occurrence of chronic postoperative pain. A lot warranted to find the best methods and strategies. of research has been performed in this subject and The results of this study certainly yield possibilities is still being performed, but this study is the first to for future research. use VR hypnosis in this procedure. This research also poses some limitations. References To begin with, the study design provided a short period of observation, during which differences 1. Puolakka, P. A. E., Rorarius, M. G. F., Roviola, M., Puolakka, T. J. S., Nordhausen, K. and Lindgren, L. might not yet become obvious between groups. 2010. Persistent pain following knee arthroplasty. Eur J The ideal timing and ‘dose’ of virtual reality is not Anaesthesiol. 27: 455-460. yet known, but current literature seems to suggest 2. Wylde, V., Beswick, A., Bruce, J., Blom, A., Howells, N. and Gooberman-Hill, R. 2018. Chronic pain after total knee that prolonged and repeated sessions invoke longer arthroplasty. EFORT Open Rev. 3: 461-470. standing analgesia (14). Furthermore, degree of 3. Reddi, D. and Curran, N. 2014. Chronic pain after surgery: flexion was used as a secondary outcome measure, pathophysiology, risk factors and prevention. Postgrad Med but this observation proved to have strong individual J. 90: 222–227; quiz 226. 4. Schiller E. Y. and Mechanic, O. J. Opioid Overdose. differences and more importantly, postoperative [Updated 2019 Dec 27]. Treasure Island (FL): StatPearls degree of flexion is mostly affected by the Publishing; 2020 Jan-. Available from : https://www.ncbi. preoperative value, (15) which was not measured nlm.nih.gov/books/NBK470415/ 5. Hoffman, H. G., Patterson, D. R., Seibel, E., Soltani, M., in this protocol. Subsequently, we did not measure Jewett-Leahy, L. and Sharar, S. R. 2008. Virtual reality pain the patients’ so called immersiveness, or the sense control during burn wound debridement in the hydrotank. of presence one can experience during VR, which Clin. J. Pain. 24 : 299-304. 6. Scapin, S., Echevarría-Guanilo, M. E., Boeira Fuculo is an important contributor to the effect. Lastly, Junior, P. R., Gonçalves, N., Rocha, P. K. and Coimbra, R. the manner in which VR was delivered has to be 2018. Virtual Reality in the treatment of burn patients: A mentioned. Two observations can be made in this systematic review. Burns. 44 : 1403-1416. respect. First of all, the VR modules used did not 7. Li, A., Montaño, Z., Chen, V. J. and Gold, J. I. 2011. Virtual reality and pain management: current trends and future allow any kind of interaction. This possibility for directions. Pain Manag. 1 : 147-157. interaction is an important contribution to the sense 8. Patterson, D. R., Jensen, M. P., Wiechman, S. A. and Sharar, of immersion, which party ensures the delivery of S. R. 2010. Virtual reality hypnosis for pain associated with recovery from physical trauma. Int J Clin Exp Hypn. 58 : effect in VR. Secondly, the environment in which the 288-300. patient experiences the VR could be a noteworthy 9. Faymonville, M. E., Laureys, S., Degueldre, C., DelFiore, factor. During the study this environment was the G., Luxen, A. and Franck, G., et al. 2000. Neural mechanisms of antinociceptive effects of hypnosis. Anesthesiology. 92 : patient’s room on the hospital ward. Seeing as this 1257-1267. is an open environment, disturbance by nurses, 10. Häuser, W., Hagl, M., Schmierer, A. and Hansen, E. 2016; visitors or neighboring patients was probable and The efficacy, safety and applications of medical hypnosis could have had a negative impact on the patients’ – a systematic review of meta-analyses. Dtsch Arztebl Int: 289-96. immersion. This could be offset by having the 11. Berlière, M., Roelants, F., Watremez, C., Docquier, M. A., physiotherapy take place in a separate room, Piette, N. and Lamerant, S., et al. 2018. The advantages of allowing no disturbance during the therapy. hypnosis intervention on breast cancer surgery and adjuvant therapy. Breast. 37: 114-118. 12. Askay, S. W., Patterson, D. R. and Sharar, S. R. 2009. Conclusion Virtual reality hypnosis. Contemp Hypn. 26: 40-47. 13. Kim, D. H., Beathe, J. C., Lin, Y., YaDeau, J. T., Maalouf, D. B. and Goytizolo, E., et al. 2019. Addition of Infiltration The VR Hypnosis group consumed less Between the Popliteal Artery and the Capsule of the piritramide than the Sham group. Unfortunately, Posterior Knee and Adductor Canal Block to Periarticular the research did not generate enough data to yield Injection Enhances Postoperative Pain Control in Total significant results with sufficient power. Therefore, Knee Arthroplasty : A Randomized Controlled Trial. Anesth. Analg. 129: 526-535. results were used to assess feasibility and generate 14. Koo, K. I., Park, D. K., Youm, Y. S., Cho, S. D. and Hwang, adaptations for a greater study. It is however, a C. H. 2018. Enhanced Reality Showing Long-Lasting unique study design in very interesting fields (VR, Analgesia after Total Knee Arthroplasty : Prospective, Randomized Clinical Trial. Sci. Rep. 8: 1-9. hypnosis and analgesia for total knee arthroplasty) 15. Farahini, H., Moghtadaei, M., Bagheri, A. and Akbarian, E. which are all being studied intensely at the moment. 2012. Factors influencing range of motion after total knee Both VR and hypnosis have already found numerous arthroplasty. Iran Red Crescent Med J. 14 : 417-421. applications in medicine, but more research is

10-Depauw.indd 77 12/01/2021 12:15

(Acta Anaesth. Belg., 2020, 71, 79-86) Topographic assessment of the medial antebrachial cutaneous nerve at the axilla relative to the axillary artery and vein using ultrasonography

B. De Breuker, N. Mintjens, L. Sermeus, M. Breebaart

Abstract : Introduction : The medial antebrachial Introduction cutaneous nerve (MACN) can be identified using ultrasound (US). Its identification can prevent unintended Brachial plexus blockade (BPB) at the axilla injury and provide selective targeting when preforming is a commonly used and safe technique to provide an axillary brachial plexus blockade (BPB). The primary surgical analgesia for procedures of the elbow, aim is to assess the topographic variation of the MACN forearm and hand. Reding described a high volume at the axilla. Secondary objectives include ; topographic blind axillary BPB technique in 1921 (1). A hundred mapping of the terminal nerves and BPB success rate analysis. years later, this has evolved greatly to a real-time Methods : After approval by local ethics committee of the ultrasound (US)-guided multiple-injection low Antwerp University Hospital, 35 patients were included volume perineural technique. These new techniques in this observational study after providing informed have higher success rates, lower complication rates, consent. The and increase patient comfort (2, 3). MACN and terminal nerves were identified at the axilla BPB techniques find their targets at the level of using dynamic US examination. Images of the plexus the roots, trunks and terminal nerves. At the axilla, the were stored and the distribution relative to the AA and brachial plexus is divided into its terminal branches AV was analyzed using a 12-section pie-chart. The of which the ulnar (UN), radial (RN), median (MN) distance between the MACN and AV was measured. A and the musculocutaneous nerve (McN) are most five-injection BPB was preformed and tested after 20 important. Recently, various studies have shown minutes using cold sensation and evaluation of motor multiple variations in the topographic arrangement function. Results : The MACN was identified in all cases and of these terminal nerves at this level (4-7). found in the upper right quadrant to the AA in 97% of Other small nerves provide sensory innervation cases and upper left quadrant to the AV in 73%. The of the upper arm and forearm as well. These mean distance between the MACN and AV was 1.3 mm include the medial antebrachial cutaneous nerve (95% CI 0.6-2.0mm). The analyzed distributions of (MACN), the medial brachial cutaneous nerve and the other terminal nerves were plotted. All nerves were intercostobrachial nerve. Although these smaller successfully blocked, including the MACN. nerves pass the axilla as well, they are not routinely Conclusions : The MACN was identified in all cases, visualized when preforming an US-guided axillary mostly superficial and medial to the BPB. They are either assumed to be blocked by AA, superficial and lateral to the AV, and adjacent to the local spread of the anesthetic injected around AV. This might help to identify the MACN during axillary the other terminal nerves, or are targeted using a BPB to either avoid injury or for selective targeting. subcutaneous infiltration. Distribution of the larger terminal nerves relative to the AA was comparable to literature. No conclusions were made about the directional topographic distribution of Bart De Breuker, MD; Nesse Mintjens, MD; Luc Sermeus, the larger terminal nerves relative MD, PhD; Margaretha Breebaart, MD, PhD to the AV. Department of Anesthesia, Universitair Ziekenhuis Antwerpen, 2650 Edegem, Belgium Corresponding author: Bart De Breuker, M.D. Universitair Keywords : Axillary artery ; axillary vein ; brachial Ziekenhuis Antwerpen, 2650 Edegem, Belgium. plexus block ; brachial plexus/anatomy & histology ; E-mail: [email protected] ultrasonography. Paper submitted on May 27, 2020 and accepted on June 04, 2020. Conflict of interest: none.

11-De Breuker.indd 79 12/01/2021 12:16 80 b. de breuker et al. Recently, high resolution US was used to identify small cutaneous nerves and assess their clinical relevance in vivo (8). The MACN is said to be easily identifiable just above the elbow or at the mid-arm level (9, 10). It can be traced up and down the arm and even be targeted for selective blocks using high resolution US (11,12). During the last two decades, multiple case reports and reviews discussed the clinical Fig. 1. — Left : The brachial plexus. Right : The cutaneous importance of the MACN, often following injury. innervation of the MACN of the right upper limb. MACN : This small nerve seems prone to iatrogenic injury medial antebrachial cutaneous nerve, McN : musculocutaneous following orthopedic, reconstructive, and vascular nerve, AxN : axillary nerve, RN : radial nerve, MN : median surgery, but many other causes have been reported nerve, UN : ulnar nerve. as well (13). In anesthesia however, its relevance and the incidence of injury have not been defined Anatomical studies have shown multiple as clearly. variations in the course of the anterior and posterior This study aims to map the topographic branches (16). The anterior branches are reported to variations of the MACN relative to the axillary run as far as the skin overlying the carpal tunnel, and artery (AA) and axillary vein (AV) at the level of the are said to communicate with the greater nerves like axilla. Secondary objectives include the assessment the median and radial nerve. Multiple perforating of topographic variation of all terminal nerves branches higher-up the upper arm have also been relative to the AA and AV, analysis of the distance reported. In these cases the MACN takes over a between the MACN and the AV, and axillary BPB large sensory part of the median brachial cutaneous success rate. nerve of the upper arm (16).

MACN anatomy and innervation Materials and methods

The work by Masear and Mian and their Patients who were scheduled for awake colleagues contributed greatly to understand the forearm surgery between 2014 and 2016 at the anatomical course of the MACN (14, 15). Antwerp University Hospital were included in this The MACN arises from the medial cord of the prospective observational single center trial. The brachial plexus, or less frequently from the lower local ethics committee of the Antwerp University trunk, containing fibers of the C8 and T1 roots Hospital, led by Prof. Dr. P. Cras, Wilrijkstraat 10 (Fig.1 left). It is not to be confused with the medial - 2650 Edegem, approved this trial on 20/10/2014. brachial cutaneous nerve, which often arises from Patients were asked to provide written informed the medial cord as well. Initially the MACN is consent. located medial to the AA. Only a few centimeters Exclusion criteria included : contraindications from its origin, it gives off its first small ramus for regional anesthesia, inability to visualize two which penetrates the fascia to innervate the skin over or more terminal nerves at the axilla, vascular the biceps brachii muscle. The MACN descends abnormalities making analysis unreliable, absence further down the ulnar side of the arm. Proximal to of supervising staff overseeing this study. the medial epicondyle it pierces the fascia together The patient was placed in supine position. An with the basilic vein at the basilic hiatus and starts IV infusion was placed in the contralateral arm. dividing into anterior and posterior branches. The Routine monitoring ; including electrocardiographic main anterior branches descend on the ulnar side of monitoring, pulse oximetry, and noninvasive blood the anterior face of the forearm all the way down to pressure measurements, was established prior to the wrist. The posterior branches pass either above the BPB procedure. The arm was placed in 90° or below the medial epicondyle to the posterior side abduction at the shoulder and 90° flexion at the of the medial forearm. elbow, as described by Winnie (17). Correspondingly, the MACN provides sensory A Flex Focus 400 US (BK medical, Denmark) innervation to a small anterior part of the upper arm with a high frequency 12-18 MHz linear probe overlying the biceps muscle, the medio-anterior was used. The frequency was set to 18 MHz and part of the cubital fossa, the ulnar side of both the depth was adjusted to 1 centimeter below the AA. anterior and posterior faces of forearm (Fig. 1 right). All blocks were preformed using a 22-Gauge, 5

11-De Breuker.indd 80 12/01/2021 12:16 topography of the macn at the axilla using us 81

Fig. 2. — Ultrasound image of the greater terminal nerves and vasculature at the axilla. The MACN was identified using dynamic examination from the elbow toward the axilla. MACN : medial antebrachial cutaneous nerve, AV : axillary vein, AA : axillary artery, u : ulnar nerve, r : radial nerve, m : median nerve, mc : musculocutaneous nerve.

cm long SonoPlex Stim needle (Pajunk, Geisingen, Fig. 3. — 12 section pie-chart used for data collection, Germany). Nerve stimulation with a stimulus numbered 1 to 12. Extra horizontal and vertical axis divisions. of 0,3mA was used as an extra safety measure to ULQ : upper left quadrant, URQ : upper right quadrant, LLQ : minimize intraneural needle tip positioning. lower left quadrant, LRQ : lower right quadrant. The probe was placed in the axilla, at the lateral border of the pectoral muscle, perpendicular performed. The nerves were targeted using a short to the axis of the upper arm. Orientation of the axis in-plane technique. Needle entry was from probe was adjusted so the lateral structures relative right to left on the US-screen corresponding with to the axillary artery, i.e. biceps muscle, appeared an inferomedial to superolateral approach. Five on the left side of the screen (Fig. 2). The AA and milliliters of Mepivacaine 1.5 % was injected the conjoint tendon of the teres major and latissimus around each terminal nerve, three milliliters were dorsi muscles were visualized. Venous structures used for the MACN. The needle was adjusted to were identified by applying and releasing pressure provide circumferential spread around each nerve. to the skin. When multiple venous structures were Twenty minutes after completion of the visualized, the axillary vein was identified using BPB procedure, motor function was examined by proximal tracing starting at mid-arm level following assessing strength during elbow flexion (McN) and the basilic vein. All terminal nerves were identified extension (RN), opposition of the thumb as honeycomb-like structures following specific (MN) and abduction of the little finger (UN). anatomical courses as described by Ranganath Sensory blockade was assessed using cold sensation (18). The MACN was identified 10 centimeters at the fifth finger (UN), the index finger (MN), the proximally to the medial epicondyle, lying adjacent dorsum of the hand (RN), the skin at the radial to the basilic vein and was tracked cranially toward (McN), and ulnar part (MACN) of the forearm. the axilla, as described by Thallaj (9). Demographic variables including age, sex, Following identification of the MACN and weight, and height were recorded. the other terminal nerves at the lower border of the conjoint tendon, a US-image was stored after digital Data analysis marking of the MACN, MN, RN, McN, UN, AA and AV (Fig. 2). Directional topographic results All collected data was entered in IBM SPSS were collected using a numeric 12-section pie-chart statistics 26 (SPSS Inc., Chicago, IL) for descriptive with the middle of the AA being used as reference analysis. Results are reported as a mean and 95% points, as described by Christophe et al (4) (Fig. 3). CI, mean and SD, frequency, or percentages. The same was done using the middle of the AV as a reference point. The distance between the AV and Results MACN was measured between the outlining of both structures. Forty patients gave informed consent for this Following disinfection and ensuring aseptic study. One patient was excluded from the study due conditions, the five-injection axillary BPB was to the presence of a double axillary artery. Four

11-De Breuker.indd 81 12/01/2021 12:16 82 b. de breuker et al. Table 1 collection. Thirty-three cases were analyzed. Summary of demographic variability in patient population Directional topographic analysis relative to the AV showed a more widespread distribution of the Summary of demography MACN. It was mostly found in section 10 and 11, in respectively 14 (42%) and 8 cases (24%). The n=35 Min. Max. Mean SD MACN was found three times in both sections 3 Age (y) 18 83 58 16 and 9, two times in sections 4 and 12, and ones in section 2. In 84% of cases the MACN was found to Weight (kg) 52 121 79 14 be superficial to the AV (Fig. 4). Height (cm) 157 187 172 9 The MN was found only in sections 9, 10, and 11 in respectively 12 (36%), 18 (55%), and 3 cases BMI (kg/m²) 21 37 27 4 (9%) (Fig. 5b). The UN showed a very widespread Min.: Minimum, Max.: Maximum, SD: Standard deviation. variation. It was found in all but two sections of the pie chart centered on the AV. In section 9 and 10 is patients were excluded due to lacking data, especially was found most frequently in respectively 7 (21%) missing demographic variables. A summary of and 6 cases (18%). In section 8 the UN appeared demographic characteristics is displayed in Table 1. four times, in all other sections less than four times. Twenty-one out of 35 cases were male. A denser distribution was found for the RN peaking In all patients the MACN was identified at the in sections 8 and 7 in respectively 14 (42%) and 9 level of the axilla. In 33 patients all four terminal cases (27%). In 76% of cases the RN was found in nerves were identified and marked. The McN could the lower left quadrant relative to the AV, no cases not be identified at the level of the axilla intwo were found in the upper quadrants. cases. The McN had the densest concentration and In two cases no AV was marked on ultrasound was found in section 9 in 30 out of 32 cases (94%). at the level of the axilla. It is unclear if this was The remaining two cases appeared in section 10. an error due to compression, hydration status of the As mentioned before, in two cases this nerve was patient, or anatomical variations. These patients not identified, one of these cases also lacked the were not excluded from the study, results relative to presence of the AV. the AA were used for analysis. In 85% of cases the distance between the outlining of the MACN and the AV was less than Topography relative to the AA three millimeters, the mean distance being 1.3mm The MACN was predominantly found in (95% CI 0.6-2.0mm). In all cases the sections 1, 2, and 3 in respectively 11 (31%), 12 MACN was found within one centimeter of (34%), and 11 cases (31%) out of 35 patients. When the AV (Table 2). combining these results, the MACN appeared in the upper right quadrant (URQ) relative to the AA in Axillary plexus block success rate 97% of cases. It only appeared once outside of this quadrant, in section 4 (Fig. 4). In all studied cases, significant motor block off The directional distribution of the greater all targeted nerves was observed within 20 minutes. terminal nerves relative to the AA are shown in No objective measurements of strength and paralysis Figure 5b. The MN was mostly found in sections 12 and 1 in respectively 13 and 10 cases. These two sections account for 66% of all studied cases. The UN was most commonly found in sections 3 and 2 in respectively 19 (54%) and 11 cases (31%). In 89% of all studied cases, it was in the URQ relative to the AA. The RN was found in 22 out of 35 cases (63%) in section 4, the McN in section 9 in 23 out of 33 cases (79%).

Topography relative to the AV Fig. 4. — Topographic distribution of the MACN relative to As mentioned before, there were two cases the AA (left) and AV (right). AA : Axillary artery, AV : Axillary lacking the AV on the stored images for data vein.

11-De Breuker.indd 82 12/01/2021 12:16 topography of the macn at the axilla using us 83

Fig. 5. — a : Topographic distribution of the terminal nerves relative to the AA (red). b : Topographic distribution of the terminal nerves relative to the AV (blue). AA : Axillary artery, AV : Axillary vein, MN : median nerve, UN : ulnar nerve, RN : radial nerve, McN : musculocutaneous nerve.

Table 2 too can be a useful guide in identifying the MACN Summary of variations in distance between the Medial Antebrachial Cutaneous at the level of the axilla. The AV however has not been used for directional topographic mapping thus Mean 1.3 its value is yet unknown. 95% Confidence Lower 0.6 interval for men Upper 2.0 Interestingly, in most cases the MACN was Median 0.7 adjacent to the AV and in all cases found within one Std. deviation 2.0 centimeter of the AV. Although anatomical studies Minimum 0.1 already reported the correlation between the MACN Maximum 7.8 and basilic vein at mid-arm level, none can be Nerve and the Axillary Vein. Mm : millimeters, SD : standard deviation, found studying the MACN at the level of the axilla. Min. : minimum, Max. : maximum. Masear et al. reported that the MACN originates between the AA and AV, and initially runs down were done. All patients had an effective sensory adjacent to the AA before crossing over to the basilic blockade to cold sensation at the medial side of the vein in the distal part of the upper arm (14). This forearm as well as in all other dermatomes. Surgical anatomical description has been adapted often in anesthesia was established in all cases. literature although we could not find any confirming studies. Our results do not support this description. Discussion Matsuda et al. described a similar course as found in our study with the MACN being adjacent to the During this trial we were able to identify the basilic vein in the entire upper arm as well (12). We MACN in all patients. Its course could be tracked believe further research is needed to explore the up to the axilla and it was found in the URQ correct relationship between the vascular structures relative to the AA in all but one case. This finding and the MACN at this level. This could become a can be a useful guide for identifying this nerve third valuable guide in identifying the MACN while during a US-guided axillary BPB procedure. When preforming an axillary BPB. reviewing available literature, we could not find any The importance of the MACN has been publications reporting topographic mapping of the reported often in surgical literature. It is a known MACN at the level of the axilla relative to the AA. subject to injury during orthopedic surgery of Likewise, the MACN was found in the ULQ the elbow, especially when involving the cubital relative to the AV in three out of four cases. It was tunnel (19, 20). Postoperative complaints include rarely found lying deep to the middle of the AV. This numbness over the medial parts of the elbow and

11-De Breuker.indd 83 12/01/2021 12:16 84 b. de breuker et al. forearm, hyperesthesia and allodynia, or patients maps studying 153 patients using the same pie-chart may even present with symptomatic neurinoma analysis (4). In their results, the MN was found in causing severe pain (21). Although multiple the ULQ in 92%, the UN in the URQ in 90%, the case reports exist on this subject, there is still no RN in the LRQ in 89% and lastly the McN in the consensus among orthopedic surgeons about the LLQ in 91%. Although more dense distributions need to identify the MACN during ulnar nerve were reported, the sections in which the terminal transposition surgery (22). Vascular surgeons nerves were found most frequently were equal or should be familiar with the MACN as well, as it adjacent to the corresponding sections found in runs adjacent to the large blood vessels used for our study. A methodical difference was noted with arteriovenous fistula formation (23). Over the last higher pressure being applied to collapse all venous few years, various techniques have been developed structures before the image was stored for analysis. to avoid this small nerve (24). In reconstructive Retzl et al. studied the topography at the axilla surgery, the MACN has successfully been used as even earlier (5). As in this trial, they also reported a cable graft for decades, for example as a digital or a more widespread variation than reported in the facial nerve graft. More recently, it has also become study by Christophe et al. It was unclear if high or of interest to provide sensibility when preforming a low pressure on the probe was applied in this study. forearm free flap phalloplasty (25). Two studies published very recently by a Spanish Other causes of MACN injury found in litera- research group also reported similar distributions ture include trauma, lipoma, positional neuropraxia, and variations as found in our trial (6, 7). Although neurogenic thoracic outlet syndrome, repetitive they also measured distances between the UN and motion stress in athletics, or other iatrogenic causes AV, it was unclear if distributions were analyzed such as venipuncture (13). Seror bundled sixteen with or without venous collapse. case reports of patients presenting with atypical No conclusions could be made about the pain and paresthesia of the forearm without motoric directional topographic maps of the terminal nerves deficit (26). All had a MACN injury diagnosed relative to the AV. The McN and MN showed dens using neurophysiological examination. Remarkably distributions relative to the AV, although this can however, no clear etiology could be found in nine be explained by their distance to the AV. No other cases. research was found mapping the terminal nerves to The contribution of the anesthetist in MACN the AV. injury has not been defined as clearly. Jung et al. A 100% BPB success rate was found during recently published two cases in which the MACN this trial. Earlier reports show lower success rates. was accidentally injured during brachial plexus Tran et al. published a study in which the US-guided blockade at the axilla (27). These patients had double, triple and quadruple injection technique of painful paresthesia, dysesthesia, and numbness axillary BPB was compared in 120 patients (28). over the medial aspect of the forearm. The authors They found non significantly differing success concluded that these were the first documented rates of respectively 97.5%, 90.0%, and 97.5%. cases of MACN injury following axillary BPB. Interestingly, they used a volume 35 milliliters of Furthermore, as the MACN seems to accompany local anesthetics, compared to 23mL in our study. the AV along the entire upper arm, it can easily be This dose reduction should decrease the risk of injured during placement of US guided upper arm Local Anesthetic Systemic Toxicity (LAST). vascular access devices such as peripherally inserted In this trial, the region innervated by the central catheters (PICC line) or midline catheters MACN was successfully blocked in all cases as well. No other clear cases of MACN injury following the selective targeting of this nerve. The were reported so far following these procedures. necessity to perform a selective infiltration of the However, we believe this complication is probably MACN can be argued. Deleuze et al. published underdiagnosed as it does not cause any motoric a study comparing an infraclavicular to a triple dysfunction and does not involve innervation of the injection axillary BPB technique, in which they hand. Seror and his colleagues had similar thoughts included the sensory blockade of the MACN on the reported incidence of MACN injury as well region as an outcome parameter (29). As in our (22). study, they found a sensory blockade of the MACN The results from this study on the directional region in close to 100% in the axillary BPB group topographic distribution of the terminal nerves (n=50). However, they reported a surgical block relative to the AA were comparable to earlier success rate following the axillary BPB in only research. Christophe et al. published topographic 88%. Kapral et al. published a similar clinical trial,

11-De Breuker.indd 84 12/01/2021 12:16 topography of the macn at the axilla using us 85 using a single injection perivascular axillary BPB level of the axilla. It is most commonly located in the technique instead (30). In the axillary BPB branch, URQ relative to the AA. It has a more widespread they reported a sensory blockade of the MACN directional topographic distribution relative to the region in 85%. These results suggest local spread AV but is found in the of the injected local anesthetics around the other ULQ in three out of fore cases and is closely terminal nerves, rendering the selective targeting adjacent to the AV in the vast majority of cases. of the MACN unnecessary. Concluding however ; Although selective targeting of the MACN should both studies used volumes of 40 milliliters of local be researched, and may seem redundant to achieve anesthetics, and they did not find a full 100% success a surgical block, its identification while preforming rate as found in this trial. Other studies including an axillary BPB can prevent accidental injury of the MACN skin region as an outcome parameter the nerve and its resulting pain syndromes and are scarce. Further research is needed to draw any diagnostic dilemmas. conclusions on the necessity of selectively targeting Therefore, we believe identification of the the MACN during an axillary BPB. MACN should become mandatory when preforming US-guided procedures at the axilla or upper arm. Limitations References Preforming this study, we found that special attention should be given to assessments relative 1. Reding M. 1921. Nouvelle Methode d’Anesthesie to both the AV and AA. The AV has not been Regionale du Membre Superieur. Trans. of la presse used before in topographic studies thus its value Medicale (France) 29:294-296. 2. Chan VW, Perlas A, McCartney CJ, Brull R, Xu D, Abbas has not yet been proven. As all veins, it is subject S. 2007. Ultrasound guidance improves success rate of to dehydration and external pressure on the US axillary brachial plexus block. Can J Anaesth 54:176-182. probe which can be challenging in identifying this 3. Ecoffey C, Oger E, Marchand-Maillet F, Cimino Y, Rannou JJ. et al. 2014. Complications associated with 27 031 structure. Secondly, as multiple venous structures ultrasound-guided axillary brachial plexus blocks: a web- are often found at the axilla, it can also be difficult based survey of 36 French centres. Eur J Anaesthesiol to correctly identify the AV. Consequently, an 31:606-610. incorrect reference point may be used. Concluding 4. Christophe JL, Berthier F, Boillot A, Tatu L, Viennet A. et al. 2009. Assessment of topographic brachial plexus nerves this trial, we cannot determine the validity of the variations at the axilla using ultrasonography. Br J Anaesth AV as a reference point for directional topographic 103:60612. mapping. 5. Retzl G, Kapral S, Greher M, Mauritz W. 2001. Ultrasonographic findings of the axillary part of the brachial The AA has been used more often as a reference plexus. Anesth Analg 92:1271-2175. for topographic maps. Even though it is less 6. Gili S, Abreo A, GoMez-Fernandez M, Sola R, Morros C, subjective to external pressure itself, the positions Sala-Blanch X. 2019. Patterns of Distribution of the Nerves Around the Axillary Artery Evaluated by Ultrasound and of the nerves relative to the AA also change when Assessed by Nerve Stimulation During Axillary Block. using high pressure on the probe. This seems to Clin Anat 32:2-8. result from tissue and nerves being pushed down 7. Silva MG, Sala-Blanch X, Marin R, Espinoza X, Arauz and rolled over the AA. Simultaneously, the middle A, Morros C. 2014. [Ultrasound-guided axillary block: anatomical variations of terminal branches of the brachial of the AA moves towards the skin, changing the plexus in relation to the brachial artery]. Rev Esp Anestesiol reference point as well. Retzl et al reported similar Reanim 61:15-20. thoughts in conclusion of their research (5). 8. Chang KV, Mezian K, Nanka O, Wu WT, Lou YM, et al. 2018. Ultrasound Imaging for the Cutaneous Nerves of the Because of the methodological differences Extremities and Relevant Entrapment Syndromes: From we found comparing our trial to beforementioned Anatomy to Clinical Implications. J Clin Med 7. literature, we were unable to correctly analyze 9. Thallaj A, Marhofer P, Kettner SC, Al-Majed M, Al- Ahaideb A, Moriggl B. 2011. Highresolution ultrasound statistical differences. accurately identifies the medial antebrachial cutaneous Seeing only 35 cases were analyzed during this nerve at the midarm level: a clinical anatomic study. Reg trial, we are unable provide any conclusions about Anesth Pain Med 36:499-501. the topography and the BPB success rate. Further 10. Moritz T, Prosch H, Pivec CH, Sachs A, Pretterklieber ML. et al. 2014. High-resolution ultrasound visualization of the research using a standard subcutaneous nerves of the forearm: a feasibility study in protocol is needed. anatomic specimens. Muscle Nerve 49:676-679. 11. Nakanishi Y, Omokawa S, Kobata Y, Shimizu T, Kira T. et al. 2015. Ultrasound-guided Selective Sensory Nerve Conclusion Block for Wide-awake Forearm Tendon Reconstruction. Plast Reconstr Surg Glob Open 3:e392. The MACN can be identified using high 12. Matsuda H, Oka Y, Takatsu S, Katsube R, Yoshida R. et al. resolution US and traced along the upper arm to the 2016. Ultrasound-guided block of selective branches of the

11-De Breuker.indd 85 12/01/2021 12:16 86 b. de breuker et al. brachial plexus for vascular access surgery in the forearm: 22. Dy CJ, Antes AL, Osei DA, Goldfarb CA, DuBois JM. a preliminary report. J Vasc Access 17:284-290. 2018. The Critical Portions of Carpal Tunnel Release, Ulnar 13. Asheghan M, Khatibi A, Holisaz MT. 2011. Paresthesia and Nerve Transposition, and Open Reduction and Internal forearm pain after phlebotomy due to medial antebrachial Fixation of the Distal Part of the Radius. J Bone Joint Surg cutaneous nerve injury. J Brachial Plex Peripher Nerve Inj Am 100:e148. 6:5. 23. Przywara S. 2016. Importance of medial antebrachial 14. Masear VR, Meyer RD, Pichora DR. 1989. Surgical cutaneous nerve anatomical variations in upper arm surgery. anatomy of the medial antebrachial cutaneous nerve. J J Neurosci Rural Pract 7:337-338. Hand Surg Am 14:267-271. 24. Przywara S, Ilzecki M, Terlecki P, Zubilewicz T. 2015. 15. Mian A, Chaudhry I, Huang R, Rizk E, Tubbs RS, Loukas Axial splitting of the medial antebrachial cutaneous nerve M. 2014. Brachial plexus anesthesia: A review of the facilitates second-stage elevation of basilic or brachial vein relevant anatomy, complications, and anatomical variations. in patients with arteriovenous fistula. J Vasc Surg 62:1353- Clin Anat 27:21021. 1356. 16. Stylianos K, Konstantinos G, Pavlos P, Aliki F. 2016. 25. Kim S, Dennis M, Holland J, Terrell M, Loukas M, Schober Brachial branches of the medial antebrachial cutaneous J. 2018. The anatomy of forearm free flap phalloplasty for nerve: A case report with its clinical significance and a short transgender surgery. Clin Anat 31:145-151. review of the literature. J Neurosci Rural Pract 7:443-446. 26. Seror P. 2004. Medial antebrachial cutaneous nerve 17. Winnie AP, Collins VJ. 1964. The Subclavian Perivascular conduction study, a new tool to demonstrate mild lower Technique of Brachial Plexus Anesthesia. Anesthesiology brachial plexus lesions. A report of 16 cases. Clin 25:353-363. Neurophysiol 115:2316-2322. 18. Ranganath A, Srinivasan KK, Iohom G. 2014. Ultrasound 27. Jung MJ, Byun HY, Lee CH, Moon SW, Oh MK, Shin H. guided axillary brachial plexus block. Med Ultrason 2013. Medial antebrachial cutaneous nerve injury after 16:246-2451. brachial plexus block: two case reports. Ann Rehabil Med 19. Benedikt S, Parvizi D, Feigl G, Koch H. 2017. Anatomy of 37:913-918. the medial antebrachial cutaneous nerve and its significance 28. Tran DQ, Pham K, Dugani S, Finlayson RJ. 2012. A in ulnar nerve surgery: An anatomical study. J Plast prospective, randomized comparison between double-, Reconstr Aesthet Surg 70:1582-1588. triple-, and quadruple-injection ultrasound-guided axillary 20. Lowe JB, 3rd, Maggi SP, Mackinnon SE. 2004. The brachial plexus block. Reg Anesth Pain Med 37:248-253. position of crossing branches of the medial antebrachial 29. Deleuze A, Gentili ME, Marret E, Lamonerie L, Bonnet cutaneous nerve during cubital tunnel surgery in humans. F. 2003. A comparison of a singlestimulation lateral Plast Reconstr Surg 114:692-696. infraclavicular plexus block with a triple-stimulation 21. Stahl S, Rosenberg N. 2002. Surgical treatment of painful axillary block. Reg Anesth Pain Med 28:89-94. neuroma in medial antebrachial cutaneous nerve. Ann Plast 30. Kapral S, Jandrasits O, Schabernig C, Likar R, Reddy B. Surg 48:154-158; discussion 8-60. et al. 1999. Lateral infraclavicular plexus block vs. axillary block for hand and forearm surgery. Acta Anaesthesiol Scand 43:1047-1052.

11-De Breuker.indd 86 12/01/2021 12:16 (Acta Anaesth. Belg., 2020, 71, 87-92) Efficacy and safety of spontaneous ventilation using dexmedetomidine-remifentanil in children undergoing direct laryngoscopy for surgical procedures : a single cohort observational study

T. Bongaerts (*), S. Maes (*), S. Aerts (*), D. Wildemeersch (*,**), V. Saldien (*,**)

Abstract : Introduction : Supraglottoplasty is a procedure Introduction performed in the management of laryngomalacia. Using direct laryngoscopy, redundant supraglottic tissue is Supraglottoplasty is a procedure performed in resected. Spontaneous ventilation is preferred to allow the management of laryngomalacia. Laryngomalacia both adequate oxygenation and a clear surgical working is an anomaly of the larynx which results in space. Because of its ability to maintain respiratory drive supraglottic airway collapse during inspiration. It and airway patency, dexmedetomidine is increasingly is the most common cause of inspiratory stridor in used during endoscopic airway procedures. However, literature regarding its use for direct laryngoscopy with neonates and infants. Stridor typically intensifies surgical intervention is limited. with feeding, agitation and in the supine position. Research goals : Evaluation of the efficacy and safety Most cases can be managed with conservative of spontaneous ventilation using dexmedetomidine- treatment, but surgical intervention is required for remifentanil in children undergoing direct laryngoscopy severe laryngomalacia, with incidences varying with surgical intervention. from 10-20%. Signs of severe disease include poor Methods : In this single cohort observational study, weight gain with failure to thrive, constant stridor, children undergoing supraglottoplasty were recruited. deep intercostal and xyphoid retraction, episodes of After inhalational induction, IV access was gained. Rapid respiratory distress, apparent life threatening events bolus dosing of dexmedetomidine 4 µg kg-1 was followed -1 -1 and cor pulmonale (1, 2). by dexmedetomidine at 2 µg kg h and remifentanil Most of the new supraglottoplastic techniques at 0.3 µg kg-1 min-1 as maintenance. During surgery, are minimally invasive endoscopic techniques, spontaneous ventilation was maintained. Results : 26 children were enrolled. Adverse airway designed at resecting redundant supraglottic tissue events were frequent : laryngospasm in 12/26 (46%), responsible for airway collapse. Common sites apnea in 10/26 (38%). However, only 5/26 (19%) include the arytenoids, the aryepiglottic folds or resulted in desaturation. Hemodynamic adverse events the epiglottis (3). In pediatric patients, general were : bradycardia 12/25 (48%), tachycardia 2/24 (8%), anesthesia is required. It is a challenging procedure hypertension 6/15 (40%) and hypotension 2/15 (13%). for both the anesthesiologist and the surgeon, as they Missing hemodynamic baseline data in some subjects have to share the same target, i.e. the airway. Goals precluded analysis. Body movement occurred in 3/26 of anesthetic airway management are maintaining (12%), cough in 2/26 (8%). Total procedure time lasted oxygenation and ventilation, while allowing opti- 53 (SD 18) min consisting of 9 (SD 3) min induction mal surgical conditions. When considering the time, 29 (SD 12) min surgical time and 16 (SD 8) min wake up time. Surgery was successful in all cases with

high surgery satisfaction scores. T. Bongaert, MD ; S. Maes, MD ; S. Aerts, MD ; D. Conclusion : The combination of dexmedetomidine- Wildemeersch, MD ; V. Saldien, MD, PhD remifentanil can be a safe and efficacious method to (*) Department of Anesthesia, Antwerp University Hospital, allow spontaneous ventilation in children undergoing 2650, Edegem, Belgium direct laryngoscopy with surgical intervention. Although (**) Faculty of Medicine and Health Science, University of Antwerp, 2610, Wilrijk, Belgium laryngospasm and apnea were frequent, serious Corresponding author : Bongaerts Toon, MD, Department of consequences such as desaturation were rare. Rapid Anesthesiology, Antwerp University Hospital, Wilrijkstraat bolus dosing of dexmedetomidine was hemodynamically 10, 2650, Edegem, Belgium. well tolerated and allowed short induction times with E-mail : [email protected] good surgical conditions. Paper submitted on Jun 4, 2020 and accepted on Jun 4, 2020. Conflict of interest : none. Keywords : Airway management ; anesthesia ; dexme- An abstract of the results was presented at the SARB online detomidine ; laryngoscopy ; laryngomalacia Graduation Day 2020 on 23-05-2020.

12-Bongaert.indd 87 12/01/2021 12:17 88 t. bongaert et al.

target population, typically neonates and infants Methods with a known endangered airway, one can easily understand the difficult challenge. Different modes of Ethics approval ventilation are being used : mechanically controlled ventilation through a small caliber tube or with This trial has been approved by the jet ventilation, intermittent apnea or spontaneous ethics committee of our institution (17/31/360, ventilation. However, mechanically controlled 04/09/2017, Antwerp University Hospital, Edegem, ventilation interferes with the surgical work space, Belgium). It was registered at ClinicalTrials.gov while intermittent apnea limits the available (NCT03426579). Written informed consent was surgical time. Therefore, spontaneous ventilation is obtained from parents or the legal guardians. For the technique of choice in experienced centers. This children who were able to read or understand study can be achieved either with inhalational anesthetics procedure, a child specific informed consent was or with total intravenous anesthesia (TIVA). provided. Dexmedetomidine is increasingly used in various types of diagnostic and therapeutic endo- Trial design scopic airway procedures, including flexible bronchoscopy, rigid bronchoscopy and direct In this single cohort observational study, laryngoscopy. It has been described in pediatric children scheduled for direct laryngoscopy with anesthesia both as the main anesthetic or as an surgical intervention between 1 month and 8 years adjunct (4-10). Dexmedetomidine is a highly old were enrolled after obtaining written informed specific α2-adrenergic agonist with sedative, anal- consent. Exclusion criteria were refusal of informed gesic, anxiolytic and amnestic properties. A key consent, known allergy for dexmedetomidine or advantage is its ability to maintain respiratory remifentanil. Children were withdrawn from the drive and airway patency with increasing levels of study if the pulse oxygen saturation was lower than sedation. Moreover, it does not show synergistic 85% at the start of anesthesia. Given the lack of effects on respiration when combined with other literature the study was set up as a pilot study. anesthetics, unlike for example remifentanil or The studied patient cohort was initially propofol. To allow surgical intervention during examined as a pilot study to evaluate the correlation direct laryngoscopy, a second anesthetic agent is between sedation depth monitoring and adverse needed. The combination of dexmedetomidine airway reflexes during intravenous anesthesia with and remifentanil has been described as a safe and dexmedetomidine and remifentanil in children effective method during flexible bronchoscopy (9, undergoing direct laryngoscopy for surgical proce- 10). dures. However, the study was prolonged and is still A high incidence of adverse airway reflexes running. Given the lack of data regarding airway remains during endoscopic airway procedures. management during supraglottoplasty, the safety Depending on the level of anesthesia, laryngo- and efficacy outcomes of the initial pilot cohort are spasm or apnea might occur. Given the lower reported in this paper. functional residual capacity and the relative higher oxygen consumption in infants and young Anesthesia children, hypoxemia might quickly follow (11). Literature regarding anesthetic management for Standard monitoring included 3-lead electro- supraglottoplasty is scarce. Based on the respiratory cardiography, non-invasive blood pressure measure- characteristics of dexmedetomidine, it might be ment, peripheral pulse-oximetry, respiratory rate an ideal anesthetic to allow adequate oxygenation and rectal temperature. and ventilation with spontaneous respiration, while Inhalational induction of anesthesia was allowing optimal surgical working conditions. performed using sevoflurane. After induction of anesthesia a 24 Gauge intravenous catheter was Research goals placed in a peripheral vein. Baseline parameters The aim of the study was to evaluate the containing heart rate, oxygen saturation and efficacy and safety of spontaneous ventilation blood pressure were noted at this timepoint, using dexmedetomidine and remifentanil in defined as T0. Children received midazolam children undergoing direct laryngoscopy with sur- 0.05 mg kg-1 intravenously. A rapid bolus dose gical intervention, more specifically in children of dexmedetomidine 4 µg kg-1 was administered undergoing supraglottoplasty for laryngomalacia. intravenously. Administration of sevoflurane was

12-Bongaert.indd 88 12/01/2021 12:17 efficacy and safety of spontaneous ventilation using dexmedetomidine-remifentanil 89 stopped and maintenance of anesthesia was normal distribution, quantitative variables are pre- accomplished with dexmedetomidine infused at sented as mean [standard deviation (SD)]. In case 2 µg kg-1 h-1 and remifentanil at 0.3 µg kg-1 min-1. of non-normal distribution, quantitative variables Dexmedetomidine and remifentanil were titrated are presented as median (25th quartile-75th quartile at the anesthesiologists discretion based on the [25th-75th)]. Qualitative variables are presented as evaluation of respiratory and hemodynamic clinical incidence [percentage (%)]. sings and parameters. Before start of surgery, direct laryngoscopy was performed by the surgeon and 1 Results spray of 10% lidocaine was delivered to anesthetize the vocal cords. Afterwards all children received a A total of 26 patients were enrolled from nasopharyngeal tube for oxygen supplementation. November 2017 until January 2020. Children had a Spontaneous ventilation was maintained. In case median age of 5 months (25th-75th 3-18,5) with a mean of laryngospasm and desaturation, intravenous weight of 6.45 kg (SD 4.62). Fourteen patients were propofol was used as rescue medication at the male and 12 were female. Baseline characteristics, anesthesiologists discretion. After surgery all including demographic characteristics, are shown in children were monitored on the pediatric intensive table 1. care unit (PICU) or in the post anesthesia care unit (PACU). Table 1 Demographics and baseline parameters Outcome measures Variable Demographic patient details were obtained a from digital patient record C2M (Cegeka Medical Age (months) 5 (3-18.5) Health Care Systems, Hasselt, Belgium). Heart Gender (n, male/female) 14/12 rate, non-invasive blood pressure, pulse oxygen Weight (kg) 6.45 (4.62)b saturation and respiratory rate were recorded at Heart rate T0 (beats/min) 117 (18) b Innovian (Draeger Medical Systems, Andover, Syst blood pressure T0 (mmHg) 81 (20) b MA, USA). Airway and hemodynamical adverse b events and procedural times (induction, surgical, Diastolic blood pressure T0 (mmHg) 46 (17) wake up and total) were assessed. Induction time Oxygen saturation T0 (%) 99 (1) b was defined from bolus dosing until placement of aMedian (25th-75th), bMean (SD). the direct laryngoscope for surgical intervention. Surgical conditions were assessed by the surgeon Table 2 using surgery satisfaction scores : 1 = extremely dissatisfied ; 2 = not satisfied but able to manage ; Surgery satisfaction scores 3 = extremely satisfied. At the end of surgery Variable Incidence (%) total cumulative dose of dexmedetomidine and Surgery satisfaction score remifentanil were calculated. Rescue propofol 1 2/26 (8%) doses and vasoactive medications were noted. Adverse events were defined as : 2 4/26 (15%) - Apnea or suspension of breathing 3 20/26 (77%) - Laryngospasm 1 = extremely dissatisfied ; 2 = not satisfied but able to manage ; 3 = - Hypoxemia as saturation (SpO2) lower than extremely satisfied. 90% for more than 30 seconds. - Body movement Table 3 - Cough Intraoperative variables - Bradycardia or tachycardia as a change of 30% beyond baseline heart rate (T0). Variable Mean (SD) - Hypertension or hypotension as a change Procedure time (min.) 53 (18) of 30% beyond baseline systolic or diastolic Surgical time (min.) 29 (12) blood pressure (T0). Induction time (min.) 9 (3) Statistics Wake up time (min.) 16 (8) The Shapiro-Wilk test was used to test Dexmedetomidine cumulative dose (µg) 40.23 (18.62) quantitative variables for normality. In case of Remifentanil cumulative dose (µg) 82.55 (56.27)

12-Bongaert.indd 89 12/01/2021 12:17 90 t. bongaert et al. Table 4 Adverse airway and hemodynamic events

Airway event Incidence (%) Hemodynamic event Incidence (%) Apnea 10/26 (38) Bradycardia 12/25 (48) Desaturation 5/26 (19) Tachycardia 2/24 (8) Cough 2/26 (8) Hypertension 6/15 (40) Laryngospasm 12/26 (46) Hypotension 2/15 (13) Body movement 3/26 (12) Atropine 6/26 (23) Propofol 2/26 (8)

Fig. 1. - Mean parameters during different time points during surgery. T0 = baseline ; T1 = lidocaine spray ; T2 = nasopharyngeal intubation ; T3 = start of surgery ; T4 = end of surgery ; T5 = awakening.

Surgery was successfully performed in all hemodynamic adverse events are summarized in 26 cases. The surgeon rated operative conditions table 4. as 3 (extremely satisfied) in 20/26 (77%), as 2 (not satisfied but able to manage) in 4/26 (15%), Discussion and 1 (extremely dissatisfied) in 2/26 (8%) cases. Total procedure time lasted a mean 53 min (SD 18) Airway management during surgical endo- consisting of a mean 9 min (SD 3) induction time, scopic airway procedures is challenging for both the a mean 29 min (SD 12) surgical time and a mean anesthesiologist and the surgeon. Debate exists about 16 min (SD 8) wake up time. Surgical satisfaction different ventilation strategies, but spontaneous scores and intraoperative variables are summarized ventilation is often preferred to allow both adequate in tables 2 and 3. oxygenation and a clear surgical working space. A Adverse airway events occurred in a number variety of inhalational and intravenous anesthetic of patients : laryngospasm in 12/26 (46%), cough regimens has been described. Nevertheless, litera- in 2/16 (8%), apnea in 10/26 (38%). Of these ture regarding spontaneous ventilation for surgical adverse airway reflexes only 5/26 (19%) resulted in interventions during direct laryngoscopy, for desaturation with only 2/26 (8%) requiring a rescue example supraglottoplasty, is scarce (5, 6, 12-15). propofol dose. Body movement occurred in 3/26 Because of its ability to maintain respiratory (12%). drive and airway patency, dexmedetomidine Recorded hemodynamically adverse events is increasingly used during endoscopic airway were : bradycardia 12/25 (48%), tachycardia 2/24 procedures, either as the main anesthetic or as an (8%), hypertension 6/15 (40%) and hypotension adjunct (4-10). However, most studies investigate 2/15 (13%). In 6/26 (23%) atropine was given for the use of dexmedetomidine during flexible or bradycardia. No vasoactive drugs were required rigid bronchoscopy. Only three case reports for tachycardia, hypertension or hypotension. described dexmedetomidine to allow spontaneous Unfortunately, some parameters at baseline where ventilation during laryngeal surgery in children (4- missing, which made further analyses impossible : 6). 11 missing blood pressures at baseline, 2 missing In our study, 26 children successfully underwent heart rates at baseline. Mean hemodynamic para- supraglottoplasty for severe laryngomalacia under meters at different time points during the procedure TIVA with the association of dexmedetomidine and are shown in figure 1. Adverse airway events and remifentanil. Adverse airway reflexes were frequent

12-Bongaert.indd 90 12/01/2021 12:17 efficacy and safety of spontaneous ventilation using dexmedetomidine-remifentanil 91 with laryngospasm in 12/26 (48%) and apnea in the timing of the baseline blood pressure. Baseline 10/26 (38%). Lower incidences of laryngospasm (0- blood pressure was measured after inhalational 15%) and apnea (5-21%) are found in other studies induction with sevoflurane, when IV access was investigating TIVA and spontaneous ventilation established. Blood pressure can be expected to be for endoscopic airway procedures (7-11, 15). A relatively hypotensive at this timepoint as compared couple of reasons might have led to these higher to an awake blood pressure measurement. Anesthesia incidences. First, in our study, laryngospasm and is often induced with only oxygen saturation and apnea were either present or absent, with no further heart rate monitoring in these types of procedures. distinction in the degree of severity. Desaturation, Since agitation can lead to worsening of inspiratory defined as oxygen saturation lower than 90% for stridor with possible respiratory distress in patients more than thirty seconds, was seen in only 5/26 with severe laryngomalacia, everything that might (19%) cases. Only 2 cases (8%) required a rescue upset the child is done only after induction. This propofol dose to overcome laryngospasm. Body includes blood pressure measurement. Therefore movement (12%) and cough (8%) was infrequent. we opted to take the blood pressure after IV access These numbers are rather low compared to other was established as baseline. studies : desaturation (8-38%), cough (26-54%) The effect of dexmedetomidine bolus dosing and body movement (10-58%) (7-11, 15). This on blood pressure is typically described as a biphasic might indicate that the majority of adverse airway pattern. An initial short hypertensive period is events witnessed in our study population were followed by a longer lasting hypotensive period. mild. This is also evidenced by the high surgical This effect is caused by peripheral vasoconstriction satisfaction rating : 20/26 (77%) and 4/26 (15%) through direct action on peripheral α2-adrenergic cases received a surgical satisfaction rating of receptors in vascular smooth muscle, followed by respectively extremely satisfied or dissatisfied but inhibition of sympathetic outflow through stimu- manageable. Only 2/26 (8%) cases were rated as lation of central α2-adrenergic receptors (18, 19). extremely dissatisfied. One would expect lower This pattern was less pronounced in our cohort, scores with severe, repetitive or prolonged episodes as seen in figure 1. Possibly, the association of of airway problems hampering surgery. Lastly, our remifentanil might have blunted the initial hyper- study population has specific characteristics. Mean tensive effect. The wash-out of sevoflurane might age was low, with most children still being infants, have counterbalanced the hypotensive period. and all subjects required surgical intervention for Rapid bolus dosing of dexmedetomidine known laryngomalacia. By definition, this group has and the association of dexmedetomidine allowed a severe form of laryngeal anomaly, making airway feasible anesthesia times. Induction lasted a management more difficult. The airway at baseline mean 9 min. This included direct laryngoscopy is more precarious in these children as compared to with lidocaine spraying of the vocal cords and a more heterogenous group studied for other airway nasopharyngeal intubation. Wake up time lasted a procedures, including older children and children mean 16 min. Compared to a mean surgical time of undergoing diagnostic procedures. Therefore they 29 min and total procedural time of a mean 53 min, can be expected to be more susceptible to adverse these are acceptable and realistic anesthesia times. airway events. Especially when considering the precarious patient Dexmedetomidine dosing is generally recom- population, which necessitates adequate time taken mended as a loading dose over 10 minutes followed to safely perform anesthesia and maintain airway by a continuous infusion. In daily practice this can control. be inconvenient and unrealistic, especially in a high- A limitation to our study is the small sample turnover setting with shorter surgical procedures. size of the investigated cohort. As little data is Rapid bolus administration has also proven to be found regarding anesthesia, airway management safe in two studies (16, 17). In our study a rapid and adverse event rates for direct laryngoscopy bolus was followed by a continuous infusion. with surgical intervention in children, it was set Dexmedetomidine dosing was hemodynamically up as a pilot study. Secondly, there were some well tolerated. Bradycardia (48%) was a frequent protocol violations, especially regarding baseline event, with half of these cases necessitating atropine blood pressure measurement. For unclear reasons, administration. With or without atropine, effects were in 11/26 patients a baseline blood pressure was transient with little effect on other hemodynamic not noted. Therefore in these cases nothing can be parameters. Hypertension (40%) also frequently said about the occurrence of either hypertension or occurred, although this might also be explained by hypotension, as this was defined as a blood pressure

12-Bongaert.indd 91 12/01/2021 12:17 92 t. bongaert et al. respectively higher or lower than baseline. Possibly, 6. Quezado Z.M., Groblewski J.C., Gelfand H.J. and since induction is one of the most critical moments Shah R.K. 2009. Dexmedetomidine and proprofol in complex microlaryngeal surgery in infants. Int J Pediatr for airway management, attention was lost to the Otorhinolaryngol.73(9):1311-2. application of the blood pressure cuff. Despite these 7. Cai Y., Li W. and Chen K. 2013. Efficacy and safety of limitations, our results show dexmedetomidine spontaneous ventilation technique using dexmedetomidine for rigid bronchoscopic airway foreign body removal in and remifentanil can be an efficacious and safe children. Paediatr Anaesth.23(11):1048-53. method to allow spontaneous ventilation in children 8. Chen K.Z., Ye M., Hu C.B. and Shen X. 2014. Dexme- undergoing direct laryngoscopy with surgical detomidine vs remifentanil intravenous anaesthesia and spontaneous ventilation for airway foreign body removal in intervention. Further research will be necessary to children. Br J Anaesth.112(5):892-7. determine the ideal intravenous agents. 9. Li X., Wang X., Jin S., Zhang D. and Li Y. 2017. The safety and efficacy of dexmedetomidine-remifentanil in children undergoing flexible bronchoscopy: A retrospective dose- Conclusion finding trial. Medicine (Baltimore).96(11):e6383. 10. Zhang H., Fang B. and Zhou W. 2017. The efficacy of Total intravenous anesthesia with dexmede- dexmedetomidine-remifentanil versus dexmedetomidine- tomidine and remifentanil can be a safe and propofol in children undergoing flexible bronchoscopy: A retrospective trial. Medicine (Baltimore).96(1):e5815. efficacious method to allow spontaneous ventilation 11. Chen L.H., Zhang X., Li S.Q., Liu Y.Q., Zhang T.Y. and in children undergoing direct laryngoscopy with Wu J.Z. 2009. The risk factors for hypoxemia in children surgical intervention. Although laryngospasm and younger than 5 years old undergoing rigid bronchoscopy for foreign body removal. Anesth Analg.109(4):1079-84. apnea frequently occurred, serious adverse events 12. Stern Y., McCall J.E., Willging J.P., Mueller K.L. and such as desaturation were rare. Rapid bolus dosing Cotton R.T. 2000. Spontaneous respiration anesthesia of dexmedetomidine was hemodynamically well for respiratory papillomatosis. Ann Otol Rhinol Laryn- tolerated and allowed short induction times. Lite- gol.109(1):72-6. 13. Werkhaven J.A. 2004. Microlaryngoscopy-airway mana- rature regarding anesthetic and airway management gement with anaesthetic techniques for CO(2) laser. during direct laryngoscopy with surgical inter- Paediatr Anaesth.14(1):90-4. vention is limited. Further research is needed to 14. Mausser G., Friedrich G. and Schwarz G. 2007. Airway management and anesthesia in neonates, infants and compare the effect of different inhalational and children during endolaryngotracheal surgery. Paediatr intravenous anesthetic regimens. Anaesth.17(10):942-7. 15. Malherbe S., Whyte S., Singh P., Amari E., King A. and Ansermino J.M. 2010. Total intravenous anesthesia and References spontaneous respiration for airway endoscopy in children- -a prospective evaluation. Paediatr Anaesth.20(5):434-8. 1. Ayari S., Aubertin G., Girschig H., Van Den Abbeele T., 16. Jooste E.H., Muhly W.T., Ibinson J.W., Suresh T., Denoyelle F., Couloignier V. and Mondain M. 2013. Damian D., Phadke A. and Davis P.J., et al. 2010. Acute Management of laryngomalacia. Eur Ann Otorhinolaryngol hemodynamic changes after rapid intravenous bolus Head Neck Dis.130(1):15-21. dosing of dexmedetomidine in pediatric heart transplant 2. Bedwell J. and Zalzal G. 2016. Laryngomalacia. Semin patients undergoing routine cardiac catheterization. Anesth Pediatr Surg.25(3):119-22. Analg.111(6):1490-6. 3. Del Do M., Song S.A., Nesbitt N.B., Spaw M.C., 17. Dawes J., Myers D., Gorges M., Zhou G., Ansermino J.M. Chang E.T., Liming B.J. and Camacho M., et al. and Montgomery C.J. 2014. Identifying a rapid bolus dose 2018. Supraglottoplasty surgery types 1-3: A practical of dexmedetomidine (ED50) with acceptable hemodynamic classification system for laryngomalacia surgery. Int J outcomes in children. Paediatr Anaesth.24(12):1260-7. Pediatr Otorhinolaryngol.111:69-74. 18. Tobias J.D. 2007. Dexmedetomidine: applications in 4. Seybold J.L., Ramamurthi R.J. and Hammer G.B. 2007. pediatric critical care and pediatric anesthesiology. Pediatr The use of dexmedetomidine during laryngoscopy, bron- Crit Care Med.8(2):115-31. choscopy, and tracheal extubation following tracheal 19. Su F. and Hammer G.B. 2011. Dexmedetomidine: pediatric reconstruction. Paediatr Anaesth.17(12):1212-4. pharmacology, clinical uses and safety. Expert Opin Drug 5. Shukry M. and Kennedy K. 2007. Dexmedetomidine Saf.10(1):55-66. as a total intravenous anesthetic in infants. Paediatr Anaesth.17(6):581-3.

12-Bongaert.indd 92 12/01/2021 12:17 (Acta Anaesth. Belg., 2020, 71, 93-104) Comparative study on different routes of intraoperative administration of Ropivacaine to reduce postoperative pain in gynecological laparoscopy - a double blinded randomized trial

B. Clinck, M. Hoffman, M. Jans, H. Coppejans, A. De Bruyn, Y. Jacquemyn, K. Wouters, S. Morrison, V. Saldien

Abstract : Both postoperative pain and side effects of Background pain therapy result in significant morbidity for patients. Opioids are associated with nausea and respiratory Inadequate pain therapy after surgery can depression, which limit their use in day hospital. For a result in a slower recovery, longer revalidation laparoscopy, local anesthetics can be administered locally and more postoperative complications. It can by infiltration of the trocar entry points or by instillation also trigger a chronic pain syndrome. In day care in the intraperitoneal cavity. Both methods have proven effective in lowering postoperative pain scores compared procedures, like most gynaecological laparoscopies to a placebo. This double blinded randomized trial are, postoperative pain can be a cause of prolonged aimed to compare these two local anesthetic techniques, hospital stay. Standard therapy with Paracetamol when used either alone or combined in a patient, with and non-steroidal anti-inflammatory drugs is often respect to postoperative pain scores and the use of rescue insufficient and usually requires addition of an pain medication after a gynaecological laparoscopic opioid (1-3). procedure. Between September 2017 and August 2019, Opioids are associated with side effects like 56 patients (ASA I or II) who came for a gynaecological nausea and respiratory depression which limits laparoscopy in day hospital were enrolled. They were their use in day hospital (3-6). The need for opioids randomly divided into three groups. Group A received limits the procedures that can be performed in both trocar site infiltration and intraperitoneal instillation day hospital setting. There has been research into of Ropivacaine ; group B received intraperitoneal other pain management techniques and consensus instillation alone and group C received only trocar site infiltration. Further anesthesia and postoperative dictates the use of multimodal or multi-faceted pain analgesia were standardized. Following surgery, a strategies. This includes the use of local anesthetics, questionnaire was completed at 1, 2, 6 and 24 hours and like Ropivacaine (7-11). again after 1 week. Pain scores during the first 24 hours For a laparoscopy, Ropivacaine can be locally and one week after surgery were not significantly different administered either by infiltration of the surgical between the three study groups (P-value of 0,55). The use entry points or by instillation in the abdominal of pain medication, however, was significantly higher cavity. These techniques have both been proven to in the study group that received both infiltration and have analgesic effects when compared to a placebo, instillation of a local anesthetic, compared to the groups but nevertheless, opioids have not been abandoned who received either instillation or infiltration (P-value of in clinical practice (2, 3, 12-25). The administration 0,027). routes have also been compared to one another with This could indicate that a maximal dose of local infiltration giving better postoperative analgesia and anesthetics distributed over more than one administration route is less effective in treating pain, than the full dose instillation giving less shoulder pain (19). of 3 mg/kg given in one area. More research is necessary B. Clinck, M.D. ; M. Hoffman, M. Jans, H. Coppejans, M.D. ; to investigate this possibility. A. De Bruyn, M.D., Y. Jacquemyn, M.D., Ph.D. ; K. Wouters, S. Morrison, M.D. ; V. Saldien, M.D., Ph.D. Keywords : Laparoscopy ; anesthetic ; infiltration ; Antwerp University Hospital Belgium, Wilrijkstraat 10, 2650 instillation ; pain. Edegem, Belgium Corresponding author : Doctor B. Clinck, Department of Anesthesia, Antwerp University Hospital Belgium, Wil- rijkstraat 10, 2650 Edegem, Belgium. E-mail : [email protected] Paper submitted on May 29, 2020 and accepted on Jun 4, 2020. Conflict of interest: none.

13-Clinck.indd 93 12/01/2021 12:18 94 b. clinck et al. In literature, a combination of the two techniques was studied in comparison to a placebo but results were contradictory (1, 26) In this double- blinded randomized clinical trial, we aimed to evaluate these two local anesthetic techniques when used alone or in combination, with respect to postoperative pain scores after a gynecological procedure and the need for rescue opioid therapy. We expected the postoperative pain scores and need for pain medication to be lower in the group of study patients who received both infiltration and instillation of a local anesthetic.

Methods and trial design

Ethical Permission

On 07/08/2017, approval was provided by the Universiteit Antwerpen Ethisch Comité, Drie Eikenstraat 655, 2650 Edegem. The committee chairperson was Cras Patrick and the acting chairperson was Michielsen Peter. The Belgian registration number was B300201733092. Fig. 2. — Patient population. Sample size study and power estimates approaches. For the group of peritoneal instillation, The sample size was calculated before study we used the study of Roy KK et al. (25). For the initiation based on the program G*Power and trocar site injection we used the study of Sugihara the results of other studies that compared similar et al. (24), and for the combination of both methods we used the study conducted by Pellicano et al. (23). To reach a statistical power of 80%, each group had to contain 18 patients. We aimed for a study population of 60 patients, 20 patients per group. Figure 1 displays the recruitment. Unfortunately, the amount of drop-outs was larger than anticipated, mostly because of the intraoperative intra-abdominal use of Hyalobarrier©. This is a polysaccharide used to prevent adhesions, of which the effects on postoperative pain are unknown[ (27, 28). The decision was made to continue with the 56 included patients. Figure 2 displays the study population as divided into the three study groups. In the screening procedure the inclusion criteria and exclusion criteria were tested. Information about the presence of preoperative pain and risk factors for postoperative nausea and vomiting were collected.

Eligibility criteria

Inclusion criteria were : (1) Elective gynaecologic laparoscopy for benign conditions in day care hospital at Antwerp Fig. 1. – Recruitment. University Hospital

13-Clinck.indd 94 12/01/2021 12:18 study on different routes of intraoperative administration of ropivacaine 95 (2) Female patient and intraperitoneal instillation, group B received (3) 18-70 years intraperitoneal instillation alone and group C (4) ASA I-II received only infiltration of the trocar incision sites. (5) informed consent signed by the patients Further intraoperative analgesia was provided by Fentanyl 2 µg/kg and a non-steroidal anti- Exclusion criteria were : inflammatory drug at the time of induction. Para- (1) Hypersensitivity to anesthetics of the cetamol 1 g was administered approximately 30 amide-type, contraindications for non-steroidal minutes before the end of the procedure. anti-inflammatory drugs Following surgery, a questionnaire on pain (2) Obesity, BMI > 35 scores, quality of recovery and opioid-related side (3) Patients with chronic use of analgesics / effects was completed at 1, 2, 6 and 24 hours and patients with chronic pain again after 1 week (addendum 1). (4) Patients with an abnormal liver function The primary objective was to compare post- (5) Pregnancy (this means we only include operative pain in patients who underwent a gyne- patients who use contraceptives, are sterilized, cological laparoscopy and were given a local dose have a negative dipstick testing or a negative serum of Ropivacaine through either port site infiltration, testing or patients who are postmenopausal) intra-abdominal instillation or a combination of (6) Surgical or anesthetic complications both techniques. The other objective of this study (7) Use of pain relieving products during the was to assess the occurrence of pain therapy related study that are not included in the study protocol side effects. (8) Concomitant participation in another study The primary outcomes were Numeric Rating that would affect the results of this study. Scale (NRS) scores (29) for pain at 1, 2, 6 and 24 hours postoperatively. The secondary outcomes Consent, compliance and withdrawal were NRS at 1 week postoperatively, the use of pain medication, occurrence of shoulder pain, First, it is important to note that the use of local itch, postoperative nausea and vomiting, need for anesthesia is standard practice for gynaecological antiemetics, problems of urination and the level of laparoscopy at the hospital. However, the optimal activity. Pain scores were obtained both in rest and administration route is not yet determined. Further during an exercise (like walking or coughing). information, related to the study protocol, was Our hypothesis was that a combination of provided to all patients preoperatively in order to trocar entry point infiltration and intraperitoneal obtain an informed consent. Participation was free instillation of Ropivacaine would result in lower and a participant could withdraw from the trial pain scores after surgery with fewer side effects like without consequences for his or her perioperative nausea, vomiting, pruritus and shoulder pain. care. All study material related to this particular patient would be destroyed. All patient data col- Three-arm double-blinded RCT lected were anonymously processed using a num- bering system. Eligible patients were recruited preoperatively by a student or an anesthetist. After having signed the Study design and setting informed consent, the participants were randomised to one of the three study groups by means of This double-blinded three-arm randomized sealed envelopes, opened by the anesthetist in the clinical trial was conducted at the University Hos- operating room. From this moment forward they pital of Antwerp, Belgium. Between September were assigned a study number which was used for 2017 and August 2019, 56 healthy women (ASA I further identification. or II) who presented for gynaecological laparoscopy Anesthetic management was standardized. in day hospital were enrolled. Routine intraoperative monitoring devices were A written informed consent was obtained used, which included ECG, pulse oximetry and from all participants at the moment of inclusion, non-invasive blood pressure. All patients were pre- which was usually on the day of the surgery. The oxygenated with 100% oxygen before induction, patients were randomly assigned to one of the which was achieved with a combination of three study groups (18-20 patients per group). Fentanyl 2µg/kg, Propofol TCI Marsh model 6µg/ Using Ropivacaine 0.75% in a total dose of 3mg/ ml (titration based on effect) and Rocuronium 0.5 kg : group A received trocar entry point infiltration mg/kg. For maintenance of anesthesia Propofol TCI

13-Clinck.indd 95 12/01/2021 12:18 96 b. clinck et al. March model 4 µg/ml (8-12 mg/kg/h) was used with All five postoperative study times, 1, 2, 6 and titration based on both effect and hemodynamic 24 hours and one week, were concluded in the data parameters. All study patients were given a non- analysis. Our primary outcome parameters were steroidal anti-inflammatory drug at the time of pain scores, in rest and during activity, at 1, 2, 6 and induction. Paracetamol (1g, IV) was administered 24 hours. Secondary outcomes consisted of NRS 30 minutes before the end of the procedure. pain scores after 1 week, the use of pain medication, At the end of the surgical procedures, Ropi- shoulder pain, pruritus, nausea, vomiting, use of vacaine, in a total dose 3 mg/kg, was administered antiemetic drugs, urinary problems, activity level according to the randomly assigned study group, and satisfaction. as indicated by the sealed envelope. Group A Data analysis patients received both intraperitoneal instillation of Ropivacaine 2 mg/kg in 200 ml of NaCl 0,9% Data were analyzed using R software. Pain and trocar entry point infiltration with Ropivacaine scores in the first 24 hours after surgery were studied 1 mg/kg in 20 ml of NaCl 0,9%. The trocar sites of using a linear mixed effects model, with fixed group B patients were infiltrated with Ropivacaine 3 effects time, study group, use of pain medication mg/kg in 20 ml of NaCl 0,9% and group C patients and the interaction between time and study group. received 3 mg/kg of Ropivacaine in 200 ml of NaCl A random intercept was added for every participant 0,9% in the abdominal cavity. to adjust for the correlation between pain scores of After surgery the patients were transferred every individual patient. Based on the AIC (Akaike’s to the recovery and later to the day hospital ward, Information Criterium) and likelihood ratio tests, where the NRS was used to assess pain scores and to time was considered categorical. When interaction guide the standard postoperative analgesia regimen. between time and group was non-significant, it was All patients were given Paracetamol 1 g every 6 removed from the model. hours. If pain scores remained higher than three, Pain scores will be affected by pain medication. additional pain therapy consisted of Tramadol 100 For that reason a correlation between those variables mg and Alizapride 50 mg. If re-evaluation after 30 was assumed and, thus, in the analysis of pain minutes showed pain scores persisting above three, scores, a correction for the use of pain medication an anesthesiologist was consulted and patients was included. received intravenous morphine. Patients were discharged from the hospital when all the discharge Results criteria were met according to standard operating procedures. Prescriptions for continuing the use of Table 1 shows a descriptive overview of the Paracetamol at home and for Naproxen 500 mg 3 results of the questionnaires. times a day were provided. Table 2 shows a relation between pain At 1, 2, 6, 24 hours and 1 week after surgery, medication and pain scores (P-value = 0,011). This a researcher contacted the patient to complete the matter was a constant throughout the statistical questionnaire. Both the patient and researcher analysis. Differences in pain scores in rest during conducting the survey postoperatively were blinded the first 24 hours postoperatively between the three to the intraoperative Ropivacaine administration study groups were not significant (P-value = 0,55). route used during surgery. All responses were Figure 3 shows a boxplot of the pain scores organized in a database of Open Clinica. found in the three study groups at the specific During the trial, there were no deviations from times. The study groups infiltration and both show this study protocol. lowest pain scores at 2 hours postoperatively. The relevance of this finding is unclear and further study Data management with larger study groups could be recommended. Pain scores in rest one week postoperatively All data was stored in Open Clinica, in a database connected to the University Hospital of Pain scores 1 week postoperatively are Antwerp. Only the researchers had access. Three compared between study groups in a linear regres- study groups were defined. Group A patients received sion model to be able to correct for the use of pain both instillation and infiltration of Ropivacaine and medication. will thus, from here on forth, be referred to as “both”. Table 3 shows no significant difference in pain Group B and C will be identified as infiltration and scores between the three study groups (P-value = instillation respectively. 0,88).

13-Clinck.indd 96 12/01/2021 12:18 study on different routes of intraoperative administration of ropivacaine 97

Table 1 Descriptive overview

1h 2h 6h 24h 1w Pain in Rest Instillation 2.60 (1.64) 2.75 (1.55) 2.65 (1.63) 2.40 (1.76) 0.89 (1.13) Infiltration 1.78 (1.77) 1.61 (1.79) 2.28 (1.74) 2.61 (2.55) 0.93 (1.16) Both 2.17 (2.04) 1.83 (2.18) 2.89 (2.70) 3.28 (2.44) 1.11 (1.78) Pijn during exercise Instillation 3.25 (1.92) 3.45 (1.47) 3.30 (1.56) 3.90 (2.02) 2.00 (2.09) Infiltration 1.94 (1.92) 2.22 (1.99) 3.22 (1.9) 4.28 (2.49) 2.00 (2.20) Both 2.61 (2.17) 2.22 (2.39) 3.78 (2.69) 4.67 (2.47) 2.39 (2.66) Shoulder pain Instillation 0/20 (0%) 4/20 (20%) 11/20 (55%) 12/20 (60%) 0/18 (0%) Infiltration 3/18 (17%) 2/18 (11%) 12/18 (67%) 8/18 (44%) 4/15 (27%) Both 4/18 (22%) 6/18 (33%) 15/18 (83%) 11/18 (61%) 1/18 (6%) Use of pain medication Instillation 3/20 (15%) 2/20 (10%) 8/20 (40%) 18/20 (90%) 7/18 (39%) Infiltration 3/18 (17%) 3/18 (17%) 6/18 (33%) 14/18 (78%) 8/15 (53%) Both 4/18 (22%) 3/18 (17%) 14/18 (78%) 18/18 (100%) 8/18 (44%) Nausea Instillation 0/20 (0%) 2/20 (10%) 2/20 (10%) 1/20 (5%) 0/18 (0%) Infiltration 2/18 (11%) 0/18 (0%) 1/18 (6%) 1/18 (6%) 2/15 (13%) Both 2/18 (11%) 2/18 (11%) 1/18 (6%) 2/18 (11%) 0/18 (0%) Vomiting Instillation 1/20 (5%) 0/20 (0%) 0/20 (0%) 0/20 (0%) 0/18 (0%) Infiltration 0/18 (0%) 0/18 (0%) 0/18 (0%) 1/18 (6%) 0/15 (0%) Both 0/18 (0%) 0/18 (0%) 0/18 (0%) 0/18 (0%) 0/18 (0%) Use of antiemetic drugs Instillation 1/20 (5%) 1/20 (5%) 1/20 (5%) 0/20 (0%) 0/18 (0%) Infiltration 1/18 (6%) 1/18 (6%) 0/18 (0%) 0/18 (0%) 1/15 (7%) Both 4/18 (22%) 3/18 (17%) 1/18 (6%) 1/18 (6%) 0/18 (0%) Pruritus Instillation 1/20 (5%) 1/20 (5%) 0/20 (0%) 2/20 (10%) 9/18 (50%) Infiltration 0/18 (0%) 0/18 (0%) 0/18 (0%) 1/18 (6%) 3/15 (20%) Both 0/18 (0%) 0/18 (0%) 0/18 (0%) 1/18 (6%) 9/18 (50%) Urinary problems Instillation 0/20 (0%) 4/20 (20%) 2/20 (10%) 2/20 (10%) 1/18 (6%) Infiltration 0/18 (0%) 1/18 (6%) 1/18 (6%) 2/18 (11%) 0/15 (0%) Both 0/18 (0%) 0/18 (0%) 1/18 (6%) 2/18 (11%) 1/18 (6%) Exercise Instillation 0/20 (0%) 7/20 (35%) 16/20 (80%) 19/20 (95%) 17/18 (94%) Infiltration 1/18 (6%) 7/18 (39%) 13/18 (72%) 16/18 (89%) 14/15 (93%) Both 1/18 (6%) 5/18 (28%) 9/18 (50%) 14/18 (78%) 17/18 (94%)

13-Clinck.indd 97 12/01/2021 12:18 98 b. clinck et al. Table 2 Pain scores in rest first 24 hours

Estimate (Std. Error) P-value (Intercept) 2.28 (0.37) < 0.001 Time (ref = 1h) 0.88 2h -0.08 (0.30) 0.79 6h 0.16 (0.31) 0.62 24h -0.01 (0.37) 0.98 Group (ref = Instillation) 0.55 Infiltration -0.51 (0.47) 0.28 Both -0.18 (0.47) 0.71 Fig. 4. — Pain scores during activity first 24 hours. Pain medication 0.79 (0.31) 0.011 Table 4 Pain scores during activity first 24 hours

Estimate (Std. Error) P-value (Intercept) 2.77 (0.40) < 0.001 Time (ref = 1h) 0.016 2h 0.06 (0.31) 0.85 6h 0.60 (0.33) 0.071 24h 1.19 (0.39) 0.003 Group (ref = Instillation) 0.56 Infiltration -0.54 (0.50) 0.28 Both -0.25 (0.50) 0.61

Fig. 3. — Pain scores in rest first 24 hours. Pain medication 0.64 (0.32) 0.050

Table 5 Table 3 Pain scores during an activity one week postoperatively Pain scores in rest 1 week after surgery Estimate (Std. Error) P-value Estimate (Std. Error) P-value (Intercept) 2.58 (0.68) < 0.001 (Intercept) 1.35 (0.40) 0.001 Group (ref = Instillation) 0.85 Group (ref = Instillation) 0.88 Infiltration -0.14 (0.81) 0.87 Infiltration -0.06 (0.48) 0.89 Both 0.34 (0.77) 0.67 Both 0.18 (0.45) 0.69 Pain medication -0.95 (0.65) 0.15 Pain medication -0.76 (0.39) 0.056

Pain scores during an activity during the first 24 Pain scores during an activity one week post- hours postoperatively operatively

As demonstrated in table 4, there is no Linear regression with covariates group and significant difference in pain scores during the pain medication showed no significant difference first 24 hours after surgery (P-value = 0,56). The in pain scores between the three groups (P-value = boxplot (Fig. 4) shows a similar pattern for study 0,85) one week after surgery, as illustrated in table group infiltration and both, with pain scores rising 5. over time, while pain scores in the instillation group remain a constant. The reason the tendency in first Use of pain medication during the first 24 hours two groups is not statistically significant could be postoperatively that the power of this study is insufficient due to too small study groups. Since the use of pain medication is a dicho- tomous outcome, a logistic mixed effect model

13-Clinck.indd 98 12/01/2021 12:18 study on different routes of intraoperative administration of ropivacaine 99

Table 6 Use of pain medication first 24 hours OR (95%CI) P-value (Intercept) 0.12 (0.04 - 0.32) < 0.001 Time (ref = 1h) < 0.001 2h 0.74 (0.25 - 2.21) 0.59 6h 6.19 (2.43 - 15.8) < 0.001 24h 73.5 (21.7 - 249) < 0.001 Group (ref = Instillation) 0.027 Infiltration 0.81 (0.29 - 2.30) 0.70 Both 3.08 (1.11 - 8.51) 0.030 Fig. 5. — Use of pain medication first 24 hours. was chosen, with fixed effects time, as categorical variable, group and interaction. A random intercept P-value = 0,030 and OR = 3,78 (95% CI 1,31 - 10,9), was added again. The use of pain medication P-value = 0,014). There was no difference between proved to be significantly different between the ‘infiltration’ and ‘instillation’ (P-value 0,70). three study groups (P-value = 0,027). A higher use of pain medication in the group ‘both’ was Use of pain medication one week postoperatively recorded, compared to the groups ‘instillation’ and ‘infiltration’ (respectively OR = 3,08 (1,11 - 8,51), Here, a Chi-square test was performed and there were no significant discrepancies between the three study groups. Table 7 Pain medication after 1 week Secondary outcomes Instillation Infiltration Both P-value Secondary outcomes were binary and consisted Pain medication 7/18 (39%) 8/15 (53%) 8/18 (44%) 0.71 of shoulder pain, pruritus, urinary problems, level

Fig. 6. — Binary secondary outcome parameters.

13-Clinck.indd 99 12/01/2021 12:18 100 b. clinck et al. 0.71 0.50 0.43 0.28 0.032 0.003 < 0.001 < 0.001 P-value OR 2.44 5.17 3.69 1.79 1.96 5.51) (1.08 - General (95%CI) (0.37 -0.5) (2.19-12.2) (1.58-8.63) (0.32-9.87) experience 0.13 0.71 0.70 0.053 < 0.001 < 0.001 < 0.001 P-value OR 7.79 38.4 0.78 0.27 Activity Ref = 2h (95%CI) (11.8-124) (3.11-19.5) (0.21-2.94) (0.07-1.02) 0.81 0.74 0.75 0.51 0.41 0.29 0.90 P-value OR 0.75 1.27 0.45 0.34 3.91) 5.68) 2.96) 2.55) (0.15- (0.28- (0.07- (0.05- Urinary Ref = 2h (95%CI) problems P-value OR Te weinig events Te Pruritus (95%CI) 0.53 0.44 0.98 0.28 0.85 0.090 0.040 0.14 P-value OR 0.63 0.25 0.17 1.04 5.13 (95%CI) (0.15-2.67) (0.05-1.24) (0.03-0.92) (0.04-28.9) (0.27-97.2) Combination Table 8 Table 0.64 0.91 0.99 0.73 0.95 0.054 0.028 0.080 P-value OR 0.61 0.08 0.03 0.98 4.41 (0-0.68) (95%CI) (0-24279) (0-19250) (0.08-4.74) (0.01-1.04) Medication P-value Overview statistical analysis binary secondary outcome parameters OR Te weinig events Te (95%CI) Vomiting 0.99 0.98 0.84 0.78 0.84 0.98 0.63 0.99 0.081 P-value OR 1.03 1.20 1.30 0.97 1.68 0.20 Nausea (95%CI) (0.17-6.26) (0.20-7.41) (0.21-8.22) (0.10-9.86) (0.03-1.21) (0.21-13.62) lue 0.18 0.12 0.91 0.68 0.66 0.062 < 0.001 < 0.001 < 0.001 P-va OR 2.18 1.06 2.92 0.82 30.73 13.88 (95%CI) (0.69-6.87) (0.35-3.24) (0.95-8.97) (0.31-2.14) (9.55-98.84) (3.69-52.15) Shoulder pain Shoulder Time (ref = 1h) (ref Time 2h 6h 24h = (ref Group Instillation) Infiltration Both Pain medication Antiemetic Medication x Group Time

13-Clinck.indd 100 12/01/2021 12:18 study on different routes of intraoperative administration of ropivacaine 101 of activity, the use of antiemetic drugs and the 20 ml of Ropivacaine 7,5% which he injects when presence of nausea and vomiting. For the two latter a finishing with his sutures. However, this is in correction for the use of antiemetic drugs was made accordance with the maximum dose of 3 mg/kg for and for the shoulder pain the pain medication was a patient with a weight of 50 kg and, based on this taken into account. In the analysis, a logistic mixed study, could be insufficient to treat patients with a effects model with random intercept was applied. higher weight. This is an interesting item for further Similarly for satisfaction (ordinal in 4 categories) an research. ordinal logistic mixed model with random intercept was used. In all models the interaction between The study design the study groups was not significant and could be abandoned. Some variables could not be analyzed Other points of discussion are the following. as there were too few events. As there was a low First, a wide variety of laparoscopic procedures appearance of nausea, vomiting and antiemetic use were included. This heterogeneity ensures that these three items were also analyzed as a combined the population in this study represents the actual variable. population in a gynecological surgery department As illustrated in table 8 we could not find but also may have influenced the postoperative a significant difference in regard to secondary pain scores. The more invasive the procedure, the outcome measurements, between the three study higher the pain scores that could be expected. The groups. Boxplot figures (Fig. 6) show a lower rate procedures were also performed by many different of shoulder pain and higher level of activity in the surgeons. This could give a bias. groups ‘infiltration’ and ‘instillation’ compared Secondly, some of the patients we enrolled for to ‘both’. These findings, however, were not the study were given intraperitoneal Hyalobarrier significant. by the gynecologist during surgery. Because of the unknown effect this has on pain scores, especially Discussion in combination with intraperitoneal instillation of sodium chloride 0,9% or Ropivacaine, depending The results on the study group, we decided to exclude these patients. No adverse events were recorded during Thirdly, NRS is a validated method to record the course of this study. As mentioned above, pain scores but pain still remains a subjective matter. some results had similar patterns or trends who, This is the first study in its genre that does not in this analysis, were not significant but might be bring a placebo into the comparison. In the sample significant if tested in a larger study population. size calculation we might have overestimated the We expected the group ‘both’ to have expected effect and therefore have worked with a better pain scores and less pain medication use sample size that is too small. As a result some of compared to the other groups or not to differ at all. the clinical relevant differences between the study Statistical analysis, however, showed that patients groups turned out not statistically significant. Despite who received both infiltration and instillation of this, we were able to demonstrate a higher use of Ropivacaine needed more pain medication in the pain medication in the study group that received first 24 hours after surgery. An explanation could be both infiltration and instillation of Ropivacaine. the lower doses of Ropivacaine used in the areas of Further research combining both administration administration when combined in one patient. The techniques does not seem useful. maximum dose of Ropivacaine of 3 mg/kg was, in the “both” study group, divided over the two Conclusion administration routes. 2 mg/kg was instilled in the abdominal cavity whereas 1 mg/kg was infiltrated at Local anesthetics, whether they are the trocar entry points. This would seem to suggest administered by infiltration, instillation or for that the given dose of the local anesthetic is more nerve blocks, have become a very important part important than the manner or than the location of multimodal pain therapy. Their use will continue where it is administered. This was also suggested to grow. More research is needed into the best by Malhotre et al.(30). administration modalities. In our hospital we always infiltrate trocar entry In this clinical trial we could not find a points with a local anesthetic. Generally this means difference in pain scores between the two tested we provide the surgeon with a sterile ampoule of administration routes of Ropivacaine, infiltration

13-Clinck.indd 101 12/01/2021 12:18 102 b. clinck et al. or instillation, nor was there a significant additional 12. A. Manan, A. A. Khan, I. Ahmad, M. Usman, T. Jamil, and beneficial effect of a combination of the two M. A. Sajid. 2020. Intraperitoneal bupivacaine as post- laparoscopic cholecystectomy analgesia. Journal of the techniques. In contrast, patients who received both College of Physicians and Surgeons Pakistan, vol. 30, no. infiltration and instillation of Ropivacaine needed 1, 9-12. significantly more pain medication in the 24 hours 13. M. Toleska et al. 2018. Efficacy of Intraperitoneal Bupiva- caine on Pain Relief After Laparoscopic Cholecystectomy. after surgery. This could indicate that a maximum PRILOZI, vol. 39, no. 1, pp. 123-129. dose of local anesthetics distributed over more than 14. [R. Sharan, M. Singh, A. Kataria, K. Jyoti, V. Jarewal, and one administration route is less effective in treating R. Kadian. 2018. Intraperitoneal instillation of bupivacaine and ropivacaine for postoperative analgesia in laparoscopic pain, than the full dose of 3 mg/kg given in one cholecystectomy. Anesthesia : Essays and Researches, vol. area. More research is necessary to investigate this 12, no. 2, p. 377. possibility. 15. S. Jain, N. Nazir, S. Singh, and S. Sharma. 2018. A prospective randomised controlled study for evaluation of high-volume low-concentration intraperitoneal bupivacaine for post-laparoscopic cholecystectomy analgesia. Indian References Journal of Anaesthesia, vol. 62, no. 2, 109-114. 16. A. Kahokehr, T. Sammour, S. Srinivasa, and A. G. Hill. 1. B. Hilvering, W. A. Draaisma, J. D. W. van der Bilt, R. M. 2011. Systematic review and meta-analysis of intra- Valk, K. E. Kofman, and E. C. J. Consten. 2011. Randomized peritoneal local anaesthetic for pain reduction after clinical trial of combined preincisional infiltration laparoscopic gastric procedures. British Journal of Surgery, and intraperitoneal instillation of levobupivacaine for vol. 98, no. 1, 29-36. postoperative pain after laparoscopic cholecystectomy. 17. R. Patel, J. C. A. Carvalho, K. Downey, M. Kanczuk, P. British Journal of Surgery, vol. 98, no. 6, 784-789. Bernstein, and N. Siddiqui. 2017. Intraperitoneal Instillation 2. K. Sevensma, T. Schleichert, C. Schwickerath, A. of Lidocaine Improves Postoperative Analgesia at Cesarean Shoemaker, and C. Miller. 2019. A randomized double Delivery. Anesthesia & Analgesia, vol. 124, no. 2, 554-559. blinded study to determine the effectiveness of utilizing 18. B. Y. Oh et al. 2016. Analgesic efficacy of ropivacaine intraperitoneal bupivacaine : Does it reduce postoperative wound infusion after laparoscopic colorectal surgery. opioid use following laparoscopic appendectomy? Ameri- Annals of Surgical Treatment and Research, vol. 91, no. 4, can Journal of Surgery, vol. 217, no. 3, 479-482. 202. 3. S. Safari, F. Rokhtabnak, S. Djalali Motlagh, M. Ghanbari 19. [S. M. Kaushal-Deep, A. Anees, S. Khan, M. A. Khan, Garkani, and A. Pournajafian. 2020. Effect of intraperitoneal and M. Lodhi. 2018. Randomized controlled study of bupivacaine on postoperative pain in laparoscopic bariatric intraincisional infiltration versus intraperitoneal instillation surgeries. Surgery for Obesity and Related Diseases, vol. of standardized dose of ropivacaine 0.2% in post-lapa- 16, no. 2, 299-305. roscopic cholecystectomy pain : Do we really need high 4. P. Bhakta et al. 2016. Incidence of postoperative nausea doses of local anesthetics – time to rethink!. Surgical and vomiting following gynecological laparoscopy : A Endoscopy, vol. 32, no. 7, 3321-3341. comparison of standard anesthetic technique and propofol 20. T. Tam, G. Harkins, L. Wegrzyniak, S. Ehrgood, A. infusion. Acta Anaesthesiologica Taiwanica, vol. 54, no. Kunselman, and M. Davies. 2014. Infiltration of bupivacaine 4,108-113. local anesthetic to trocar insertion sites after laparoscopy : 5. M. T. Ganter et al. 2014. The length of stay in the post- A randomized, double-blind, stratified, and controlled trial. anaesthesia care unit correlates with pain intensity, nausea Journal of Minimally Invasive Gynecology, vol. 21, no. 6, and vomiting on arrival. Perioperative Medicine, vol. 3, no. 1015-1021. 1, p. 10. 21. [S. Loizides, K. S. Gurusamy, M. Nagendran, M. Rossi, G. P. 6. H. S. Smith, E. J. Smith, and B. R. Smith. 2012. Post- Guerrini, and B. R. Davidson. 2014. Wound infiltration with operative nausea and vomiting. Annals of palliative medi- local anaesthetic agents for laparoscopic cholecystectomy. cine, vol. 1, no. 2, 94-102. Cochrane Database of Systematic Reviews, vol. 2014, no. 7. “Management of acute perioperative pain - UpToDate.” 3. John Wiley and Sons Ltd, 12-Mar-2014. [Online]. Available : https://www.uptodate.com/contents/ 22. G. J. Choi, H. Kang, C. W. Baek, Y. H. Jung, and D. R. Kim. management-of-acute-perioperative-pain. [Accessed : 01- 2015. Effect of intraperitoneal local anesthetic on pain Mar-2020]. characteristics after laparoscopic cholecystectomy. World 8. R. Chou et al. 2016. Management of postoperative pain : A Journal of Gastroenterology, vol. 21, no. 47, p. 13386. clinical practice guideline from the American pain society, 23. M. Pellicano, F. Zullo, C. di Carlo, E. Zupi, and C. Nappi. the American society of regional anesthesia and pain 1998. Postoperative pain control after microlaparoscopy in medicine, and the American society of anesthesiologists’ patients with infertility : A prospective randomized study. committee on regional anesthesia, executive committee, Fertility and Sterility, vol. 70, no. 2, 289-292. and administrative council. Journal of Pain, vol. 17, no. 2, 24. [M. Sugihara et al. 2018. Does local infiltration anesthesia 131-157. on laparoscopic surgical wounds reduce postoperative 9. H. Kehlet and J. B. Dahl. 2003. Anaesthesia, surgery, and pain? Randomized control study. Reproductive Medicine challenges in postoperative recovery. Lancet, vol. 362, no. and Biology, vol. 17, no. 4, 474-480. 9399. Elsevier Limited, 1921-1928. 25. K. K. Roy, M. Subbaiah, M. Naha, S. Kumar, J. B. Sharma, 10. K. S. Ladha, E. Patorno, K. F. Huybrechts, J. Liu, J. P. and N. Jahagirdar. 2014. Intraperitoneal bupivacaine for Rathmell, and B. T. Bateman. 2016. Variations in the use pain relief after minilaparoscopy in patients with infertility. of perioperative multimodal analgesic therapy. Anesthe- Gynecology and Obstetrics, vol. 289, no. 2, 337-340. siology, vol. 124, no. 4, 837-845. 26. N. Gupta, V. Dadhwal, and S. Mittal. 2012. Combined 11. T. O’Brien et al. 2017. European Pain Federation posi- intraperitoneal instillation and port site infiltration of local tion paper on appropriate opioid use in chronic pain anaesthetic (bupivacaine) for postoperative analgesia in management. European Journal of Pain (United Kingdom), women undergoing daycare diagnostic gynaecological vol. 21, no. 1, 3-19. laparoscopy. European Journal of Obstetrics and Gyne-

13-Clinck.indd 102 12/01/2021 12:18 study on different routes of intraoperative administration of ropivacaine 103 cology and Reproductive Biology, vol. 161, no. 1. Elsevier Minimally Invasive Gynecology, vol. 25, no. 7. Elsevier Ireland Ltd, 109-110. B.V., pp. 1194-1216. 27. G. Ahmad, H. O’Flynn, A. Hindocha, and A. Watson. 2015. 29. S. Safikhani et al. 2018. Response scale selection in adult Barrier agents for adhesion prevention after gynaecological pain measures : results from a literature review. Journal of surgery. Cochrane Database of Systematic Reviews, vol. Patient-Reported Outcomes, vol. 2, no. 1. 2015, no. 4. John Wiley and Sons Ltd, 30-Apr-2015. 30. N. Malhotra, C. Chanana, K. K. Roy, S. Kumar, V. Riwari, 28. S. Farag, P. F. Padilla, K. A. Smith, M. L. Sprague, and S. and J. B. Sharma. 2007. To compare the efficacy of two E. Zimberg. 2018. Management, Prevention, and Sequelae doses of intraperitoneal bupivacaine for pain relief after of Adhesions in Women Undergoing Laparoscopic operative laparoscopy in gynecology. Gynecology and Gynecologic Surgery : A Systematic Review. Journal of Obstetrics, vol. 276, no. 4, pp. 323-326.

Addendum 1 : Questionnaire

Comparative study on the intraoperative administration of Ropivacaine in the context of pain relief after a gynaecological laparoscopy, a double blind randomized trial v1 (25062017) 2017-2018 Content : – Information sheet – Questionnaires To be completed after surgery at the following times : – 1h after surgery → at ……………. Hour – 2h after surgery → at ……………. Hour – 6h after surgery → at ……………. Hour – 24h after surgery → date : … / …./ ………… at ……………. Hour – ○ On this day you will be contacted to give us your answers of the first 24h 1 week after surgery → date : … / … / …………. at ………….. Hour – ○ On this day you will be contacted again Thank you!

Medical questionnaire clinical trial Patient number: Research centre: University Hospital Antwerp Title research project: Comparative study on the intraoperative administration of Ropivacaine in the context of pain relief after a gynaecological laparoscopy, a double blind randomized trial

INFORMATION SHEET

Dear Mrs, Dear Ms, We thank you for your participation in our study and we hope your surgery went well. In this folder you’ll find a number of questionnaires. The questions are identical but the time they have to be completed is different for each sheet. We have a questionnaire for 1h, 2h, 6h, 24h and 1 week after surgery. We ask you to complete these questionnaires as correctly as possible and to respect the moments stated on the individual sheets. The first page might already be filled in by the nurses of the recovery room. We would like to explain a little more about the pain questions. You will be asked about your pain experience at rest and during an effort (like coughing). You must quote this experience on a scale from 0 to 10 with 0 is no pain and 10 is the worst possible pain, the maximum of pain you can theoretically feel. You are not allowed to use decimals. The figure below could help you score your pain.

13-Clinck.indd 103 12/01/2021 12:18 104 b. clinck et al.

24 after surgery you will be contacted by a member of our research group. This person will ask you about the answers on page 3-6. After a week, you will be contacted a second time so we can include the last page in our research. We sincerely thank you for your participation in our study and wish you a good recovery after the surgery.

TO COMPLETE 1H AFTER SURGERY

1) PAIN Quote your pain on a scale from 0 to 10 with 0 = no pain and 10 = the worst pain (see information sheet)

– Pain score during rest score – Pain score during an effort (like coughing) score

Do you experience shoulder pain? Yes No Did you take any medication to lower your pain? Yes No

- If Yes, which pain killer and what dose? ………………………………………………..…… ……………………………………………………..

2) NAUSEA Do you experience nausea? Yes No Have you vomited? Yes No Did you take any medication for your nausea Yes No

- If Yes, which medicament and what dose? ……………………………………………………… ……………………………………………………… 3) OTHER Do you experience itching? Yes No Do you have trouble peeing/urinating? Yes No Have you already done an activity? (walking, cycling, …) Yes No

What is your general experience? Very bad Bad Neutral Good Very good

13-Clinck.indd 104 12/01/2021 12:18 (Acta Anaesth. Belg., 2020, 71, 105-112) The systemic inflammatory response to cardiopulmonary bypass: a narrative review

B. Keunen (*), M. Van de Velde (*,**), L. Al Tmimi (*,**)

Abstract : The cardiopulmonary bypass (CPB) is a the inflammation to improve clinical outcomes standard technique in many cardiac interventions. It may after CPB, with variable degrees of success. The elicit a systemic inflammatory response that, fortunately, present review will provide an overview of the often remains subclinical and resolves quickly after the pathophysiology of SIRS after CPB and its possible CPB stops. However, in some patients, this systemic therapeutic options in adult patients. inflammatory response may be extreme and progress to a systemic inflammatory response syndrome (SIRS). The latter may give rise to significant adverse events Pathophysiology of SIRS and even death. Possible provocations for this syndrome include a contact of blood with the artificial CPB-surface, Warren and colleagues stated that inflammation ischemia-reperfusion injury, endotoxemia and surgical is a non-specific response of the body to harmful trauma. stimuli (3). Although this response aims to be Many authors have tried to curb this inflammation protective, in some instances, an exaggerated and improve clinical outcomes, with variable degrees response damages the host more than the insult. of success. This narrative review provides readers Cardiac surgery on CPB and prolonged anesthesia with a brief and up-to-date knowledge concerning are important triggers of this inflammatory reaction the pathophysiology and therapeutic options of this (2). The inflammatory response after CPB can be syndrome. divided into two fundamental phases, the “early” and “late” (3). Key words : Heart lung machine ; cardiopulmonary Early phase bypass ; systemic inflammatory response syndrome ; SIRS. The early phase starts with the surgical trauma and the contact of blood with the synthetic material of the CPB-circuit (2, 3, 4) as demonstrated in figure Background 1. The contact and complement system are almost instantly activated. These two systems stimulate Since its introduction in 1953, the cardio- leukocytes, endothelium, and platelets to interact pulmonary bypass (CPB) has become a standard and produce cytokines, oxygen-free radicals, and technique used in many cardiac interventions (1). inducible nitric oxide synthase (iNOS) (3). It helps the cardiac surgeon to perform complex Furthermore, white blood cells, endothelium, reconstructions in a bloodless operating field and platelets express adhesion molecules like while providing the patient with physiological support. However, the CPB may be associated with several adverse reactions such as activation of the B. Keunen, MD ; M. Van de Velde, MD, PhD, EDRA ; L. Al tmimi, MD, PhD complement system, impaired hemostasis, trauma to (*) Department of Anesthesiology, University Hospitals blood components, and the systemic inflammatory Leuven, Leuven, Belgium response syndrome (SIRS). Key triggers for SIRS (**) Department of Cardiovascular Sciences, KULeuven, Leuven, Belgium are the contact of blood with the artificial CPB- Corresponding author : Dr. Bram Keunen, Department of surface, ischemia-reperfusion injury, endotoxemia, Anesthesiology, University Hospitals Leuven, Herestraat and operative trauma. This activates the innate 49, 3000 Leuven, Belgium and the adaptive immune system together with E-mail : [email protected] the hemostatic cascades and results in significant Paper submitted on May 31, 2020 and accepted on June 04, adverse events with major organ dysfunction and 2020. even death (2). Many authors have tried to restrict Conflict of interest : none.

14-Keunen.indd 105 12/01/2021 12:18 106 b. keunen et al. CPB. Subsequently, we will explain, in detail, the role of these cellular and protein systems on the development of SIRS. After surgical trauma, the contact system activates factor XII, proenzyme of factor XIIa. Factor XIIa is a crucial initiator of multiple pathways (3) : Essentially, factor XIIa activates the intrinsic coagulation pathway (Fig. 1). It also converts high molecular weight kininogen (HK) into bradykinin. The latter produces vasodilatation and non-vascular smooth muscle contraction. Moreover, factor XIIa cleaves prekallikrein into kallikrein, which is an essential activator of neutrophils, fibrinolysis, and a positive feedback loop to increase factor XIIa (2, Fig. 1. — Pathophysiology of the early phase of Systemic 3, 7). Inflammatory Response Syndrome (SIRS) after cardio- Simultaneously with the contact system, the pulmonary bypass (CPB). complement system is activated mainly due to the interaction of blood with the synthetic material selectins, integrins and cellular adhesion molecules of the CPB-circuit. Normally, there are three (CAMs). These adhesion molecules provide conceivable pathways in complement induction : “rolling” and “binding” of white blood cells to the the classic, alternative and mannose-binding lectin endothelium, after which they transmigrate and pathway. The alternative pathway is probably the release their inflammatory granules in non-vascular most crucial in the development of SIRS after CPB tissues with a resulting end-organ inflammation (3, 8, 9). Activation of this pathway leads to the (2, 3, 5, 6), as demonstrated in figure 2. Likewise, formation of a membrane attack complex (MAC) the hemostatic and fibrinolytic systems become and the activation of neutrophils (2, 3). Many overactivated (Fig. 1), leading to a consumptive proteins of the MAC, such as C3a and C5a, are coagulopathy and micro-emboli (3, 7). potent anaphylatoxins and chemoattractants. They All the processes mentioned above represent a induce capillary leakage and vasodilatation (2, 3, 8). brief overview of the early phase of the SIRS after Furthermore, C3a stimulates platelet aggregation,

Fig. 2. — Pathophysiology of extravasation of white blood cells in non-vascular tissues with release of their inflammatory granules. The horseshoe-shaped lines are a schematic representation of selectins and CAMs. E: Endothelial cells, CAMs: cellular adhesion molecules, P: platelets, RBC: red blood cell, WBC: white blood cell.

14-Keunen.indd 106 12/01/2021 12:18 sirs after cardiopulmonary bypass 107 while C5a induces neutrophilic aggregation and the disruption of endothelial integrity and the pro- binding of white blood cells to the endothelium (9). inflammatory environment (2, 3, 8). This activation of the complement and contact Platelets are massively activated by the CPB, system leads to a complex interaction of white blood resulting in a 30-50% decrease in counts and cells, platelets, and endothelium (Fig. 2). White substantial functional impairment (2, 3). When blood cells are initially diminished in count due to activated, platelets connect to fibrinogen and lead hemodilution but they increase substantially during to bridge formation. They adhere to each other and and after CPB. This increase is mainly due to the monocytes by selectins. Afterwards, they release activation of neutrophils and monocytes, while the their granules, which contain chemoattractants, total count of lymphocytes is sharply reduced (2, 3). coagulation proteins, and vasoactive substances

Neutrophils are the principal cytotoxic effector like thromboxane A2. The thrombocytopenia and cells. Activation of neutrophils after the initiation of increased bleeding times return to normal within CPB could be attributed to the following pathways. 4-12 hours after initiation of the CPB (3). First, through activation of the complement and Besides this inflammatory reaction, there is an contact systems. Second, by direct interaction with overactivation of the coagulation and fibrinolytic endothelial cells. Last, other mediators such as system. An extensive discussion of the coagulation thrombin, tumor necrosis factor a (TNF-a), heparin, system would lead us beyond the scope of this endotoxins, and histamine play an essential role (3) review. However, we highlight factor IIa (thrombin) (5). After transmigration through the endothelium, formation by either the intrinsic or the extrinsic the neutrophils release cytotoxic granules in pathway. During CPB, the intrinsic pathway is extravascular tissues and initiate a respiratory burst. thought to be activated mainly by contact with the This results in increased microvascular permeability, synthetic surfaces, while the extrinsic pathway interstitial edema, thrombosis and parenchymal cell particularly by trauma (3). death (3, 4, 5). Of note, thrombin has numerous essential On the other hand, monocytes are activated functions in the pathophysiology of SIRS after more slowly than neutrophils and exert their effects CPB. It converts fibrinogen to fibrin, activates blood mainly during the early postoperative period. They platelets and stimulates endothelial cells. Besides produce inflammatory cytokines and play a vital its hemostatic effects, thrombin is considered as a role in hemostasis by forming platelet-monocyte crucial inflammatory mediator (3, 9). conjugates and producing tissue factor (3). The fibrinolytic system, at last, limits the In contrast to the neutrophils and monocytes, excessive formation of thrombus. Plasminogen is the concentration of lymphocytes declines during activated by tissue plasminogen activator (tPA) to CPB. This decline is also associated with a plasmin. Plasmin cuts down fibrin strands into small diminished phagocytosis capacity and an inhibition degradation products with thrombus resolution. In of the monocytes’ ability to present antigens to physiological circumstances, tPA is slowly released lymphocytes. The result is a weakened cellular by the endothelium of damaged vessels. However, immune response with higher infection risks that in cardiac surgery, the release of tPA continues may remain for three to seven days postoperatively extensively around the pericardial wound, which (3, 8). leads to an overactivation of the fibrinolytic system. As mentioned before, the leukocytes interact Moreover, the CPB aggravates this phenomenon by with endothelial cells and platelets after stimulation reintroducing previously aspirated blood (3). by pro-inflammatory cytokines. The endothelial The early phase processes diminish during cells start to express selectins and CAMs (Fig. 2). CPB. This is called a “passivation” process and Selectins are low-affinity cell adhesion molecules is due to the covering of the CPB-inner circuit by that mediate the “rolling” of platelets and leukocytes. human proteins, rendering it more biocompatible In addition, CAMs bind neutrophils, monocytes, (3). and platelets more covalently to the endothelium than selectins and facilitate the transmigration Late phase to non-vascular tissues (3, 6). Furthermore, in physiological situations, endothelial cells influence Three factors contribute to the late phase : the vascular tone and capillary permeability by ischemia-reperfusion injury, endotoxemia, and releasing multiple mediators such as nitric oxide heparin-protamin interactions. (NO), endothelin-1, and noradrenaline (3). This effect is intensely disordered during CPB due to

14-Keunen.indd 107 12/01/2021 12:18 108 b. keunen et al. Ischemia-reperfusion injury ischemia induced by the aortic clamp particularly affects the heart and lungs. Second, the lungs are the As we already know, the aortic cross- only organs to receive the entire cardiac output and clamp induces ischemia in the heart and lungs its inflammatory reaction. Finally, neutrophils are (2, 3, 10). Reperfusion after release of this clamp trapped in the small pulmonary vascular circulation provokes a second inflammatory response with with significant congestion (14). increased capillary permeability, interstitial At last, other frequent complications of the edema, leukocytosis, coagulopathy and end-organ SIRS are postoperative delirium, acute kidney dysfunction (2, 3, 10). injury, intestinal hypoperfusion, liver failure, and even multi-organ failure with death (2, 3). Endotoxins Therapeutic interventions Endotoxins are lipopolysaccharide molecules from the cell wall of Gram-negative bacteria. Multiple therapies have been investigated Splanchnic vasoconstriction leads to a resulting for the prevention of SIRS after CPB. However, ischemia of enteric mucosa which permits gut few studies have shown clinically significant translocation of these Gram-negative bacteria. The implications. Therefore, we restricted our review endotoxins activate complement via the alternative to the following therapies : corticosteroids, statins, pathway and stimulate the release of cytokines and aprotinin, and technical adaptations. NO (2, 3, 9, 11). Corticosteroids Heparin-protamin interactions Notable hope was established on corticosteroids Heparin is used to prevent clotting during CPB. as potent anti-inflammatory drugs. Before 2009, most It potentiates the functions of antithrombin III and, literature suggested corticosteroids might attenuate by this, indirectly inhibits the effects of thrombin. the increase of inflammatory markers and possibly However, this still leads to progressive thrombin improve pulmonary gas exchange with a reduced generation with a consumptive coagulopathy need for postoperative inotropic support (16). Ho (3). Reversal of heparin with protamine induces et al. demonstrated in a meta-analysis that low- activation of the classic pathway of complement, dose corticosteroid treatment diminished the risk of leading to further inflammation (3, 8, 9). POAF and mechanical ventilation. Unfortunately, the available studies at that moment were small Clinical consequences of SIRS in sample size with significant heterogeneity (17). Another meta-analysis performed by Dieleman The CPB always elicits a systemic inflammatory et al. confirmed the reduction of POAF (-40%) response that, fortunately, often remains subclinical but failed to show a significant reduction in and resolves quickly after stopping of the CPB. cardiopulmonary complications or mortality. With However, in more extreme cases, a SIRS develops. the availability of beta-blockers and amiodarone SIRS’ clinical consequences may be characterized with fewer side-effects and excellent efficiency in by fever and variable vital organ dysfunction. the treatment of POAF, they stated no indication for A common cardiac complication of SIRS corticosteroids and highlighted the need for large is postoperative atrial fibrillation (POAF). This randomized controlled trials (RCT) (13). Later on, can lead to several adverse events like stroke, the effects of corticosteroids for the prevention of low cardiac output syndrome and hemodynamic systemic inflammation have been examined in two instability. All these events may prolong length of well designed and powered trials : the DECS and stay in the intensive care unit (ICU) and hospital (3, SIRS-trial (16, 18). 12, 13). Furthermore, SIRS often induces pulmonary In the DECS-trial (16), almost 4500 patients dysfunction with lung edema, pulmonary infections, received either 1 mg kg-1 dexamethasone or and prolonged duration of mechanical ventilation placebo, after the induction of anesthesia. The (3, 14, 15). primary endpoint was a composition of death, The lungs, more than any other organs, seem myocardial infarction, stroke, renal or respiratory to be vulnerable to CPB-related injury. Apostolakis failure within 30 days. The study failed to show et al. suggested three possible explanations : As the a significant reduction of this primary endpoint CPB provides circulatory support to all organs, the (7% vs. 8.7%, p-value 0.07). Subgroup analysis,

14-Keunen.indd 108 12/01/2021 12:18 sirs after cardiopulmonary bypass 109 however, revealed a statistically significant benefit both are corticosteroids, they have different phar- in younger patients (age <65 years) (relative risk macodynamic profiles. Dexamethasone has a potent (RR), 0.65 ; 95% CI, 0.44-0.96 ; P=.03) and a anti-inflammatory effect with less glucocorticoid trend towards harm in older patients (age >80 secondary effects than methylprednisolone (22). years). Also, there was a consistent pattern of an Furthermore, the different dosing schemes render it improved pulmonary condition in all patients, difficult to compare both trials. However, we still which translated into a shorter ICU and hospital believe that there is a potential role of dexamethasone length of stay. Furthermore, there was a significant in young patients undergoing cardiac surgery with reduction in postoperative infections and delirium. CPB. Yet, this needs to be further investigated in a Finally, the incidence of PAOF was similar between well-designed RCT. groups. Of note, four years later, the same authors demonstrated in a retrospective study a higher Statins incidence of late re-thoracotomy due to tamponades in the dexamethasone-group (19). Statins are selective, competitive inhibitors The SIRS trial (18) enrolled 7507 high risk of 3-hydroxy-3-methyl-glutaryl-coenzyme A re- patients with EuroSCORE ≥ 6 (European System ductase and are mainly used for lowering low for cardiac Operative Risk evaluation Score) (20) density lipoprotein-cholesterol (23, 24). However, to receive either methylprednisolone, 250 mg statins also exert a cholesterol-independent, anti- at the induction of anesthesia plus 250 mg at the inflammatory effect after cardiac surgery on initiation of CPB, or placebo. They concluded that CPB, a characteristic that was already known in methylprednisolone in this high-risk group did not atherosclerotic diseases (24). decrease the composite risk of death, myocardial Several studies investigated the biochemical injury, stroke, renal failure, and respiratory failure influences of statins on systemic inflammation after at 30 days. However, they noticed a significantly CPB. In 2009, Morgan et al. demonstrated in a increased risk of myocardial injury in the steroid systematic review that statins reduce postoperative group. Patients with signs of myocardial injury had pro-inflammatory cytokines. Nevertheless, these a significantly increased relative risk of death at results should be interpreted with caution due to 30 days compared with those without myocardial the limited numbers of patients in the included injury. A posthoc analysis showed that this increase studies. Moreover, some of these studies had in myocardial injury was due to non-Q-wave methodological weaknesses (25). Chee et al. stated infarctions. They stated that the DECS trial did not that the principal anti-inflammatory properties of retain this significant side effect due to an absence of statins on CPB are exerted through inhibition of the mandate measurement of biomarkers. Furthermore, neutrophil transendothelial migration. Furthermore, no reduction in POAF was noticed (18). they found preoperative high-dose statin therapy A recent meta-analysis (including the resulted in significantly lower levels of troponins DECS and SIRS-trial) by Dvirnik and colleagues and urinary neutrophil gelatinase-associated lipo- concluded that steroid administration had an calin, demonstrating less cardiac and renal injury unclear impact on mortality and increased the risk (24). The clinical benefit, however, remains a of myocardial injury. They also found a reduction topic for discussion. Several trials have shown in POAF. The latter was interpreted with caution the potential role of statins in reducing POAF like since the two large trials found no difference. The the ARMYDA-3 trial (26). Two hundred patients infection risk and length of stay in ICU and hospital scheduled for elective cardiac surgery on CPB, were significantly lower (21). with no previous history of statin treatment or atrial Nonetheless, in our opinion, a few funda- fibrillation, were included. The authors concluded mental differences need to be highlighted between that atorvastatin 40 mg/day, initiated seven days the DECS- and the SIRS-trial. First, as already before surgery, significantly reduced the incidence mentioned, the SIRS-trial included patients only of POAF with a shorter hospital length of stay (26). with an EuroSCORE ≥ 6, which is a different Kuhn et al. confirmed these findings in a meta- population than those patients who would possibly analysis of more than 90.000 patients undergoing benefit from the intervention (young patients cardiac surgery. However, most studies were with a robust inflammatory response) along with retrospective and a significant part of the included the DECS trial. Second, the SIRS-trial used 500 patients underwent off-pump coronary bypass mg of methylprednisolone instead of 1 mg kg-1 grafting (23). Deghani et al. found a statistically dexamethasone in the DECS-group. Although significant reduction of POAF after cardiac valve

14-Keunen.indd 109 12/01/2021 12:18 110 b. keunen et al. surgery when administering 40 mg atorvastatin In 2013, Meybohm et al. added nuance in a 3 days preoperatively and continuing 5 days meta-analysis that aprotinin might be associated postoperatively (27). with increased mortality in low- and intermediate- Less evidence is available regarding all other risk cardiac interventions but not in high-risk cardiac clinical outcomes. The meta-analysis of Kuhn et surgery. Moreover, they suggested aprotinin might al. concluded preoperative statins resulted in lower probably be beneficial in high-risk cardiac surgery mortality rates and favorable findings concerning as it reduces the risk of transfusion and bleeding stroke, atrial fibrillation and length of hospital stay complications (33). (23). Again, the majority of their available data were based on retrospective studies and comprised Minimally invasive extracorporeal circulation off-pump coronary surgery. The prospective RCTs systems (MiECC) of Pierri, Chee and Dehghani et al. were not able to retain a significant improvement of non-POAF- Over the years, several technical adaptations related outcomes (24, 27, 28). have been introduced to ameliorate the outcome We would therefore recommend the advice after CPB. Landis et al. published in 2014 an of Landis et al in 2014 : “In patients receiving overview of available therapies and stated level statins, it is advisable to continue statins because A evidence for minimized extracorporeal circuits they appear to decrease perioperative morbidity and and biocompatible surface coating with divergent mortality. However, in patients who were not on results (29). statins before surgery, it is not clear if perioperative In 2016, the Minimal invasive Extra-corporeal statin therapy offers any benefit” (29). Furthermore, Technologies international Society (MiECTiS) large randomized trials would be preferable as there published a position paper defining the principle is currently no evidence of an optimal combination “MiECC” (34). They defined MiECC as a “combined of statins and dosing schemes. strategy of surgical approach, anesthesiologic and perfusion management which is not limited to the Aprotinin CPB circuit alone. The main components include Aprotinin is a nonspecific serine protease a closed CPB circuit ; biologically inert blood inhibitor. It was first tested in the 1980’s as an anti- contact surfaces ; reduced priming volume ; a cen- inflammatory drug (30). However, the initial trials trifugal pump ; a membrane oxygenator ; a heat showed an important reduction in perioperative exchanger ; a cardioplegia system ; a venous bubble bleeding, which led to its widespread use in trap/venous air removing device and a shed blood cardiac surgery. Nonetheless, it certainly possesses management system”. By integrating this system, important anti-inflammatory mechanisms. As it they found significantly reduced blood loss, lower inhibits kallikrein and plasmin, the contact and incidences of POAF and better preserved renal and fibrinolytic system are attenuated. As a consequence, cardiac function in comparison to normal heart- the expression of cytokines and CAMs are reduced. lung machine (Class of recommendation I, level Furthermore, it also prevents proteolysis of the of evidence A). Furthermore, they found evidence important thrombin receptor on platelets, thereby of an attenuated inflammatory response and better inhibiting platelet activation which indicates pro- preservation of neurocognitive function (class of and anti-thrombotic effects (30). recommendation II, level of evidence B). Last, In 2006, Mangano and coworkers published they indicated a trend towards lower mortality and an observational study which stated almost a subclinical protective effects on end-organ function doubling of the rates of cardiac, renal and cerebral by improved microvascular organ perfusion (34). complications using aprotinin instead of other The result concerning lower rates of POAF lysine analogues (31). There were however some through using minimally invasive extracorporeal methodological concerns and senior commentators technology was further confirmed by Ranucci and suggested awaiting the Canadian BART’s trial. This colleagues (35). Besides, they noted a trend towards trial concluded “a possibility of modest reduction better lung function and less acute kidney injury in the risk of massive bleeding”, but stated “the and highlighted the multi-factorial nature of these consistent negative mortality trend associated with complications renders it difficult to ameliorate aprotinin precludes its use in cardiac surgery” (32). outcomes with only one intervention. Nonetheless, However, no difference was made between low-, along them, “the existing body of evidence is in intermediate and high-risk interventions and cardiac favor of MiECC” (35). revascularizations were the main type of surgery.

14-Keunen.indd 110 12/01/2021 12:18 sirs after cardiopulmonary bypass 111

Summary systemic inflammatory response. Ann Thorac Surg. 63: 277-284 7. Esper S., Subramaniam K. and Tanaka K. 2014. The pathophysiology of SIRS after CPB is quite Pathophysiology of cardiopulmonary bypass: Current complex. It comprises an early and late phase. The strategies for the prevention and treatment of anemia, “early” phase is induced by operative trauma and coagulopathy, and organ dysfunction. Semin Cardiothorac Vasc Anesth. 18: 161-176 contact of blood with the CPB. The complement and 8. Holmes I., Connolly N., Paull D., Hill M., Guyton S. and contact system stimulate leukocytes, endothelium Ziegler S., et al. 2002. Magnitude of the inflammatory and platelets to produce cytokines and oxygen- response to cardiopulmonary bypass and its relation to adverse clinical outcomes. Inflamm Res. 51: 579-586 free radicals. This induces the transmigration of 9. Wan S., Leclerc J and Vincent J. 1998. Inflammatory leukocytes in non-vascular tissues with resulting end- response to cardiopulmonary bypass. Mechanisms Involved organ inflammation. Furthermore, the hemostatic and Possible Therapeutic Strategies. Ann Thorac Surg. 66: 676-692 and fibrinolytic systems are overactivated with 10. Durandy Y. 2014. Minimizing Systemic Inflammation consumptive coagulopathy and microemboli. These During Cardiopulmonary Bypass in the Pediatric Popu- early phase reactions attenuate as the duration of lation. Artif Organs. 38: 11-18 11. Kats S., Schönberger J., Brands R., Seinen W. and van the CPB increases. A “late” inflammatory reaction Oeveren W. 2011. Endotoxin release in cardiac surgery with starts after releasing the aortic clamp. It is caused by cardiopulmonary bypass: Pathophysiology and possible ischemia-reperfusion injury, endotoxemia, heparin/ therapeutic strategies. An update. Eur J Cardio-thoracic protamine reactions, and the further release of Surg. 39: 451-458 12. Epting C., McBride M., Wald E. and Costello J. 2015. cytokines and NO. Pathophysiology of Post-Operative Low Cardiac Output Major therapeutic hope was established on Syndrome. Curr Vasc Pharmacol. 14: 14-23 corticosteroids. Unfortunately, the two major trials 13. Dieleman J., van Paassen J., van Dijk D., Arbous M., Kalkman C. and Vandenbroucke J., et al. 2011. Prophylactic (DECS and SIRS trial) were not able to show a corticosteroids for cardiopulmonary bypass in adults. positive impact of steroids in the general population. Cochrane Database Syst Rev. 5: 1-156 However, we do believe that dexamethasone is 14. Apostolakis E., Filos K., Koletsis E. and Dougenis D. 2010. Lung dysfunction following cardiopulmonary bypass. J a potential benefit for young patients with a Card Surg. 25: 47-55 robust inflammatory capacity, which deserves 15. Fiorentino F., Jaaly E., Durham A., Adcock I., Lockwood further investigation. At that moment, a mandate G. and Rogers C., et al. 2019. Low-frequency ventilation during cardiopulmonary bypass for lung protection: A measurement of cardiac biomarkers would also be randomized controlled trial. J Card Surg. 34: 385-399 advisable. 16. Dieleman J., Nierich A., Rosseel P., van der Maaten J., Hof- Regarding all other investigated therapies, land J. and Diephuis J., et al. 2012. Intraoperative High- Dose Dexamethasone for Cardiac Surgery. JAMA. 308: minimal invasive extracorporeal circuit, aprotinin 1761-1767 (specifically in high-risk surgery) and statins have 17. Ho K. and Tan J. 2009. Benefits and risks of corticosteroid shown potential benefits to clinical outcomes in prophylaxis in adult cardiac surgery a dose-response meta- various studies. Future studies should investigate analysis. Circulation. 119: 1853-1866 18. Whitlock R., Lamy A., Vincent J., Pogue J., Pettit S. and the role of using a multimodal approach to diminish Chrolavicius S., et al. 2015. Methylprednisolone in patients the inflammatory response and improve patients’ undergoing cardiopulmonary bypass (SIRS): A randomised, outcome after CPB. double-blind, placebo-controlled trial. Lancet. 386: 1243- 1253 19. van Osch D., Nathoe H., Bunge J., Dieleman J., Boasson References M. and van Dijk D., et al. 2015. Intraoperative High- Dose Dexamethasone in Cardiac Surgery and the Risk of Rethoracotomy. Ann Thorac Surg. 100: 2237-2242 1. Gibbon J. 1954. Application of a mechanical heart and lung 20. Nashef S., Roques F., Sharples L., Nilsson J., Smith C. apparatus to cardiac surgery. Minn Med. 37: 171-185 and Goldstone A., et al. 2012. Euroscore II. Eur J Cardio- 2. Paparella D., Yau. T and Young E. 2002. Cardiopulmonary thoracic Surg. 41: 734-745 bypass induced inflammation: Pathophysiology and 21. Dvirnik N., Belley-Cote E., Hanif H., Devereaux P., Lamy treatment. An update. Eur J Cardio-thoracic Surg. 21: 232- A. and Dieleman J., et al. 2018. Steroids in cardiac surgery: 244 a systematic review and meta-analysis. Br J Anaesth. 120: 3. Warren O., Smith A., Alexiou C., Rogers P., Jawad N. 657-667 and Vincent C., et al. 2009. The Inflammatory Response 22. Samuel S., Nguyen T. and Choi H. 2017. Pharmacologic to Cardiopulmonary Bypass: Part 1-Mechanisms of Characteristics of Corticosteroids. J Neurocritical Care. 10: Pathogenesis. J Cardiothorac Vasc Anesth. 23: 223-231 53-59 4. Asimakopoulos G. 2001. Systemic inflam-mation and 23. Kuhn E., Liakopoulos O., Stange S., Deppe A., Slottosch cardiac surgery: An update. Perfusion. 16: 353-360 I. and Choi Y., et al. 2014. Preoperative statin therapy in 5. Hall R., Stafford M., Rocker G. and Uk M. 1997. The cardiac surgery: A meta-analysis of 90 000 patients. Eur J Systemic Inflammatory Response to Cardiopulmonary Cardio-thoracic Surg. 45: 17-26 Bypass: Pathophysiological, Therapeutic, and Pharmaco- 24. Chee Y., Watson R., McCarthy J., Chughtai J., Nölke logical Considerations. Anesth Analg. 85: 766-782 L. and Healy D. 2017. High dose statin prophylaxis in 6. Boyle E., Pohlman T., Johnson M. and Verrier E. 1997. cardiopulmonary bypass related surgery: Clinical utility. J Endothelial cell injury in cardiovascular surgery : The Cardiothorac Surg. 12: 1-10

14-Keunen.indd 111 12/01/2021 12:18 112 b. keunen et al. 25. Morgan C., Zappitelli M. and Gill P. 2009. Statin prophylaxis to Cardiopulmonary Bypass: Part 2-Anti-Inflammatory and inflammatory mediators following cardiopulmonary Therapeutic Strategies. J Cardiothorac Vasc Anesth. 23: bypass: A systematic review. Crit Care.13: 1-11 384-393 26. Patti G., Chello M., Candura D., Pasceri V., D’Ambrosio A. 31. Mangano D., Tudor I. and Dietzel C. 2006. The risk and Covino E., et al. 2006. Randomized trial of atorvastatin associated with aprotinin in cardiac surgery. N Engl J Med. for reduction of postoperative atrial fibrillation in patients 354: 353-365 undergoing cardiac surgery: Results of the ARMYDA-3 32. Fergusson D., Hébert P., Mazer D., Fremes S., MacAdams (Atorvastatin for Reduction of MYocardial Dysrhythmia C. and Murkin J., et al. 2008. A Comparison of Aprotinin After cardiac surgery) study. Circulation. 114: 1455-1461 and Lysine Analogues in High-Risk Cardiac Surgery. N 27. Dehghani M., Kasianzadeh M., Rezaei Y. and Sepehrvand N. Engl J Med. 358: 687-696 2015. Atorvastatin Reduces the Incidence of Postoperative 33. Meybohm P., Herrmann E., Nierhoff J. and Zacharowski Atrial Fibrillation in Statin-Naive Patients Undergoing K. 2013. Aprotinin May Increase Mortality in Low and Isolated Heart Valve Surgery. J Cardiovasc Pharmacol Ther. Intermediate Risk but Not in High Risk Cardiac Surgical 20: 465-472 Patients Compared to Tranexamic Acid and ε-Aminocaproic 28. Pierri M., Crescenzi G., Zingaro C., D’Alfonso A., Capestro Acid – A Meta-Analysis of Randomised and Observational F. and Scocco V., et al. 2016. Prevention of atrial fibrillation Trials of over 30.000 Patients. PLoS One. 8: 1-7 and inflammatory response after on-pump coronary artery 34. Anastasiadis K., Murkin J., Antonitsis P., Bauer A., Ranucci bypass using different statin dosages: a randomized, M. and Gygax E., et al. 2016. Use of minimal invasive controlled trial. Gen Thorac Cardiovasc Surg. 64: 395-402 extracorporeal circulation in cardiac surgery: principles, 29. Landis C., Brown J., Fitzgerald D., Likosky D., Shore- definitions and potential benefits. A position paper from Lesserson L. and Baker R., et al. 2014. Attenuating the the Minimal invasive Extra-Corporeal Technologies systemic inflammatory response to adult cardiopulmonary international Society (MiECTiS). Interact Cardiovasc bypass: A critical review of the evidence base. J Extra Thorac Surg. 22: 647-662 Corpor Technol. 46: 197-211 35. Ranucci M. and Baryshnikova E. 2019. Inflammation and 30. Warren O., Watret A., de Wit K., Alexiou C., Vincent C. coagulation following minimally invasive extracorporeal and Darzi A., et al. 2009. The Inflammatory Response circulation technologies. J Thorac Dis. 11: 1480-1488

14-Keunen.indd 112 12/01/2021 12:18 (Acta Anaesth. Belg., 2020, 71, 113-119) Rapid sequence induction and intubation in children revisited : a narrative systematic review on current literature with emphasis on airway management

A. Dionys, J. Lauweryns

Abstract : Ever since it is first described, rapid anesthesiologists (4). The main reason for this sequence induction and intubation has been the heterogeneity is the increasing questioning of induction technique of choice for patients at risk for both the effectiveness and adverse events of this pulmonary aspiration. It’s a well-known and well technique. When reviewing current literature, many described technique in adult patients. However, in concerns and potential disadvantages are brought to children, data regarding the use of ‘classic’ rapid light. sequence induction is scarce. When applying it The pediatric use of this practice, in particular, in children, it could be potentially harmful. This is heavily debated. The past three decades, airway narrative review summarizes current literature management strategy in treating children has regarding the use of ‘classic’ rapid sequence evolved. Initially, anesthesiologists even recom- induction in children, controversies in its use and mended awake tracheal intubation in children potential benefits and adverse events. at risk for pulmonary aspiration, although this technique has now become obsolete (5). Currently, Keywords : Pediatric ; rapid sequence induction/intuba- the heterogeneity in practical approach in children tion ; pulmonary aspiration, hypoxemia. is even greater as in adults, thus raising questions about the feasibility of using the same guidelines for Introduction adult and pediatric patients.

Mendelson (1) first described the deleterious Methodology effects of pulmonary aspiration in 1946 and recognized the need for a safe anesthesia induction A literature search was conducted in database technique which could prevent these occurrences. of Pubmed, using the following keywords : “pedia- After adding the introduction of succinylcholine tric”, “rapid sequence induction and intubation”, in 1951 and the use of cricoid pressure in 1961 (2), “emergency airway management”, “adverse events” it was Stept and Safar in 1970 (3) who described a and “risks”. Each abstract was screened and included new technique, which they called ‘Rapid Sequence or excluded by judgment on title and relevance to Induction and Intubation’, a fifteen-step method this review. Articles who weren’t written in English of induction and intubation, designed to secure were excluded. The references for these studies the airway quickly while actively reducing the were then searched for any relevant articles not risk of aspiration. It consisted of preoxygenation, previously identified in our primary search. This induction with a predetermined dose of thiopental literature is hereby summarized, focusing on airway followed by succinylcholine, application of cricoid management (Fig. 1). pressure at loss of consciousness, avoidance of positive pressure ventilation, and finally tracheal intubation with a cuffed tube before removal of A. Dionys, M.D. ; J. Lauweryns, M.D. the cricoid pressure. Without questioning and only Department of Anesthesiology, University Hospital Leuven, 3000 Leuven, Belgium empiric recommendations available, this has been Corresponding author : Annelies Dionys, Department of the technique of choice for induction of anesthesia Anesthesiology, University Hospital Leuven, Herestraat 49, in patients who are at risk of pulmonary aspiration. 3000 Leuven, Belgium. Although clear guidelines on practice in E-mail : [email protected] adult patients exist, there seems to be a significant Paper submitted on May 31, 2020 and accepted on Jun 4, 2020. heterogeneity in its practical approach among Conflict of interest : none.

15-Dionys.indd 113 12/01/2021 12:19 114 a. dionys & j. lauweryns Following these reviews, an inquiry into the usefulness of classic RSI in children was initiated, led by a large group of pediatric anesthesiologists. Stedeford et al. (8) conducted a survey in south- west England in 2007, asking both pediatric and non-pediatric anesthetists their preferred technique in performing RSI in children. Their results showed that classic RSI was still performed by the majority of anesthetists. However, it was often modified for infants, especially by more recently trained consultants, by avoiding cricoid pressure and application of gentle intermittent mask ventilation. In 2020, Klucka et al. (4) performed an online international survey in 56 countries to evaluate clinical practices of RSI, both in adult and pediatric patients. The survey was completed by 1921 members of the European Society of Anesthesiology Fig. 1. — Prisma flow diagram. From : Moher D, et al. and showed a significant heterogeneity in perfor- 2009. Preferred Reporting Items for Systematic Reviews mance of RSI. and MetaAnalyses: The PRISMA Statement. PLoS Med 6: In the search for greater patient safety, common e1000097. practice promotes a standard of care essential to optimize a patient’s outcome. In this case, outcome Results is not only measured by successfully placing an endotracheal tube, but also by minimizing compli- Who, what, where and how? cations associated with RSI. Rapid sequence induction and intubation is Why, then, do we see such heterogeneity in the well-described in literature for adult patients, and is practice of RSI? a standard procedure in patients who are at risk for pulmonary aspiration of gastric content. In children, Controversies however, by our knowledge, clear guidelines do not exist on how to perform safe, rapid sequence Overall, pediatric rapid sequence induction induction and intubation. The advice which is not a very common procedure, which makes exists now is extrapolated from adult guidelines. it a difficult subject to gather data for, and most ‘Classic’ RSI (rapid sequence induction) consists of data we currently have available is derived from several consecutive steps, such as preoxygenation, retrospective studies. Such studies have several application of cricoid pressure, induction with a specific limitations, such as selection and reporting predetermined dose of anesthetic followed by a bias, in particular if data is retrospectively derived neuromuscular blocking agent, and avoidance of from charts. When reported via chart review, mask ventilation during the apneic period. adverse events often seem to be underestimated in In 2004, a literature review was performed comparison with reports from video review (9, 10). regarding recommendations on pediatric RSI (6), To form a comprehensive idea of the incidence which brought emphasis to protecting the patient’s of common complications during performance of airway from gastric content aspiration by refraining classic RSI in children, Litman et al. (11) performed from manually ventilating the patient and by a retrospective cohort analysis of children who applying cricoid pressure throughout the intubation underwent a rapid sequence induction. The study procedure. focused on 1070 children aged between 3-12 years, One year later, Zelicof-Paul et al. (7) proposed who underwent RSI in a classical manner, with a potential protocol for pediatric RSI, serving to the avoidance of ventilation during apneic period, summarize existing data. This summary included application of cricoid pressure and endotracheal the use of cricoid pressure and the practical goal of intubation performed within one minute after placement of an endotracheal tube 60 seconds after induction. Most common complications were administration of a neuromuscular blocking agent, hypoxemia, bradycardia, hypotension and difficult without manually ventilating the patient during the intubation. Transient desaturation occurred in 3,6 % period of apnea. of these children, of whom 1,7 % experienced

15-Dionys.indd 114 12/01/2021 12:19 rapid sequence induction and intubation in children revisited 115

severe hypoxemia (SpO2 < 80 %). It was found capacity. This is why, when compared to adults, that children weighing less than 20 kg had an children will have a reduced tolerance to apnea and increased risk of experiencing severe hypoxemia. will quickly desaturate. Unfortunately, the study did not include children Age seems to play a role in how quickly < 3 years of age, although this age is known to be at children desaturate. Younger children desaturate particular risk for the development of complications faster than older children (15, 16). In the youngest related to RSI. This study demonstrated that a age group, the time to desaturate may even be as classic RSI technique is not a reliable approach to short as the time required to intubate. Desaturation minimize the incidence of complications associated during laryngoscopy in infants is not uncommon. with the induction of anesthesia in children with a Cricoid pressure full stomach. These findings correspond to the incidence Cricoid pressure is the application of external of complications during anesthetic management of pressure to the cricoid cartilage during induction children with post-tonsillectomy hemorrhage (12). of anesthesia. The rationale being that the upper Children needing surgical reintervention due to esophagus is occluded by compression between post-operative hemorrhage are at risk for pulmonary the trachea and the cervical vertebrae. The cricoid aspiration of gastric contents, for which they cartilage has a narrow portion at the anterior and underwent a classic rapid sequence induction. Here, the upper esophageal sphincter and cricopharyngeal the most common adverse events were hypoxemia, muscle at the posterior site. Cricoid pressure is bradycardia, and hypotension, as well as difficult designed to occlude the upper esophageal lumen intubation. and augment the normal sphincter, preventing passive reflux of gastric contents. Aspiration In 1961, Sellick introduced cricoid pressure Pulmonary aspiration during induction of “to control regurgitation of gastric or esophageal general anesthesia may occur through active contents until intubation with a cuffed endotracheal regurgitation of gastric contents into the upper tube was completed” (2). Sellick’s maneuver rapidly airway. These are active processes mostly elicited by became an indispensable component of the RSI. laryngoscopy under light anesthesia and incomplete Cricoid pressure, or the Sellick maneuver, muscle paralysis. originally was described in adults. Earliest results Clinical data in children suggest that regurgi- in children go back to 1972, when Salem et al. tation or aspiration occurs almost exclusively performed studies on infant cadavers (17). It was when depth of anesthesia and muscle relaxation found that “firm” cricoid pressure was effective in are incomplete (13). In this study, nearly all cases occluding the esophagus and preventing reflux of of pulmonary aspiration occurred in patients who saline from esophagus to the pharynx. Even in 1993, gagged or coughed during airway manipulation or Moynihan et al. (18) concluded that appropriate during induction of anesthesia in which paralysis application of cricoid pressure prevents gastric with muscle relaxants was either not present or was insufflation in infants and children. The correct insufficient to prevent a gag or cough. Some studies technique for the Sellick maneuver is the application even show pulmonary aspiration to be a rather rare of a 10 N pressure on the cricoid cartilage before event in children (14). induction of anesthesia and further increase of the Is the use, then, of RSI in children even needed? pressure to 30 N after the induction. However, over the past few years, both the Apnea and time to desaturate benefits and the manners of performing this In an apneic patient, the time for oxygen maneuver are being questioned. Despite wide desaturation depends on the adequacy of preoxy- acceptance, there is limited evidence to support genation, functional residual capacity of the lungs, the effectiveness of its use. Recent editorials are and oxygen consumption of the body. Children have questioning the lack of evidence and the increase in a reduced functional residual capacity and thus have airway complications as a result of its use (19). both a reduced oxygen reserve and a high oxygen demand, so that they will deplete their oxygen What can we do differently? stores more quickly. Besides these factors, they Gentle mask ventilation are less likely to cooperate during preoxygenation. They experience increased small airway closure The practice of routinely avoiding bag and mask secondary to a high closing capacity to functional ventilation during RSI is based on the hypothesis

15-Dionys.indd 115 12/01/2021 12:19 116 a. dionys & j. lauweryns that its use would result in gastric insufflation and was reported (21). Furthermore, there was no increase the risk of regurgitation and subsequent incidence of pulmonary aspiration during induction, aspiration. Avoidance of bag and mask ventilation laryngoscopy, and throughout the further course of does hasten the onset and severity of hypoxemia anesthesia. Results indicate that intermittent gentle during RSI. mask ventilation prior to endotracheal intubation Overall, during RSI, prevention of pulmonary support stable cardiorespiratory conditions for gastric aspiration has always been more important securing the airway in children with an expected or than protection from hypoxemia. However, due to suspected full stomach. the aforementioned reasons, the majority of children Hypertrophic pyloric stenosis is one of the most will desaturate much faster than adults in the common surgical conditions affecting neonates and absence of ventilation and oxygenation. In clinical young infants. It causes gastric outlet obstruction reality, classic RSI will often induce iatrogenic and gastric fluid retention which increases the risk hypoxia, bradycardia, hypotension and subsequent of aspiration during the induction of anesthesia. high stress levels in anesthesia providers (20). High In 2019, in infants presenting for pyloromyotomy, stress levels are likely to trigger unsafe actions. anesthetic induction with controlled RSI was Subsequently hastily performed laryngoscopy can associated with significantly less hypoxemia without induce coughing and straining, which impedes an observed increase in aspiration events (22). The laryngoscopy and induces regurgitation and aspi- classic RSI cohort had over 70 % more hypoxemic ration. events relative to the controlled RSI. In neonates, Considering that hypoxia during induction is controlled RSI led to an even greater absolute a major cause of cardiac arrest and weighing this risk reduction of hypoxemia. Results suggest that risk against the risk of pulmonary gastric aspiration, neonates may gain an even greater benefit from the ‘controlled RSI’ technique with intermittent controlled RSI. gentle mask ventilation has become increasingly In 2020, a retrospective audit in children recommended by many authors. with post-tonsillectomy bleeding (23) showed less Gentle mask ventilation is ventilation with incidence of hypoxemia and difficult laryngoscopy

a limited inspiratory pressure ( < 12-15 cmH2O), without increasing occurrence of respiratory which can lead to effective ventilation and complications when gentle bag-mask ventilation oxygenation, without insufflating the stomach. The was used between induction and tracheal intubation. inclusion of such pressure limited ventilation in the Abandoning cricoid pressure RSI algorithm is described as ‘controlled RSI’ or ‘modified RSI’. Do we know where to press? Correct identi- In the light of adding gentle mask ventilation to fication of the cricoid cartilage in young children the algorithm, an infant simulator-based evaluation is difficult using landmarks only (24). Incorrect of this modified algorithm was set up in 2009 (20). application of cricoid pressure could even be more The sequence was modified : preoxygenation, rapid harmful by hindering laryngoscopy, causing airway induction of adequate hypnosis and profound muscle obstruction and difficulties with airway mana- paralysis, gentle mask ventilation with a maximum gement.

airway pressure of 12 cmH2O, laryngoscopy, and Could it interfere with airway management? finally intubating when deep anesthesia and full Cricoid pressure may cause distortion of the upper muscular blockade are present. Results showed airway, making airway management unnecessarily that use of this modified sequence caused a more difficult. It may interfere with visualization of lower incidence of unsafe actions such as hasty the larynx or mask ventilation. In adults, it causes laryngoscopy, or forced mask ventilation with high an increase in difficult intubation rate. Because of airway pressures or esophageal or endobronchial anatomical reasons, this might even occur more in intubation, as compared to classic RSI. It also children. reduced providers’ stress levels. The modified Indeed, when applied with forces well below sequence was aimed to prevent hypoxemia and to the recommended 30 Newton, cricoid pressure reduce timing-related pressure. will cause significant compression and distortion Subsequently, an analysis was made on the of the airway in a child (25). There seems to be a incidence of respiratory and consecutive hemo- higher incidence of difficult airways and a longer dynamic complications when this modified RSI was intubation time. The trachea of infants and young used. Compared to an earlier study on classic RSI in children are more easily compressed compared with children, a lower incidence of oxygen desaturation those of adults.

15-Dionys.indd 116 12/01/2021 12:19 rapid sequence induction and intubation in children revisited 117 Does the practice achieve its desired effects? In 2019, Gaget et al. (32) performed a pro- Theoretically, cricoid pressure during RSI could spective observational trial to assess its usefulness prevent regurgitation of gastric contents by in infants with pyloric stenosis, scheduled for occluding the esophagus with the posterior cricoid pyloromyotomy. Their results suggested that quali- cartilage ring. So, the effectiveness of this maneuver tative ultrasound assessment may be a simple may be limited if it interferes with the position of tool for the choice of most appropriate anesthetic the esophagus behind this cartilage ring. In 2010, induction technique in these cases. a study was performed on the rate and degree of esophageal displacement from the airway at the level Discussion and conclusion of the cricoid cartilage (26). Lateral displacement of the esophagus occurred at a significantly greater Rapid sequence induction is a procedure that rate (45 %) in children younger than 8 years of age often leads to high stress levels (20), especially , compared with children aged 8-17 years (15 %), when performed in pediatric patients. In general, which may result in ineffective cricoid pressure. higher stress levels lead to unsafe actions and more Little data exists about the effectiveness of this complications. Clinical guidelines could help to practice in children. In adults, there does not seem reduce this stress and increase patient safety. to be evidence to either support or discourage the Although there are clear guidelines on how to use of RSI (27). One large pediatric observational perform a rapid sequence induction in adult patients, study from 2018 showed that use of cricoid pressure there appears to be a significant heterogeneity in its during induction was not associated with a lower practical use among anesthesiologists. The main occurrence of regurgitation (28). reason for this heterogeneity seems to be the rising questions of this technique regarding benefits, Apneic oxygenation practical approaches and possible adverse events. Apneic oxygenation is the application of high- When it comes to performing RSI in pediatric flow oxygen during the apneic period. During apnea, patients, scarce data exist in literature. Most advice reduced barometric pressure in the alveoli creates about pediatric RSI is extrapolated from guidelines a pressure gradient and subsequent increased gas for adult RSI. flow between upper and lower airways. By applying Historically, preventing pulmonary aspiration oxygen during the apneic period, the oxygen content during induction has been emphasized. Key in the lower airways is increased. elements in this prevention have always been the In recent years, apneic oxygenation or appli- use of cricoid pressure and refraining from mask cation of high-flow nasal oxygen cannulae (HFNC) ventilation. during the apneic period, is being tested as a Over the last few years, controversies are possible measure to reduce hypoxemia during RSII starting to put the use of classic RSI in pediatric and prolong the time available for laryngoscopy and patients into question. Children, especially of tracheal intubation. Until now, studies performed in younger ages, have a higher risk of experiencing adult patients are promising. hypoxemia and subsequent hemodynamic compli- When used during RSI in critically ill children, cations during classic RSI. When comparing the apneic oxygenation seems to be well-tolerated incidence of hypoxemia with the incidence of without causing adverse events (29). However, pulmonary aspiration, the risk of hypoxemia seems results of benefit seem to be conflicting. In 2015, to weigh through. it was not associated with a decrease in incidence, When it comes to the use of cricoid pressure, duration or depth of desaturation (30), but in 2019 there are conflicting results in literature. In general, it did seem to reduce the incidence of hypoxemia most papers are pointing towards cricoid pressure (31). not being efficient in children. Besides not being efficient, it can also cause distortion of the airway Gastric ultrasound and make airway management unnecessary difficult. Theoretically, preoperative ultrasound assess- When suggesting guidelines for pediatric ment of the gastric contents could be a useful point- RSI, we should keep in mind that priorities are of-care tool to guide choice of anesthetic induction different relative to adult patients. We cannot copy technique. By measuring the volume of the gastric adult guidelines to pediatric use without relevant content, the risk of pulmonary aspiration can be adjustments. Children are more likely to experience estimated and this can guide in choosing the method hypoxemia and subsequent hemodynamic compli- of anesthesia induction. cations in an apneic period. In literature, pulmonary

15-Dionys.indd 117 12/01/2021 12:19 118 a. dionys & j. lauweryns aspiration almost exclusively occurs when depth of biomedical reference librarians of the KU Leuven anesthesia and muscle relaxation are inadequate. Libraries – 2Bergen – learning Centre Désiré Collen Hence, prevention of hypoxemia and the provision (Leuven, Belgium), for their help in conducting the of optimal intubation conditions is the essential systematic literature search. characteristic for a safer RSI technique. On the other hand, when emphasizing the need to maintain oxygenation during induction, we References must not forget to prevent pulmonary aspiration. 1. Mendelson C.L. 1946. The Aspiration of Stomach Contents Is it sufficient to provide effective anesthesia and Into the Lungs During Obstetric Anesthesia. Am J Obstet complete muscle paralysis to avoid aspiration of Gynecol. 52: 191-205. gastric contents? An absence of evidence of RSI 2. Sellick B.A. 1961. Cricoid Pressure to Control Regurgitation decreasing the risk of aspiration does not mean it of Stomach Contents During Induction of Anaesthesia. Lancet. 2(7199): 404-406. could not be effective. 3. Stept W.J. and Safar P. 1970. Rapid Induction/Intubation We suggest adding the use of gentle mask for Prevention of Gastric-Content Aspiration. Anesth ventilation to the sequence of rapid sequence Analg. 49(4): 633-636. 4. Klucka J., Kosinova M., Zacharowski K., De Hert S., induction in children, especially in the youngest Kratochvil M., Toukalkova M. and Stoudek R., et al. 2020. ones. After all, in 1961, Sellick (2) stated that Rapid Sequence Induction: An International Survey. Eur J “during cricoid pressure, the lungs may be ventilated Anaesthesiol. 37(6): 435-442. 5. Cook-Sather S., Tullock H., Cnaan A., Nicolson S., Cubina by intermittent positive pressure ventilation without M., Gallagher P. and Schreiner M. 1998. A Comparison of risk of gastric distension”. Awake Versus Paralyzed Tracheal Intubation for Infants There is a relationship between the maximum With Pyloric Stenosis. Anesth Analg. 86(5): 945-951. airway pressure used during mask ventilation and 6. Bledsoe G. and Schexnayder S. 2004. Pediatric Rapid Sequence Intubation: A Review. Pediatr Emerg Care. 20(5): the amount of air insufflated in the stomach. When 339-344. mask ventilation is considered easy, lower airway 7. Zelicof-Paul A., Smith-Lockridge A., Schnadower D., Tyler S., Levin S., Roskind C. and Dayan P. 2005. Controversies pressures are used (< 15 cm H2O) and these rarely in Rapid Sequence Intubation in Children. Curr Opin cause stomach inflation. Applying cricoid pressure Pediatr. 17(3): 355-362. is likely to increase the maximum airway pressure 8. Stedeford J. and Stoddart P. 2007. RSI in Pediatric by compressing the airway and these patients Anesthesia – Is It Used by Nonpediatric Anesthetists? A Survey From South-West England. Paediatr Anaesth. tend to have more air in the stomach than patients 17(3): 235-242 considered easy to ventilate, therefore we can 9. Rinderknecht A., Dyas J., Kerrey B., Geis G., Ho M. and use gentle ventilation in associated with cricoid Mittiga M. 2017. Studying the Safety and Performance pressure. If ventilation proves to be difficult after of Rapid Sequence Intubation: Data Collection Method Matters. Acad Emerg Med. 24: 411-421. the application of cricoid pressure, it should be 10. Kerrey B., Rinderknecht A., Geis G., Nigrovic L. and released to avoid increasing airway pressures. Mittiga M. 2012. Rapid Sequence Intubation for Pediatric In addition, we suggest adding preoperative Emergency Patients: Higher Frequency of Failed Attempts and Adverse Effects Found by Video Review. Ann Emerg gastric ultrasound to assess the amount of volume Med. 60(3): 251-259. in the stomach to assess the risk of aspiration. By 11. Gencorelli F., Fields R. and Litman R. 2010. Complications doing this preoperatively, we can refrain from rapid During Rapid Sequence Induction of General Anesthesia in Children: A Benchmark Study. Paediatr Anaesth. 20(5): sequence induction in children who are not at risk 421-424. for gastric regurgitation and aspiration. 12. Fields R., Gencorelli F. and Litman R. 2010. Anesthetic In summary, there is still lack of consensus Management of the Pediatric Bleeding Tonsil. Paediatr and limited evidence on potential benefits and Anaesth. 20(11): 982-986. 13. Warner M.A., Warner M.E., Warner D.O., Warner L.O. and harms regarding use of classic RSI in children. Warner E.J. 1999. Perioperative Pulmonary Aspiration in Currently, there are more papers with evidence Infants and Children. Anesthesiology. 90(1): 66-71. pointing towards the ineffectiveness of classic RSI 14. Borland L., Sereika S., Woelfel S., Saitz E. Carrillo P., Lupin J. and Motoyama E. et al. 1998. Pulmonary Aspiration in in children. Pediatric Patients During General Anesthesia: Incidence More prospective studies need to be done to and Outcome. J Clin Anesth. 10(2): 95-102. gain clearer results, which would present ethical 15. Patel R., Lenczyk M., Hannallah R. and McGill W. 1994. Age and the Onset of Desaturation in Apnoeic Children. difficulties. Can J Anaesth. 41(9): 771-774. 16. Xue F., Luo L., Tong S., Liao X., Deng X. and An G. 1996. Acknowledgments Study of the Safe Treshold of Apneic Period in Children during Anesthesia Induction. J Clin Anesth. 8: 568-574. 17. Salem M., Wong A. and Fizzotti G. 1972. Efficacy of Cricoid The author wishes to thank Thomas Vanden- Pressure in Preventing Aspiration of Gastric Contents in driessche, Kristel Paque and Krizia Tuand, the Paediatric Patients. Br J Anaesth. 44(4): 401-404.

15-Dionys.indd 118 12/01/2021 12:19 rapid sequence induction and intubation in children revisited 119 18. Moynihan R., Brock-Utne J., Archer J., Held L. and Esophagus and Its Potential Influence on the Effectiveness Kreitzman T. 1993. The Effect of Cricoid Pressure on of Sellick Maneuver in Children. 26(10): 722-725. Preventing Gastric Insufflation in Infants and Children. 27. Neilipovitz D. and Crosby E. 2007. No Evidence for Anesthesiology. 78(4): 652-656. Decreased Incidence of Aspiration After Rapid Sequence 19. Brock-Utne J. 2002. Is Cricoid Pressure Necessary? Induction. Can J Anaesth. 54(9): 748-764. Paediatr Anaesth. 12(1): 1-4. 28. Kojima T., Harwayne-Gidansky I., Shenoi A., Owen E., 20. Eich C., Timmermann A., Russo S., Cremer S., Nickut A., Napolitano N., Rehder K. and Adu-Darko M. et al. 2018. Strack M. and Weiss M. et al. 2009. A Controlled Rapid- Cricoid Pressure During Induction for Tracheal Intubation Sequence Induction Technique for Infants May Reduce in Critically Ill Children: A Report From National Unsafe Actions and Stress. Acta Anaesthesiol Scand. 53(9): Emergency Airway Registry for Children. Pediatr Crit Care 1167-1172. Med. 19(6): 528-537. 21. Neuhaus D., Schmitz A., Gerber A. and Weiss M. 2013. 29. Mortimer T., Burzynski J., Kesselman M., Vallance J. Controlled Rapid Sequence Induction and Intubation - An and Hansen G. 2016. Apneic Oxygenation during Rapid Analysis of 1001 Children. Paediatr Anaesth. 23(8): 734- Sequence Intubation in Critically Ill Children. J Pediatr 740. Intensive Care. 5: 28-31. 22. Park R., Rattana-Arpa S., Peyton J., Huang J., Kordun A., 30. Overmann K., Boyd S., Zhang Y. and Kerrey B. 2019. Cravero J. and Zurakowski D. 2019. Risk of Hypoxemia Apneic Oxygenation to Prevent Oxyhemoglobin Desatu- by Induction Technique Among Infants and Neonates ration During Rapid Sequence Intubation in a Pediatric Undergoing Pyloromyotomy. Anesth Analg. [Epub ahead Emergency Department. Am J Emerg Med. 37(8): 1416- of print]. 1421. 23. Kemper M., Buehler P., Schmitz A., Gysin C., Nicolai T. 31. Vukovic A., Hanson H., Murphy S., Mercurio D., Sheedy and Weiss M. 2020. Classical versus Controlled Rapid C. and Arnold D. 2019. Apneic Oxygenation Reduces Sequence Induction and Intubation in Children with Hypoxemia During Endotracheal Intubation in the Pediatric Bleeding Tonsils (a retrospective audit). Acta Anaesthesiol Emergency Department. Am J Emerg Med. 37(1): 27-32. Scand. 64:41-47. 32. Gagey A., de Queiroz Siqueira M., Desgranges F., Combet 24. Allen L., Engelhardt T. and Lendrum R. 2014. Do Not S., Naulin C., Chassard D., and Bouvet L. 2016. Ultrasound Know Where to Press? Cricoid Pressure in the Very Young. Assessment of the Gastric Contents for the Guidance of the Eur J Anaesthesiol. 31(6): 333-334. Anaesthetic Strategy in Infants With Hypertrophic Pyloric 25. Effect of Cricoid Force on Airway Calibre in Children: A Stenosis: A Prospective Cohort Study. Br J Anaesth. 116(5): Bronchoscopic Assessment. Br J Anaesth. 104(1): 71-74. 649-654. 26. Dotson K., Kiger J., Carpenter C., Lewis M., Hill J., Raney L. and Losek J. 2010. Alignment of Cricoid Cartilage and .

15-Dionys.indd 119 12/01/2021 12:19

(Acta Anaesth. Belg., 2020, 71, 121-128) A prospective randomized double-blind trial of the efficacy of bilateral lumbar erector spinae plane blocks on the 24h morphine consumption after posterior lumbar inter-body fusion surgery : preliminary results

O. De Fré (*), D. Van Aken (**), B. Versyck (***), L. Sermeus (****), V. Saldien (*,*****), M.B. Breebaart (*,*****)

Abstract : Objective : This study tested the hypothesis Keywords : Pain, postoperative ; analgesia ; lumbosa- that the 24-h postoperative morphine consumption is cral region ; surgery ; morphine. significantly lower in patients undergoing posterior lumbar inter-body (PLIF) surgery with lumbar erector spinae plane blocks (ESPB) compared to sham blocks. Introduction Background : Spine surgery is associated with considerable postoperative pain, which can be challenging to Posterior spine surgery ranks among the most treat. Locoregional techniques suitable for spine surgery painful surgical procedures and can be challenging should cover the dorsal root of the spinal nerves at to treat (1, 2). Musculoskeletal pain in posterior the levels where surgery is performed. The ESPB is a spine surgery arises from iatrogenic mechanical locoregional technique with promising results and was damage, intraoperative retraction and partial devas- recently described at the thoracic level. cularisation and denervation to structures such as Methods : This prospective randomized double-blind bone, ligament, muscles, intervertebral disks and multicenter study with preliminary results randomly zygopophyseal joints. In addition, neuropathic pain allocated 20 of the 80 planned adult patients (according arises from compression and damage to nerve roots to our power analysis), who underwent elective 1 or 2-level PLIF surgery during general anesthesia to exiting the spinal canal and sometimes damage to one of two groups as follows : bilateral ESPB (20 mL the spinal cord itself (3, 4). The intensity of pain 0.25% levobupivacaine) or bilateral sham block (20 mL is directly proportional to the number of vertebrae NaCl 0.9%). Primary endpoint was 24-h postoperative involved in the surgery (1). Peripheral as well morphine consumption. Secondary endpoints included 72-h morphine consumption, intraoperative sufentanil

dosage, postoperative pain scores at regular time O. De Fré, MD ; D. Van Aken, MD ; B. Versyck, MD, intervals, time to first postoperative mobilisation and the PhD ; L. Sermeus, MD, PhD ; V. Saldien, MD, PhD ; MB Quality of Recovery-40 survey score. Breebaart, MD, PhD Results : A bilateral ESPB did not decrease morphine (*) Dpt of Anesthesia, Antwerp University Hospital, 2650 consumption in the first 24 hours postoperative [ESPB Edegem, Belgium. (**) Dpt of Anesthesia, AZ Klina, 2930 Brasschaat, Belgium. group : 23.9 (15) mg vs Sham group : 20.8 (17.1) mg, (***) Dpt of Anesthesia, AZ Turnhout, 2300 Turnhout, p = 0.73]. Likewise, groups did not differ significantly Belgium and Dpt of Anesthesia, Catharina Ziekenhuis, 5623 in morphine consumption at 72 hours postoperative, EJ Eindhoven, The Netherlands. intraoperative sufentanil dosage, postoperative pain (****) Dpt of Anesthesiology, Cliniques universitaires Saint scores at regular time intervals, time to first mobilisation Luc, Brussels, Belgium. (*****) Faculty of Medicine and Health Science, University of and Quality of Recovery-40 survey scores at day one and Antwerp, 2610 Wilrijk, Belgium. three postoperative. Corresponding author : De Fré Olivier, MD, Dpt of Discussion and Conclusion : A bilateral ESPB did not Anesthesiology, Antwerp University Hospital, Wilrijkstraat prove to be superior to bilateral sham blocks in PLIF 10, 2650 Edegem, Belgium surgery regarding reducing morphine requirements, E-mail : [email protected] pain scores, PONV and earlier mobilisation in this small Paper submitted on May 31, 2020 and accepted on Jun 4, 2020. preliminary study. More research with larger study Conflict of interest : none. populations is necessary to be able to draw definitive This study was presented as an abstract on the SARB Graduation conclusions on this subject. Day Meeting 23/05/2020. Funding : The study was sponsored using a Belgian Association of Regional Anaesthesia research grant.

16-De Fré.indd 121 12/01/2021 12:19 122 o. de fré et al. as central sensitization further contributes to the sumption postoperatively compared to a sham block development of increased pain (5). with 20 ml of NaCl 0.9% after posterior lumbar In spine surgery, postoperative pain can often interbody fusion surgery. Secondary objectives be severe and difficult to treat, requiring high doses include the effect of the ESPB on numeric rating of narcotics, certainly if an unimodal approach is scale (NRS) pain scores in rest at different fixed used to achieve pain control (1, 2). Many caregivers time intervals, time to first mobilization in chair are reluctant to prescribe liberal doses of opioids to (in hours), time to first walk 20 meters (in hours), achieve adequate analgesia as this may be associated total morphine dose requirement during first 72 with side effects such as respiratory depression, postoperative hours (in milligrams) and the Quality sedation, and nausea. Many techniques (e.g., of Recovery 40 score (QoR-40) at day 1 and day 3 epidural catheters, spinal and epidural morphine, after surgery. or local infiltration) have been combined in order to decrease opioid consumption after spinal surgery Methods (5, 6). These neuraxial techniques have side effects and are not routinely used. Local anesthetic wound This was an investigator-initiated prospective infiltration is often performed, unfortunately with randomized double-blind multicentre pilot trial. short-lived effect (7). The study was performed in accordance with The sensation of back pain can be divided the Declaration of Helsinki (Fortaleza, Brazil ; in segmental type of pain via nociceptors and non October 2013) and Good Clinical Practice (GCP) segmental type of pain sensation trough sympathic guidelines. The study has been approved by the and parasympathic innervation. Both types of ethics committee on 24/09/2018 at Antwerp Uni- innervation pass through the dorsal ramus nerve to versity Hospital, Wilrijk, Belgium, and AZ Klina reach the spinal cord via spinal nerves. A blockade Hospital, Brasschaat, Belgium (reference : of the dorsal ramus nerve theoretically has an effect B300201837508). The chairperson of the ethics on both segmental and non-segmental type of committee was professor doctor Patrick Cras. The innervation (8, 9). For this reason, the dorsal ramus trial was prospectively registered at at https://www. nerve block has been proposed as a technique for clinicaltrials.gov (reference : NCT03825198) and acute postoperative and chronic back pain control was monitored by the clinical trial center of the (10, 11). In a recent series of case reports, a bilateral Antwerp University Hospital. block of the lumbar dorsal ramus nerve showed Patients scheduled for elective 1 or 2 level improved pain scores and reduced morphine posterior lumbar inter-body fusion surgery in AZ consumption after spine surgery (12). The erector Klina Hospital and Antwerp University Hospital spinae plane block (ESPB) is a relatively new were asked for written informed consent by a member truncal interfascial plane block originally described of the Anesthesiology department. Recruitment by Forero et al. at the thoracic level for chronic pain. occured during the preoperative consultation ; it It involves the deposition of local anesthetics in a opened on July 9, 2019 ; and was paused on January plane between the erector spinae muscle and the 31, 2020. Inclusion criteria were as follows : (1) transverse process resulting in considerable spread American Society of Anesthesiologists (ASA) in cranio-caudal and medial-lateral directions physical status of 1-3, (2) age of 18-75 years, and (13). Cadaveric studies have confirmed consistent (3) normal liver and renal function. Exclusion staining of the dorsal rami with variable effect on criteria were as follows : (1) body mass index the ventral rami (14-16). To our knowledge only (BMI) of less than 18 or BMI of more than 35 ; two small randomized controlled trials concerning a (2) allergy to one or more medications used in the lumbar erector spinae plane block in posterior spine study, including epinephrine, levobupivacaine, surgery have been published just after the start of dexamethasone, propofol, sufentanil, rocuronium, this trial reporting effective analgesia by blocking ketorolac, morphine, ketamine, dehydrobenzperidol, the dorsal rami of spinal nerves as they run trough ondansetron, and alizapride ; (3) chronic strong the interfascial plane where the local anesthetics are opioid use (>3 administrations per week) ; (4) con- deposited (17, 18). traindications to a regional anesthetic technique ; In this multi-center prospective randomized and (5) patient refusal or no informed consent or double blind placebo controlled clinical trial with both. preliminary results we wanted to investigate if Patients were assigned consecutive numbers bilateral ESPB with 20 ml levobupivacaine 0.25% upon inclusion in the study. These numbers were can provide a reduction in 24-hour morphine con- randomly allocated (1:1) to the ESPB or the sham

16-De Fré.indd 122 12/01/2021 12:19 efficacy of bilateral lumbar erector spinae plane blocks on the 24h morphine consumption 123 block group by using a web-based randomization 1 g IV around the clock and by patient-controlled system, QMinim. QMinim uses stratified rando- intravenous analgesia (PCIA) with morphine at a mization, and stratification was carried out according concentration of 1 mg/mL and dehydrobenzperidol to site, gender, and levels of surgery. Online 0.05 mg/mL. The PCIA pump was programmed as randomization was carried out by an independent follows : no continuous infusion, a bolus dose of anesthesiologist who also prepared the medication. 1.5-mg morphine, a lockout interval of 8 min, and The ESPB study medication was 20 mL of an hourly limit of 7.5 mg. If pain management on levobupivacaine 0.25% (Chirocaine, AbbVie, Lake the PACU was inadequate, defined as a numeric Bluff, IL, USA). The sham block group received rating scale (NRS) pain score of more than 3 (0 = 20 mL of NaCl 0.9% (B. Braun, Melsungen, no pain to 10 = worst imaginable pain]), additional Germany). All investigators, staff, and patients were boluses of 1-mg morphine IV were administered blinded to the treatment groups. Both solutions and by the PACU nurses with a total additional dose syringes appeared identical. of morphine limited to 0.15 mg/kg. If pain mana- All patients received bilateral ESPB. These gement with morphine remained inadequate, an IV blocks were placed preoperatively in a separate ketamine (Ketalar, Pfizer, New York, NY, USA) block room with ultrasound after obtaining bolus (0.2 mg/kg) was administered. intravenous (IV) access and application of standard The primary endpoint was the morphine ASA monitoring. The blocks were executed as consumption during the first 24 h postoperatively described by Forero et al. (13) and modified for in milligrams and was determined from the PCIA the lumbar level. The patient was placed in the pump. As secondary endpoint, the total morphine lateral or sitting position. A curved array probe or consumption in milligrams, during the first 72 a high-frequency linear probe, depending on the postoperative hours, were extracted out of the PCIA body mass index (BMI) of the patient, was placed pump. Pain scores at rest were assessed with the in longitudinal alignment, 2-3 cm lateral to the NRS score (0 = no pain, 10 = worst imaginable pain) vertebral column. The transverse processes of the and tested at regular time intervals : at the time of vertebrae at the level of surgery, the erector spinae inclusion, in the PACU (T0 = arrival in PACU, T0 + muscle, and the psoas muscle were identified. In 15 min, T0 + 30 min), and on the ward (twice daily : case of two-level surgery, the transverse process of morning and evening until postoperative day 3). the upper level was considered as the target. A 5- Time to first mobilization to a chair (in hours since or 8-cm 22-gauge ultrasound needle was inserted T0) and time to first walk of 20 m (in hours since with an in-plane technique in a cephalad-to-caudal T0) were noted in the patients’ study diary. The direction until bone contact with the top of the QoR-40 scores were calculated from the responses transverse process was made. After slight retraction to a standard questionnaire at postoperative days 1 of the needle, 20 mL of the study medication was and 3. The QoR-40 is a widely used and extensively injected behind the erector spinae muscle. The same validated measure of quality of recovery. It is a procedure was repeated on the contralateral side. 40-item questionnaire on quality of recovery from General anesthesia – propofol 2-3 mg/kg, sufen- anesthesia that has been shown to measure health tanil 15 μg, and rocuronium 0.5 mg/kg – was then status after surgery (19, 20). Other endpoints include induced in a standardized way. After tracheal intu- the intraoperative sufentanil dose (in microgram), bation, anesthesia was maintained with sevoflurane need for rescue analgesia, postoperative nausea and and intraoperative analgesia provided with sufentanil vomiting score according to hospital protocol, need increments at the discretion of the attending anes- for rescue postoperative anti-emetics, time to first thesiologist. At the end of surgery, patients received meal and possibility to defecate within 72 hours acetaminophen 1 g IV, ketorolac 0.5 mg/kg IV postoperatively. (maximum of 30 mg), and a morphine loading SPSS software version 26 (SPSS, Chicago, dose (0.1 mg/kg) IV to manage postoperative pain. IL, USA) was used for statistical analysis. The Patients were extubated in the operating theatre and distribution of variables was evaluated for normality admitted to the Post-Anesthesia Care Unit (PACU). using the Shapiro-Wilk test of normality. Continues All patients received dexamethasone 5 mg IV as variables were reported using mean ± standard postoperative nausea and vomiting prophylaxis, deviation, categorical variables were reported as supplemented, if necessary, by ondansetron 4 mg IV counts or percentages. Continuous variables were and further with alizapride 50 mg IV as rescue. analysed using Independent samples t test if they Postoperative pain in the PACU and on the ward were normally distributed or by Mann-Whitney was treated with regular doses of acetaminophen U test if otherwise. Categorical variables were

16-De Fré.indd 123 12/01/2021 12:19 124 o. de fré et al. analysed using the Chi-square test or Fisher’s exact test. A two sided p value < 0.05 was considered significant. Our sample size calculation was based on data from a randomized controlled trial comparing the effect of systemic infused lidocaine with placebo on the 24-h morphine requirement in posterior lumbar arthrodesis by Dewinter et al (21). Since prior to the start of this study no information regarding the effect of bilateral lumbar ESPB on morphine consumption with posterior spine surgery was known, we considered a 25% reduction in PCIA morphine consumption to be clinically relevant. To calculate the sample size, we assumed a mean of 51-mg morphine with standard deviation of 19 mg (he mean morphine consumption for the placebo group of the above-mentioned trial (21)), a type 1 failure risk of 5%, and a type 2 failure risk of 20%. Thirty-five patients were required in each group to detect a 25% reduction in morphine equivalent over 24 h. The sample size calculation was based on an independent samples t test. We planned to include 80 patients in total to compensate for potential dropouts and uncertainty in predicting the actual standard deviation. However, an interim analysis Fig. 1. — Consort Flow Chart. after 20 patients was performed because the study was paused on the January 31, 2020 owing to the Table 1 COVID-19 pandemic. Demographic and operative characteristics of study patients

Results Sham Group ESPB Group P (n = 11) (n = 6) value Age (years) 53.4 ± 15.8 58.2 ± 5.9 0.50* Between July 9, 2019 and the January BMI 26.8 ± 3.4 26.3 ± 4.5 0.82* 31, 2020 Fifty-four consecutive patients were Gender (male/female) 4/7 2/4 1^ assessed for eligibility. After the January 31, 2020 ASA score (I/II/III) 9/2/0 2/4/0 0.11^ the study was paused because of the COVID-19 Surgical Level (I/II) 11/0 3/3 0.03^ pandemic. Thirty-four patients were excluded Preoperative NRS score 4.5 ± 2.7 3.3 ± 2.2 0.40° before randomization as they did not meet inclusion Length of Surgery (min) 140 ± 42.6 184.3 ± 72.6 0.13* criteria (n = 21), declined to participate (n = 2) or Site (UZA/KLINA) 4/7 2/4 1^ Expected NRS score 7.7 ± 1.5 7.5 ± 1.8 0.59° for logistical reasons (n = 11). Eligible patients for Values are presented as number or mean +/ standard deviation; ESPB, this study are shown in a Consolidated Standards of erector spinae plane block; M, male; F, female; ASA, American Society Reporting Trials (CONSORT) flow chart (Fig. 1). of Anesthesiologists; *Independent sample t test ° Mann-Whitney U Twenty subjects were randomly allocated using the test ^ Fisher’s Exact Test. Qminim software system to one of the study groups. Study intervention was discontinued in two subjects of surgery differed significantly between both during follow-up. One subject withdrew informed groups (p <0.05) with patients in the ESPB group consent just after randomization. The other undergoing more extensive surgery. (Table 1). patient suffered from a spontaneous intracerebral No significant difference in mean cumulative bleeding for which study follow-up was cancelled. morphine consumption during the first 24 hours was One patient’s data was excluded from analysis found between patients in the ESPB group versus given outlier data owing to the need for revision patients in the sham block group (23.9 ± 15.0 mg surgery just after the follow-up period. Patient vs 20.8 ± 17.1 mg, p = 0.73). Likewise, groups did demographics, ASA classification, length of surgery, not differ significantly in morphine consumption at hospital sites, preoperative and expected pain scores 72 hours postoperative, postoperative pain scores at were comparable for both groups. Extensiveness regular time intervals, time to first mobilization, time

16-De Fré.indd 124 12/01/2021 12:19 efficacy of bilateral lumbar erector spinae plane blocks on the 24h morphine consumption 125 Table 2 Cumulative Morphine Consumption and QoR-40 score at 24h and 72h postoperative Sham Group ESPB group P value* (n = 11 ) (n = 6 ) Mean 24 h cumulative morphine consumption (mg) 20.8 ± 17.1 23.9 ± 15.0 0.73 Mean 72 h cumulative morphine consumption (mg) 48.5 ± 34.6 60.2 ± 47.8 0.59 Mean day 1 postop. QoR-40 score 162.0 ± 21.0 167.3 ± 24.1 0.64 Mean day 3 postop. QoR-40 score 175.8 ± 18.6 169.7 ± 14.4 0.49 Values are presented as mean ± standard deviation. ESPB, erector spinae plane block; QoR-40, Quality of Recovery 40 score. *Independent T sample test or Mann-Whitney U test depending on normality.

Table 3 Comparison of Numeric Rating Scale Pain scores at postoperative time points Sham Group ESPB group P value* (n = 11) (n = 6) NRS At arrival on PACU ward (T0) 3.9 ± 3.4 4.2 ± 4.2 0.73 NRS at T0 + 15 minutes 5.2 ± 3.0 6.0 ± 3.1 0.66 NRS at T0 + 30 minutes 5.1 ± 2.8 6.5 ± 1.5 0.35 NRS at morning postop. day 1 3.7 ± 2.0 3.2 ± 1.8 0.66 NRS at evening postop. day 1 4 ± 2.8 4 ± 2.6 1 NRS at morning postop. day 2 3.6 ± 2.5 4.2 ± 1.7 0.59 NRS at evening postop. day 2 3.1 ± 2.6 3.6 ± 1.3 0.59 NRS at morning postop. day 3 2.9 ± 2.0 3.3 ± 1.4 0.63 NRS at evening postop. day 3 2.9 ± 1.8 3.75 ± 1.9 0.57 Values are presented as mean ± standard deviation. NRS, Numeric Rating Scale; PACU, Post-Anesthesia Care Unit; T0, Time of arrival at PACU ward. *Independent T sample test.

Table 4 Other endpoints Sham Group ESPB group P value (n = 11) (n = 6) Intraoperative Sufentanil dosage (µg) 25.9 ± 9.4 24.2 ± 13.2 0.59* Ketamine rescue analgesia (yes/no) 1/10 0/6 1° Highest PONV score ( 0/1/2/3) 10/1/0/0 5/0/0/1 0.3^ Rescue anti-emetic (yes/no) 4/7 2/4 1° Time to first mobilisation to chair (hours) 35.4 ± 19.8 45.9 ± 24.4 0.44* Time to first walk > 20 meter (hours) 45.3 ± 19.0 58.7 ± 13.9 0.31¨ Time to first meal (hours) 24.2 ± 13.1 19.2 ± 16.9 0.5¨ Defecation within 72h postop.(yes/no) 3/8 3/3 0.6° Values are presented as numbers or mean ± standard deviation. PACU, Post-Anesthesia Care Unit; PONV, Post- Operative Nausea and Vomiting. *Mann-Whitney U test, °Fisher’s exact test, ^Chi-square test, ¨Independent Student T Test.

to first walk 20 meters and Quality of Recovery-40 Discussion survey scores at day one and three postoperative. (Tables 2, 3). In this small prospective randomized double No significant difference was observed blind placebo controlled clinical trial with preli- between both groups regarding tertiary endpoints : minary results bilateral lumbar erector spinae plane intraoperative sufentanil requirement, necessity of blocks in posterior lumbar fusion surgery failed ketamine as rescue analgesia, postoperative nausea to show a significant benefit regarding morphine and vomiting score, necessity of a rescue anti-emetic, consumption, quality of recovery, pain scores, the time to eat first meal and the percentage of subjects incidence of postoperative nausea and vomiting, the able to defecate within 72 hours postoperatively. need for rescue analgesics and time to ambulation. (Table 4).

16-De Fré.indd 125 12/01/2021 12:19 126 o. de fré et al. Recently ultrasound-guided interfascial plane and 30 minutes observation monitoring periods blocks emerged as new methods of administering to account for the time to reach systemic local local anesthetics to specific anatomic locations to anesthetic peak concentrations (27). There have provide analgesia. Local anesthetics are injected been anecdotal reports suggesting lower limb in interfascial planes which act as potential spaces. weakness presumably due to spread of the local The injectate spreads along the path of the least anesthetic to the lumbosacral plexus (28). resistance to block neural structures away from The observations of this preliminary results the injection site. There is much uncertainty study are in contrast with similar small studies, regarding mechanism of action, effectiveness and published only after the start of this trial, which efficacy (22). The ESPB is a relatively new truncal have reported positive outcomes regarding pain interfascial plane block originally described by scores and opioid requirement. A retrospective Forero et al. at the thoracic level for chronic pain. study by Ueshemi et al. demonstrated lower pain It involves the deposition of local anesthetics in a scores and less fentanyl consumption after lumbar plane between the erector spinae muscle and the spine surgery in patients receiving an ESPB (29). transverse process resulting in considerable spread Singh et al. published a randomized controlled in cranio-caudal and medial-lateral directions trial reporting reduced pain scores and lower (13). Cadaveric studies have confirmed consistent morphine requirement in patients undergoing single staining of the dorsal rami with variable effect on level lumbar laminectomy or single level lumbar the ventral rami (14-16). The ESPB should provide disk herniation receiving bilateral ESPB (17). A effective analgesia by blocking dorsal rami of spinal randomized controlled trial by Yayik et al. comparing nerves as they run trough the interfascial plane bilateral lumbar ESPB to no block in lumbar where the local anesthetics are deposited (23). decompression surgery demonstrated significant Potential benefits of the lumbar ESPB include reduced pain scores in rest and movement, lower the ease of performance with clear landmarks for tramadol use and longer time to first analgesic (18). ultrasound anatomy. The technique is inherently A potential reason why this preliminary results trial safe, as the target site for injection is a muscular found no significant analgesic benefit, in contrast to plane and there is practically no risk for mechanical the above mentioned studies, could be the different nerve contact. Other benefits include the possible type of surgery patients underwent. In the study of reduction in perioperative opioid consumption. Yayik et al (18). patients had decompression surgery The ESPB can be performed in patients under without fusion, while patients in the randomized anticoagulant therapy or with coagulopathies (24). controlled trial by Singh et al (17). underwent Furthermore, hemodynamic instability due to sym- only single level laminectomy or single level disk pathetic blockade, as with epidural and spinal herniation, whereas patients in our study had single anesthesia, occurs rarely. The most likely explanation or double lumbar posterior fusion surgery with is that the actual mass of local anesthetic reaching instrumentation, which can be considered more the epidural space or ventral rami is insufficient complex and more extensive. A second reason to produce a clinically-detectable effect (25). Up might be that the frequency of single and double until now no interference with evoked potentials, a level surgery in the ESPB group in this preliminary technique used to detect injuries or ischemia of the results trial differed significantly from the sham neural pathway during surgery, has been described block group. More extensive surgery was required in the literature (23, 26). in the ESPB group (p = 0.03). As mentioned in Possible risks consist primarily of local anes- the introduction, the intensity of pain is directly thetic systemic toxicity. Since substantial doses are proportional to the number of vertebrae involved in considered necessary, there is a clinically significant the surgery, which might explain the lack of benefit risk for local anesthetic systemic toxicity, as in this study (1). Third, the study of Singh et al (17) with any high volume fascial block. The third used a stronger concentration of local anesthetic American Society of Regional Anesthesia and Pain (bupivacaine 0.5%) compared to levobupivacaine Medicine Practice Advisory on local anesthetic 0.25% in our trial while using identical volumes systemic toxicity advices to add epinephrine to (20mL each side) for the ESPB. This might also the local anesthetic mixture in high volume local explain the analgesic benefit in their study, however anethetic blocks to reduce systemic absorption and with a potential increase in risk of local anesthetic maximum local anesthetic plasma concentrations. systemic toxicity. Other recommendations include the use of diluted This preliminary results study has limitations. concentrations of less cardiotoxic local anesthetics The major limitation is the small sample size. For

16-De Fré.indd 126 12/01/2021 12:19 efficacy of bilateral lumbar erector spinae plane blocks on the 24h morphine consumption 127 this reason, no conclusions can be made regarding 4 Sharma S, Balireddy RK, Vorenkamp KE and Durieux the primary outcome. It cannot be excluded that a ME. 2012. Beyond opioid patient-controlled analgesia: a systematic review of analgesia after major spine surgery. significant effect on the primary outcome may be Reg Anesth Pain Med.37(1):79-98. unveiled after inclusion of a larger patient number. 5 Bajwa SJ and Haldar R. 2015. Pain management following For the same reason no conclusions can be made spinal surgeries: An appraisal of the available options. J Craniovertebr Junction Spine.6(3):105-10. regarding secondary or tertiary outcomes. In this 6 Benyahia NM, Verster A, Saldien V, Breebaart M, Sermeus study we opted not to place catheters in the bila- L and Vercauteren M. 2015. Regional anaesthesia and teral ESPB sites. The use of a catheter in the inter- postoperative analgesia techniques for spine surgery - a review. Rom J Anaesth Intensive Care.22(1):25-33. fascial plane has the advantage of continuously 7 Perera AP, Chari A, Kostusiak M, Khan AA, Luoma AM administering analgesics over a certain period of and Casey ATH. 2017. Intramuscular Local Anesthetic time, but it also has the potential for catheter related Infiltration at Closure for Postoperative Analgesia in problems such as local site inflammation, catheter Lumbar Spine Surgery: A Systematic Review and Meta- Analysis. Spine (Phila Pa 1976).42(14):1088-95. migration and accidental removal. Uncertainty 8 Edgar MA. 2007. The nerve supply of the lumbar regarding administering local anesthetics in the intervertebral disc. J Bone Joint Surg Br.89(9):1135-9. disrupted tissue plane and the risk of infection were 9 Higuchi K and Sato T. 2002. Anatomical study of lumbar spine innervation. Folia Morphol (Warsz).61(2):71-9. considered important reasons to choose for a single 10 Miyakoshi N, Shimada Y, Kasukawa Y, Saito H, Kodama H shot technique. In this pilot trial epinephrine, as and Itoi E. 2007. Total dorsal ramus block for the treatment suggested by the third American Society of Regional of chronic low back pain: a preliminary study. Joint Bone Spine.74(3):270-4. Anesthesia and Pain Medicine Practice Advisory on 11 Williams MG, Rigney B, Wafai A and Walder A. 2019. Are local anesthetic systemic toxicity, was not added to dorsal ramus nerve blocks the solution to postoperative the local anesthetic solution to prevent unblinding lumbar spine surgery pain? J Spine Surg.5(2):245-50. by potentially observing a rise in heart rate. Another 12 Al-Alami A, Abou El Ezz A and Kassab F. 2015. Ultrasound Guided Dorsal Ramus Nerve Block for Reduction of limitation is the absence of information regarding Postoperative Pain in Patients Undergoing Lumbar Spine preoperative pain medication, which might have Surgery: A Case Series Imaging Study. Middle East J influenced our results. The confounding effect is Anaesthesiol.23(2):251-6. 13 Forero M, Adhikary SD, Lopez H, Tsui C and Chin KJ. probably rather small as patients taking profound 2016. The Erector Spinae Plane Block: A Novel Analgesic amounts of opioids were excluded. Arguably, Technique in Thoracic Neuropathic Pain. Reg Anesth Pain patients on large amounts of opioid prescriptions Med.41(5):621-7. 14 Aponte A, Sala-Blanch X, Prats-Galino A, Masdeu J, probably experience more benefit from a multimodal Moreno LA and Sermeus LA. 2019. Anatomical evaluation analgesia technique compared to patients on of the extent of spread in the erector spinae plane block: a limited doses or no opioids. Lastly, no information cadaveric study. Can J Anaesth.66(8):886-93. regarding sensory blockade was gathered due to the 15 Breebaart MB, Van Aken D, De Jong L, Michielsen J, Versyck B, Nassauw L. 2019. The spread of the ultrasound- chosen blinding method. guided injectate after a lumbar and thoracic erector spinae plane block. A cadaveric study. Acta Anaesth Belg. Conclusion 70(4):191-6. 16 Yang HM, Choi YJ, Kwon HJ, O J, Cho TH and Kim SH. 2018. Comparison of injectate spread and nerve No conclusion regarding the effect of bilateral involvement between retrolaminar and erector spinae erector spinae plane blocks on 24h morphine con- plane blocks in the thoracic region : a cadaveric study. Anaesthesia.73(10):1244-50. sumption in PLIF surgery could be drawn because 17 Singh S, Choudhary NK, Lalin D and Verma VK. 2019. of the small sample size in this preliminary results Bilateral Ultrasound-guided Erector Spinae Plane Block study. More research with larger study populations for Postoperative Analgesia in Lumbar Spine Surgery: A is necessary to be able to draw definitive conclusions Randomized Control Trial. J Neurosurg Anesthesiol. 18 Yayik AM, Cesur S, Ozturk F, Ahiskalioglu A, Ay on this subject. AN, Celik EC and Karaavci NC. 2019. Postoperative Analgesic Efficacy of the Ultrasound-Guided Erector References Spinae Plane Block in Patients Undergoing Lumbar Spinal Decompression Surgery: A Randomized Controlled Study. 1 Lai LT, Ortiz-Cardona JR, Bendo AA. 2012. Perioperative World Neurosurg.126:e779-e85. pain management in the neurosurgical patient. Anesthesiol 19 Gornall BF, Myles PS, Smith CL, Burke JA, Leslie K, Clin.30(2):347-67. Pereira MJ and Bost JE, et al. 2013. Measurement of quality 2 Nielsen RV, Fomsgaard JS, Dahl JB, Mathiesen O. 2014. of recovery using the QoR-40: a quantitative systematic Insufficient pain management after spine surgery. Dan Med review. Br J Anaesth.111(2):161-9. J.61(5):A4835. 20 Myles PS, Weitkamp B, Jones K, Melick J, Hensen S. 3 Datta G, Gnanalingham KK, Peterson D, Mendoza N, O’Neill 2000. Validity and reliability of a postoperative quality of K, Van Dellen J, McGregor A, et al. 2004. Back pain and recovery score: the QoR-40. Br J Anaesth.84(1):11-5. disability after lumbar laminectomy: is there a relationship 21 Dewinter G, Moens P, Fieuws S, Vanaudenaerde B, Van de to muscle retraction? Neurosurgery.54(6):1413-20 ; discus- Velde M and Rex S. 2017. Systemic lidocaine fails to improve sion 20. postoperative morphine consumption, postoperative reco-

16-De Fré.indd 127 12/01/2021 12:19 128 o. de fré et al. very and quality of life in patients undergoing posterior 26 Melvin JP, Schrot RJ, Chu GM and Chin KJ. 2018. Low spinal arthrodesis. A double-blind, randomized, placebo- thoracic erector spinae plane block for perioperative controlled trial. Br J Anaesth.118(4):576-85. analgesia in lumbosacral spine surgery: a case series. Can J 22 Elsharkawy H, Pawa A and Mariano ER. 2018. Interfascial Anaesth.65(9):1057-65. Plane Blocks: Back to Basics. Reg Anesth Pain Med. 27 Neal JM, Barrington MJ, Fettiplace MR, Gitman M, 43(4):341-6. Memtsoudis SG, Morwald EE and Rubin DS, et al. 2018. 23 Chin KJ, Dinsmore MJ, Lewis S and Chan V. 2019. Opioid- The Third American Society of Regional Anesthesia and sparing multimodal analgesia with bilateral bi-level erector Pain Medicine Practice Advisory on Local Anesthetic spinae plane blocks in scoliosis surgery: a case report of Systemic Toxicity: Executive Summary 2017. Reg Anesth two patients. Eur Spine J. Pain Med.43(2):113-23. 24 Tsui BCH, Fonseca A, Munshey F, McFadyen G, Caruso 28 Selvi O and Tulgar S. 2018. Ultrasound guided erector TJ. 2019. The erector spinae plane (ESP) block: A pooled spinae plane block as a cause of unintended motor block. review of 242 cases. J Clin Anesth.53:29-34. Rev Esp Anestesiol Reanim.65(10):589-92. 25 Tulgar S, Selvi O, Senturk O, Serifsoy TE and Thomas 29 Ueshima H, Inagaki M, Toyone T and Otake H. 2019. DT. 2019. Ultrasound-guided Erector Spinae Plane Efficacy of the Erector Spinae Plane Block for Lumbar Block : Indications, Complications, and Effects on Acute Spinal Surgery : A Retrospective Study. Asian Spine J. and Chronic Pain Based on a Single-center Experience. 13(2):254-7. Cureus.11(1):e3815.

16-De Fré.indd 128 12/01/2021 12:19 (Acta Anaesth. Belg., 2020, 71, 129-133) Comparison between the different pharmacokinetic models of Propofol target controlled infusion : a narrative review

K.W. Abdul-Latif (*), M. Van de Velde (*,**), L. Al Tmimi (*,**)

Abstract : Since the introduction of the target-controlled clearance are often not included due to the complex infusions (TCI) in 1996, researchers are actively trying mathematical relationship of these co-variates with to implement the most suitable simulating programs the dose (4). When giving an appropriate dose of a that can predict drug concentrations efficiently, in the drug, the ultimate goal is a specified clinical effect, intravenously anesthetized patients. These infusions use for which a precise therapeutic concentration at the a selected set of pharmacokinetic parameters, which is site of drug action is required. This dose-response incorporated into a programable infusion pump, known as a ‘’model’’. In recent practice, only two models are relationship (Fig. 1) can be categorized into three routinely used for Propofol TCI. This is because their parts : the pharmacokinetic phase, which represents appropriate target concentrations are proved by clinical the relationship between the dose administered and trials. These models are known as Marsh and Schnider. plasma concentration, the pharmacodynamic phase The TCI technique is becoming a cornerstone in anes- that outlines the relationship between effect organ thesia practice. Besides, different TCI models for concentration and clinical effect, and the third part delivering analgesics and sedative medications have is the coupling between pharmacokinetics and been developed. Different pharmacokinetic Propofol TCI dynamics. models are available in adults, such as Marsh, Schüttler, Historically, there have been two methods to and the Schnider model. Also, there are several Propofol administer intravenous medications during anes- TCI models adapted for use in children such as Paedfusor, thesia : bolus dose or continuous infusion. Bolus Kataria, and the Short model. Notwithstanding, each doses are usually injected with a hand syringe, model has its limitation. The latter based on which patient’s data the model equation and algorithm were while infusions are given with an infusion pump. derived. A new method for intravenous administration We conclude that there is no evidence to promote the use of medications was introduced in 1996, called of one model of Propofol TCI over another and all have Target-controlled infusion (TCI) (4). The TCI demonstrated safe for use in clinical practice, specifically technique is becoming a cornerstone in anesthesia in adults. However, anesthesiologists should use the practice. Besides, different TCI-models (for delive- TCI model which they are most familiar with. Future ring analgesics and sedative medications) have been investigation should give more attention to validate a developed, such as the Marsh (5), the Schüttler (6), model for severely obese patients and also the pediatric and the Schnider model (7). Also, there are several population. Propofol TCI models adapted for use in children such as Paedfusor, Kataria, and the Short model Keywords : Propofol ; target controlled infusion ; phar- (8). Notwithstanding, each model has its limitation. macokinetic model. The latter based on which patient’s data the model equation and algorithm were derived. In the pre-

Introduction K.W. Abdul-Latif, MD ; M. Van de Velde, MD, PhD, EDRA ; L. Al tmimi, MD, PhD. The discovery of Propofol and the shorter- (*) Department of Anesthesiology, University Hospitals acting opioid remifentanil in recent decades (1,2) Leuven, Leuven, Belgium. have enabled the development of technique of total (**) Department of Cardiovascular Sciences, KU Leuven, Leuven, Belgium. intravenous anesthesia (TIVA) in which anesthesia Corresponding author : Dr. Khalid Abdul-Latif, Department is administered exclusively through intravenous of Anesthesiology, University Hospitals Leuven, Herestraat route (3). To calculate the dose of an intravenous 49, 3000 Leuven, Belgium anesthetic, the weight of the patient is required and E-mail : [email protected] considered the only independent variable. While Paper submitted on May 31, 2020 and accepted on Jun 4, 2020. other variables such as patient’s age, sex, or creatinine Conflict of interest : none.

17-Latif.indd 129 12/01/2021 12:21 130 k.w. abdul-latif et al. sent article, we would like to provide anesthesia care providers an overview of the available pharmacokinetic models of Propofol TCI and its main advantages, pitfalls, and clinical applications.

Target-controlled infusions

Maintenance of a steady state of brain concentration of the drug is essential for the efficiency of the TCI. This state is achieved by implementing complex pharmacokinetic equations that describe Fig. 1. — The three-compartment pharmacokinetic model. the distribution of drugs between the various body Adapted from Al-Rifai, Z et al. (20) V1−3 = compartment compartments and allow for adjustments of target volumes, K12 = the rate constant between V1 and V2, K21 concentration to reach the desired clinical effect. = the rate constant between V2 and V1, etc. K10 = the rate constant for drug elimination from the central compartment. The idea of TCI system existed since 1968 ke0 is the rate constant for equilibration between plasma and when Kruger-Thiemer (9) marked out a theoretical effect-site concentrations. approach to reach and maintain a steady-state of plasma concentration of a drug. This principle is further investigated by many researchers like Vaughan-Tucker (10) and Schwilden. (11) Since The microprocessor calculates how much then, many algorithms have been published for bolus dose should be given and at how much rate targeting plasma concentrations (12-16) and for should the infusion runs subsequently to ensure targeting effect-site concentrations. (17,18) maintaining the desired plasma drug concentration. The TCI system is programmed with pharma- This calculation is frequently repeated (usually cokinetic models that mathematically describe every 10 seconds) so that the rate of drug-infused the processes of drug delivery and elimination, can be adjusted to maintain the targeted plasma as demonstrated in figure 1. TCI system uses the concentration. If this desired plasma concentration principle of Bolus/Elimination/Transfer (BET) is subsequently increased, then the system gives an (19) to reach a constant plasma level of the drug. additional bolus to the central compartment (V1), When a bolus is given, first the central compartment and the infusion rate is further increased to maintain (V1) would be filled with the drug, subsequently, the new targeted plasma concentration. While if the the infusion is used to compensate firstly for the desired plasma concentration is decreased, then the redistribution to the “vessel-rich” (V2) and the system stops the infusion for a calculated period “vessel-poor” (V3) compartments and secondly for till the plasma concentration decreases to the new drug elimination from the central compartment (by desired value, then the infusion is re-started at a metabolism and renal excretion) as described by k10 lower infusion rate. which is the rate constant for drug elimination from the central compartment. If the three compartments TCI Propofol models reach a steady-state (after long infusion time), then the infusion rate slows to match only the elimination A pharmacokinetic model is a mathematical rate. model through which the plasma concentration of a drug can be predicted after a bolus dose or an TCI systems infusion of varying duration. These models are derived by measuring the plasma concentration of The elemental components of the TCI system a drug after a bolus or infusion. So far, no evidence are a user interface, a microprocessor, and an promotes the use of one model over another, and infusion pump. The microprocessor controls the all have demonstrated to be safe in clinical practice, user interface by using suitable pharmacokinetic specifically in adults, and they have similar accuracy software. The TCI system incorporates an infusion and stability in predicting the plasma and effect-site pump with an infusion rate of up to 1200 ml/h. concentrations. Usually, the system warns the user of problems such Marsh (5) and Schnider (7) models are the as line obstruction and the syringe being almost main two models used. The Marsh model is first empty, through a visual and audible alarm system. published in 1991, and was based on Gepts (18) (20) three-compartmental model, and was derived on

17-Latif.indd 130 12/01/2021 12:21 propofol tci models, narrative review 131 very small groups of patients (18 patients). Typi- concentration targeting ; otherwise, a small insuf- cally, it targets plasma concentration but can target- ficient bolus might be given. effect site. The required variables are the total body It appears that, although there can be a weight (TBW) and the age (which is not used in rationale for choosing the Schnider model for the the calculation), which must be >16 years. It uses elderly and the obese, the most relevant issue is that variable compartment sizes but a bigger central the anesthesiologist uses the model that he is most compartment. There is also a modified version of the familiar with and titrates it to the individual patient. Marsh model, which uses a shorter plasma/effect- At present, two models are commonly used in site time constant. Therefore, its use is preferable in pediatric TCI pumps. Kataria et al. (22) developed a patients at higher risk of overdose. By using the total three compartmental model for Propofol in children body weight, this model overestimates induction in using three different pharmacokinetic modeling obese patients and, overall, has a faster induction techniques in an extended group of children due to the larger calculated central compartment, between 3 and 11 years. This model has fixed k10, which results in a larger loading dose. while V1-V3 has a linear correlation with weight. The other adult TCI model is the Schnider The Kataria model is intended to be used in children Model. It was derived in 1998 in a study that age 3–16 years and with a weight of >15 kg. The included 24 volunteers. Typically, it targets the Paedfusor model (23) was adapted from one of the effect-site concentration but can target the plasma preliminary models developed by Schüttler. In the concentration. It uses age, height (to calculate Paedfusor model, the central compartment volume lean body mass), and total body weight. The size and clearance have a non-linear correlation with of the vessel-rich compartment is variable and weight as age exceeds 12 years. Weight is also used is influenced only by age, and being smaller with here to calculate the elimination constant k10. advancing age. This model is limited to Body Mass Effect-site targeting Index (BMI) of less than 42 for males and less than 35 for females to prevent huge compartment sizes Plasma and effect-site concentration equili- being calculated. Schnider model reduced the rate brium rate rely on factors which consist of those of adverse events by giving a smaller initial bolus. control the rate of transfer of the drug to the effect- Overall, less Propofol is used. site (e.g. cardiac output and cerebral blood flow) in The major difference between these three- addition to the other properties of the medication compartment pharmacokinetic models is the estima- itself that determine the rate of crossing the blood- ted volume of the central compartment V1 as it is brain barrier (e.g. lipid solubility and the degree 15.9 liter in the Marsh model against 4.27 liter in of ionization). A rate constant can calculate the Schnider for a 70 kg person. The Schnider model time course of plasma to effect-site equilibration uses this figure for all patients (fixed V1and V 3) correctly called the keo (20). Estimating the keo (19) while The Marsh model ignores age and scales can be performed by measuring the clinical effects V1−3 at linear pattern to patient’s total body weight of the medication, e.g., using the Bi-spectral index (TBW), (20) thus an identical bolus is given to all system. If the plasma levels of a group of subjects patients of a given body mass for any chosen plasma are known, measurements can be used to estimate concentration. These weight-adjusted bolus doses keo. do not apply to obese, elderly, or unwell patients. In a plasma-targeted TCI system, the user The Schnider model calculates a gender-specific specifies the plasma concentration. The effect-site lean body mass (LBM) and uses it to adjust the concentration passively follows, with a time delay elimination rate constant k10. determined by keo or time to peak effect (TTPE) In practice, less initial plasma concertation of the medication in use. When the keo is used in is chosen when a critically ill or an old patient is combination with pharmacokinetic models, it would anesthetized using Marsh TCI model and later be possible to target the effect-site concentration. adjust it accordingly. The Marsh model has a faster By targeting the effect-site, the system adjusts induction due to the larger calculated initial bolus plasma concentration to reach the target effect- dose, while the Schnider model has a reduced rate site concentration as fast as possible. Choosing a of adverse events by giving a smaller initial bolus keo value to use is especially crucial in effects-site due to the small fixed V1. Thus, lower doses of targeting to determine the overshoot and undershoot Propofol are required to achieve a given plasma of the plasma concentration used to drive the concentration. Caution from awareness should be effect-site concentration as the target effect-site taken when using the Schnider model for plasma concentration increases or decreases. (19)

17-Latif.indd 131 12/01/2021 12:21 132 k.w. abdul-latif et al. As the direct measurement of the concentration References at the effect-site is not possible and, in most cases, the plasma concentration is also unknown. 1. Thompson, K. A., and Goodale, D. B. 2000. The recent development of propofol (DIPRIVAN®). Intensive Care However, the time course of changes in effect-site Med, Suppl., 26(4). concentration can be estimated from measurements 2. Minto, C. F., Schnider, T. W., and Shafer, S. L. 1997. of clinical effects, e.g., using the Bi-spectral index Pharmacokinetics and Pharmacodynamics of Remifen- system. If the plasma levels of a group of subjects tanil. Anesthesiology. 86(1), 24-33. 3. Anderson, B. J., and Houghton, J. 2019. Total Intravenous are known, pharmacodynamic measurements can Anesthesia and Target-Controlled Infusion. In : A Practice be used to estimate keo. (8) The rate constant is of Anesthesia for Infants and Children, Elsevier, p. 177- used for mathematical estimation of the time course 198.e3 4. Struys, M. M. R. F., De Smet, T., Glen, J. B., Vereecke, of equilibrium between the subject’s plasma and the H. E. M., Absalom, A. R., and Schnider, T. W. 2016. The effect-site. History of Target-Controlled Infusion. Anesth Analg. 5. Marsh, B., White M., Morton, N., and Kenny, G. N. C. Summary 1991. Pharmacokinetic model driven infusion of propofol in children. Br J Anaesth 67(1), 41-48. The administration of Propofol using target- 6. Schüttler, J., and Ihmsen, H. 2000. Population pharmaco- controlled infusions (TCI) for total intravenous kinetics of propofol : A multicenter study. Anesthesiology. anesthesia (TIVA) is a favorite technique for anes- 92(3), 727-738. 7. Schnider, T. W., Minto, C. F., Gambus, P. L., Andresen, thesia. The safety and efficacy of TCI have a large C., Goodale, D. B., Shafer, S. L., and Youngs, E. J. body of evidence. Modern open TCI systems allow 1998. The influence of method of administration and Propofol to be used with a choice of pharmacokinetic covariates on the pharmacokinetics of propofol in adult models and effect-site or plasma targeting. Generally, volunteers. Anesthesiology. 88(5), 1170-1182. 8. Anderson, B. J., Bagshaw, O. 2019. Practicalities of Total there is no much conclusive evidence showing the Intravenous Anesthesia and Target-controlled Infusion in superiority of any particular implemented model or Children. Anesthesiology. Lippincott Williams and Wilkins. method of effect-site targeting. Largely accepted 9. Krüger-Thiemer, E. 1968. Continous intravenous infusion and multicompartment accumulation. Eur J Clin Pharmacol. that anesthesiologists best use the model with which 4(3), 317-324. they are most familiar. 10. Vaughan, D. P., and Tucker, G. T. 1975. General theory for Using Marsh model, the target concentrations rapidly establishing steady state drug concentrations using are appropriate for younger patients while it is two consecutive constant rate intravenous infusions. Eur J Clin Pharmacol. 9(2–3), 235-238. associated with hemodynamic instability in elderly 11. Schwilden, H. 1981. A general method for calculating patients. This is because of both pharmacokinetic the dosage scheme in linear pharmacokinetics. Eur J Clin and pharmacodynamic changes with age. Modern Pharmacol. 20(5), 379-386. 12. Alvis, J. M., Reves, J. G., Govier, A. V., Menkhaus, P. open TCI systems allow Propofol to be used with a G., Henling, C. E., Spain, J. A., and Bradley, E. 1985. choice of pharmacokinetic models and effect-site or Computer-assisted continuous infusions of fentanyl during plasma targeting. The Marsh model has been shown cardiac anesthesia : Comparison with a manual method. to under-predict plasma concentrations of Propofol Anesthesiology 63(1), 41-49. 13. Crankshaw, D. P., Boyd, M. D., and Bjorksten, A. R. 1987. in the elderly and known not to perform any Plasma drug efflux – A new approach to optimization of adjustments for age. As the Schnider model adjusts drug infusion for constant blood concentration of thiopental infusion rates following the age of the patient, thus it and methohexital. Anesthesiology. 67(1), 32-41. 14. Jacobs, J. R. 1990. Communications Algorithm for Opti- provides a strong argument to use this model in the mal Linear Model-Based Control with Application to elderly and in critically ill patients in whom smaller Pharmacokinetic Model-Driven Drug Delivery. IEEE bolus doses will be given after target increases, and Transact Biomed Engin. 7(1), 107-109. 15. Shafer, S. L., Siegel, L. C., Cooke, J. E., and Scott, J. C. this may improve hemodynamic stability and safety. 1988. Testing computer-controlled infusion pumps by The studies from which the Marsh and simulation. Anesthesiology 68(2), 261-266. Schnider pharmacokinetic models were developed 16. Tavernier, A., Coussaert, E., D’Hollander, A., and Can- did not include severely obese patients. Thus, target- traine, F. 1987. Model-based pharmacokinetic regulation in computer-assisted anesthesia. An interactive system : controlled infusions should be used with caution in CARIN. Acta Anaesthesiol Belg. 38(1), 63-68. severely obese patients, regardless of which model 17. Shafer, S. L., and Gregg, K. M. 1992. Algorithms to rapidly or effect-site implementation method is used. Future achieve and maintain stable drug concentrations at the site of drug effect with a computer-controlled infusion pump. J investigation should give more attention to validate Pharmacokin Biopharmac. 20(2), 147-169. a model for severely obese patients. 18. Jacobs, J. R., and Williams, E. A. 1993). Algorithm to Control “Effect Compartment” Drug Concentrations in Acknowledgments Pharmacokinetic Model-Driven Drug Delivery. IEEE Transact Biomed Engin. 40(10), 993-999. I would like to thank my promotor and co- 19. Absalom, A. R., Mani, V., De Smet, T., and Struys, M. M. promotor for their remarkable efforts and guidance. R. F. 2009. Pharmacokinetic models for propofol- Defining

17-Latif.indd 132 12/01/2021 12:21 propofol tci models, narrative review 133 and illuminating the devil in the detail. Anesth Analg. 22. Kataria, B. K., Ved, S. A., Nicodemus, H. F., Hoy, G. R., Oxford University Press. Lea, D., Dubois, M. Y., … Shafer, S. L. 1994. The pharma- 20. Al-Rifai, Z., and Mulvey, D. (2016). Principles of total cokinetics of propofol in children using three different data intravenous anaesthesia : basic pharmacokinetics and analysis approaches. Anesthesiology 80(1), 104-122. model descriptions. BJA Education, 16(3), 92-97. 23. Absalom, A., Amutike, D., Lal, A., White, M., and Kenny, 21. Gepts, E., Camu, F., Cockshott, I. D., and Douglas, E. J. G. N. C. 2003. Accuracy of the “Paedfusor” in children 1987. Disposition of propofol administered as constant rate undergoing cardiac surgery or catheterization. Br J Anaesth. intravenous infusions in humans. Anesthesia and Analgesia, 91(4), 507-513. 66(12), 1256-1263.

17-Latif.indd 133 12/01/2021 12:21

(Acta Anaesth. Belg., 2020, 71, 135-141) Performance analysis of an ischemic stroke management protocol, a single centre study

D. De Meirsman (*), S. De Meirsman (**), R. De Meirsman (**), M. Sabbe (*), P. Vanelderen (***), D. Mesotten (***) , S. Van Boxstael (***)

Abstract : Objective : Stroke is one of the leading causes Introduction of increased morbidity and mortality. “Time equals brain”. Therefore, our aim was to compare the door-to- Stroke is defined by the World Health Organi- CT time (D2CT) and door-to-needle time (DNT) of the zation as a clinical syndrome of rapidly developing Stroke Phase II TARGET protocol of the American Heart focal or global neurological disturbances, lasting Association (AHA) with our newly implemented stroke more than 24 hours. Stroke classification has two management protocol. main divisions : ischemic and hemorrhagic. Eighty Methods : Data were collected before, during and after implementation of the new stroke protocol in a referral percent (80%) of stroke patients presenting at the hospital for advanced stroke treatments. A total of emergency department are experiencing an ischemic 317 eligible patients were included. D2CT and DNT stroke (1). were registered according to standard clinical practice. It is the leading cause of long-term neurological Descriptive analysis was done on three time periods : impairment and the second most common cause of period 1 (before) (2017-2018), period 2 (during) (2018- mortality worldwide (2). The most important factor 2019) and period 3 (after) (2019-2020). Stroke patients in the management of acute ischemic stroke is time were subdivided in four groups according to the treatment to treatment. Patients lose around a hundred ninety plan as patients may have been referred for specific thousand brain cells and around fifty times as many treatments : conservative (CON), thrombolysis (LYS), nerve connections every minute since the onset of thrombectomy (EMT) and combination (COM). Data are their stroke. The sooner treatment is initiated, the presented as median and interquartile range (IQR). greater the proportional benefit and the more brain Results : Period 1 showed a mean D2CT of 40 min for CON, 21min for LYS, 20min for EMT and 17min for tissue could be saved (3). COM and a mean DNT of 48min for LYS and 33min The risk of having an ischemic stroke is for COM. Mean D2CT was 43 min for CON, 31min for correlated with patient related factors, such as LYS, 20min for EMT and 18min for COM and DNT diabetes mellitus, arterial hypertension, smoking, was 60min for LYS and 45min for COM in Period 2. family history of stroke, dyslipidemia, physical Period 3 had a mean D2CT of 35min for CON, 21min inactivity, large and small vessel disease, heart for LYS, 25min for EMT and 16min for COM and DNT problems. Associated neurological symptoms can was 32min for LYS and 20 min for COM. D2CT times be transient which makes the presentation of stroke remained consistently higher in patients, who eventually omnifarious (4). underwent conservative therapy. Conclusion : D2CT did not meet the AHA criteria (25min). However, the DNT has always fallen within D. De Meirsman, MD ; S. De Meirsman, MD ; R. De Meirsman, the AHA recommendations (60min). Continuous efforts MD ; M. Sabbe, MD ; P. Vanelderen, MD ; D. Mesotten, need to be made towards improving D2CT and sustaining MD ; S. Van Boxstael, MD DNT times. (*) Emergency Department, University Hospitals Leuven, Belgium (**) Department of Anesthesiology, University Hospitals Leuven, Belgium Keywords : Stroke ; emergency department ; hospital ; (***) Critical Care Department, Ziekenhuis Oost-Limburg, thrombectomy ; NIHSS ; door-to-needle-time ; door-to- Belgium CT time. Corresponding author : Sam Van Boxstael, MD. Critical Care Department, Ziekenhuis Oost Limburg, Schiepse Bos 6, 3600 Genk, Belgium. E-mail : [email protected]

Paper submitted on June 5, 2020 and accepted on June 5,2020. Conflict of interest : none.

18-De Meirsman.indd 135 12/01/2021 12:22 136 d. de meirsman et al. Systemic hypoperfusion causes ischemia due to inadequate cerebral blood flow, usually evoked by decreased perfusion pressures in frail patients (1). Different treatment modalities have been recommended for ischemic stroke patients (5). The use of intravenous thrombolysis is the standard treatment recommended by the American Heart Association (AHA/ASA) for patients presenting within the first three hours of clinical symptoms. Evidence based research, e.g. : NINDS, ATLANTIS, ECASS, ECASS-III, have shown that the administration of tissue plasminogen activator (tPA) given within a time window of three to four hours was better than conservative care (3, 6, 7). Recently endovascular mechanical thrombectomy has become part of the standard treatment for patients who experienced an acute ischemic stroke (8), granted that treatment was initiated within 6 hours after onset (9) (10). Stroke has a major impact on the quality of life (QOL) of patients, their family and caregivers (11). Effort should be made to shorten any delays Fig. 1. — Flowchart – Genk Stroke Database. TIA : Transient in the initiation of treatment, because time equals ischemic attack, CON : Conservative therapy, COM : Com- bination of therapy (Thrombolysis + Thrombectomy), EMT : brain tissue (10). The TARGET : Stroke Phase Endovascular mechanical thrombectomy, LYS : Thrombolysis. II campaign (TARGET) of the American Heart Association strives to improve the operational flow in stroke care management. They evaluated the consistent with the start and implementation of the implementation of a stroke management protocol protocol on the ED. Throughout this period, data as a strategy to improve the initial assessment were collected retrospectively by a study nurse of a suspected stroke patient in the emergency who manually checked all patient files. The second department. According to TARGET, an efficient period (12/2018-12/2019) was considered as a stroke care management protocol requires : a door- control year to evaluate the protocol. At the start of to-needle time (DNT) of less than 60 minutes, a the second period, data were prospectively collected door-to-computed tomography (D2CT) time of with automatically generated data extraction less than 25 minutes, with neuroimaging analysis from the electronic patient file with standardized and interpretation in less than 45 minutes (12). definitions of each data point. In the third period Moreover, effective stroke care depends on a well- (1/2020-Present) a review and final implementation functioning team, from the emergency department of the protocol was done. to the neurologist and the interventional radiologist The inclusion criteria for eligibility were (13). Our objective was to compare our D2CT and patients with a documented ischemic stroke pre- DNT to the times suggested by TARGET. senting themselves within the time window of 3.5 hours. The exclusion criteria were hemorrhagic Methods stroke, TIA’s or stroke mimics e.g. Todd’s paresis In order to assess the efficacy of our stroke We retrospectively analyzed prospectively management protocol, our primary endpoints collected data, stored in a protected database, consisted of door-to-CT (D2CT) and door-to-needle specifically on stroke care in a single center. This (DTN) times for the different treatment groups center is a tertiary non-academic hospital with in comparison with the internationally accepted 55000 patients presenting annually to the emergency TARGET times. These treatment groups consisted department (ED). This study was approved by the of conservative (CON), thrombolysis (LYS), endo- hospital ethics committee (19/0059U). The results vascular mechanical thrombectomy (EMT) and are reported according to STROBE guidelines a combination of thrombolysis and endovascular (14). We distinguished three periods in our analysis mechanical thrombectomy (COM). An effective (Fig. 1). The first period (11/2017-11/2018) is stroke protocol yields a D2CT time within 25

18-De Meirsman.indd 136 12/01/2021 12:22 performance analysis of an ischemic stroke management protocol 137 minutes and a DTN time within 60 min for 75% of the final analysis, 126 patients in the first period, eligible patients (12). The electronic data software 134 patients in the second period and 35 in the third program was used to retrieve the D2CT and DTN period were included. times. We refer to the appendix for a detailed report Demographic data (Table1) between different of the stroke protocol. time periods did not differ. 48% (n=142) were Statistical analyses were conducted with treated conservatively and had a median NIHSS GraphPad Prism (v5, GraphPad, San Diego, USA). of 4 (3-12). 52% (n=153) received an intervention. Prior to analysis, normality of the data was assessed 28% (n=83) received only intravenous tPA, in 13% using Shapiro-Wilk test. As the time, in minutes, (n=39) thrombectomy was performed and in 11% from door to CT and door to needle were not (n=31) a combination. Patients receiving intravenous normally distributed, box plots were constructed to tPA alone had a mean age of 74±13 years, 52% evaluate the evolution of our stroke protocol over (n=43/83) were men, and the median NIHSS was 11 time with regard to D2CT and DTN times for the (5-15). Patients receiving thrombectomy alone had different treatment groups. All data was anonymized a mean age of 70±15 years, 49% (n=19/39) were before analysis. men, and the median NIHSS was 18 (9-22). Patients receiving both therapies had a mean age of 76±14 Results years, 63% (n=20/31) were men, and the median NIHSS was 16 (10-20). In total 747 patients were examined for potential For a general overview of the primary outcome eligibility (Fig. 1). Following the exclusion criteria, results we refer to table 2 and figure 2. 438 patients were eligible for analysis. Another 121 During the first period, in the CON group 26% ischemic stroke patients were excluded as they did (n=10/38) was within the allowed time for D2CT not present within the 3.5 h time window. In total, (25min) and 74% (n=65/88) in the other groups. 317 patients were included in the analysis, another Interestingly, in the COM group 100 % was reached 22 participants were lost due to incomplete data. In (n=21/21). The goal for the DNT (60 min) was

Table 1 Patient characteristics and demographic data

Variable Conservative Thrombolysis Thrombectomy Combination Total Patients characteristics N=142/295 N=83/295 N=39/295 N=31/295 N=295 Age Yrs±SD 71 ±16 74 ±13 70 ±15 76 ±14 73 ±14 Gender Female n % 51 36% 40 48% 20 51% 11 37% 122 41% Male n % 91 64% 43 52% 19 49% 20 63% 173 59% Race White n % 141 99% 83 100% 39 100% 31 100% 294 99% Other n % 1 1% 0 0% 0 0% 0 0% 1 1% Smoking Yes n % 33 23% 18 22% 7 17% 6 20% 64 22% Alcoholic Yes n % 0 0% 7 9% 2 4% 0 0% 9 3% Hypertension Yes n % 63 44% 68 82% 29 75% 24 77% 184 62% Diabetic Yes n % 44 31% 15 18% 4 11% 2 6% 65 22% Cholesterol Yes n % 47 33% 37 45% 11 29% 5 17% 100 34% Obesitas Yes n % 26 18% 21 25% 5 14% 2 6% 54 18% Family History Yes n % 21 15% 19 23% 5 14% 2 6% 47 16% Prior Stroke Yes n % 18 13% 9 11% 4 11% 4 13% 35 12% NIHSS Score (IQR) 4 3-12 11 5-15 18 9-22 16 10-20 12 4-17 Yrs : years ; SD : standard deviation ; NIHSS : National Institutes of Health Stroke Scale. Data are presented as median and interquartile range (IQR) for NIHSS-score, as an absolute number with the percentage of the whole or for years as an average with standard deviation.

18-De Meirsman.indd 137 12/01/2021 12:22 138 d. de meirsman et al. Table 2 Comparison of door-to-CT time and door-to-needle time in the acute ischemic stroke patient over time

Variable Conservative Thrombolysis Thrombectomy Combination Total Period 1: 11/2017-11/2018 N=38/126 N=42/126 N=25/126 N=21/126 N=126 D2CT Min 40 (24-71) 21 (12-38) 20 (13-33) 17 (9-20) 23 (15-41) D2CT <25min n % 10 26% 27 64% 17 68% 21 100% 75 59%

N=42/63 N=21/63 N=63 DNT Min 48 (34-73) 33 (28-45) 42 (32-65) DNT <60min n (%) 23 55% 15 71% 38 60% Period 2: 12/2018-12/2019 N=83/134 N=30/134 N=12/134 N=9/134 N=134 D2CT Min 43 (29-67) 31 (16-42) 20 (12-28) 17 (9-25) 35 (21-52) D2CT <25min n % 16 19% 12 40% 7 58% 7 78% 42 31% N=30/39 N=9/39 N=39 DNT Min 60 (37-71) 45 (31-74) 57 (36-70) DNT <60min n % 13 43% 6 67% 19 49% Period 3: 1/2020 – 5/2020 N=21/35 N=11/35 N=2/35 N=1/35 N=35 D2CT Min 35 (23-62) 21 (12-34) 26 (16-35) 16 * 29 (23-62) D2CT <25min n % 5 22% 6 55% 1 50% 1 100% 13 37% N=11/12 N=1/12 N=12 DNT Min 32 (19-49) 20 * 25 (20-44) DNT <60min n % 8 73% 1 100% 9 75% D2CT : door-to-CT time, Min : Minutes, DNT : Door-to-needle time, * : insufficient data to calculate interquartile range. Data are presented as median and interquartile range (IQR) or as an absolute number/total number (n/N) with the percentage % of the whole.

Fig. 2. — Comparison of door-to-CT time and door-to-needle time in the acute ischemic stroke patient over time.

reached for 60% (n=38/63) in the LYS and COM Every effort should be made to shorten any delays groups. in the initiation of treatment, because time saves During the second period, 19% (n=16/83) was brain tissue (12). within the TARGET interval for CON and 51% During the first period, DNT almost met (n=26/51) in the other groups. The goal for DNT the stroke phase II TARGET guidelines’ times. in the LYS and COM groups was reached for 49% In contrast, there was an increased DNT in the (n=19/39). second period. Lastly, in the third period DNT were Lastly, in the third period 22% (n=5/21) was online with the current guidelines. Overall D2CT within the D2CT interval for the CON group and targets were reached for approximately 70% of the 57% (n=8/14) for the other groups. The goal for treatment groups (LYS-EMT-COM) whereas this DNT was reached for at least 75% (n=9/12). was only 22% for the conservatively treated groups. The first period represents the implementation Discussion of the protocol. Several publications described pathways leading to prioritization of stroke Stroke has an important impact on the quality patients for imaging and emphasized the benefits of life of patients, their family and caregivers (11). of screening tools and a systematic approach,

18-De Meirsman.indd 138 12/01/2021 12:22 performance analysis of an ischemic stroke management protocol 139 e.g. : use of an ACT-FAST triage protocol, early addition, as the CT-scanner is situated in the back of warning of the stroke team and direct interpretation our ED a fast assessment by our emergency doctor, of brain imaging (12, 15, 16). During this first whilst the patient waits on the EMS transport period, a multi-disciplinary team of ED physicians, gurney, was done, saving time bypassing the ED radiologists, neurologist, interventional radiologist bed. Only limited data exist evaluating this potential and anesthetists established a stroke pathway using time saving using this method (24). these published key strategies. Moreover, effort Finally, moving tPA start-up at the ED CT- was put into education of staff members, nurses, scanner significantly improved DNT. Even to faci- residents and trainees. Also, posters reporting the litate administration, we standardized order sets new stroke protocol were put up in the ED to draw and constructed weight specific charts limiting attention. Combining these efforts resulted in D2CT dosing errors (25). The sample size of this group and DNT of approximately 60% within target time. was gathered in a shorter time period, resulting in The second period was an evaluation of the fewer eligible patients, partially also influenced by protocol. The lack of improvement of the D2CT and the COVID crisis. Nevertheless, the results already DTN times were somewhat surprising. The increase reflect a noticeable difference in comparison with in median times was likely due to a combination data from the second period. of factors. While the data were now prospectively Although we already use some recommended gathered, feedback on suboptimal performance was strategies, some future implementations will be not yet established. Essedeeg et al demonstrated made such as : single-call activation systems, pre- that improvement of DTN times only happens after mixing tPA, prompt data feedback (25). However, feedback was given (17, 18). In fact, a dedicated team these are still riddled with logistical side notes. must meet on a regular basis to review the hospital’s We are conscious of several limitations in our protocol and, to reflect on clinical outcomes leading study. First of all, it was a single center study and to recommendations of strategic improvements (12). the results cannot be generalized, due to intrinsic Secondly, an interdisciplinary stroke team has a variations among stroke centers. We report only major impact on time efficiency on the ED (12). differences in time as a result of various interventions In our center, members providing acute stroke care in the early stroke care and highlight only a relevant changed frequently, due to mandatory rotations of part of stroke care in the ED. Our study is mainly residents and staff members. Thirdly, the effects of descriptive. While we focused on the DNT and D2CT augmented awareness and of subsequent strategies, times, we did not focus on functional outcome or probably started to fade (19). Lastly, switching from overall mortality. A limit of 3,5h was chosen, taking manual retrospective data collection by a study nurse into account the average door-to-needle time of 60 to prospective data collection from the standard min, starting from ED presentation, to remain below clinical practice with an automated data generation the accepted international 4,5h range of therapeutic system, following clearly defined criteria could also eligibility in acute stroke care. Sample size of the 3th partly explain the time differences between first and group was limited in time and number and unequal second period. compared to the first two periods, this affects the In the third period, a protocol review and reliability and may cause a higher variability, which final implementation was performed. The above- leads to caution in interpreting the results. Longer mentioned shortcomings were tackled, and adapted observation periods are needed to determine if these strategies were implemented to reduce D2CT and results are reliable for long term or can be further DNT time. We started with a standardized EMS improved. prehospital stroke notification. Numerous studies have demonstrated that a pre-hospital warning can Conclusion reduce D2CT and DTN times (16, 20-22) At the same time, we began transferred stable EMS patients with The Emergency Department has an important a potential stroke immediately from the ambulance role in the management of acute stroke. After to the CT scanner. This was based on Meretoja et implementation of our stroke management protocol, al., who looked at EMS personnel evaluating acute the TARGET : Stroke Phase II DNT goal within stroke victims on route to the hospital to certify 60 min was reached for 75% of eligible patients. “last-known-well” time and document a NIHSS- In contrast, a D2CT within 25 minutes was only score in the ambulance, permitting the EMS to reached for 37% of eligible patients. In our hospital, transport patients directly from the ambulance to the additional new strategies will be necessary to further CT scanner, reducing their DNT and D2CT (23). In improve stroke patient flows.

18-De Meirsman.indd 139 12/01/2021 12:22 140 d. de meirsman et al. References 14. Von Elm E., Altman DG., Egger M., Pocock SJ., Gøtzsche PC., Vandenbroucke JP. 2007. The Strengthening the 1. Louis Caplan. 2009. Part 1 : General principles, chapter 3 : Reporting of Observational Studies in Epidemiology Diagnosis and the clinical encounter. In : Caplan’s Stroke : (STROBE) statement : guidelines for reporting observa- a clinical approach, 4th edition. Philadelphia. Elsevier Inc., tional studies. Lancet. 370:1453-1457. P. 66. 15. Schwamm LH., Pancioli A., Acker JE., Goldstein LB., 2. Wolfe CD. 2000. The impact of stroke. Br Med Bull. Zorowitz RD. and Shephard TJ. et al. 2005. Recom- 56:275-286. mendations for the establishment of stroke systems of care : 3. Emberson J., Lees KR., Lyden P.,Blackwell L., Albers recommendations from the American Stroke Association’s G. and Bluhmki E. et al. 2014. Effect of treatment delay, Task Force on the Development of Stroke Systems. Stroke age, and stroke severity on the effects of intravenous 36:690-703. thrombolysis with alteplase for acute ischaemic stroke : a 16. Saver JL., Smith EE., Fonarow GC., Reeves MJ., Zhao X. meta-analysis of individual patient data from randomised and Olson DM. et al. 2010. The “Golden Hour” and Acute trials. Lancet. 384:1929-1935. Brain Ischemia. Stroke. 41:1431-1439. 4. Boehme AK., Esenwa C., Elkind MSV. 2017. Stroke Risk 17. Ghrooda E., Alcock S., Jackson AC. 2012. Improvement Factors, Genetics, and Prevention. Circ Res. 120:472-495. in thrombolytic therapy administration in acute stroke with 5. Lisa LP., Jughters A., Kerckhofs E. 2013. The effectiveness feedback. Can J Neurol Sci. 39:789-792. of different treatment modalities for the rehabilitation of 18. Duncan K., Pozehl B. 2000. Effects of performance unilateral neglect in stroke patients : a systematic review. feedback on patient pain outcomes. Clin Nurs Res. 9:379- NeuroRehabilitation. 33:611-620. 401. 6. Hacke W., Donnan G., Fieschi C., Kaste M., Von Kummer 19. Buehler R., Griffin D., Ross M. 1994. Exploring the R. and Broderick JP. et al. 2004. Association of outcome “planning fallacy” : Why people underestimate their task with early stroke treatment : pooled analysis of ATLANTIS, completion times. J. Pers. Soc. Psychol. 67:366-381 ECASS, and NINDS rt-PA stroke trials. Lancet. 363:768- 20. Bae HJ., Kim DH.,Yoo NM., Choi JH., Huh JT. and Cha JK., 774. et al. 2010. Prehospital Notification from the Emergency 7. National Institute of Neurological Disorders and Stroke rt- Medical Service Reduces the Transfer and Intra-Hospital PA Stroke Study Group. 1995. Tissue plasminogen activator Processing Times for Acute Stroke Patients. J Clin Neurol. for acute ischemic stroke. N Engl J Med .333:1581-1587. 6:138-5. 8. Yang P., Zhang Y., Zhang L., Zhang Y., Treurniet KM. and 21. Abdullah AR., Smith EE., Biddinger PD., Kalenderian D., Chen W., et al. 2020. endovascular thrombectomy with Schwamm LH. 2009. Advance Hospital Notification by or without Intravenous alteplase in acute stroke. N Engl J EMS in Acute Stroke Is Associated with Shorter Door-to- Med. 382:1981-1993 Computed Tomography Time and Increased Likelihood 9. Goyal M., Menon BK., van Zwam WH., Dippel DWJ., of Administration of Tissue-Plasminogen Activator. Mitchell PJ. and Demchuk AM. et al. 2016. Endovascular Prehospital Emergency Care. 12:426-431. thrombectomy after large-vessel ischaemic stroke : a meta- 22. Lin CB., Peterson ED., Smith EE., Saver JL., Liang L., analysis of individual patient data from five randomised Xian Y. et al. 2012. Emergency medical service hospital trials. Lancet. 387:1723-1731. prenotification is associated with improved evaluation and 10. Powers WJ., Derdeyn CP., Biller J., Coffey CS., Hoh BL. treatment of acute ischemic stroke. Circ Cardiovasc Qual and Jauch EC. et al. 2015 American Heart Association/ Outcomes. 5:514-522. American Stroke Association Focused Update of the 2013 23. Meretoja A., Strbian D., Mustanoja S., Tatlisumak T., Guidelines for the Early Management of Patients With Lindsberg PJ., Kaste M. 2012. Reducing in-hospital delay Acute Ischemic Stroke Regarding Endovascular Treatment. to 20 minutes in stroke thrombolysis. Neurology. 79:306- Stroke 46:3020-3035. 313. 11. Chuluunbaatar E., Chou YJ., Pu C. 2016. Quality of life 24. Cone DC., Cooley C., Ferguson J., Harrell AJ., Luk JH. of stroke survivors and their informal caregivers : A and Martin-Gill C., et al. 2017. Observational Multicenter prospective study. Disabil Health J. 9:306-312. Study of a Direct-to-CT Protocol for EMS-transported 12. Fonarow GC., Smith EE., Saver JL., Reeves MJ., Hernandez Patients with Suspected Stroke. Prehospital Emergency AF. and Peterson ED. et al. 2011. Improving door-to-needle Care. 22:1-6. times in acute ischemic stroke : the design and rationale 25. Xian Y., Smith EE., Zhao X., Peterson ED., Olson DM. for the American Heart Association/American Stroke and Hernandez AF. et al. 2014. Strategies used by hospitals Association’s Target : Stroke initiative. Stroke. 42:2983- to improve speed of tissue-type plasminogen activator 2989. treatment in acute ischemic stroke. Stroke. 45:1387-1395. 13. Chugh C. 2019. Acute Ischemic Stroke : Management 26. Zhao H., Pesavento L., Coote S., Rodrigues E., Salvaris P., Approach. Indian J Crit Care Med. 23:S140-S146. Smith K. et al. 2018. Ambulance Clinical Triage for Acute Stroke Treatment : Paramedic Triage Algorithm for Large Vessel Occlusion. Stroke. 49:945-951.

18-De Meirsman.indd 140 12/01/2021 12:22 performance analysis of an ischemic stroke management protocol 141 APPENDIX 1 - STROKE PROTOCO

All patients with a tentative diagnosis of stroke were assessed for eligibility, including referrals, out of hospital and in hospital patients. When patients presented to our triage with a suspected stroke, the triage nurse conducted an ACT-FAST score for early detection of possible stroke patients. The ACT-FAST algorithm looks for signs of stroke (unilateral arm drift to stretcher <10 seconds, severe language deficit, gaze deviation, hemineglect or the eligibility of stroke mimics (e.g. hypoglycemia) (26). When ACT-FAST was positive a secondary survey was done in order to activate the stroke management protocol by an emergency physician. The decision was based on an estimation of onset of symptoms and a short summary of the patient history. On activation of the protocol the triage nurse placed an IV-access, a pre-assembled stroke kit was opened which included : laboratory test tubes, venous blood gas for point of care testing, pre-ordered CT application-forms and a NIHSS-scoring sheet. Parameters were taken. The emergency physician then contacted the department for neurology and radiology to notify them of a possible acute stroke patient. Meanwhile the CT-scanner was reserved, and transport was organized. For all patients a prospective data gathering was started as soon as possible. Data collection was done via a survey, which collected different demographic data. This was done either by the patient themselves or by their family and was supplemented by an extensive patient history recollected from the hospital records. Baseline characteristics included : age, sex, race, language, smoking, alcohol use, hypertension, diabetic, dyslipidemia, obesitas, prior stroke events, NIHSS, premorbid medical and functional status and time of stroke onset were gathered. A neurologist at the bedside evaluated all patients. Specific radiologic stroke scores (ASPECT-scores, collateral scores and location of occlusion) were collected from radiologic reports and added to the database with the help of a study nurse. In our center radiologists ensured an immediate imaging protocol 24/7. After scanning the patient returned to the emergency department for further assessment and appropriate therapy was given.

Stroke protocol Gen

Triage Nurse, ED physician Stroke patient Location: Triage Phone ED physician (4444) (first assessment – phone radiology (4605)) Control of past medical history – Kidney function: eGFR <50ml/min= NaHCO3 1.4% 500ml Phone neurologist Assessment of vital signs, IV-line placement, blood samples, CVA-kits (lab tubes, NIHSS, Follow-up) Wake-up, posterior stroke, allergy: consider MRI i/o Angio CT ↓ ED physician, Radiologist Location: Ct-scanner 1.Assesment ED: Hemorrhagic stroke? 2. Assessment radiologist: ischemic stroke – stand-by activation of interventional radiologist ↙ ↘ Location: shock room ED Location: shock room ED Assessment neurologist: Assessment neurologist: No endovascular therapy ± start thrombolysis endovascular therapy ± start thrombolysis -call interventional radiologist: call-off stand-by -call interventional radiologist: timing procedure -Call ED physician: organize anaesthesia ↓ ↓ Location: stroke unit Location: shock room ED Admission stroke unit -Call anaesthetist (5099/6666) ↑ 2-lumen peripheral IV-catheter Arterial line by staff ↓ Location: IRCC HD stable and -call ER (1580) extubation -Transport patient ↑ ↓ Anaesthetist ICU Patient mechanically ventilated to ICU

18-De Meirsman.indd 141 12/01/2021 12:22

(Acta Anaesth. Belg., 2020, 71, 143-152) Anesthesia in patients with Duchenne muscular dystrophy : a retrospective analysis of all Duchenne muscular dystrophy patients undergoing general and locoregional anesthesia in a tertiary referral center

E. Degryse (*), N. Goemans (**), G. Dewinter (*), M. Verhaegen (*), M. Van de Velde (***)

Abstract : Background : Duchenne muscular dystrophy Introduction (DMD) is the most common severe myopathy in children. They often require surgical interventions under Duchenne muscular dystrophy (DMD) is an general or locoregional anesthesia. These patients remain autosomal recessive, X-linked dystrophinopathy challenging cases for the anesthesiologist as they are prone that affects boys with an incidence varying to develop complications perioperatively. Anesthesia- between 2 to 3 per 10.000 male births (1). Due to induced rhabdomyolysis (AIR) and hyperkalemic cardiac an abnormality in the gene encoding for the cell arrest are the most feared complications, and clearly associated with the use of succinylcholine and potent membrane protein dystrophin, its absence leads to volatile agents. Intravenous agents are favored, but also a continuous degeneration of striate muscle fibers carry a risk of causing rhabdomyolysis. replacing these with fibrofatty tissue (1,2). Clinically Objectives : We reviewed different types of perioperative this is translated into affected boys suffering from complications of patients with DMD and tried to correlate proximal muscle weakness over impaired mobility them to their likely cause, be it underlying pathology, to full wheelchair dependency. Parallel to these surgical complication or anesthetic intervention or agent. manifestations they develop respiratory failure as Methods : We performed a retrospective medical record well as cardiac failure, leading to premature death analysis of anesthetic interventions in confirmed DMD in the third or fourth decade (3, 4). patients from January 1970 through December 2019. Due to the progressive nature of this neuro- Results : We included 105 DMD patients, collectively muscular disease, this specific patient population undergoing a total of 185 anesthetic interventions often needs a variety of surgical interventions and (volatile n=78, intravenous n=107). Complications arose in 20.5% (n=38) of overall anesthetic interventions. The is therefore widely exposed to anesthesia and its mainstay could be attributed to surgical or preexisting perioperative risks. medical conditions, or a combination of both. But in two More specifically, since the 1960s case reports cases rhabdomyolysis developed. One in the intravenous describe the occurrence of acute rhabdomyolysis group, the other in the volatile group. The latter was and hyperkalemic cardiac arrest with the use of complicated with a fatal hyperkalemic cardiac arrest. An succinylcholine and/or halogenated volatile agents overall rate of difficult tracheal intubations was revealed in 19.2% of cases. Conclusions : There is no anesthetic technique with E. Degryse, MD ; N. Goemans, MD, PhD ; G. Dewinter, MD, a zero risk for AIR. Potent volatile agents should be PhD ; M. Verhaegen, MD, PhD ; M. Van de Velde, MD, avoided whenever possible. When they are necessary, PhD, EDRA switch to an intravenously maintained anesthesia should (*) Dpt of Anesthesiology, University Hospitals of the KU be made. However, use of propofol infusions in these Leuven, 3000 Leuven, Belgium (**) Dpt of Pediatric Neurology, University Hospitals of the children requires at least caution. KU Leuven, 3000 Leuven, Belgium (***) Dpt of Cardiovascular Sciences, Dept. of Anesthesiology Keywords : Duchenne muscular dystrophy ; intra-oper- and University Hospitals of the KU Leuven, 3000 Leuven, ative complications ; postoperative complications ; anes- Belgium thesia induced rhabdomyolysis ; hyperkalemic cardiac Corresponding author : Ellen Degryse, MD, Dpt of Anes- thesiology, University Hospitals of the KU Leuven, arrest. Herestraat 49, 3000 Leuven, Belgium. E-mail : [email protected]

Paper submitted on Jun 8, 2020 and accepted on Jun 9, 2020. Conflict of interest : none.

19-Degryse.indd 143 12/01/2021 12:22 144 e. degryse et al. in DMD patients (5). This acute rhabdomyolysis adheres to the ICH - GCP principles (International has to be differentiated from the clinical entity Conference on Harmonization Guidelines on malignant hyperthermia as there is absence of the Good Clinical Practice), the latest version of the hypermetabolic state preceding this rhabdomyolysis Declaration of Helsinki, the Oviedo Convention and cardiac arrest (6). An important difference that on Human Rights and Biomedicine and applicable has led to the introduction of the term anesthesia- laws and regulations. induced rhabdomyolysis. On the other side, the recognition of this difference is of utmost Patient identification importance as these two clinical entities require different initial therapeutic management and this The department of pediatric neurology at the will have an important influence on survival. UZL keeps a record of all their DMD patients. Their The previously mentioned case reports database contains 210 confirmed DMD patients have discouraged the use of volatile agents in and data were collected from January 1st 1970 up favor of intravenous agents (propofol, ketamine, to December 31st 2019. All of these records were benzodiazepines and barbiturates). An ideal agent scrutinized for this study, to map out which of these for total intravenous anesthesia, however, doesn’t patients had been anesthetized, in what way and if seem to exist as these agents also carry a risk of any complications arose. causing rhabdomyolysis (7). A number of case All DMD patients recorded by the department reports questioned their safety profile in DMD of pediatric neurology (n=210) were potentially patients (8-10), propofol included. Rhabdomyolysis eligible for inclusion in our database, but not is only one clinical feature of the so-called propofol every single patient got withheld for inclusion. infusion syndrome (PRIS), otherwise characterized The following exclusion criteria were applied : by an unexplained myocardial failure and metabolic (a) Patients whose original medical file did not get acidosis in the setting of prolonged, high dose transferred into our hospital’s electronic patient propofol infusion (>4 mg kg-1 h-1 for > 48h) (11). record system (KWS, Klinisch WerkStation), (b) The use of locoregional anesthetic techniques Patients without any mention of an anesthetic inter- in DMD patients has been used without major vention in their medical file and (c) Patients that got complications, although caution is required for pos- included in the pediatric neurology database outside sibly increased sensitivity to local anesthetics (12). of the time window described previously. Unfortunately, regional anesthesia is not always A PRISMA-flow diagram (Fig. 1) shows possible. patient selection through inclusion and exclusion The University Hospitals Leuven (UZL) are a criteria, and the reasons for exclusion. tertiary referral center for DMD. As a tertiary referral Preoperative anesthetic assessment documents center, we deal with a broad spectrum of pathologies were mainly used to extract demographic data (sex, and have access to a large database of DMD patients. Various types of anesthetic techniques and agents are used on a routine basis. The aim of this study is to analyze all anesthetic records of DMD patients that underwent general or locoregional anesthesia in search of differences in complication rate between patients who received volatile or intravenous hypnotics, review the anesthetic management strategies and give recommendations on the optimal anesthetic management in this specific patient population.

Methods

Study design

This retrospective medical record analysis was approved on January 20th, 2020 by the Ethical Fig. 1 — PRISMA-flow diagram, depicting the selection of Research Committee (ERC) of the University patient population following the different criteria and hereby Hospitals Leuven (Chairperson prof. dr. Minne mapping the number of records identified, included, excluded, Casteels, reference number S63489). The ERC and the reasons for exclusion.

19-Degryse.indd 144 12/01/2021 12:22 anesthesia and duchenne muscular dystrophy 145 age, weight, length, ASA classification), regular Demographics drugs and an up-to-date medical history. Pulmonary function tests (forced vital capacity (FVC), forced A total of 185 anesthetic interventions expiratory volume in 1 second (FEV1) and FEV1/ performed in 105 DMD patients over a timespan of FVC ratio) and transthoracic echocardiograms were 50 years were analyzed. Among the whole group, reviewed when available. 60.5% of these interventions took place at an age Intraoperative and postoperative anesthesia records were reviewed in all patients. The following Table 1 data were extracted : type of surgery and anesthesia Patient and anesthetic characteristics (inhalational, intravenous, locoregional techniques Total Volatile intravenous or combinations), anesthetic agents used (hypnotics, anesthetics anesthetics neuromuscular blockers, local anesthetics), intuba- (n=78) (n=107) tion reports, type of airway used, duration of Age at surgery (years) surgery, inotropic or vasopressor need, estimated 0-10 31 26 blood loss, transfusion need and volume, summary 11-20 43 69 of postoperative stay on recovery or intensive care. 21-30 2 11 Procedures were divided into 2 groups ac- 31-40 2 1 cording to the type of anesthesia performed : use Mean demographics of a volatile agent or use of an intravenous agent. Mean age (years) ± SD 11.7 ± 6.5 14.3 ± 6.3 In search of a possible causative relation between Mean body weight (kg) ± SD 36.5 ± 16.2 45.5 ± 19.2 Mean length (cm) ± SD 130.2 ± 23.8 141.7 ± 21.4 the use of volatile anesthetics and perioperative ASA score at time of surgery complications (hemodynamic, respiratory, infec- ASA 1 10 6 tious or other), we pooled isolated volatile tech- ASA 2 22 25 niques and combination techniques (induction with ASA 3 41 63 a potent volatile agent and maintenance with an ASA 4 5 10 intravenous agent) in a single group. ASA 5 0 3 Volatile agent We performed a subgroup analysis of all patients Sevoflurane 55 that underwent locoregional anesthesia. Of note Sevoflurane induction – 22 is that these locoregional techniques were always propofol maintenance performed as an adjunct to a general anesthesia with Halothane 1 volatile and/or intravenous anesthetics. Therefore, Nitrous oxide 33 9 patients included in this subgroup are also included Intravenous agent Propofol - 103 in both main groups. Ketamine - 22 Midazolam - 2 Results Neuromuscular blockers Succinylcholine 0 0 Nondepolarizing muscle Identification of patient population 13 32 relaxants Rocuronium 2 19 Based on the previously mentioned inclusion Cisatracurium 10 3 and exclusion criteria, we ended up with a total Mivacurium 1 6 of 110 DMD patients, collectively undergoing a Vecuronium 0 4 total of 213 anesthetic interventions (Fig. 1). Of Airway management these 213 cases, preoperative anesthetic evaluation Preexisting endotracheal tube 1 1 documents, intraoperative anesthetic charts and Endotracheal intubation 28 54 Difficult intubation 6 14 surgical summary documents were evaluated. Use of muscle relaxants at 3 4 Cases were additionally excluded according to time of difficult intubation the following criteria : (a) No anesthetics were Laryngeal mask 36 26 administered (anesthetist remained stand-by) (n=1), Spontaneous airway 5 8 (b) No complete dataset of intraoperative (type of No information on airway 2 4 ICU admission anesthesia) and postoperative (recovery or intensive Elective 15 41 care unit) anesthetic records could be found (n=27). Unplanned 1 0 Eventually, the data that were included came Readmission 2 4 from 105 DMD patients that collectively underwent ASA = American Society of Anesthesiologists, SD = Standard a total of 185 anesthetic interventions. Deviation, ICU = Intensive Care Unit.

19-Degryse.indd 145 12/01/2021 12:22 146 e. degryse et al. range of 11 to 20 years, followed by the age group Table 3 0 to 10 years, in which 30.8% of procedures were Preoperative test results according to the type of anesthetic counted (Table 1). agent used More than half of the included DMD patients Volatile Intravenous Preoperative test results (n=104 [56.2%]) were categorized as ASA class (n=78) (n=107) 3 at the time of surgery. 60.6% (63/104) of these Median VC (% of predicted) ± SD 74 ± 29.5 58 ± 26.9 ASA 3 patients were anesthetized with an isolated (range) (7-119) (14-119)

intravenous technique, compared to 39.4% (41/104) Median FEV1 (% of predicted) ± 75 ± 27.2 56 ± 26.8 receiving a volatile based general anesthesia (Table SD (range) (7-116) (15-115) 1). A more balanced preference of choice of Median FEV1 (% of VC) ± SD 95.5 ± 10.1 94 ± 11.7 (range) (74-117) (34.5-117) anesthetic technique was found in the ASA 2 group 45 67 (n=47) : 53.2% (25/47) received only intravenous Normal TTE (N, %) (57.7%) (62.6%) anesthetics, versus 46.8% (22/47) receiving volatile 15 23 Pathological TTE (N, %) anesthetics. (19.2%) (13.1%) 18 13 ASA 1 patients were predominantly anesthe- No TTE (N, %) tized with an isolated volatile technique (n=10/16 (23.1%) (15.9%) [62.5%]). VC = Vital Capacity, FEV1 = Forced Expiratory Volume in 1 second, When averaging our demographics in function TTE = Transthoracic Echocardiography. of type of anesthetic agent used, mean age was higher in the intravenous group (Table 1). procedures, only volatile agents were used. In 22 An overview of the most common comor- cases, a potent volatile agent was used for induction bidities can be found in Table 2, together with the and maintenance was done with intravenous preoperative test results (Table 3). hypnotics (Table 1). The majority of surgery was performed in the Table 2 age group ranging from 11 to 20 years (112/185). Patient and surgical characteristics Out of these 112 procedures, 69 (60,5%) were Overall anesthetic procedures 185 performed using an intravenous anesthetic agent, Comorbid conditions 43 (38.4%) using a potent volatile anesthetic agent. Cardiomyopathy 32 In the age group ranging from 0 to 10 years (30,8% Restrictive lung disease 31 of procedures), volatile and intravenous agents were Non-invasive ventilation 20 used in a similar proportion (54.4% and 45.6% Scoliosis 49 Chronic corticoid use 108 respectively) (Table 1). Wheelchair dependent 44 Propofol is the most commonly used hypnotic Type of surgery for total intravenous anesthesia in our DMD patients. Spinal fusion 49 Similar results were found in a study by Muenster et Muscle biopsy 44 al (13). Sevoflurane was used almost exclusively in Orthopedic 36 volatile anesthetic techniques (Table 1). Dental care 15 Abdominal - Gastrointestinal 14 A more specific overview on the variety Cardiovascular 13 anesthetic agents used, is found in table 1. Other (thoracic, ophtalmologic, neurosurgery) 11 Locoregional technique Otolaryngologic 4 In this database of DMD patients the use of Type of surgery locoregional techniques was very limited. In 10 patients, undergoing a spinal fusion, the surgeon The mainstay of surgery performed was major placed an epidural catheter. After a bolus of local spinal surgery (26.4%), followed by muscle biopsy anesthetic, given by the anesthetist, the catheter (23.8%) and orthopedic interventions (19.5%) was connected with a patient controlled epidural (Table 2). analgesia pump. Details on which local anesthetics had been used were not available. General anesthesia One caudal block was performed in a 4-year- old patient for a leg muscle biopsy. 107 of 185 anesthetic interventions were Unfortunately, we found no details on types performed with a total intravenous technique. In 56 of local anesthetic used. Although the use of these

19-Degryse.indd 146 12/01/2021 12:22 anesthesia and duchenne muscular dystrophy 147 Table 4 Patient and surgical characteristics of patients undergoing spinal fusion

Volatile (n=12) Intravenous (n=37) Overall spinal fusions (n=49)

14.5 ± 1.1 16.1 ± 2.5 15.7 ± 2.4 Mean age (y) at surgery ± SD (range) (13-17) (12-22) (12-22) 51.8 ± 12.8 57.3 ± 14.0 55.9 ± 13.9 Mean body weight (kg) ± SD (range) (24-77) (35-109.8) (24-109.8) Population ASA2 0.083 0.081 0.082 Population ASA3 0.917 0.811 0.837 Population ASA4 0 0.108 0.082 49 ± 19.5 50 ± 20.8 49 ± 20.5 Median VC (% of predicted) ± SD (range) (30-106) (22-105) (22-106) 52 ± 21.4 55 ± 20.3 53 ± 20.7 Median FEV1 (% of predicted) ± SD (range) 34-116) (24-113) (24-116) 93 ± 10.7 93 ± 5.9 93 ± 7.7 Median FEV1 (% of VC) ± SD (range) (79-113) (77-102) (77-113) Normal TTE (n, %) 6 (50%) 24 (64.9%) 30/49 (61.2%) Pathological TTE (n,%) 6 (50%) 12 (32.4%) 18/49 (36.7%) No TTE (n, %) 0 1 1/49 (2.0%) 265.8 ± 46.6 320.4 ± 77.1 307 ± 74.5 Duration (min) (200-345) (190-510) (190-510) 2002.5 ± 922.3 2144.3±1456.2 2117 ± 1372 Estimated blood loss (ml) (150-3000) (700-8000) (150-8000) Nber perioperative complications 9 16 25 Nber of difficult intubations 3 6 9 Nber of difficult intubations despite use muscle relaxants 2 3 5

techniques in our hospital was very limited, no were already intubated and no information on complications related to its use were reported. difficulty was noted. In 20/102 intubations (19,6%) the anesthetist experienced some difficulties and Neuromuscular blockers and its reversal agents resorted to using intubating bougie/stylet, video laryngoscopy or fiberoptic intubation. At the In 45 of 185 anesthetic procedures (24.3%) moment of difficult intubation, muscle relaxants non-depolarizing neuromuscular blockers were had been given in 35% (Table 1). In 45% of reported used (Table 1) : rocuronium (n=21), cisatracurium intubation difficulty (9/20, despite the use of muscle (n=13), mivacurium (n=7) and vecuronium (n=4). In relaxants in 5 of them), planned surgery was spinal 33 of these 45 procedures (73.3%) a single dose was fusion. given at induction to facilitate intubation conditions, after which no further top-ups were given. Spinal fusions In two patients only, neuromuscular blocker Due to progressive loss of muscle strength, reversal agents (Sugammadex) were used. No DMD patients eventually develop a pronounced obvious reason for or complications with its use weakness of trunk and paraspinal muscles, leading were mentioned. The main reason for the limited to a worsening thoracic scoliosis with deteriorating use of neuromuscular blocker reversal agents was pulmonary function and need for surgical correction duration of surgery or continued sedation and (16, 17). More than a quarter (26.4%) of our ventilation post-surgery. reviewed surgical procedures were spinal fusions As depolarizing neuromuscular blocker in- (Table 2). The mean age at fusion was 15.7 years duced rhabdomyolysis is a known complication with a mean body weight 55.9 kg and mean height in DMD patients (14, 15), no depolarizing neuro- 157.2 cm. Most of these DMD patients had some muscular blocking agents were found to have degree of restrictive lung disease and 18 of 49 DMD been used in any of the 185 reviewed anesthetic patients had a pathological echocardiography. In interventions. 83.7%, they had an ASA physical status 3. The mean duration of surgery was 307 minutes and Difficult tracheal intubation the mean estimated blood loss 2117 ml (Table 4). On a total of 185 procedures 102 tracheal These findings are comparable to those found by intubations were performed. In 2 procedures patients Duckworth et al. (16) and Turturro et al. (17).

19-Degryse.indd 147 12/01/2021 12:22 148 e. degryse et al. In 18 procedures (36.7%) patients were hemo- Table 5 dynamically supported with vasopressors, more Overview of anesthesia-related perioperative complications specifically norepinephrine. Perioperative complications – Volatile Intravenous DMD patients are at greater risk of peri- anesthesia-related anesthetics anesthetics operative complications (14, 16). In the subgroup of Respiratory failure 1 4 patients undergoing spinal fusion, 25 perioperative Postoperative (upper) airway 1 3 complications arose (51%). obstruction Cardiac failure 1 5 Intensive care unit admission Cardiac dysrhythmias (bradycardia) 1 1

Post major surgery (spinal fusion, cardiotho- Postoperative airway obstruction requiring racic surgery, complicated abdominal surgery) prolonged intubation and ventilation was identified DMD patients were electively admitted to the in 4 cases (1 in the volatile group, 3 in the intravenous intensive care unit (n=56). group), 3 of which were extensive spinal fusions Patients who stayed on ICU preoperatively in prone positioning (possibly contributing to (n=6) were readmitted on ICU afterwards. One airway edema). Difficult intubation circumstances patient needed an unplanned ICU admission were noted in 2 cases. Macroglossia was present after developing a hyperkalemic cardiac arrest in preoperatively in the other 2 cases. All patients were recovery following uncomplicated anesthesia for successfully extubated within a mean of 36 hours placement of a Port-à-Cath (Table 1). postoperatively. One patient in the volatile group died of Perioperative complications heart failure 18 days post left ventricular assist device placement for end stage heart failure All intra- and postoperative complications (LVEF 12.8 %) with pulmonary hypertension (51 (taking place in theatre, recovery or intensive care) mmHg) despite high dose hemodynamic support were looked into. A perioperative complication with norepinephrine, dobutamine, and enoximone was defined as an event deviating from a normal infusion and nitric oxide inhalation. perioperative course and clearly documented so. Five patients in the intravenous group under- As anesthetists, apart from the occurrence going spinal fusion, were started on vasopressor of acute vital complications such as respiratory and/or inotropic support (norepinephrine, dobu- failure (hypoventilation, upper airway obstruction) tamine). 3 of these patients had a reduced left ven- and hemodynamic failure (cardiac failure due to tricular systolic function preoperatively (mean left acute on chronic decompensation or arrhythmias), ventricular ejection fraction 40,3%). Preoperative we were especially interested in the occurrence cardiac function was normal in the other 2 patients. of complications that might be directly linked to In all cases, intraoperative blood loss was high, our anesthetic agents such as anesthesia induced stressing a weak heart even further. Propofol rhabdomyolysis (AIR). infusion may have had an additional negative Of the 185 anesthetic interventions that got inotropic effect on the stressed heart. scrutinized, a perioperative complication arose in Sinus bradycardia occurred intraoperatively 38 cases (20.5%). In some cases, more than one in 2 patients. In one patient, sevoflurane was used different type of complication arose (overall amount for induction and maintenance. Propofol wat used complications n=43). Perioperative complications for induction and maintenance in the other. High were identified in 17 cases in the volatile group, concentrations of both anesthetic agents may cause counting 78 anesthetic interventions (21.8%). Peri- depression of sympathetic cardioaccelerator tone operative complications were found in 21 cases with a suppressive effect on heart rate. Lowering in the intravenous group, counting 107 anesthetic agent concentrations and an atropine bolus quickly interventions (19.6%) (Table 5). restored normal heart rate. Respiratory failure was reported in 5 patients, The mainstay of the anesthesia-related com- in 4 of them perioperatively to undergoing major plications described above can also be attributed spinal surgery. 4 patients developed CO2 retention. to the nature and extensiveness of surgery (e.g. 1 patient developed increased oxygen requirement major blood loss with coagulopathy in spinal fusion needing intensified respiratory physiotherapy. Only surgery), preexisting medical conditions specific one of these patients received a volatile anesthetic to DMD patients (skeletal muscle weakness with intraoperatively. restrictive lung disease, cardiomyopathy) or a

19-Degryse.indd 148 12/01/2021 12:22 anesthesia and duchenne muscular dystrophy 149

Table 6 Case 2 Overview of not anesthesia-related perioperative complications The medical history of this 21-year-old

Perioperative complications – not Volatile Intravenous patient revealed restrictive lung disease (Vital anesthesia related anesthetics anesthetics Capacity 20%), malnutrition (Body Mass Index Hemodynamic instability caused 11.4) and wheelchair dependency for 10 years at 0 2 by SIRS time of surgery. Preoperative cardiac evaluation Hemodynamic instability caused by 2 0 only revealed a limited septal hypokinesia. Total septic shock intravenous anesthesia was used during gastrostomy Bleeding 5 5 Pain problem (nociceptive, placement (Target Controlled Infusion with target 3 2 neuropathic) plasma concentration of 0.6 µg ml-1 maximum, Urinary disorder (retention, 2 1 duration 35 minutes). The intervention was un- polyuria) eventful, and he was referred to the ward. One Infection (wound, urinary tract) 2 1 day after placement, he developed an abdominal Air embolism 0 1 pain syndrome with myoglobinuria (dark urine). SIRS = Systemic Inflammatory Response Syndrome Rhabdomyolysis (heme dipstick 3+, serum CK 40100 U L-1, AST 906 U L-1, ALT 218 U L-1, combination of both. As such, these complications LDH 2519 U L-1, normal preoperatively) and cannot solely be attributed to the type of anesthetic metabolic and respiratory acidosis were noted. He agent used. was admitted on ICU where he was treated with An overview of not anesthesia-related peri- fluid resuscitation, alkalization, ion correction and operative complications is listed in table 6. sustained nightly BiPAP support. He was able to Rhabdomyolysis occurred in 2 cases. One return to the ward 24 hours after ICU admission. in the volatile group, the other in the intravenous AST = aspartate aminotransferase, normal group. range ≤ 37 U L-1 ALT = alanine transaminase, normal range ≤ Case 1 41 U L-1 LDH = lactate dehydrogenase, normal range A boy, aged 11, was scheduled for placement 157-272 U L-1 of a Port-à-Cath. He was on chronic corticoid BiPAP = Bilevel Positive Airway Pressure treatment and was still ambulatory (10 minutes) at time of surgery. Anesthesia was maintained Discussion with sevoflurane. Surgery was uneventful. After extubation, the boy was in pain. Moments after Duchenne muscular dystrophy is the most admission in recovery, he became cyanotic with common severe myopathy in children (4). Because deteriorating respiration and irresponsiveness. The of advances in supportive therapy, the lifespan of patient was resuscitated. A tension pneumothorax DMD patients has been extended to sometimes was excluded. Blood gas analysis revealed a even four decades (3, 4). This means a single DMD potassium level of 8.7 mmol/L and a metabolic and patient often undergoes a multitude of diagnostic

respiratory acidosis (pH 6.839, pCO2 81.7 mmHg and surgical procedures, each of them requiring and lactic acid 25 mmol L-1 (0.5-2.2)). Serum some sort of anesthetic support. Despite dealing creatinine kinase (CK) was up to 58660 U L-1 (≤ with them more frequently, these patients remain 190 U/L) and urine dipstick was highly positive (3+) challenging cases for the anesthesiologist as they for red blood cell heme. Extracorporeal membrane are prone to develop perioperative complications oxygenation (ECMO) and renal replacement (16). Some of these complications, like anesthesia- therapy was initiated. Despite maximal therapy, induced rhabdomyolysis and hyperkalemic cardiac he died 4 days later. Off note, one year earlier the arrest, are directly linked to anesthetic agents patient underwent an uneventful muscle biopsy as succinylcholine and volatile anesthetics. The under sevoflurane induction. We had no data on safe use of intravenous agents in DMD patients, anesthesia agents used in three other previous however, is less clear. In this retrospective medical procedures under general anesthesia. record analysis, we analyzed the anesthetic records of DMD patients receiving general or locoregional anesthesia. Besides rigorously describing patient, surgical and anesthetic characteristics, we searched

19-Degryse.indd 149 12/01/2021 12:22 150 e. degryse et al. for differences in complication rate between patients zymes could have fit this syndrome. In addition, who received volatile or intravenous hypnotics. We malnutrition is a potential risk factor for development compared these findings with available literature of PRIS (27). On the other hand, propofol was and give recommendations on the optimal anesthetic administered at low dose and during a short period management in this specific patient population. of time. Moreover, at the time of presentation of In this study the mainstay of procedures was clinical features, propofol had been discontinued performed in the age group ranging from 11 to for 24 hours. A PRIS-mediated rhabdomyolysis 20 years. As most DMD patients lose the ability alone seems unlikely, but we cannot exclude a to walk and become wheelchair dependent at the contribution to it. Muenster et al. (13) concluded age of ten (1), contractures in hips, knees and that propofol is not associated with the occurrence feet are common. Core muscle weakness leads to of rhabdomyolysis or unexplained fever in DMD up progressive scoliosis. Both conditions often require to date. In contrast, AIR was reported repeatedly in surgery. Mean age at contracture surgery and spinal association with the use of different potent volatile fusion surgery was 12. 4 and 15.7 years respectively. agents (19-24, 28-30). In our institution, sevoflurane In the age group ranging from 0 to 10 years, the was used almost exclusively in volatile anesthetic most commonly performed procedure was muscle techniques. biopsy, as this is often the way the disease first gets Apart from rhabdomyolysis, we reported a diagnosed. These findings are comparable to those variety of anesthesia-related perioperative com- made by Segura et al. (18). plications (respiratory failure, upper airway ob- Out of 185 revised procedures, we found two struction, cardiac failure and dysrhythmias). These cases of rhabdomyolysis. The first patient underwent complications could not be directly linked to the type an uneventful sevoflurane based general anesthesia. (volatile or non-volatile) of anesthetic agent used but No muscle relaxants were given. Only in the could be attributed to the nature and extensiveness recovery room, he developed a fatal hyperkalemic of surgery, to preexisting medical conditions cardiac arrest. Sevoflurane was thought to have specific to DMD patients or a combination of both. triggered rhabdomyolysis as other causes for the Complication rate between patients who received eventual cardiac arrest were excluded. Literature volatile or intravenous hypnotics was similar suggests that in AIR, there is a very wide range in (respectively, 21.8% versus 19.6%). time to onset of symptoms after exposure to a volatile We had too little information to make a trigger (isoflurane, sevoflurane, cyclopropane, statement about the safe use of local anesthetics halothane) (19-24). There is also a wide variation and locoregional anesthesia techniques, although in concentrations of a potent volatile anesthetic that generally its use is accepted to be safe in DMD can cause AIR, making it an idiosyncratic reaction patients (12). However, dosing of local anesthetics (19-24). A previously uncomplicated volatile based should be done carefully as increased sensitivity anesthesia is no guarantee that AIR will not occur at to local anesthetics is described in DMD (12). The a new exposure to a volatile anesthetic. use of ultrasound has caused a general increase in The second patient developed rhabdomyolysis popularity of locoregional anesthetic techniques and myoglobinuria, without cardiac arrest, nor in the past decades. Hopefully, more data will be cardiac dysrhythmias. Triggering events for rhabdo- available soon to evaluate its safety profile in this myolysis like seizures, infections, ionic disturbances patient population. However, young patient age, the and ischemia were absent. As there was only a mild nature of surgery and often pronounced orthopedic hyperglycemia (193 mg dL-1) and endocrinologic deformities in these DMD patients may limit its use. evaluation for diabetes mellitus was negative, dia- Airway management in DMD patients is often betic ketoacidosis was excluded. The presence challenging. Macroglossia, obesity, limited man- of ketones (dipstick 4+) and metabolic acidosis dibular mobility and orthopedic deformities like (with normal lactic acid) was assumed to be related scoliosis and contractures are factors contributing to preoperative fasting. Indeed, rhabdomyolysis to a limited mouth opening and thus a more difficult has been described in ketoacidosis in dystrophic airway anatomy (31). Our study revealed an overall patients (in this case body mass index 11.4) who rate of difficult tracheal intubations in 19.2% are fasted (25). It is unlikely that propofol triggered of cases. This is a much higher rate compared to rhabdomyolysis in this case. reported incidences in other retrospective re- Considering the variety in presenting clinical views (Segura et al. (18) 4.2% ; Muenster et al. features of PRIS (26), the presence of metabolic (13) 4-7.5%). One explanation could be that our acidosis, rhabdomyolysis and elevated liver en- institution is a university hospital, teaching many

19-Degryse.indd 150 12/01/2021 12:22 anesthesia and duchenne muscular dystrophy 151 anesthesia trainees, each on a different level on the risk of inducing AIR in these cases. Keep in mind, learning curve for endotracheal intubation. Another however, that in the second case rhabdomyolysis explanation could be that in our study population, still occurred despite the sole use of intravenous neuromuscular blockers were only used in 24% of agents. anesthetic procedures (and in only 35% of difficult This malnourished, dystrophic DMD patient intubations, 7/20). In contrast, use of muscle in ketoacidosis was fasted preoperatively, further relaxants was much higher in the study performed depleting his carbohydrate storages, making him by Segura et al. (18) (49.6%) and Muenster et an excellent substrate for development of rhabdo- al. (13) (58.8-100%). As muscle relaxants are myolysis. So the use of propofol infusions in these recommended to facilitate tracheal intubation (32), children requires at least caution. Limiting fasting a less restrictive use might end up in reducing our time and providing carbohydrate supplementation rate of 19.4%. But, the use of muscle relaxants in the form of a dextrose infusion may potentially in this specific population remains a concern for reduce the risk. the anesthetist. DMD patients are reported to It seems there is no anesthetic technique with a have an increased sensitivity to non-depolarizing zero risk for AIR. Every anesthetic intervention in a neuromuscular blockers (like vecuronium and DMD patient demands awareness about the potential mivacurium) (33), particularly in more advanced for AIR and should be approached with extra stages of the disease (34). For rocuronium, several vigilance at every step of way, from preoperative studies showed delayed onset time to maximum evaluation to referral to ward postoperatively. block and a prolonged duration of action, whilst a similar peak effect was found in DMD patients References compared with healthy controls (35-36). This altered 1. Sussman M. 2002. Duchenne Muscular Dystrophy. J. Am. response makes neuromuscular monitoring of depth Acad. Orthop. Surg. 10:138-151. of block a must (32). As there is an increased risk 2. Bushby K.M., Gardner-Medwin D., Nicholson L.V., for residual neuromuscular block, reversal with Johnson M.A. and Haggerty I.D., et al. 1993. The clinical, genetic and dystrophin characteristics of Becker muscular sugammadex is probably recommended following dystrophy. II. Correlation of phenotype with genetic and administration of steroidal neuromuscular blockers protein abnormalities. J. Neurol. 240:105–112. in DMD patients (32). As sugammadex has a 3. Finsterer J. and Cripe L. 2014. Treatment of dystrophin favorable safety profile and causes fast reversal of cardiomyopathies. Nat. Rev. Cardiol. 11(3):168-179. 4. Emery A.E.H. 2002. The muscular dystrophies. Lancet. steroidal non-depolarizing neuromuscular blockers 359:687-695. (37-39), a more liberal use of this pair of products 5. Grietens J., Thiessen S.E., Dewinter G., Veekman L. in DMD patients should be considered. and Van de Velde M. 2019. Acute rhabdomyolysis in dystrophinopathies during anesthesia : a systematic review. This is in sharp contrast with the widely Acta Anaesth. Belg. 70:55-66. accepted contraindication for using succinylcholine 6. Gray R.M. 2017. Anesthesia-induced rhabdomyolysis or in DMD patients (5). We found no signs of its use malignant hyperthermia : is defining the crisis important? Pediatr. Anesth. 27:490-493. in our database. 7. Warren J.D., Blumbergs P.C. and Thompson P.D. 2002. The limitations of this study are inherent to a Rhabdomyolysis : a review. Muscle Nerve. 25:332-347. retrospective analysis, since our results are based 8. Wang J.M. and Stanley T.H. 1986. Duchenne muscular on the presence and accuracy of medical records of dystrophy and malignant hyperthermia – two case reports. Can. J. Anesth. 33(4):492-497. patients. Crucial information about the perioperative 9. Irwin M.G. and Henderson M. 1995. Cardiac arrest during course might have gone missing. major spinal scoliosis surgery in a patient with Duchenne’s In conclusion, we have revised the anesthetic muscular dystrophy undergoing intravenous anaesthesia. Anaesth. Intens. Care. 23:626-629. records of 185 procedures in 105 DMD patients. 10. Schmidt G.N., Burmeister M.A., Lilje C., Wappler F. and We described two cases of rhabdomyolysis, one of Bischoff P. 2003. Acute heart failure during spinal surgery which was complicated with a fatal hyperkalemic in a boy with Duchenne muscular dystrophy. Br. J. Anaesth. 90(6):800-804. cardiac arrest after sevoflurane anesthetic. This 11. Kam P.C.A. and Cardone D. 2007. Propofol infusion case adds to the abundance of case reports on syndrome. Anaesthesia. 62:690-701. AIR in DMD patients. It is our opinion that vola- 12. Pérez-Ferrer A., Gredilla E., De Vincente J., Reinoso F., tile anesthetic agents should be avoided in this Garcia J. and Duran P. 2005. Bloqueo motor completo con ropivacaína epidural en el postoperatorio de un paciente population whenever possible. When volatile agents con distrofia muscular de Duchenne. Rev. Soc. Esp. Dolor. are necessary (difficult airway, difficult IV access), 12(1)46-49. prompt switch to intravenous agents should be made 13. Muenster T., Mueller C., Forst J., Huber H. and Schmit H.J. 2012. Anaesthetic management in patients with Duchenne once airway or IV access is secured. The safety muscular dystrophy undergoing orthopedic surgery : a advantages of a gas induction outweigh the small review of 232 cases. Eur. J. Anaesthesiol. 29:489-494.

19-Degryse.indd 151 12/01/2021 12:22 152 e. degryse et al. 14. Morris P. 1997. Duchenne muscular dystrophy : a challenge 28. Miller E.D., Sanders D.B., Rowlingson J.C., Berry F.A., for the anaesthetist. Paediatr. Anaesth. 7:1-4. Sussman M.D. and Epstein R.M. 1978. Anesthesia-induced 15. Gronert G.A. 2001. Cardiac arrest after succinylcholine. rhabdomyolysis in a patient with Duchenne’s muscular Anesthesiology. 94:523-529. dystrophy. Anesthesiology. 48(2):146-148. 16. Duckworth A.D., Mitchell M.J. and Tsirikos A.I. 2014. 29. Poole T.C., Lim T.Y.J., Buck J. and Kong A.S. 2010. Incidence and risk factors for post-operative complications Perioperative cardiac arrest in a patient with previously after scoliosis surgery in patients with Duchenne muscular undiagnosed Becker’s muscular dystrophy after isoflurane dystrophy. Bone Joint J. 96:943-949. anaesthesia for elective surgery. Br. J. Anaesth. 104(4):487- 17. Turturro F., Rocca B., Gumina S., De Cristofaro R., 489. Mangiola F. and Maggiona N., et al. 2005. Impaired primary 30. Simpson R.S. and Van K. 2013. Fatal rhabdomyolysis hemostasis with normal platelet function in Duchenne following volatile induction in a six-year-old boy with muscular dystrophy during highly-invasive spinal surgery. Duchenne muscular dystrophy. Anaesth. Intens. Care. Neuromuscular Disord. 15:532-540. 41(6):805-807. 18. Segura L.G., Lorenz J.D., Weingarten T.N., Scavonetto F., 31. Morris P. 1997. Duchenne muscular dystrophy : a challenge Bojani K., Selcen D. and Sprung J. 2013. Anesthesia and for the anaesthetist. Paediatr. Anaesth. 7:1-4. Duchenne or Becker muscular dyysttrophy : review of 117 32. Plaud B., Baillard C., Bourgain J.L., Bouroche G., anesthetic exposures. Pediatr. Anesth. 23:855-864. Desplanque L., and Devys J.M., et al. 2020. Guidelines on 19. Nathan A., Ganesh A., Godinez R.I., Nicolson S.C. and muscle relaxants and reversal in anaesthesia. Anaesth. Crit. Greeley W.J. 2005. Hyperkalemic cardiac arrest after Care Pain Med. 39:125-142. cardiopulmonary bypass in a child with unsuspected 33. Ririe D.G., Shapiro F. and Sethna N.F. 1998. The response Duchenne muscular dystrophy. Anesth. Analg. 100:672- of patients with Duchenne’s muscular dystrophy to 674. neuromuscular blockade with vecuronium. Anesthesiology. 20. Chalkiadis G.A. and Branch K.G. 1990. Cardiac arrest 88:351-354. after isoflurane anaesthesia in a patient with Duchenne’s 34. Ihmsen H., Schmidt J., Schwilden H., Schmitt H.J. and muscular dystrophy. Anaesthesia. 45:22-25. Muenster T. 2009. Influence of disease progression on the 21. Girshin M., Mukherjee J., Clowney R., Singer L.P. and neuromuscular blocking effect of mivacurium in children Wasnick J. The postoperative cardiovascular arrest of a and adolescents with Duchenne muscular dystrophy. 5-year-old male : an initial presentation of Duchenne’s Anesthesiology. 110:1016-1019. muscular dystrophy. Pediatr. Anesth. 16:170-173. 35. Muenster T., Schmidt J., Wick S., Forst J. and Schmitt H.J.

22. Kelfer H.M., Singer W.D. and Reynolds R.N. 1983. 2006. Rocuronium 0.3 mg/kg (ED95) induces a normal peak Malignant hyperthermia in a child with Duchenne muscular effect but an altered time course of neuromuscular block dystrophy. Pediatrics. 71:118 in patients with Duchenne’s muscular dystrophy. Pediatr. 23. Boba A. 1970. Fatal postanesthetic complications in two Anesth. 16:840-845. muscular dystrophic patients. J. Pediatr. Surg. 5:71-75. 36. Wick S., Muenster T., Schmidt J., Forst J. and Schmitt 24. Kleopa K.A., Rosenberg H. and Heiman-Patterson T. 2000. H.J. 2005. Onset and duration of rocuronium-induced Malignant hyperthermia-like episode in Becker muscular neuromuscular blockade in patients with Duchenne dystrophy. Anesthesiology. 93:1535-1537. muscular dystrophy. Anesthesiology. 102:915-919. 25. Hopkins P.M. 2010. Anaesthesia and the sex-linked 37. De Boer H., Van Esmond J., Booij L., and Driessen J. 2009. dystrophies : between a rock and a hard place. Br. J. Reversal of rocuronium-induced profound neuromuscular Anaesth. 104(4):397-400. block by sugammadex in Duchenne muscular dystrophy. 26. Hemphill S., McMenamin L., Bellamy M.C. and Hopkins Pediatr. Anesth. 19:1226-1228. P.M. 2019. Propofol infusion syndrome : a structured 38. Galvez J.L., Luis C., Ordonez A., Falco E., Mazzinari G., literature review and analysis of published case reports. Br. Roble D. and Vila M. 2011. Sugammadex, efficacious in J. Anaesth. 122(4):448-459. reversing a neuromuscular block in a woman with Becker 27. Mirrakhimov A.E., Voore P., Halytskyy O. Khan M. and muscular dystrophy. Rev. Esp. Anestesiol. Reanim. 58(10): Ali A.M. 2015. Propofol infusion syndrome in adults : a 614-617. clinical update. Crit. Care Res. Pract. 260385. 39. Kim J.E. and Chun H.R. 2017. Rocuronium-induced neuromuscular block and sugammadex in pediatric patient with Duchenne muscular dystrophy. Medicine. 96(13):pe6456.

19-Degryse.indd 152 12/01/2021 12:22 (Acta Anaesth. Belg., 2020, 71, 153-158) Comparison of analogue and digital PCIA, pilot study

M. Weyns, V. Saldien

Abstract : The patient controlled administration of various reasons (4). It has been widely proven to intravenous morphine has been studied widely and be a safe and effective acute treatment option (5). proven its efficacy. The administration devices however Making the (opioid) intravenous analgesia delivery have changed in the course of their use and the impact a patient-controlled process has been widely proven of these devices has not been studied even though there to be an effective way to both reduce pain scores are severe differences in patient feedback between and to reduce opioid dosage. the commercially available devices. It is suggested by the auther that, since it has been proven that patient We performed a literature search to review satisfaction increases while the opioid dose remains which types of pumps were used to deliver the similar, part of the satisfaction is dependend on auditive opioids using both MeSH as non-MeSH terms. or visual cues provided by the PCIA systems inducing a While there are numerous studies comparing placebo effect. In this pilot study we performed a post- different dosages of medication, we were unable to hoc analysis of data, comparing the evolution of reported uncover a study that compared two types of patient pain scores through the numeric rating scale (NRS) by controlled intravenous analgesia devices. However, patient between an analogue (Vygon Freedom 5) and there is a large difference between the commercially a digital (CADD Solis VIP ambulatory Pump, Smiths available systems, both in use and in price. In most Medical) PCIA pump for the administration of the same Belgian hospitals the standard of care is to use a morphine/dehydrobenzperidol solution during a 5 day digital PCIA infusion pump since it provides a period after surgery. No difference opioid dosage was commercially easily accessible device that can be considered, only a difference in NRS. Data analysis was performed through independent T-Testing on every used for both intravenous and epidural analgesia. measurement point, after checking for normality, using Maintenance of these devices is relatively easy, the IBM SPSS statistics software. Both groups showed however not inexpensive. The economic cost of the a similar downward trend as is expected in a post- electronic devices on the other hand is significant operative trajectory. Although a difference in trendline since the main, reusable electronic, device has was obtained, no significant result was obtained between a high initial cost, there is also the battery cost two groups. As was expected since this study served as a and disposable catheter part to consider. Other pilot study to determine the scale of future research. delivery devices are available that are analogue and disposable.1 Keywords : Analgesia, patient-controlled ; analgesia ; Since the PCIA pump is capable of reducing patient controlled / methods ; analgesics, opioid / admi- both the pain scores, improving the patient satis- nistration & dosage ; morphine ; infusion pumps. faction and reduce the delivered opioid dosage, it should be argued that part of the pain relief is based Introduction on placebo effects provided by the delivery system (6, 7). Therefore, we suspect that part of the placebo Post-operative analgesia has known rapid effect relies on auditive, visual and haptic cues that evolution and development of new materials. While the effectiveness and cost-efficiency for epidural

delivered analgesia in combination with continuous M. Weyns, MD ; V. Saldien, PhD wound infiltration has been widely studied (1-3). Dpt of Anesthesia, Universitair Ziekenhuis Antwerpen, There is a wide range of patients not eligible for Belgium. Corresponding author : M. Weyns, Dpt of Anesthesia, epidural analgesia. Patient controlled intravenous Universitair Ziekenhuis Antwerpen, Belgium. analgesia (PCIA) with parenteral morphine admi- E-mail : [email protected] nistration has an important place in the post- Paper submitted on Jun 9, 2020 and accepted on Jun 11, 2020. operative analgesia in most hospitals. Even after Conflict of interest : none. initializing epidural analgesia, there is a frequent Presentation of this study in the form of an abstract occurred on transfer to intravenous pain medication because of the Online SARB graduation day meeting on 23 may 2020.

20-Weyns.indd 153 12/01/2021 12:23 154 m. weyns & v.saldien are provided by the delivery system. This has not been investigated by previous researchers.

Objective

Since no comparative study exists, we performed a pilot study comparing two different types of intravenous administration devices, both commercially available for efficacy, to provide data for a power calculation for future research.

Method Fig. 1. — Graphical representation of the findings, also a statistically insignificant linear regression is shown for both systems. The pilot study was conducted at the center where the authors are active. Post-hoc data was analyzed of two pump systems. The first pump Statistical analysis was performed using the system was obtained from Vygon (Vygon Freedom IBM SPSS v22 software suite. First control of 5 model 7301.10 PCA kit), the second pump normality was performed by usage of the Shapiro- system was obtained from CADD systems [CADD Wilk test. Testing for homogeneity invariance was Solis VIP ambulatory Pump, Smiths Medical ). performed by using the Levene Test. An independent Both devices were loaded with the same dose of T Test was performed on every point of observation morphine (1 milligram per milliliter (mg ml-1)] to average the NRS score results set against the time and dehydrobenzperidol (DHBP ; 0.05 mg ml-1). since admission to the ward. No other confounders Patients agree to post-hoc data evaluation upon were considered for this study as this was a pilot admission to the ward, prior to their surgery. No study and patient subgroups were considered be too additional informed consent was obtained. Data small to yield significant results. The obtained data registry started from arrival of our patients in their was also insignificant to consider the amount of respective wards after surgery and a limited stay in morphine delivered to every patient as an outcome the anesthesia recovery ward. Observations were parameter. After obtaining a mean on every point of performed by nursing staff of the ward, based on measurement, a linear plot line was calculated using the Numeric Rating Scale (NRS) as is the standard the Microsoft Office Excell, build 12730.20270, of care at our hospital during the first five 24-hour to provide graphical representation of the dataset periods of medication administration after return obtained through the SPSS software (Fig. 1). from the anesthesia recovery ward. Patients with a stay in the intensive care unit or a prolonged stay Results in the anesthesia recovery ward, defined as more than 6 hours postoperative, were rejected as the The test of normality using the Shapiro-Wilk reliability of the data could not be guaranteed and test resulted in regular values >0.05, meaning other confounding medication could be present. This violation of normality assumption, as is shown is because it is a common practice on the intensive in Table 1. This is because there are relatively care ward to provide a continuous administration of few possible NRS scores and few patients in both morphine and DHBP through the electronic pump, groups. Since the independent T Test is considered which is not registered in the patient follow-up as robustly performant when in violation of charts. normality, testing was still performed on the later Data from 73 patients was obtained over a observation points. period from April 2018 through May 2018 for the Since the Levene test showed equal variance analogue pump and from April 2019 through May distribution across all observation points, only these 2019 for the digital pump. In the Vygon group 8 values were retained in the independent T Test. patients were rejected before analysis because of a Table 2 shows the means and standard deviations stay in the intensive care unit, leading to a group for on each observation point for both pumps. Table 3 statistical analysis of n = 33 patients. In the CADD shows the independent T Test for each measurement Solis group 15 patients were rejected because of between the analogue and digital PCIA pump. As their stay in the intensive care unit, leading to a is shown in the table, only on measurement D2-3 group for statistical analysis of n = 17 patients. (third measurement on the third day) statistically

20-Weyns.indd 154 12/01/2021 12:23 comparison of analogue and digital pcia, pilot study 155

Table 1 Discussion Test for normality using the Shaprio-Wilk Test Although no overall statistically significant Tests of Normality result was obtained in comparing both study groups, Shapiro-Wilk this study was meant serve as the basis for a power Statistic df Sig. calculation for a future, larger scale, observational D0-1 0,729 4 0,024 D0-2 0,729 4 0,024 study to compare an analogue and digital delivery D0-3 0,63 4 0,001 method for efficacy in patient-controlled pain relief. D0-4 0,863 4 0,272 After performing statistical analysis, both groups D0-5 0,729 4 0,024 showed a similar downward linear evolution, though D0-6 0,729 4 0,024 only on one observation was there a statistically D1-1 0,729 4 0,024 significant difference between an analogue and a D1-2 0,63 4 0,001 digital pump, favoring the analogue pump. D1-3 0,63 4 0,001 D1-4 0,791 4 0,086 When considering the one significant obser- D1-5 0,729 4 0,024 vation point, we can only make assumptions as to D1-6 0,729 4 0,024 why there is a difference, since no other outcome D2-1 0,729 4 0,024 parameter was considered. There is no measurement D2-2 0,729 4 0,024 of delivered morphine and/or DHBP nor is there a D2-3 0,945 4 0,683 registration patient opinion as to their experience D2-4 0,945 4 0,683 D2-5 0,801 4 0,103 with the PCIA pump. We can suggest that pain D2-6 0,945 4 0,683 relief was possibly compromised because of the D3-1 0,945 4 0,683 many alarms on the digital pump, e.g. for battery or D3-2 0,729 4 0,024 reservoir depletion, intravenous line obstruction or D3-3 . 4 . near-obstruction and so on. This might interfere with D3-4 . 4 . an already compromised sleep pattern, resulting in D3-5 0,63 4 0,001 D3-6 0,63 4 0,001 higher NRS values. D4-1 0,945 4 0,683 Another possible explanation for the results D4-2 0,63 4 0,001 is that the audiovisual feedback when pressing the D4-3 0,63 4 0,001 digital PCIA pump, e.g. the mechanical noise and D4-4 0,729 4 0,024 the light on the pump handle, is absent when the D4-5 0,863 4 0,272 lockout period is still in effect. This is not the case D4-6 0,863 4 0,272 D5-1 0,993 4 0,972 in the analogue PCIA pump, which provides the D5-2 0,945 4 0,683 same haptic feedback regardless of the amount of D5-3 0,801 4 0,103 chamber filling. The aforementioned placebo effect D5-4 0,895 4 0,406 is therefore not broken, the patient can assume he D5-5 0,729 4 0,024 is being dosed, while in reality he is only pressing D5-6 0,729 4 0,024 a button. When regarding possible points of improve- significant difference is shown (M= 1.097 ; SD = ment, neglecting the obvious lack in scale in this 0.497, t = 2.205, p = 0.033), with a higher NRS pilot study, we could incorporate confounders for score on the digital PCIA pump, thus favoring type of surgery/trauma, for this would be expected the analogue PCIA pump for pain relief. On no to impact the evolution of pain and its relief, since other observation could a statistically significant not all patients follow the same linear pattern of difference be shown. pain improvement. When abdominal surgery has It is worth noting that towards the later passed, one expects a linear improvement of the observations, few patients remained in either study pain with no further events. In orthopedic surgery group, with only 3 patients completing the full 5 however, we expect to see a rise in NRS scores 24 hour observation periods. This is because the when mobilization and/or physiotherapy occurs. administration of IV morphine through PCIA pumps Another issue with this study, is that we did not take for those patients had ceased, however no distinction into account the amount of delivered morphine, for is made between patients being discharged from the the objective of this study was merely to show a ward or still being admitted, but no longer needing difference in NRS score, not in amount of morphine patient controlled morphine delivery. delivered. This might provide an interesting sub- analysis in a future study.

20-Weyns.indd 155 12/01/2021 12:23 156 m. weyns & v.saldien Table 2 Mean and standard deviation for both groups on every point of observation, Mean value is the mean value for the observed NRS score

Type Pomp N Mean Std. Deviation Std. Error Mean Digital Pump 17 3,29 2,494 ,605 D0 - 1 Analogue Pump 33 2,21 2,408 ,419 Digital Pump 17 2,94 2,135 ,518 D0 -2 Analogue Pump 31 2,29 1,637 ,294 Digital Pump 17 2,24 1,855 ,450 D0-3 Analogue Pump 28 2,71 2,052 ,388 Digital Pump 16 2,44 1,548 ,387 D0-4 Analogue Pump 26 2,31 1,594 ,313 Digital Pump 17 2,29 1,404 ,340 D0-5 Analogue Pump 25 2,40 1,443 ,289 Digital Pump 15 2,13 1,356 ,350 D0-6 Analogue Pump 25 2,72 1,768 ,354 Digital Pump 17 2,29 1,829 ,444 D1 - 1 Analogue Pump 26 2,58 1,770 ,347 Digital Pump 17 1,94 1,819 ,441 D1 -2 Analogue Pump 25 2,32 1,547 ,309 Digital Pump 17 2,59 1,839 ,446 D1-3 Analogue Pump 27 2,22 1,311 ,252 Digital Pump 17 2,65 1,730 ,420 D1-4 Analogue Pump 27 2,30 2,109 ,406 Digital Pump 17 2,18 1,741 ,422 D1-5 Analogue Pump 24 2,46 1,841 ,376 Digital Pump 17 2,06 1,713 ,415 D1-6 Analogue Pump 25 2,00 1,500 ,300 Digital Pump 17 2,41 1,770 ,429 D2 - 1 Analogue Pump 23 2,00 1,567 ,327 Digital Pump 17 2,06 1,560 ,378 D2 -2 Analogue Pump 24 2,00 1,560 ,319 Digital Pump 17 2,88 1,965 ,477 D2-3 Analogue Pump 28 1,79 1,371 ,259 Digital Pump 16 2,38 1,360 ,340 D2-4 Analogue Pump 24 2,08 1,666 ,340 Digital Pump 15 1,60 1,639 ,423 D2-5 Analogue Pump 25 2,20 1,658 ,332 Digital Pump 15 1,33 1,447 ,374 D2-6 Analogue Pump 27 1,96 1,285 ,247 Digital Pump 16 1,31 1,352 ,338 D3 - 1 Analogue Pump 24 1,88 1,569 ,320 Digital Pump 15 1,93 2,187 ,565 D3 -2 Analogue Pump 24 1,54 1,351 ,276 Digital Pump 15 1,93 2,658 ,686 D3-3 Analogue Pump 24 1,71 1,122 ,229 Digital Pump 14 1,21 1,122 ,300 D3-4 Analogue Pump 23 1,70 1,428 ,298 Digital Pump 13 2,38 2,987 ,828 D3-5 Analogue Pump 23 1,52 1,201 ,250 Digital Pump 13 1,62 1,502 ,417 D3-6 Analogue Pump 24 1,58 1,213 ,248 Digital Pump 12 1,33 1,497 ,432 D4 - 1 Analogue Pump 20 1,15 1,137 ,254 Digital Pump 12 1,42 1,564 ,452 D4 -2 Analogue Pump 22 ,91 ,868 ,185 Digital Pump 11 1,55 1,809 ,545 D4-3 Analogue Pump 22 1,00 ,816 ,174 Digital Pump 10 1,10 1,449 ,458 D4-4 Analogue Pump 17 1,18 ,951 ,231 Digital Pump 10 1,20 1,476 ,467 D4-5 Analogue Pump 18 1,17 1,150 ,271

20-Weyns.indd 156 12/01/2021 12:23 comparison of analogue and digital pcia, pilot study 157

Digital Pump 9 1,00 1,118 ,373 D4-6 Analogue Pump 19 1,21 1,548 ,355 Digital Pump 5 1,00 1,000 ,447 D5 - 1 Analogue Pump 17 1,47 1,375 ,333 Digital Pump 5 1,40 1,342 ,600 D5-2 Analogue Pump 17 1,53 1,375 ,333 Digital Pump 4 2,00 1,414 ,707 D5-3 Analogue Pump 16 1,63 1,628 ,407 Digital Pump 3 2,00 1,732 1,000 D5-4 Analogue Pump 15 1,67 1,175 ,303 Digital Pump 3 2,33 2,517 1,453 D5-5 Analogue Pump 15 1,13 ,915 ,236 Digital Pump 3 1,33 1,155 ,667 D5-6 Analogue Pump 11 1,45 1,293 ,390

Table 3 Test on every observation point with Equal variance assumed; CI: confidence interval Levene’s Test for t-test for Equality of Means Equality of Variances Sig. Std. Error 95% CI Observation point F Sig. T Mean Difference (2-tailed) Difference Difference Lower Upper D0-1 0,586 0,448 1,487 0,143 1,082 0,728 -0,381 2,545 D0-2 1,013 0,319 1,181 0,244 0,651 0,551 -0,458 1,760 D0-3 0,068 0,795 -0,786 0,436 -0,479 0,609 -1,707 0,749 D0-4 0,176 0,677 0,259 0,797 0,130 0,501 -0,883 1,143 D0-5 0,000 0,997 -0,236 0,815 -0,106 0,449 -1,013 0,801 D0-6 0,326 0,571 -1,103 0,277 -0,587 0,532 -1,663 0,490 D1-1 0,385 0,538 -0,506 0,616 -0,283 0,559 -1,413 0,847 D1-2 1,903 0,175 -0,725 0,472 -0,379 0,522 -1,434 0,677 D1-3 2,511 0,121 0,771 0,445 0,366 0,475 -0,592 1,324 D1-4 0,051 0,822 0,574 0,569 0,351 0,611 -0,882 1,584 D1-5 0,007 0,933 -0,494 0,624 -0,282 0,571 -1,436 0,873 D1-6 2,150 0,150 0,118 0,907 0,059 0,499 -0,950 1,068 D2-1 0,806 0,375 0,778 0,442 0,412 0,529 -0,660 1,484 D2-2 0,023 0,880 0,119 0,906 0,059 0,495 -0,942 1,059 D2-3 0,621 0,435 2,205 0,033 1,097 0,497 0,094 2,100 D2-4 0,396 0,533 0,582 0,564 0,292 0,501 -0,723 1,306 D2-5 0,289 0,594 -1,113 0,273 -0,600 0,539 -1,692 0,492 D2-6 2,535 0,119 -1,454 0,154 -0,630 0,433 -1,505 0,245 D3-1 0,040 0,842 -1,172 0,249 -0,563 0,480 -1,534 0,409 D3-2 1,352 0,252 0,694 0,492 0,392 0,565 -0,752 1,536 D3-3 4,397 0,043 0,368 0,715 0,225 0,612 -1,015 1,465 D3-4 0,157 0,694 -1,074 0,290 -0,481 0,448 -1,392 0,429 D3-5 7,867 0,008 1,231 0,227 0,863 0,701 -0,562 2,288 D3-6 0,052 0,821 0,071 0,944 0,032 0,454 -0,890 0,954 D4-1 1,578 0,219 0,392 0,698 0,183 0,468 -0,772 1,138 D4-2 5,389 0,027 1,224 0,230 0,508 0,415 -0,337 1,352 D4-3 5,756 0,023 1,203 0,238 0,545 0,453 -0,379 1,470 D4-4 1,941 0,176 -0,166 0,869 -0,076 0,460 -1,025 0,872 D4-5 1,190 0,285 0,066 0,948 0,033 0,502 -0,998 1,065 D4-6 0,425 0,520 -0,364 0,719 -0,211 0,579 -1,400 0,979 D5-1 1,977 0,175 -0,707 0,488 -0,471 0,666 -1,859 0,918 D5-2 0,008 0,930 -0,186 0,854 -0,129 0,696 -1,581 1,322 D5-3 0,215 0,648 0,421 0,679 0,375 0,891 -1,497 2,247 D5-4 0,923 0,351 0,419 0,681 0,333 0,796 -1,354 2,020 D5-5 6,206 0,024 1,536 0,144 1,200 0,781 -0,456 2,856 D5-6 0,129 0,726 -0,146 0,886 -0,121 0,828 -1,925 1,683

20-Weyns.indd 157 12/01/2021 12:23 158 m. weyns & v.saldien

Conclusion 2. Liu S, Richman J, Thirlby R and Wu C. 2006. Efficacy of continuous wound catheter delivering local anesthetic for postoperative analgesia : a quantitative and qualitative No statistically significant difference could systematic review of randomized controlled trials. J Am Coll be shown between the analogue PCIA pump and Surg 203: 914-923 the digital PCIA pump for the administration of 3. Ochroch EA, Troxel AB, Frogel JK and Farrar JT. 2007. The inﬂuence of race and socioeconomic factors on patient Morphine/DHBP for hospitalized patients using acceptance of perioperative epidural analgesia. Anesth Analg the available data. The data can be used for power 105: 1787-1792 calculation of a future comparison study between 4. Ready L. 1999.Acute pain : lessons learned from 25 000 patients. Reg Anesth Pain Med 24: 499-505 both delivery mechanisms. 5. Walder B, Schafer M, Henzi I and Tramer MR. 2001. Efficacy and safety of patient-controlled opioid analgesia for References acute postoperative pain. A quantitative systematic review. Acta Anaesthesiol Scand 45: 795-804 1. Beaussier M, El’Ayoubi H and Schiffer E, et al. 2007. 6. Ballantyne JC, Carr DB, Chalmers TC, Dear KB, Angelillo Continuous preperitoneal infusion of ropivacaine provides IF and Mosteller F. 1993. Postoperative patient-controlled effective analgesia and accelerates recovery after colorectal analgesia. Meta-analysis of initial randomized control trials. surgery : a randomized, double-blind, placebo-controlled J Clin Anesth. 5:182-193. study. Anesthesiology 107: 461-468 7. Chumbley GM, Hall GM and Salmon P.1999. Why do patients feel positive about patient-controlled analgesia? Anaesthesia. 54:386-9.

20-Weyns.indd 158 12/01/2021 12:23 (Acta Anaesth. Belg., 2020, 71, 159-167) Comparing ultrasound-guided retroclavicular versus infraclavicular brachial plexus blockade : a retrospective analysis of safety and efficacy for distal upper limb surgery

A. Beaumond, S. Coppens

Abstract : Background : The ultrasound-guided coracoid Introduction approach to the infraclavicular block is the most common, but needle visualization can be challenging due to a steep Advantages of multimodal analgesia including angle. The retroclavicular approach enhances needle regional anesthesia for surgery on limbs have been imaging as the needle can be kept parallel to the ultrasound widely described and are largely accepted. The probe. The main drawback of this approach is a blind spot brachial plexus arises from C5 to T1 in the neck, caused by acoustic shadowing when passing below the then descends and runs under the clavicle through collarbone. This ensues potential injury to neurovascular structures, especially the suprascapular nerve which can the axilla down to the arm. Various techniques for lie in the needle trajectory. This study aimed to determine brachial plexus blockade are possible to render the incidence of long term neurological complications surgical anesthesia and analgesia of the lower arm. using both approaches. Secondary objectives were the The infraclavicular block has suggested advantages incidence of short term adverse events, block success rate for procedures from the elbow to the hand. The and opioid consumption. brachial plexus is arranged as three cords (lateral, Methods : We conducted a retrospective cohort study medial and posterior) around the axillary artery at the reviewing charts of all patients who received an level of the infraclavicular space. An effective block infraclavicular block for distal arm surgery between can be achieved with a single U-shaped injection of 01/01/2016 and 31/12/2018, with inclusion of follow- local anesthetic to surround the artery despite not up information up until 1 year after surgery. Exclusion accurately identifying the three cords, avoiding the criteria were unavailable medical records, and loss to need for multiple needle passes. The block can also follow-up with their attending surgeons immediately after the procedure. Statistical analysis was performed be performed without having to abduct or move the using SigmaStat version 4.0 (SYSTAT software inc.). arm, which is more comfortable for patients with an Results : 155 patients were included, 94 in the coracoid injured limb or restrictive mobility of the shoulder/ group and 61 in the retroclavicular group. One patient had elbow joint. A significant decrease in tourniquet persistent neurological symptoms possibly associated pain has been reported with an infraclavicular block with the retroclavicular block (1,6%) compared to two compared with other blocks. And lastly, it allows patients in the coracoid group (2,1%), which was not reliable continuous catheter placement, seeing as the statistically significant (p-value 0,703). New onset and chest wall musculature helps securing the catheter persistent neurological symptoms were all associated and prevents its dislodgement without excessively with ulnar neuropathy. Secondary objectives did not restricting arm movement (1-3). show a significant statistical difference. There are various approaches to deposit local Conclusion : Primary and secondary objectives were anesthetic around the infraclavicular brachial plexus, similar in both groups. This study unfortunately does not have the power to validate the safety and the efficacy of the retroclavicular approach. We could however not find one case of suprascapular nerve injury in the 61 patients who received an infraclavicular plexus block using the A. Beaumond, MD ; S. Coppens, MD retroclavicular approach. It can be assumed that cases Department of Anesthesiology, University Hospitals Leuven, 3000 Leuven, Belgium with ulnar neuropathy were not a consequence of the Corresponding author : Dr. S. Coppens, Department of block but rather due to surgery or trauma. Anesthesiology, University Hospitals Leuven, Herestraat 49, 3000 Leuven, Belgium. Keywords : Anesthesia and analgesia ; brachial plexus E-mail : [email protected] block ; regional anesthesia ; upper extremity ; ultraso- Paper submitted on Jun 23, 2020 and accepted on Jun 26, 2020. nography ; interventional. Conflict of interest : none.

21-Beaumond.indd 159 12/01/2021 12:23 160 a. beaumond & s. coppens the most common being the ultrasound-guided paresthesia during block performance and a shorter coracoid approach. The transducer probe is placed time needed to perform the retroclavicular block (9, in a parasagittal plane below the coracoid process 10). to identify the axillary artery and the different cords Grape et al. compared it to the supraclavicular of the brachial plexus. Scanning medially allows brachial plexus block, showing an identical success you to locate the pleura in order to maintain a safe rate while providing similar pain relief, albeit with distance. The needle is then inserted just below reduced needling time for the supraclavicular plexus the clavicle, in a plane from ventral to dorsal and block (11). cephalo-caudal, aiming towards the posterior aspect The main drawback of the retroclavicular of the axillary artery. Needle visualization can be approach is the acoustic shadowing when passing challenging due to the depth of the plexus and below the collarbone, making the needle invisible. limited space for the transducer to be placed below This causes a hazard of damaging neurovascular the clavicle, resulting in a steep angle of the needle structures when advancing the needle blindly (2, 4). beneath the clavicle. Due to the inability to compress The posterior or retroclavicular approach vascular structures underneath the clavicle it seems to the infraclavicular block was first described reasonable to proceed with caution in case of by Hebbard and Royse (5). In this approach, the coagulopathy. ultrasound probe is placed inferior to the collarbone A recent cadaver study calls attention to and medial to the coracoid process, identifying the the potential risk of suprascapular nerve injury. different cords and axillary vessels in a parasagittal Dissection following the retroclavicular block plane, while confirming adequate distance from showed this nerve was systematically found in the the ribs and pleura. The needle is inserted in the path of the needle posterior to the clavicle, directly supraclavicular fossa sufficiently posterior to the touching the needle in three out of six cases, and lateral part of the clavicle, usually 1-2cm, and medial in close proximity in the remaining cases. The to the insertion of the trapezius muscle. It is wise nerve supplies motor branches to the supraspinatus to assess the length of the surface distance between and infraspinatus muscle and sensory branches the insertion point and the probe, which should be to the shoulder joint. Injury may manifest as about 3 to 4 cm, before inserting the needle. The weakness with external rotation and abduction of needle is then advanced cephalo-caudally, passing the humerus, or pain localized to the superior or just underneath the collarbone while staying aligned posterolateral aspect of the shoulder. To reduce the with the probe, until the needle tip becomes visible, risk of iatrogenic nerve injury, it is best to exclude aiming underneath the artery. With a different any external rotation of the shoulder using nerve needle entry point above the clavicle, it is possible stimulation while advancing the needle in the to maintain the needle parallel to the ultrasound acoustic shadow underneath the clavicle (12-14). probe, which enhances needle visualization as the This raises concern about the safety of this ultrasound beams are reflected perpendicularly. particular approach. However, there has yet to be This theoretically improves the safety of the block a report in literature of suprascapular nerve injury as the tip of the needle can be steered towards the following the retroclavicular approach. At our objective under direct visibility, avoiding damage center, we have been regularly performing the to neurovascular and thoracic structures. Additional retroclavicular block for a number of years, with advantages have been described, notably avoiding positive outcomes similar to the aforementioned the cephalic vein, the acromial branch of the thoraco- published research. As far as we know, there are acromial artery and the lateral cord, which all lie no published studies looking at the long term in the needle trajectory of the classical coracoid complication rate of this retroclavicular approach. approach (6, 7). This is a retrospective cohort study that aims to review This novel approach has been validated by a charts of patients who received an infraclavicular number of studies. Charbonneau et al. described block over a period of 2 years for a variety of its feasibility, concluding that the approach offers a distal arm surgery, in particular to determine the high success rate and low needling time (8). risk of short and long term complications using the Two studies comparing this approach to the retroclavicular approach compared to the coracoid coracoid approach showed a similar success rate approach. and satisfaction but with a significantly better visualization of needle tip and shaft. Ozturk et al.’s study even demonstrated fewer needle passes, lower

21-Beaumond.indd 160 12/01/2021 12:23 a retrospective analysis of safety and efficacy for distal upper limb surgery 161

Materials and Methods block, confirming correct documentation of block by reviewing reports and imaging, reviewing This manuscript is a retrospective cohort study operative anesthetic charts, reviewing drug charts in and adheres to the applicable STROBE guidelines. the 24h period after surgery, and finally reviewing The trial was conducted in compliance with the postoperative visits with the attending surgeon and/ principles of the Declaration of Helsinki (current or visits with physiatrist or any contacts suggesting version), the principles of GCP and in accordance persistent neurological injury or other possible side with all applicable regulatory requirements. The effects related to the regional anesthesia up to one protocol and related documents were submitted and year after surgery. It was expected that we would approved by the Clinical Trial Center and the Ethics need to evaluate the medical records of 250 patients Committee. in order to identify all subjects who were eligible The study procedures were limited to reviewing for inclusion in this study. and collecting information of already existing The primary objective of this study was to medical records. The study involves all patients who determine the incidence of persistent neurological received an infraclavicular brachial plexus block for symptoms after 1 year associated with the retro- distal upper limb surgical procedures, performed clavicular approach compared to the coracoid over a period of two years at University Hospitals approach. Leuven. All cases were included receiving their The secondary objectives were the incidence initial surgery between 01/01/2016 and 31/12/2018, of short term adverse events : new onset of with inclusion of follow up information on all neuropathy (described as new neuropathy in the subjects up until 1 year after surgery. immediate postoperative period lasting up to one Eligibility criteria were males and females who month), vascular puncture or hematoma, Horner’s underwent distal upper limb surgical procedures with syndrome, local anesthetic systemic toxicity an infraclavicular block, using the retroclavicular or (LAST), pneumothorax and infection at injection the coracoid approach, between the aforementioned site. Other objectives were to assess the efficacy period, and had at least one postoperative visit with of this technique for distal upper limb surgical their attending surgeon. Exclusion criteria were procedures and postoperative analgesia, defined as unavailable medical records, and loss to follow up block success rate and opioid consumption during with their attending surgeons immediately after the the first 24 hours starting from surgery, using procedure. cumulative morphine milligram equivalents and We decided not to exclude patients with a pre- satisfaction scores. existing neurological deficit of the distal arm or Statistical analysis was performed using patients who underwent amputation surgery from SigmaStat version 4.0 (SYSTAT software inc.). analysis to broaden the patient population in an effort Descriptive data are presented as mean ± standard to look at suprascapular nerve injury in particular. deviation for continuous variables and number (n) In the University Hospitals of Leuven, having a pre- with percentages for nominal variables. Comparative existing neurological deficit does not necessarily data were tested for a normal distribution with the exclude patients from receiving locoregional Shapiro-Wilk test. Normally distributed variables anesthesia. The decision is individualized as some were compared using the independent sample t-test. patients may benefit from a locoregional anesthetic The Mann-Whitney Rank Sum Test was used for technique rather than other anesthetic or analgesic nonparametric variables. The chi-squared test and options, under the condition that the preoperative Fisher’s exact test were used for nominal data neurological status is scrutinously documented where appropriate. A probability value of p < 0.05 prior to any intervention. Only new onset of was considered to indicate statistical significance. neuropathy was taken into account for this study, excluding pre-existing neuropathy or neuropathy Results due to amputation. Data collection included all of the following : Between January 2016 and December reviewing the existing regional anesthesia database 2018, 203 patients were identified in the regional for all infraclavicular blocks during the designated anesthesia database, whose records were con- period, identifying patients personal information sequently screened. 155 patients were available for (name, dates of birth and patient’s hospital analysis. 48 patients were excluded, 37 of those identification numbers) and date of surgery with received a different block than was documented in corresponding date of the infraclavicular plexus the database, 9 patients figured twice on the list, and

21-Beaumond.indd 161 12/01/2021 12:23 162 a. beaumond & s. coppens Table 1 Table 2 Baseline characteristics Type of surgery

RA (n = 61) CA (n = 94) RA CA Age 50 ± 19 50 ± 18 Male 29 (48%) 44 (47%) (n = 61) (n = 94) Elbow osteosynthesis (open reduction Female 32 (52%) 50 (53%) 27 (44,3%) 46 (48,9%) BMI 26,3 ± 5,6 25,6 ± 4,6 and internal fixation) ASA I 17 (28%) 28 (30%) Elbow arthroplasty (radial or total) 12 (19,7%) 16 (17%) Elbow arthrolysis ± removal of osteosyn- ASA II 34 (56%) 52 (55%) 6 (9,8%) 13 (13,8%) ASA III 10 (16%) 13 (15%) thesis material Smoker 13 (21%) 23 (24%) Distal radial and/or ulnar osteosynthesis 6 (9,8%) 5 (5,3%) Amputation (ranging from distal forearm Diabetes Mellitus 5 (8%) 8 (9%) 4 (6,6%) 4 (4,3%) to fingers) Pre-existing neuropathy 9 (15%) 13 (14%) Surgical management of traumatic crush With continuous catheter 52 (85%) 78 (83%) 3 (4,9%) 4 (4,3%) or saw injury Single shot bolus 9 (15%) 16 (17%) Tendon/ligament reinsertion (distal Values are given as the number and percentage or mean ± standard biceps tendon, lateral ulnar collateral 1 (1,6%) 3 (3,2%) deviation. RA : retroclavicular approach, CA : coracoid approach, BMI: ligament) body mass index, ASA : American society of Anesthesiologists. Resection of elbow tumor (bone and/or 0 2 (2,1%) soft-tissue) 7 patients did not have a single follow-up surgical Radial and/or ulnar osteotomy 1 (1,6%) 0 Wound debridement 0 1 (1,1%) consultation. Arteriovenous fistula surgery 1 (1,6%) 0 Demographic patient data, preoperative characteristics and continuous catheter versus single shot Values are given as the number and percentage. RA : retroclavicular approach, CA : coracoid approach. bolus of local anesthetic are shown in Table 1. There were no significant differences in age, sex, BMI, ASA score or comorbidities related to peripheral Primary and secondary outcomes are listed neuropathy. in Table 3. New onset and persistent neurological The different types of surgery which were symptoms found after reviewing the data were all the indication for the infraclavicular block are associated with ulnar neuropathy, excluding pre- figured in Table 2. These were primary procedures existing neuropathy or neuropathy due to amputation. or revisions. Sometimes these procedures were Looking at the primary outcome, one patient had combined with a concomitant second procedure on persistent neurological symptoms after 1 year a different anatomical location. possibly associated with the retroclavicular block

Table 3 Primary and secondary outcomes

RA (n = 61) CA (n = 94) p-value Persistent neurological symptoms associated with 1 (1,6%) 2 (2,1%) p = 0,703 block Short term complications New onset neuropathy 2 (3,3%) 2 (2,1%) p = 0,939 Vascular puncture or hematoma 0 2 (2,1%) p = 0,676 Horner’s syndrome 1 (1,6%) 1 (1,1%) p = 0,676 LAST 1 (1,6) 0 p = 0,833 Pneumothorax 0 0 - Infection at injection site 0 0 - Combination with GA (planned) 47 (77%) 73 (78%) LRA only 14 (23%) 21 (22% Successful block 11 (79% of LRA only) 19 (90% of LRA only) p = 0,369 Failed block (conversion to GA or rescue LRA) 3 (21% of LRA only) 2 (10% of LRA only) Cumulative 24-hour morphine consumption 41,4 ± 33,8 39,3 ± 34,1 p = 0,794 with catheter 45,4 ± 34,8 44,3 ± 34,7 p = 0,990 single shot 18,5 ± 13,1 14,8 ± 14,3 p = 0,530

Values are given as the number and percentage or mean ± standard deviation. Cumulative morphine consumption is expressed in morphine milligram equivalents. RA: retroclavicular approach, CA : coracoid approach, LAST : local anesthetic systemic toxicity, GA : general anesthesia, LRA : locoregional anesthesia.

21-Beaumond.indd 162 12/01/2021 12:23 a retrospective analysis of safety and efficacy for distal upper limb surgery 163 (1,6%) compared to two patients in the coracoid group (2,1%), which is not statistically significant to exclude the possibility that the difference is due to random sampling (p-value 0,703). Two cases of new onset ulnar neuropathy and one of Horner’s syndrome were found in both groups, two vascular punctures occured in the coracoid group, and one possible local anesthetic toxicity was observed in the retroclavicular group, none of which showed a significant statistical difference. The vast majority of patients underwent a combined general anesthetic with the use of the Fig. 1. — Flowchart of neurological complications. EMG : infraclavicular block, which makes it difficult to electromyography. assess block success. There was no statistical significance of block success (p = 0,369) in the small In the retroclavicular group, the two cases of sample of patients who underwent the surgical new onset ulnar neuropathy remained persistent. procedure solely under locoregional anesthetic, One developed persistent neuropathy in the coracoid counting 79% and 90% for the retroclavicular and group, and one was unfortunately lost to follow- coracoid approach respectively. up after her first postoperative surgical visit. A Cumulative morphine consumption, expressed flowchart of the cases with neurological symptoms in morphine milligram equivalents, showed similar is summarized in figure 1. results in both groups, with a marked decrease in the The ulnar nerve is a terminal branch of the group who only received a single shot bolus of local medial cord of the brachial plexus. In both the anesthetic. P-values were 0,794 ; 0,990 and 0,530 coracoid or the retroclavicular approach, the for the total, catheter group and single shot group medial cord is lying opposite to the needle entry respectively. point, making it improbable to be injured during No satisfaction were noted or found in any of an infraclavicular block. Furthermore, when depo- medical records. siting an U-shaped injection of local anesthetic around the axillary artery, the needle usually does Discussion not come near the vicinity of the medial cord. It thus seems highly unlikely that the cases of There was a relatively high rate of persistent ulnar neuropathies were due to the locoregional neurological symptoms following surgery, lasting anesthetic, but rather due to trauma or surgery. at least more than 6 months, excluding pre-existing All of these cases were following either neuropathy or neuropathy due to amputation. These elbow arthroplasty, osteosynthesis or arthrolysis. were all patients who developed neuropathy of the Ulnar nerve damage can be acute injury-related, ulnar nerve, ranging from numbness with or without iatrogenic during operative treatment, or present paresthesia for the majority, to weakness with or itself in a delayed manner postoperatively, stressing without muscle atrophy in a minority of cases. the importance of a careful clinical evaluation of Four cases in the retroclavicular and three the peripheral nerve function prior to any operation cases in the infraclavicular, totaling 7 patients or locoregional anesthesia. Ulnar neuropathy has of the 155 (4,5%), developed persistent ulnar been cited as a frequent complication ranging neuropathy due to elbow surgery, which were all from 3% to 12% of internal fixations and total confirmed with electromyography. On top of this, elbow replacements. Perioperative manipulation one patient had persistent ulnar neuropathy in of the ulnar nerve during elbow surgery is almost the retroclavicular group and two patients in the inevitable, sometimes requiring decompression, coracoid group, all of which developed after the isolation and/or transposition of the nerve. Surgery- surgery date. Unfortunately these three patients did induced lesions are often caused by traction on the not undergo an electromyography in the following nerve, direct damage or compression from a retractor, months, whereby this persistent neuropathy could or indirectly due to a prolonged high-pressure not be excluded as a potential consequence of the pneumatic tourniquet time. Late presentation of block. neuropathy is possible secondary to postoperative swelling or scarring, and the nerve may also be

21-Beaumond.indd 163 12/01/2021 12:23 164 a. beaumond & s. coppens injured due to migration of fixation material or the using this approach might be more hazardous seeing forming of osteophytes and heterotopic ossification as the collarbone impedes adequate compression. (15-18). There was one case of possible local anesthetic Even though 3 patients with persistent ulnar systemic toxicity in the retroclavicular group. This neuropathy did not undergo an electromyography, did not occur during the placement of the block, it can be assumed these were not a consequence of but 48 hours later. The patient described nausea the block but rather due to surgery or trauma. In and dizziness, combined with numbness and particular, no records of suprascapular nerve injury paresthesia on the chin. The patient was a male were found in any of the 61 patients who received individual weighing 81 kilograms and measuring the retroclavicular approach. 186 centimeters. He had a continuous infusion of International reported incidence of postopera- ropivacaine 0.2% at a rate of 5 milliliters per hour tive neurologic adverse events after peripheral combined with a 5 milliliters bolus with a lockout nerve blockade ranges from 0,18% to 3% for of 60 minutes. He received a total of 450 milligrams symptoms lasting longer than 5 days, and 0.04% to over the course of 48 hours. The continuous catheter 0.09% for symptoms lasting longer than 6 months infusion was stopped, and his symptoms resolved (19-21). We should therefore be cautious to draw without further intervention after a few hours. A lot conclusions in this relatively small patient popu- of factors could have contributed to his symptoms lation. Suprascapular nerve injury might have a and it is not clear if this was an actual case of local more subtle clinical presentation, and it might not anesthetic toxicity, but fortunately he recovered be recognized by the attending physician if it is not without any serious symptoms. The infraclavicular evaluated specifically. In any case, we found no plexus block offers reliable placement of an major increased risk of nerve injury as is preconized indwelling catheter for postoperative analgesia with by the cadaver study of Sanchetti et al. (12). dislodgement rates that seem similar using either There were two cases (2,1%) of vascular approaches, with the advantage of a technically puncture in the coracoid group. In the first case easier placement using the retroclavicular approach. blood was aspirated on the first pass, but had an (6) But due to the positioning of the catheter around uncomplicated block after a short period of com- the axillary artery, the rate of systemic absorption pression. The second case had an accidental arterial of local anesthetic might be higher, theoretically placement of the catheter, which was consequently putting patients at risk of toxicity. However, removed, and developed a large hematoma des- reported numbers on systemic toxicity following pite compressive therapy. Even though there an infraclavicular block (0.2%) are not higher than is no statistical significance compared with the with other peripheral nerve blocks (0.7%) (1). retroclavicular group, a possible explanation might The rate of other short term complications was be the reduced visualization of the needle tip, low, in line with literature on the infraclavicular making it theoretically more a risk of piercing block. One case of ipsilateral Horner’s syndrome structures. It has been proven that the retroclavicular was found in both groups. The percentage of cases approach offers a significantly better visualization following infraclavicular nerve blockade is much of needle tip and shaft (9,10). Although this does lower than with blocks above the collarbone. It is not auto-matically translate to fewer complications, benign and the effect lasts for the duration of the guiding the needle and tip under direct visualization blockade. But since local anesthetic can travel along allows the anesthetist to see and navigate away the brachial plexus sheath up to the stellate ganglion, from sensitive structures and monitor the spread of there is a possibility of phrenic nerve blockade as local anesthetic, which most consistently excludes well using the infraclavicular block, which has more intravascular injection (22, 23). vital consequences. While being rare, it has been Another explanation could be the location of reported and should always be kept in mind using the cephalic vein and the acromial branch of the this block. There were no cases of pneumothorax or thoraco-acromial artery in the needle path of the injection site infection in both groups (1, 24). coracoid approach as opposed to the retroclavicular Retrospective analysis of block success was approach, which could mean an increased risk of made difficult, since a vast majority of the cases vascular puncture with the classic approach (6, 7). received a general anesthetic as well. This was due However it must be noted that the blind spot in the to various reasons : prone positioning, patient’s and/ retroclavicular approach does not allow appreciation or surgeon’s preference, prolonged operation time of any vascular structures along the trajectory of the or concomitant surgery on a different anatomical block below the clavicle, and a vascular puncture location (bone marrow graft, osteosynthesis on

21-Beaumond.indd 164 12/01/2021 12:23 a retrospective analysis of safety and efficacy for distal upper limb surgery 165 a second limb,...). The remaining population who accounted for as well since some procedures were underwent surgery without general anesthesia performed without regional anesthesia, with patients was unfortunately very small, counting 14 and 21 receiving a rescue block only postoperatively. There patients and a success rate of 79% and 90% for the was no standard postoperative protocol, and some retroclavicular and coracoid groups respectively. patients received a systematic dose of tramadol This difference was not statistically significant (p = at certain hours independent of their pain scores. 0,369) with a trend towards the coracoid approach, Sometimes the catheter was removed preemptively but no conclusions can be made due to the limited to assess the risk of compartment syndrome at the sample. Possible reasons for failure of the block request of the surgical team. In both groups, there were emergency procedures, an abridged wait was a marked lower consumption in the single between block and surgery, and a language barrier. shot bolus population compared to the patients Delayed anesthesia of the musculocutaneous who received a continuous catheter. Cumulative nerve using the retroclavicular approach has been morphine milligram equivalents were also similar reported in other studies. The needle passes further in both subgroups. These results seem contradictory, away from the lateral cord in this technique compared however patients receiving a catheter were also the to the coracoid approach, where the lateral cord lies patients getting extensive and painful procedures. in the needle path. This lowers the risk of injury to No satisfaction were noted or found in any of this cord, but is also the probable explanation to the medical records. This underscores the need for more delay in onset of anesthesia of the musculocutaneous rigorous follow-up. Regular audits or more specific nerve, which usually branches early from the lateral post-anesthesia investigation seems in order. This cord. Depositing some local anesthetic at the level could also increase the involvement of anesthesia in of the lateral cord could improve the time needed to perioperative care. block this nerve (6, 9, 25, 26). There are of course serious limitations to this As with every intervention, it is important to study. The population sample is small, especially know its limitations, and consequently select only when looking for rare outcomes of complications patients who could benefit from the block. Patients after regional anesthesia. As mentioned before, with a high body mass index might prove a chal- the used records were not designed for this study, lenge due to a deep localization of the cords with resulting in an important heterogeneity between increasingly difficult needle visualization using the procedures and perioperative care. As with any coracoid approach. Unfortunately, these patients retrospective study, some data might be missing often also have an anatomically challenging supra- or wrongfully documented. One of the exclusion clavicular fossa that might limit the possibility of criteria was loss to follow-up without a first the retroclavicular approach. Patients with a short surgical visit, but this resulted in patients who were neck, deformed or highly angulated clavicle, a full lost to follow-up after this visit being included. supraclavicular fossa or thin chest wall are less well Unfortunately not every patient was seen after 1 suited for this approach, as it might prove a challenge year, potentially meaning an underestimation of the to pass the needle underneath the clavicle. Similar rate of complications. This study also shows the to a subclavian catheter insertion, adducting the arm need for better record keeping and more rigorous might offer some help in increasing visualization follow-up. This cannot be emphasized enough! and passing below the collarbone. This makes the While the retroclavicular approach has proven block very suitable for emergency procedures or to offer better needle visibility, we should be cautious pain relief in patients with an injured limb (3, 8, 25, about having a false sense of safety. Despite the fact 27). that it has become the universal standard of care for Cumulative opioid consumption during the first peripheral nerve blockade, ultrasound guidance has 24 hours starting from surgery was calculated using not been able to prove a significant effect on the cumulative morphine milligram equivalents (MME). incidence of postoperative neurologic symptoms. Results in both groups were similar, ranging from Ultrasound is only a tool, and it should be used 41,4 ± 33,8 MME for the retroclavicular and 39,3 with full knowledge of its limitations. Standard ± 34,1 MME for the coracoid group, which showed monitoring, with use of neurostimulation and no significant difference (p = 0,794). However, ultrasound guidance should always be combined seeing as this was a retrospective study, there was while performing peripheral nerve blocks. (28) an important heterogeneity in type of surgery and postoperative care, resulting in large standard deviations. Intraoperative usage of sufentanil was

21-Beaumond.indd 165 12/01/2021 12:23 166 a. beaumond & s. coppens

Conclusion 6. Sutton E.M., Bullock W.M. and Gadsden J. 2015. The retroclavicular brachial plexus block: additional advantages. Reg Anesth Pain Med. 40: 733-734. The infraclavicular plexus block offers several 7. Soeding P. and Eizenberg N. 2009. Review Article: advantages compared to other upper limb alter- Anatomical Considerations for Ultrasound Guidance for natives, notably a single injection block and reliable Regional Anesthesia of the Neck and Upper Limb. Can J Anaesth. 56: 518-533. catheter placement. The retroclavicular approach to 8. Charbonneau J, Fréchette Y, Sansoucy Y and Echave P. this block offers a useful alternative to maintain the 2015. The Ultrasound-Guided Retroclavicular Block: A ultrasound probe parallel to the needle, resulting in Prospective Feasibility Study. Reg Anesth Pain Med. 40: 605-609. better visibility of shaft and tip. 9. Ozturk N.K. and Kavakli A.S. 2017. Comparison of the The downside of this approach is the acoustic coracoid and retroclavicular approaches for ultrasound- window when passing beneath the collarbone and guided infraclavicular brachial plexus block. J Anesth. 31: the possible location of the suprascapular nerve in 572-578. 10. Blanco A.F.G., Laferrière-Langlois P., Jessop D., D’Aragon the needle path. F., Sansoucy Y. and Albert N. et al. 2019. Retroclavicular Neurological and other characteristic compli- vs Infraclavicular block for brachial plexus anesthesia: a cations of locoregional anesthesia of the upper limb, multi-centric randomized trial. BMC Anesthesiol. 19: 193. 11. Grape S., Pawa A., Weber E., and Albrecht E. 2019. block success and postoperative analgesia were Retroclavicular vs supraclavicular brachial plexus block for similar using both approaches. This retrospective distal upper limb surgery: a randomised, controlled, single- study unfortunately does not have the power blinded trial. Br J Anaesth. 122: 518-524. 12. Sancheti S.F., Uppal V., Sandeski R., Kwofie M.K. and to validate the safety and the efficacy of the Szerb J.J. 2018. A cadaver study investigating structures retroclavicular approach to the infraclavicular block encountered by the needle during a retroclavicular approach compared to the classic coracoid approach. We to infraclavicular brachial plexus block. Reg Anesth Pain could however not find one case of suprascapular Med. 43: 752-755. 13. Boykin R.E., Friedman D.J., Higgins L.D. and Warner J.J. nerve injury in the 61 patients who received an 2010. Suprascapular neuropathy. J Bone Joint Surg Am. 92: infraclavicular plexus block using the retroclavicular 2348-2364 approach. Large, randomized multicentric studies 14. Leurcharusmee P., Layera S., Finlayson R.J. and Tran D.Q., 2019. Retroclavicular approach to infraclavicular are needed to further analyze the safety and efficacy brachial plexus block: a logical conundrum. Reg Anesth of this approach. The University Hospitals of Pain Med. rapm-2018-100133. Leuven is a teaching center, where most blocks 15. Gschwend N., Simmen B.R. and Matejovsky Z. 1996. Late complications in elbow arthroplasty. J Shoulder Elbow are performed by anesthetists in training. Both Surg. 5: 86-96. approaches offer an easy and feasible option with 16. Voloshin I., Schippert D.W., Kakar S., Kaye E.K. a satisfying learning curve. They are useful skills to and Morrey B.F. 2011. Complications of total elbow have in the repertory of the locoregional anesthetist, replacement: a systematic review. J Shoulder Elbow Surg. 20: 158-168. presuming an appropriate selection of patients is 17. Obert L., Ferrier M., Jacquot A., Mansat P., Sirveaux made. F. and Clavert P. et al. 2013. Distal humerus fractures in patients over 65: complications. Orthop Traumatol Surg Res. 99: 909-913. 18. Shin R. and Ring D. 2007. The ulnar nerve in elbow References trauma. J Bone Joint Surg Am. 89: 1108-1116. 19. Sites B.D., Taenzer A.H., Herrick M.D., Gilloon C., 1. Chin K.J., Alakkad H., Adhikary S.D. and Singh M. Antonakakis J., Richins J. and Beach M.L. 2012. Incidence 2013. Infraclavicular brachial plexus block for regional of local anesthetic systemic toxicity and postoperative anaesthesia of the lower arm. Cochrane Database Syst Rev. neurologic symptoms associated with 12,668 ultrasound- 8: CD005487. guided nerve blocks: an analysis from a prospective clinical 2. Hadžić A., Atchabahian A., Vandepitte C. and Lopez registry. Reg Anesth Pain Med. 37: 478-482. A.M. 2017. Ultrasound-Guided Infraclavicular Brachial 20. Brull R., McCartney C.J., Chan V.W. and El-Beheiry H. Plexus Block In: Textbook of regional anesthesia and acute 2007. Neurological complications after regional anesthesia: pain management, 2nd edition. p574-579. McGraw-Hill contemporary estimates of risk. Anesth Analg. 104: 965- Medical. 974. 3. Luftig J., Mantuani D., Herring A.A. and Nagdev A. 2017. 21. Barrington M.J., Watts S.A., Gledhill S.R., Thomas R.D., Ultrasound-guided retroclavicular approach infraclavicular Said S.A. and Snyder G.L., et al. 2009. Preliminary results brachial plexus block for upper extremity emergency of the Australasian Regional Anaesthesia Collaboration: a procedures. Am J Emerg Med. 35: 773-777. prospective audit of more than 7000 peripheral nerve and 4. Sites B.D., Brull R., Chan V.W., Spence B.C., Gallagher plexus blocks for neurologic and other complications. Reg J., Beach, M.L., et al. 2007. Artifacts and pitfall errors Anesth Pain Med. 34: 534-541. associated with ultrasound-guided regional anesthesia: Part 22. Sites B.D. and Brull R. 2006. Ultrasound guidance in II: a pictorial approach to understanding and avoidance. peripheral regional anesthesia: philosophy, evidence-based Reg Anesth Pain Med. 32: 419-433. medicine, and techniques. Curr Opin Anaesthesiol. 19: 630- 5. Hebbard P. and Royse C. 2007. Ultrasound guided posterior 639. approach to the infraclavicular brachial plexus. Anaesthesia. 23. Brull R., Perlas A., Cheng P. and Chan V. 2008. Minimizing 62:539. the risk of intravascular injection during ultrasound-guided

21-Beaumond.indd 166 12/01/2021 12:23 a retrospective analysis of safety and efficacy for distal upper limb surgery 167 peripheral nerve blockade. Anesthesiology. 109: 1142- musculocutaneous nerve from the brachial plexus: 1145. implications for infraclavicular nerve blocks. Anesth 24. Gentili M.E., Deleuze A., Estèbe J.P., Lebourg M. Analg. 114: 1131-1133. and Ecoffey C. 2002. Severe respiratory failure after 27. Beh Z.Y., Hasan M.S. and Lai H.Y. 2016. Ultrasound- infraclavicular block with 0.75% ropivacaine: a case report. guided retroclavicular block (aka posterior approach J Clin Anesth. 14: 459-461. infraclavicular block): anatomical variation of the clavicle 25. Uppal V., Kalagara H.K. and Sondekoppam R.V. 2018. limits block feasibility. Reg Anesth Pain Med. 41: 658-659. Tips and tricks to improve the safety of the retroclavicular 28. Neal J.M. 2016. Ultrasound-guided regional anesthesia brachial plexus block. Am J Emerg Med. 36: 1107-1108. and patient safety: update of an evidence-based analysis. 26. Pianezza A., y Nedeo A.S., Chaynes P., Bickler P.E. Reg Anesth Pain Med. 41: 195-204. and Minville V. 2012. The emergence level of the

21-Beaumond.indd 167 12/01/2021 12:23

(Acta Anaesth. Belg., 2020, 71, 169-178) Measurement of ACT via I-STAT after administration of ad-random Heparin

B. Vanwijnsberghe, H. De Ruyter, F. De Somer

Abstract : Background : During vascular surgical Introduction procedures a fixed dose of heparin is usually given. This fixed dose is rarely based on significant evidence to reach Problem Statement a target ACT. Literature has shown that these heparin doses are often inaccurate, leading to extra dosing, Vascular surgery is one of the fields of surgery excessive testing or even major complications to reach which has advanced progressively over the last few this target. Physical characteristic-based dosing is rarely years. This is thanks to the advances in different used in vascular surgery. Objectives/design : Using a retrospective observational medical branches, especially laboratory fields. study our goal was to determine whether we could get Laboratory testing is now possible point-of-care a weight-, BSA- or EBV-based dose of heparin to have (POC) and this has made vascular surgery and a narrow ACT target and whether the impact of Heparin anesthesia much safer and more reliable. It has made Resistance is important in this matter. Another goal was battling hypercoagulable states convenient and has to see if basic laboratory data like white blood cell count helped in the form of a prophylaxis medication for and platelet count would influence this ACT and in turn, clot prevention. our required dose of heparin. One of these advances is the POC analyzer for Methods : Data was collected from pre-operative labo- Activated Clotting Time (ACT). At our institution, ratory measurements and anesthesiology charts. Using the Department of Anesthesiology, Ghent Univer- Multivariate Linear Regression, we tried to determine the sity, the Abbott I-STAT® ACT is used. dosing, based on weight, BSA and EBV. The standard When looking up heparin dosage in literature, deviation of all HSI can give us an idea of its effect on ACT values in different people and the possibility of a fixed dose of 5000 IU is usually given atthe using it in clinical practice if the spread is narrow. beginning of peripheral vascular surgery. Some Setting : Monocentric study at Ghent University Hospital institutions tend to use a dose per weight unit of 50- Results : Data was collected from 450 patients, of which 75 IU kg-1, but this is exceptional. We could say that 115 patients met inclusion criteria. In 8/115 patients, Heparin dose is at random. ACT was lower than 200 s and in 16/115 patients ACT During peripheral vascular surgery procedures was higher than 250 s, possibly inducing complications. (arterial and venous dilatations, endarterectomies, In all cases a semi-fixed dose was used, usually 5000 IU, endoprostheses placement) we also tend to use adjusted according to weight to 2500 or 7500 IU. There a (semi-)fixed dose of heparin, usually 5000 IU. was no significant correlation between weight, BSA or A change in dose to 7500 IU or 2500 IU is only EBV and ACT. HSI in all variables demonstrated a wide considered if the patient’s weight suggests we SD (resp. 33.4%, 32.7% and 33.5%). Furthermore, there change doses, but this is based on gut feeling was no obvious effect from WBC or platelets on the ACT target. and has no scientific background. Literature (and Conclusion : Although in this study multiple cases with clinical experience) has shown these doses are often ACT under- or overshooting could be demonstrated after semi-fixed heparin dosing, these results show that weight-based, BSA-based or EBV-based heparin dosing B. Vanwijnsberghe, MD ; H. De Ruyter, MD ; F. De Somer, is not recommended. Data was also insufficient to show MD a direct response from WBC or platelet count correction. Department of Anesthesia, Ghent University Hospital, 9000 Gent, Belgium. Corresponding author : Dr. Bram Vanwijnsberghe, Ghent Keywords : Anesthesiology ; Heparin ; general sur- University Hospital, Corneel Heymanslaan 10 - 9000 Gent, gery ; anticoagulants. Belgium. E-mail : [email protected]

Paper submitted on Jun 24, 2020 and accepted on July 1, 2020 Conflict of interest : None

22-Vanwijnsberghe.indd 169 12/01/2021 12:24 170 b. vanwijnsberghe et al. too high or too low, which can lead to insufficient Other studies have shown Estimated Blood coagulation, bleeding or other major complications Volume (EBV) to be even better correlated with (1-5). Even in patients with similar physiological anticoagulation tests. One study demonstrated that characteristics, a widespread difference in ACT- dosing heparin based on EBV vs based on weight, values can be seen. reduced the use of heparin and bleeding events. We This difference in heparin response is called calculated the Lean Body Mass (LBM) using the Heparin Sensitivity (Index) (HSI) or Heparin Janmahasatian equation, since this is the active mass Resistance (HR). in heparin pharmacokinetics. We used this LBM to Although in the late 80’s there was doubt calculate the EBV using the Nadler Equation (9-11). about the use of heparin during vascular procedures, Because different brands of ACT measurement studies have shown that microthrombi can be devices were available at our institution, and these present and lead to early graft failure if no heparin is different devices all have their own ACT-values, used (1-6). Temporary clamping of arterial vessels unification to one device was imminent. This would can cause significant thrombus formation. reduce differences in ACT measurements across Because measurement of ACT-levels was not same blood samples, as these differences in values available in the 80’s, and laboratory aPTT values can lead to different treatment and consequently are difficult to get fast enough, in most cases there possible major complications. was no knowledge about the coagulation status of To determine which clinical and laboratory the patient (1, 5-7). tests were needed, literature was consulted. Anti- Most of the heparin protocols that were thrombin/platelet and fibrinogen levels could give found in literature were directly extrapolated from useful information. Other studies determining clini- interventional cardiology procedures without ever cal factors that would influence ACT, also showed doing any RCT studies. When reasoning, extra- platelet count to be influencing. Since our standard polation cannot be done without further action, as laboratory does not include antithrombin and both vessels are of a different caliber and many fibrinogen in every patient (due to high cost) and used wires/catheters have a different diameter. this is a retrospective study, only platelet count can Some institutions also aim for higher ACT be taken to analyze (3). values (> 300 s) (8). Still, these protocols were Other studies showed infection (and especial- adopted into peripheral vascular surgery by many ly white blood cell count) had an impact on anti- institutions and studies. Recent studies however coagulation, so white blood cell count could in- tried to quantify the best ACT-level for vascular fluence the heparin dose requirements. interventions, mostly based on the occurrence of Since our pre-op laboratories are very limited complications with high ACT values. They related and first results were not satisfying, we quickly bleeding complications outside of the depicted cut- determined that extra tests on other lab results off values. Main conclusions from this large registry would be useless. were that a higher total heparin dose (>60 IU kg-1) In no studies age or gender had any influence and a peak procedural ACT of >250 s were strong on ACT, so this was not used in our study. predictors of significantly increased postprocedural bleeding events. The technical and procedural Objectives success was high and did not differ between the described groups with higher or lower heparin dose As little is known about quantifying heparin or peak ACT. Deduced from these results, it was dose during peripheral vascular procedures, and it is strongly suggested that, during Peripheral Arterial still debated whether weight-based heparin dosage Interventions (PAI), a body-weight-dependent or fixed dosage is the way to go, it was our goal to dose of up to 60 IU kg-1 should be administered, get a retrospective look on the semi-fixed heparin while the ACT should have a target peak value of doses and concomitant ACT-values in peripheral < 250 s. An ACT level of 200-250 s should give vascular surgery. Moreover, we studied whether enough anticoagulation to diminish the risk for heparin resistance has a large enough impact to microthrombi during these peripheral procedures. abolish weight, BSA or EBV as a significant factor. In these studies, the relationship between weight If not, we can raise the question if it’s better to use and ACT is not studied but it is suggested that a weight, BSA or EBV-based dosing in clinical setting dose of > 60 IU kg-1 would give an ACT of > 250 s. and potential prospective study? Another question In the results a wide spread of ACT values can be we can raise is whether pre-anesthesia clinical seen, in both low and high dose heparin (2, 5). and laboratory data have a significant influence on

22-Vanwijnsberghe.indd 170 12/01/2021 12:24 measurement of act via i-stat after administration of ad-random heparin 171 patients’ sensitivity to heparin, so we can use this to get an even more accurate heparin dose? As stated before, which data to collect, was chosen after consulting literature. We tend to think Rationale that weight and length (BSA) influence heparin dose requirements and wanted to show this correlation ACT measurement has only recently started so we could more accurately dose our heparin for being used in vascular surgery to guide the heparin desired ACT-values. In some studies EBV seemed doses and adjust if needed. Although accurate to be better correlated. Since LBM is the active measurements can be done with new devices like distributive mass in heparin pharmacokinetics, EBV the Abbott I-STAT®, we still use a semi-fixed dosage was calculated from this weight. of heparin, without looking at any aspect of the Janmahasatian equation was used to determine patient’s characteristics, whether it be laboratory Lean Body Mass. Nadler’s equation was used to or clinical values. It is semi-fixed, since we do use calculate Estimated Blood Volume : higher doses if patients have a higher than normal weight (and lower dose in low weight). This is Janmahasatian equation : (9270*Weight) however based on gut feeling and no cut-off weight Men : LBM (kg) = was ever determined. With this retrospective study, Weight (6680+(216* ( ) we would like to see whether we can get a dose, ( Height ) 2 in which we can accurately predict the ACT-value, 100 making peripheral vascular surgery at our institute (9270*Weight) safer, as fewer complications would arise, and in Women : LBM (kg) = turn, fewer cost for the patient both emotionally (in Weight (6780+(244* ( ) severe complications) and economically. This all ( Height )2 with little extra effort. Accurate prediction will also 100 lead to fewer blood sampling, fewer use of I-STAT® cartridges, which is economically interesting for the Nadler’s equation : hospital. It would also be less time-consuming, a Men : EBV(L)=(0.3669 × Height^3 ) + (0.03219 × factor often forgotten, since only certain OR nurses Weight) + 0.6041 ® can/are allowed to operate these I-STAT devices, Women : EBV(L)= (0.3561 × Height^3) + (0.03308 many of whom have different OR jobs and are × weight) + 0.1833 usually not immediately available. But above all, it leads to safer anesthesia and surgery. For measuring HSI, the following formulas were Experimental Approach/ Methods used : (ACT (3min) - ACT (start) HSI (Weight)= Data was collected by retrospectively looking ( Heparin Bolus Dose ) weight at cases in vascular surgery at the Ghent University or st (ACT (3min) - ACT (start) Hospital over a period of 9 months from 1 of HSI (BSA)= st ( Heparin Bolus Dose ) November 2017 until 31 of July 2018. Length and BSA weight at the time of operation was taken from the or (ACT (3min) - ACT (start) pre-operative questionnaires each patient is given HSI (EBV) ( Heparin Bolus Dose ) prior to surgery. Platelet counts and white blood cell EBV counts were gathered from pre-operative laboratory testing the night before surgery. ACT measurements Platelet count and white blood cell count and heparin doses were taken from anesthesia were gathered to see if this would influence dose charts. requirements, as studies have shown that infection Inclusion criteria were wide, since it was a can influence coagulation and as stated before, retrospective study and we wanted a complete platelets can influence Heparin Resistance. image of how we were doing in using heparin in We also wanted to use AT III, since a lot of our patients. discussion is going on about the effect of AT III in Patients should be 18 years or older, getting low dose heparin, but unfortunately, this is not in peripheral vascular surgery with general anesthesia our pre-operative laboratory assessment. On the and the use of heparin, where ACT was measured at other hand, if we could show that weight-based different times and targeted at 200-250 s. heparin is enough to determine how much heparin

22-Vanwijnsberghe.indd 171 12/01/2021 12:24 172 b. vanwijnsberghe et al. should be given, we could possibly derive that ATIII indeed does not influence HSI significantly. In total, there were 445 cases of Peripheral Vascular Surgery meeting our inclusion criteria (every case in which the aim for ACT was 200-250 s). In 20 cases, there was no anesthesia file scanned. In 75 cases, clinical data (weight, height, Platelets or WBC) was missing. From these 350 patients, 210 patients were excluded because baseline ACT or ACT after first heparin dose was not measured. Of the data of the remaining 140 patients, 25 were thrown out because it was data from a patient who Fig. 1. — Numbers of patients treated with respective heparin already was included (undergoing 2nd surgery). dose of 7500 IU (column 1) – 5000 IU (column 2) – 3250 IU Eventually 115 patients were included in (column 3) – 2500 IU (column 4). A summary column was the study. In only 65 of these, ACT was again added to the right. Green represents ACT values in the required range. Orange represents too low ACT. Red, too high ACT. measured after 60 minutes or at the end of surgery. Approval of the Ehtics Committee was obtained on 3/1/2019, Commissie voor Medische Ethiek – Ghent University,Chairman : Prof. Dr. D. Matthys, differences, where gut feeling was the biggest BE Registration Nr : B670201838394, National decision maker and no weight cut-off was chosen. Registry : FAGG – Research and Development An overview can be seen in figure 1. In 8 patients (7%), the first heparin dose Flow chart for data collection was insufficient and extra heparin had to be given See Flowchart 1 in Figures (usually at the 8-10-minute marker after first dose). In most cases, surgery is well underway by then and catheters are usually already in place. In 16 of 115 (13.9%) patients, the ACT value was higher than 250. 12 of 16 patients with an ACT higher than 250 (75%) were given 7500 IU. If we use the guidelines of 60 IU kg-1, a person should only receive 7500 IU if he weighs over 125 kg (107 kg if 70 IU kg-1 is used). There was no patient included with this weight, or a higher weight, in our study (7 patients weighed more than 107 kg). 18 other patients, weighing less than 107 kg, received 7500 IU and were still in the safe range. There was only one person weighing more than 107 kg who received 7500 IU and was in the safe range. In 5/7 cases with patients weighing more than 107 kg, 5000 IU was given, again showing the randomness of higher heparin dosage. Only in one instance of these 5, extra heparin was needed to reach ACT >200 s. One patient weighing 108 kg, received 7500 IU and still needed 2500 IU extra to reach an ACT of > 200 s. Another patient weighing only 86 kg, needed another 2500 IU after 7500 IU. Results Figure 1 shows the different doses of Heparin and the respective ACT outcome. It is obvious that In 80 patients, heparin dose was 5000 IU. In 32 most patients received 5000 IU of heparin, although instances, heparin dose was 7500 IU. In 2 instances many patients, even though not overweight, were it was 2500 IU and in one case 3750 IU was given. given 7500 IU and were still in the “safe” range of Doses were mostly chosen because of weight ACT.

22-Vanwijnsberghe.indd 172 12/01/2021 12:24 measurement of act via i-stat after administration of ad-random heparin 173

Fig. 3. — Baseline ACT of 115 patients. b

not let us see which medication was given at the time of or prior to surgery. Since we use a semi-fixed dosage of 5000 to 7500 IU of Heparin, the X-axis on figure 2a, b and c shows a wide spread of heparin IU kg-1, which is contributory to our study. Still, the difference in ACT values is very widespread, and no significant link between heparin dose and weight can be drawn. A practice that is used in cardiac surgery, where the heparin dose is determined by the weight of the -1 c patient (300-400 IU kg ). The results were gathered using linear regression analysis. Because BSA might be more related to effective heparin dosage, this was also calculated and set out in a model in figure 2b. Again, there is a wide spread in heparin dosage per m² of BSA and a widespread in subsequent ACT values. We can perform the same analysis on ACT values based on Estimated Blood Volume (EBV), since studies have shown better correlation and less bleeding events, seen in figure 2c. Fig. 2. — ACT values of patients at start (Baseline ACT) and All these models look fairly the same and on ACT value 3 minutes after administration of heparin. ACT values are set out against heparin dose per kilogram of weight first sight, no apparent difference in spread can be in a, against BSA in b and against EBV in c. The slope of one seen. To have a better view of this, the slope of each line represents the Heparin Resistance (or HSI) of a patient. A line represents the HSI of each patient. To compare spread of HSI can already be seen. these different methods, we can put the HSI of every case in a numerical graph and look at the numbers Heparin dose per weight unit (kg) was set out of the spread. and is shown in figure 2. This shows a wide spread of HSI, which can It is obvious that there is a wide spread of be seen in figure 4a (per kg), figure 4b (per2 m BSA) baseline ACT, seen on the Y-axis in figure 2a and and figure 4c (per L EBV). Mean values for HSI quantified in figure 3. A Gaussian distribution can based on weight were 1.2058 (SD : +/- 0.4039, be seen in baseline ACT values. With these ACT 33.4% of mean). based on BSA 0.02964 (SD : +/- baselines we should ask ourselves if all patients 0.009715, 32.7% of mean). Based on EBV 0.07623 were heparin naive at the start of surgery, potentially (SD : +/- 0.02555, 33.5% of mean). already at bleeding risk when they receive extra To determine if WBC or platelet count had heparin. Unfortunately, our technical system does any effect on the change in ACT, we looked at the

22-Vanwijnsberghe.indd 173 12/01/2021 12:24 174 b. vanwijnsberghe et al.

Fig. 5. — Correction of HSI over a wide range of WBC count and platelet (Thrombo) count. Based on Heparin doses by weight unit (kg), BSA or EBV in respectively a, b and c. Note that at higher platelet counts, the ACT seems to be rising, but in both high and low WBC and platelet count the deviation is too large to make a statement. WBC values were between 5000 and 10000. An Fig. 4. — HSI count based on Heparin bolus dose per weight increase can be seen from 15000 and higher but unit (kg) in a. Per BSA unit (m²) in b. Per EBV unit (L) in c. too little data was collected to determine if this was a significant change. Platelet count also had change in HSI corrected with the pre-op value of no significant effect on the HSI values. Again, a both WBC and platelet count. prospective study with more patients would have to As this is a retrospective study, we wanted be set up. to see if standard laboratory tests would suffice in In figure 5b, the same WBC count and platelet enhancing our prediction of ACT, without need for count are set out against HSI based on BSA. further, more expensive lab testing. A prospective Again, no significant effect can be seen, and data study will be set up in the future where more lab is insufficient to determine whether high WBC and tests can be done preoperatively to determine other platelet count might have any effect. factors. In figure 5c, WBC and platelet count are set Figure 5a shows that HSI (based on weight) out against ACT based on EBV in L, with again no was not changed by correction with WBC and that significant effect of both variables. data is insufficient to determine whether high WBC Of course, the effect of platelet inhibitors like count can change ACT significantly. Most of the acetylsalicylic acid, or even ADP-R inhibitors like

22-Vanwijnsberghe.indd 174 12/01/2021 12:24 measurement of act via i-stat after administration of ad-random heparin 175 clopidogrel, also has an effect. Our system did not significant correlation between body weight and allow to see which patients were taking these drugs ACT after multivariable regression (OR 0.92 ; 95% at the time of surgery or whether platelet inhibitors 0.87-0.98 ; P = 0.008). Something we also could not were stopped 10 days prior to surgery. reproduce (18). Other studies tried to EBV as a variable Discussion on which heparin could be dosed. Based on the findings of one study which demonstrated rapid These results show that weight, BSA and EBV therapeutic anticoagulation and less bleeding events have little to no predictive value in sensitivity to in continuously administered heparin, we added heparin and thus our goal to get a fixed dose per unit calculation of our doses based on EBV using the (kg, m² or L) was not reached. Nadler equation (6, 11). Again, our study could In 16 of 115 patients the ACT was higher not reproduce better results using EBV as the only than 250 s, which in other studies has shown that factor. Variation in HSI was still too spread out. this significantly increases the risk for bleeding It seems HSI variation in the population is too complications (8, 11). A dosage of 60-70 IU kg-1 as large to let weight, BSA or EBV be the only factor suggested by multiple guidelines to be the absolute in deciding heparin dosage. maximum for peripheral vascular surgery would Based on these numbers, there is no significant mean a dose of 4800 IU for a patient weighing 80 difference between HSI based on weight, HSI based kg. In some instances, we observed doses of up to on BSA or HSI based on EBV. All SD seem to be 80 or 90 IU kg-1 in our data. evenly widespread (resp. 33.4%, 32.7% and 33.5%). From figure 1 we can conclude that the doses One retrospective study from 1998 by Ferguson of heparin administered, are partially random. Only et al. using high dose heparin (130 IU kg-1, target 1/7 patients weighing more than 107kg received >300s) compared the effect of weight and BSA 7500 IU and were in the desired ACT-range. 4/7 in PTCA and demonstrated similar results (33% were also in the desired ACT-range, but only vs 32% spread of SD in resp. Weight and BSA). received 5000 IU. This study also demonstrated slight but significant Other studies have shown that weight or BSA correlation between heparin response and weight (r is poorly correlated with HSI or ACT (12,13). One = 0.37) and heparin response and BSA (r = 0.36). study showed that obese patients often tend to get Still, Ferguson et al. did not recommend weight overdosed if weight is used as a dose value. In our based dosing, as the correlation was only slightly study, 12 of 16 patients with an ACT higher than 250 significant (15). (75%) were given 7500 IU. If we use the guidelines No study showed the spread of HSI when of 60 IU kg-1, a person should only receive 7500 IU if dosing is based on EBV. Our result of 33.5% (SD % he weighs over 125 kg (107 kg if 70 IU kg-1 is used). of mean) suggests no gain in the use of EBV. It should There was no patient included with this, or a higher be considered that EBV is already an estimation and weight in our study (7 patients had a weight more Nadler’s equation has its own limitations and errors than 107 kg). In 5/7 cases with patients weighing which could also influence our results significantly. more than 107 kg, only 5000 IU were given, again Since there was a wide spread of baseline showing the randomness of higher heparin dosage. ACT, we can ask ourselves if all patients were Only in one instance of these 5, extra heparin was heparin naive, this could suggest a lower dose of needed to reach ACT >200 s. heparin is needed to reach a certain ACT level. We One patient weighing 108 kg, received 7500 can also ask ourselves if a higher baseline ACT or IU and still needed 2500 IU to reach an ACT of > if patients already receiving heparin, would have a 200 s. difference in Heparin Response, possibly indicating One study concluded that a fixed dose of 5000 less need for heparin, regardless of the provided IU helped achieve ACT >200 s in almost all patients, target. Literature has already shown that heparin but “most patients were outside recommended half-life lengthens when heparin dose is increased. target interval”, meaning they had an ACT of 250 or ACT peak and half-life after multiple heparin doses higher (14). In our study this was not the case, as a was not evaluated in any study and the effect is 5000 IU dose only gave an ACT of higher than 250s unknown. in 4/115 cases. To provide a more consistent and In 8.6% of patients, an extra dose of heparin predictable heparinization, this study recommended was needed, which means that an extra blood a weight-based bolus dose of heparin of 60-70 IU sample and cartridge was needed for another test. kg-1”. Unlike our study, this study also showed a In most instances however (77%) the first dose of

22-Vanwijnsberghe.indd 175 12/01/2021 12:24 176 b. vanwijnsberghe et al. heparin was enough to reach the target interval like AT III, fibrinogen, CRP/WBC and sedimen- (87% in patients receiving 5000 IU vs 56% in the tation could maybe show a more significant 7500 IU group). effect of these factors. Also, since liver function Unfortunately, our study could not show can have a significant impact on coagulation, a any significant link between weight and BSA for future study could include measurements of AST/ determining the heparin dosage. As expected, ALT and their impact on ACT. It is still unclear Heparin Resistance seems to have a great effect whether liver disorders have an effect on baseline on target ACT reached and is too widespread in coagulation or on heparin sensitivity (22). the population. Other studies (most of them done There are a lot of limitations in this study. in Percutaneous Coronary Interventions (PCI) with Obviously, since this is a retrospective study, there higher dose heparin and target ACT) also showed is a very wide inclusion criteria of patients. A irregular coagulation in most patients after weight- randomized controlled trial would give more insight based dosing, some showing only slight significant in to whom would be included in this study. This correlation (8, 16, 17). study did not make any difference in which vascular In our study, platelet count did not significantly procedure was done, ranging from dilatations to influence the heparin sensitivity. In one study with > placement of endoprostheses. Both have their own 500 patients undergoing Cardiopulmonary Bypass impact on coagulation. Little impact is however (CPB), Heparin Resistance was alleviated by giving expected at the beginning of surgery, where no patients AT III. It could show that platelet count ( intervention/manipulation is done, and target ACT > 300000) had a significant influence on Heparin is 200-250 s. Resistance (18). In our results a higher platelet count Secondly, ACT devices have their own limi- led to a non-significant higher HSI, however, more tations. A comparison was done by the university of patients with high platelet count should be included alberta where they compared i-STAT to Hemochron to have a significant meaning, since only 1 patient devices. The difference in measurements in different in our study had a platelet count > 300000, and only brands and the standard deviation of results of one 2 patients had a platelet count < 100000. Another specific brand can have a significant effect on ACT- study demonstrated that AT III made patients more values, although in clinical setting, these differences susceptible to heparin and one in-vitro study of 2000 are probably not as significant. The coefficient showed that addition of AT III increases sensitivity variation in the I-STAT is only 3.1% when targeting to supplemental heparin as reflected by significantly between 90 and 400s (23, 24). prolonged ACT (19), whereas contradictory studies As stated by Wallach et al., ACT is influenced showed that preoperative AT III-levels had no by numerous factors, including hemodilution, significant influence on HSI (12, 20). This suggest temperature of the blood sample, environmental other more complex factors are at work. Again, a lot temperature, blood sample volume and timing of of the studies are not specific to vascular surgery, blood sample and many others (25). most of them are based on cardiac surgery and Since this is a retrospective study, we cannot patients going on Cardio-Pulmonary Bypass or be certain about the similarity of these factors in our Percutaneous Coronary Interventions with higher patients. We are taught to take a blood sample at doses of heparin and higher ACT values. the three-minute marker, but many factors influence This also means that a future prospective study this : a nurse needs to be available and as this is at with AT III measurement could be interesting to the beginning of surgery, both anesthesiologists and determine its effect on these ACT values. nurses usually have many things on their hands, No other studies have shown the effect of delaying the sampling. In a future prospective study, white blood cell count on ACT values, but since blood sampling could be done at the same time in many studies state that infection has an influence every patient. Since peak effect of heparin on ACT on coagulation status, we wanted to see if we could values is marked at 60-90 s, this could also be correlate this with the WBC count (21). Obviously, changed in a future study, since at the three-minute since at our institution, infectious patients are mark, the effect is already in a downward slope usually postponed from surgery, not a lot of patients and could have already dropped. Since half-life of had high WBC count, so it was difficult to really heparin is 60 to 90 minutes, this would only have determine if this would have any effect on ACT a slight but clinically insignificant influence. In values. future study, when subsequent ACT measurements Again, a prospective study with more patients, and heparin doses are measured, half-life and peak pre-operative (or PERI-operative) laboratory values effect of heparin will have to be considered.

22-Vanwijnsberghe.indd 176 12/01/2021 12:24 measurement of act via i-stat after administration of ad-random heparin 177 It is difficult in a retrospective study to assess approach could reduce periprocedural bleeding and the effect of laboratory outliers, as we tend to should be investigated as a possible alternative. An postpone surgery for these patients to determine the important matter in Anti-Xa monitoring is the long cause of these abnormalities. This makes studying duration of measurement and the cost for analyzing on ACT values and their influencing factors difficult one sample (27, 28). as people with high CRP and WBC values (in When it comes to heparin, aPTT has long infections) are usually postponed. been the standard of use during measurement of And last, surgeons also tend to use heparin anticoagulation. Although literature has shown during vascular surgery procedures, apart from our ACT has a bad correlation with aPTT, it is still intravenous dose of heparin, surgeons use solutions not recommended in intra-operative setting due to of heparin in saline (usually 100 UI ml-1). Studies inability to measure aPTT point-of-care. One study have shown these can significantly influence demonstrated a Pearson correlation of 0.17 when systemic clotting times (studies were done using ACT was correlated with aPTT. aPTT values) and thus, will also have an effect on A retrospective, single center, cohort study measured ACT, although it is difficult to determine by Vandiver et al compared anti-Xa and aPTT how much of localized heparin is taken up sys- monitoring of IV heparin in the treatment of temically (26). pulmonary embolism and deep vein thrombosis. They compared 88 patients where aPTT was used to Conclusions target vs 98 where aXa was used. The study showed that the use of anti-Xa assay-based monitoring Since our study could not demonstrate a protocol resulted in statistically significantly higher significant correlation between weight, BSA or percentage of within-range monitoring values than EBV and HSI, and HSI was too widespread, the aPTT protocol (69% vs 41%). In addition, we cannot recommend weight-, BSA- or EBV- patients in the anti-Xa group reached therapeutic based dosing based on these results. Too many levels sooner and received fewer dose adjustments variables influence HSI and future study will than those in the aPTT group (29-31). have to be done with numerous variables to get Maybe in future study a TEG measurement a better view on the factors influencing HSI. might give more information on how heparin For now, a fixed dose of heparin with frequent ACT should be dosed. This would give information about controls and adjustments are needed to get a full ACT values, alongside other information regarding view of our patients’ anticoagulation status. coagulation. One possibility to get a better grip on this As stated before, although these measurements status, would be to change our way of measuring are difficult point-of-care, they might be useful in anticoagulation. As stated before, ACT devices have future study setting to get better results and possibly an inherent variation, which means that for every show better correlations with weight, BSA, BV or test on the same blood sample, there is a possibility other variables. for error. ACT devices also have a variation when Future study in a controlled setting with they are compared which each other. Especially at accessibility to multiple laboratory data and more higher ACT-values (>400-600), the I-Stat tends to accurate anticoagulation status measurement is deviate to higher values when compared to other needed to get more accurate heparin dosing in devices. At ranges lower than 400 s, there was no peripheral vascular surgery. definitive bias between brands (23, 24). Anti-Factor Xa (anti-Xa) measurement is a possible future alternative to ACT. Since recent References studies show heparin dose can be reduced when 1. Gertz S.D., Kurgan A., Wajnberg R.S. and Nelson E. 1979. using anti-Xa measurement as opposed to ACT Endothelial cell damage and thrombus formation following measurement. Studies have also shown that ACT temporary arterial occlusion : Effects of pretreatment with has a correlation with anti-Xa of r = 0.77 (using aspirin or heparin. J Neurosurg. 50 : 578-586. 2. Kasapis C., Gurm H.S., Chetcuti S.J., Munir K., Luciano Spearman correlation) and an inaccuracy rate of A. and Smith D., et al. 2010. Defining the optimal 9.4%, which can have a significant effect on ACT degree of heparin anticoagulation for peripheral vascular values and concurrent treatment. Use of ACT as interventions: insight from a large, regional, multicenter opposed to anti-Xa measurement also showed more registry. Circulation: Cardiovasc Interv. 3(6) : 593-601. 3. Muster I., Haas T., Quandt D., Kretschmar O. and Knirsch intra-operative bleeding events during endovascular W. 2017. Factors influencing ACT after intravenous bolus aneurysm repair. We further speculate that this administration of 100 IU/kg of unfractionated heparin

22-Vanwijnsberghe.indd 177 12/01/2021 12:24 178 b. vanwijnsberghe et al. during cardiac catheterization in children. Clin Appl 18. Kawatsu S., Sasaki K., Sakatsume K., Takahara S., Thromb/Hemost. 23(7): 740-747. Hosoyama K. and Masaki N., et al. 2018. Predictors of 4. Veerhoek D., Groepenhoff F., van der Sluijs M., de Wever heparin resistance before cardiovascular operations in J., Blankensteijn J. and Vonk A., et al. 2016. Heparin adults. Ann Thoracic Surg. 105(5): 1316-1321. responsiveness during arterial vascular surgery: is a fixed 19. Levy J.H., Montes F., Szlam F., Hillyer C.D. 2000. The in heparin dose adequate? J Cardiovasc and Vasc Anesth. 30: vitro effects of antithrombin III on the activated coagulation S21-S2. time in patients on heparin therapy. Anesth Analg. 90(5): 5. Wiersema A.M., Watts C., Durran A.C., Reijnen M.M., Van 1076-1079. Delden O.M. and Moll F.L., et al. 2016. The use of heparin 20. Garvin S., Fitzgerald D., Muehlschlegel J.D., Perry T.E., during endovascular peripheral arterial interventions: a Fox A.A. and Shernan S.K., et al. 2010. Heparin dose synopsis. Scientifica. Article ID 1456298. response is independent of preoperative antithrombin 6. Williams N., Broe P., Burke P., Meagher E., O’Donoghe activity in patients undergoing coronary artery bypass graft C. and Otridge B., et al. 1989 Heparin kinetics in vascular surgery using low heparin concentrations. Anesth Analg. surgery. Eur J Vasc Surg. 3(6): 493-496. 111(4): 856. 7. Quigley F.G., Jamieson G.G., Lloyd J.V. and Faris I.B. 21. Levi M., Keller T.T., van Gorp E. and ten Cate H. 2003. 1989. Monitoring of heparin in vascular surgery. J Vasc Infection and inflammation and the coagulation system. Surg. 8(2): 125-127. Cardiovasc Res. 60(1): 26-39. 8. Ndrepepa G. and Kastrati A. 2015. Activated clotting 22. Sette H., Hughes R., Langley P., Gimson A. and Williams time during percutaneous coronary intervention: a test for R. 1985. Heparin response and clearance in acute and all seasons or a mind tranquilizer? Am Heart Asso. 8(4): chronic liver disease. Thromb Haemos. 53(03): 591-4. e002576. 23. Tremey B., Szekely B., Schlumberger S., François D., Liu 9. Baker M.S., Skoyles J.R., Shajar M., Skinner H., Richens N. and Sievert K., et al. 2006. Anticoagulation monitoring D. and Mitchell IM. 2005. Can lean body mass be used to during vascular surgery: accuracy of the Hemochron® low reduce the dose of heparin and protamine for obese patients range activated clotting time (ACT-LR). BJA: Br J Anaesth. undergoing cardiopulmonary bypass? J Extracorp Technol. 97(4): 453-459. 37(2): 153. 24. Schussler J.M., Aguanno J.J., Glover E.N., Vish N.A., 10. Janmahasatian S., Duffull S.B., Ash S., Ward L.C., Byrne Wissinger L.A. and Schumacher J.R., et al. 2003. Com- N.M. and Green B. 2010. Quantification of lean bodyweight. parison of the i-STAT handheld activated clotting time Clin Pharmacokin. 44(10): 1051-1065. with the Hemochron activated clotting time during and 11. Gillette M.A., Franck A., Reeder D.N., Knott A. and after percutaneous coronary intervention. Am J Heat Card. Frohnapple D.J. 2015. Unfractionated heparin anticoa- 91(4):464-466. gulation using estimated blood volume based dosing versus 25. Snyder L.M., 2019. Wallach’s interpretation of diagnostic weight-based dosing in a Veteran population. Am J Health- tests: Lippincott Williams & Wilkins. Ch. 10: 550-555. System Pharm. 72(23, Suppl. 3): S171-S176. 26. Cox G.W., Runnels S., Hsu H.S.-H. and Das S.K. 1992. A 12. Esposito R.A., Culliford A.T., Colvin S.B., Thomas S.J., comparison of heparinised saline irrigation solutions in a Lackner H., Spencer F.C. 1983. The role of the activated model of microvascular thrombosis. Br J Plast Surg. 45(5): clotting time in heparin administration and neutralization 345-348. for cardiopulmonary bypass. J Thor Cardiovasc Surg. 27. Dieplinger B., Egger M., Luft C., Hinterreiter F., 85(2): 174-185. Pernerstorfer T. and Haltmayer M, et al. 2018. Comparison 13. Hong X., Shan P.R., Huang W.J., Zhu Q.L., Xiao F.Y., Li S., between activated clotting time and anti-activated factor et al. 2016. Influence of body mass index on the activated X activity for the monitoring of unfractionated heparin clotting time under weight-based heparin dose. J Clin Lab therapy in patients with aortic aneurysm undergoing an Analysis. 30(2): 108-113. endovascular procedure. J Vasc Surg. 68(2): 400-407. 14. Nissborg E., Wahlgren C-M. 2016. Anticoagulant Effect 28. Reyher C., Würfel C., Lindhoff-Last E., Meybohm P., of Standard Dose Heparin During Peripheral Endovascular Zacharowski K. and Moritz A., et al. 2016. Antifactor Xa Intervention. Ann Vascular Surg. 60: 286-292. Activity for the Management of Anticoagulation during 15. Ferguson J., Fathy H.W., Le D., Thomakos N. and Wilson Cardiac Surgery. Thor Cardiovasc Surg. 64(06): 494-500. J. 1998. The Effect of Body Weight and Body Surface 29. Vandiver J.W. and Vondracek T.G. 2013. A comparative Area Correction on the Distribution of the ACT Response trial of anti-factor Xa levels versus the activated partial to Bolus Doses of Heparin for PTCA. J Invasive Cardiol. thromboplastin time for heparin monitoring. Hosp Pract. 10(6): 318-322. 41(2):16-24. 16. Soleimannejad M., Aslanabadi N., Sohrabi B., Shamshir- 30. Yie K., Chon S.-H.D. and Na C.-Y. 2016. Activated clotting garan M., Separham A. 2014. Activated clotting time level time test alone is inadequate to optimize therapeutic with weight based heparin dosing during percutaneous heparin dosage adjustment during post-cardiopulmonary coronary intervention and its determinant factors. J resuscitational extracorporeal membrane oxygenation Cardiovasc Thoracic Res. 6(2): 97. (e-CPR). Perfusion. 31(4): 307-315. 17. Veerhoek D., Groepenhoff F., van der Sluijs M., de Wever 31. Atallah S., Liebl M., Fitousis K., Bostan F. and Masud J., Blankensteijn J. and Vonk A., et al. 2017. Individual dif- F. 2014. Evaluation of the activated clotting time and ferences in heparin sensitivity and their effect on heparin activated partial thromboplastin time for the monitoring anticoagulation during arterial vascular surgery. European of heparin in adult extracorporeal membrane oxygenation Journal of Vascular and Endovascular Surgery. 54(4):534- patients. Perfusion. 29(5): 456-461. 41.

22-Vanwijnsberghe.indd 178 12/01/2021 12:24 (Acta Anaesth. Belg., 2020, 71, 179-184) Local infiltration of anesthetics gives better results for pain and comfort scores then patient controlled epidural anesthesia in total knee arthroplasty

S. Pennynck (*), A.F. Kalmar (**), P. Van Overschelde (***), P. Byn (***), H. Van Overschelde (**)

Abstract : Background : Local infiltration of anesthetics and delayed revalidation, which might lead to (LIA) is an upcoming alternative method for analgesia a prolonged hospital stay. It could also increase following total knee arthroplasty (TKA). It could replace the risk for thromboembolic events and impaired patient controlled epidural anesthesia (PCEA). The wound healing (1, 2). Adequate perioperative anal- advantages of LIA are faster mobilization, less risks gesia gives less surgical stress which results in less compared to placement of an epidural catheter, and it postoperative morbidity and faster revalidation and increases patient care in a quantitative and qualitative recovery. In addition, there is a smaller chance of way. On the other hand, PCEA provides prolonged analgesia postoperatively. The aim of this study was developing a chronic pain syndrome (3). Despite to compare the effects of LIA and PCEA on pain and the importance of adequate perioperative analgesia, comfort scores during the seven days following TKA. up to fifty percent of patients receiving TKA Methods : Two weeks after hospital dismissal, a experience severe postoperative pain (1). questionnaire was sent to 116 patients who underwent The ideal method for analgesia has a long TKA. Patients were asked for their pain scores on the time of action, acts fast, has no adverse effects, day of surgery (=DO), two days after (=D2) and 7 days and should be easy to administer. It provides early after (=D7). They were also asked about their general mobilization (4, 5) and thereby decreases the number comfort, postoperative nausea and vomiting (PONV), of physical therapy sessions (6), thromboembolic vertigo and comfort urinating. The primary endpoint was events and infections (7) and possible formation pain at D7, secondary endpoints were pain on D0 and D2, of arthrofibrosis (8). These risks increase with a and comfort scores. prolonged hospital stay. Muscle weakness and Results : Data of 81 patients were analyzed, 40 were provided with PCEA and 41 patients received LIA. Pain orthostatic hypotension are also contributing factors scores on D7 were significantly better in the LIA-group (9). (18 [0-75] vs 38 [0-100] for LIA and PCEA, respectively In local infiltration of anesthetics (LIA), (p=0.01)). Pain scores on D2 were also significantly peripheral nerve block and wound infiltration with better in the LIA group, whereas no difference in pain local anesthetics are combined. Although benefits of scores was seen on D0. There was less PONV on D0 and LIA were already reported in 2008, the exact location less discomfort urinating on D7 in the LIA group. Other of the infiltration sites has not yet been described scores showed no significant differences between the two (10). The distribution of the local anesthetics is groups. Conclusion : LIA showed in lower pain scores and improved comfort scores, compared to PCEA. Seth Penninck, M.D. ; Alain F. Kalmar, MD, PhD ; Philippe Van Overschelde, MD ; Pieter Byn, MD ; Henk Van Overschelde, MD Keywords : Arthroplasty ; replacement ; anesthesia, (*) Dpt of Anesthesia, Ghent University Hospital, Ghent, epidural ; infiltration local anesthesia. Belgium. (**) Dpt of Anesthesia and ICM, Maria Middelares Hospital, Ghent, Belgium. Introduction (***) Dpt of Orthopedics, Maria Middelares Hospital, Ghent, Belgium. Corresponding author : Seth Pennynck, Department of On a yearly base, thousands of total knee Anesthesia, Ghent University Hospital, C. Heymanslaan 10, arthroplasty (TKA) procedures take place in 9000 Gent, Belgium. Belgium. TKA demands adequate analgesia per- Email : [email protected] and postoperatively. Inadequate analgesia can lead Paper submitted on Aug 5, 2020 and accepted on Aug 10, 2020. to postoperative discomfort, slower mobilization Conflict of interest : none.

23-Pennynck.indd 179 12/01/2021 12:25 180 s. pennynck et al. described in a neuroanatomically cadaver study where specific sensible locations were found (11). When using LIA, mastering the neuroanatomy of Fig. 1. — Visual Analogue Score (VAS) which was included the knee is a prerequisite. Only when injections are in the questionnaire. placed correctly, optimal analgesia can be achieved. The aim of this study was to compare pain and comfort scores between patient controlled epidural anesthesia (PCEA) and LIA with specific attention for the neuroanatomy of the knee. We hypothesized that LIA would provide better pain scores postoperatively.

Materials and methods

The study was approved by the ethical committee of Maria Middelares hospital, Ghent, Belgium (reference number PBH/nm/2015.52). Fig. 2. — Infiltration sites of the local anesthetics. Written informed consent was obtained from all patients. Reporting has been done conform to the Table 1 STROBE guidelines. Questions asked in the questionnaire Patient selection 1. How did you experience your pain on the day of surgery? All patients, aged ≥18 years, undergoing 2. How did you experience your pain the first day after surgery? 3. How did you experience your pain 7 days after surgery? TKA between 01-01-2014 and 01-03-2015 were 4. How did you experience the preparation 2 hours prior to surgery? eligible. Patients with allergy to non-steroid anti- 5. How was your general comfort in the days following surgery? inflammatory drugs (NSAID) or local anesthetics 6. In case of a second TKA, would you choose the same analgesic were excluded. Also revisions and partial knee method? 7. To what extent was it comfortable for you to urinate in the week replacements were excluded. Two weeks after following surgery? hospital dismissal, a questionnaire (table 1) was sent 8. To what extent did you experience vertigo in the days following through mail. Patients were asked about pain scores surgery? 9. To what extent were you nauseous in the days following surgery? on the day of surgery (=DO), two days after (=D2) and 7 days after (=D7). They were also asked about their general comfort, postoperative nausea and Anesthesia protocol and the local infiltration vomiting (PONV), vertigo and comfort urinating. A visual analogue scale (VAS) (Fig. 1) was included A strict protocol was used to provide anesthesia. and the patients were asked to draw a line on the Either LIA with spinal anesthesia (LIA-group) or included scale in proportion with their answer. This combined spinal-epidural (CSE) anesthesia (PCEA- scale had a length of 100 mm, which made it easier group) was used. Patients in the LIA-group received to calculate the scores afterwards. A stamped and spinal anesthesia with 3 ml levobupivacaine 0.25%, addressed envelope was included to make it easy for whereas patients in the PCEA-group received spinal the patient to return the questionnaire. anesthesia with 3 ml levobupivacaine 0.5%. General After collecting all questionnaires, comple- anesthesia was induced in all patients with propofol teness was checked. Incomplete or inaccurate 2-3 mg/kg, sufentanil 10 µg and dexamethasone questionnaires were left out of the analysis. VAS- 5 mg. Patients also received parecoxib 40 mg and scores were qualified and put into an Excel file. paracetamol 1 g. When registering the data, it was not known whether Infiltrations in the LIA-group were done with the patient received LIA of PCEA. After unblinding, a moving needle technique to prevent intravenous no more adjustments were made in the data file. injection of the mixture. They were placed with Primary endpoint was the pain score on D7. 50ml syringes with 18G needles with a length of Secondary endpoints were pain scores on D0 4 cm. and D2, as well as PONV, experience during the The injections were given in two phases (Fig. preparations, comfort in the first days, comfort 2), based on the description of Kerr (10). The surgeon urinating at D7 and their favor for the same pain applied the technique. No catheter was left behind. management in future TKA. Posterior, proximal and medial structures were

23-Pennynck.indd 180 12/01/2021 12:25 epidural anesthesia in total knee arthroplasty 181 infiltrated with a mixture of 140 ml ropivacaine 2 Postoperative management mg/ml, ketolorac 30 mg and 10 µg/ml epinephrine. To guarantee a good spread of the mixture, When patients from the PCEA-group arrived following technique was used by the surgeon. at the post anesthesia care unit (PACU) a bladder After completing all osseous incisions, the catheter was placed by the nurses. Pain medication posterior tissue of the knee in 90° flexion was was given to all patients following a strict in- infiltrated (Fig. 2a) : hospital protocol. Paracetamol 1g was given each 1. Through the posterior capsule of the knee, 6 hours intravenously of orally, parecoxib each 12 with a medial direction the posterior side of hours until 36 hours after surgery. In case of pain the femur was found and 40 ml of the mixture tramadol 100 mg with alizapride 50 mg was given was applied every 6 hours and if necessary, also piritramide 20 2. Through the posteromedial capsule another 7 mg intramuscular. After 36 hours all analgesics were ml was injected given orally with either Zaldiar® (paracetamol 325 3. Through the posterolateral capsule 3 ml was mg/tramadol 37,5 mg) and paracetamol 500 mg, or given paracetamol 1 g every 6 hours and diclofenac 75 mg After cementation of all components (Fig. 2b), twice a day. the synovial and muscle tissues were infiltrated : Analgesics were stopped when the patients 4. Medial synovial and capsular structures : 3x5 did not feel any need for it anymore. Passive ml mobilization was initiated at D0 in both groups. 5. For the vastus medialis : 15 ml In the LIA group, patients were allowed to 6. In the adductor channel also 15 ml walk with crutches at D1 every 2-3 hours. In the 7. In the suprapatellar pouch another 5 ml was PCEA-group, active mobilization started on D2 infiltrated because epidural analgesia might cause orthostatic To prevent dermal necrosis, subcutaneous infil- hypotension and motoric block. tration of local anesthetics was performed without During hospital stay, patients were guided by epinephrine in the mixture. Following infiltrations a physical and occupational therapist, which was completed the LIA continued throughout the rehabilitation process 8. In the medial subcutaneous tissue 6x5 ml was after hospital dismissal. injected 9. In the proximal subcutaneous tissue in medial Statistical analysis direction 20 ml was infiltrated While infiltrating all these structures, the A pilot study showed a mean (± standard neuroanatomy of the knee was considered, both the deviation) pain score of 40 (± 30) and 25 (± 20) in posterolateral as the anteromedial innervation of the the PCEA and LIA group, respectively. A difference knee. of 15% between both techniques was considered In the PCEA-group, patients did not receive clinically relevant. We used a Wilcoxon-Mann- any additional local anesthetics intraoperatively. Whitney test with an α-error probability of 0.05 PCEA was provided from the moment mobility had and a power of 0.8 to calculate a sample size of returned, with a mixture of levobupivacaine 1.25 39 patients in each group. Considering a non- mg/ml, sufentanil 0.25 µg/ml and clonidine 1.125 responders rate of 35%, 60 persons per group were µg/ml. It ran at 4 ml/h continuously with 4 ml boli approached. To assess the normality of continuous if necessary with a limit of 50 ml/4 hours and an variables a Kolmogorov–Smirnov test was used. interval of 20 min between two boli. To assess the difference between the two groups, a Wilcoxon-Mann- Whitney U test was used. P < 0.05 Surgical technique is considered statistically significant.

Both groups received comparable techniques Results (12). A standard medial parapatellar incision was used and all components were cemented. To prevent The flow diagram of the study inclusion is swelling, a compressive bandage was used for presented in figure 3. Of the 124 patients who venous and lymphatic compression. No drains were were assessed for eligibility, eight of them did not left behind in the wounds. Surgery was performed meet the inclusion criteria (6 revisions, 2 partial by two surgeons. One surgeon used LIA, the other TKA). Of the remaining 116 patients, 83 returned surgeon treated patients receiving PCEA. their questionnaire. Because of inaccuracy, two

23-Pennynck.indd 181 12/01/2021 12:25 182 s. pennynck et al.

Fig. 4. — Boxplot of the main findings, the boxplots show the minimum and maximum value, median value, and mean value (-). * indicates statistical difference (P<0.05). Fig. 3. — Flow diagram according to the Consolidated Standard of Reporting Trials (CONSORT) statement. We asked for pain scores even after discharge Table 2 from the hospital until day seven postoperatively. Patient characteristics and VAS scores (scale 0-100) This could be interesting, as there are no other studies comparing both techniques on the longer PCEA group LIA group P-value term. An earlier study comparing LIA and PCEA Number of patients 41 40 found no difference between both in the first three Male/female 31/10 30/10 days postoperatively. In that study, a catheter was Age 63 [44-84] 67 [40-84] left behind until the first day postoperatively (13). Pain at D0 18 [0-98] 21 [0-100] 0.56 A major review (14) identified 12 important Pain at D2 21 [0-100] 20 [0-81] <0.01* parameters for the outcome after orthopedic Pain at D7 38 [0-100] 18 [0-75] 0.01* surgery : mortality, length of hospital stay, waiting Experience during time until surgery, complications, readmissions for 13 [0-100] 21 [0-100] 0.66 preparations revisions, mobility, quality of life, pain, daily living Comfort score first 20 [0-97] 20 [0-77] 0.05 activities, medication use, location of living and days costs. An adequate perioperative analgesic strategy Favour same pain 17 [0-95] 0 [0-100] 0.93 management could influence some of these parameters. Comfort urinating D7 0 [0-100] 0 [0-41] <0.01* The cornerstone of postoperative pain mana- Vertigo first days 0 [0-100] 0 [0-97] 0.12 gement should consist of a multimodal, balanced analgesia with a minimal risk for complications. Nausea at D0 0 [0-100] 0 [0-93] <0.01* It should promote early rehabilitation and have These variables are reported as absolute number or as median a minimal cost. Central sensitization should be [minimum-maximum]. *P<0.05. avoided in order to prevent the evolution to chronic questionnaires were refused. Eventually 81 ques- pain syndrome (15). tionnaires were analyzed (41 belonged to the LIA- Besides paracetamol and NSAID, other medi- group and 40 to the PCEA-group). cations and techniques exist to threat postoperative Patient characteristics and reported VAS pain in TKA. Of all these, opioids, femoral nerve scores are presented in table 2. Figure 4 depicts the block, adductor channel block, epidural anesthesia main findings of the study. The primary endpoint and LIA are most frequently used. All these was reached, pain scores on D7 were significantly techniques have specific pros and cons. In the better in the LIA-group. Pain scores on D2 were present study, we compared LIA with PCEA as also significantly better in the LIA group, whereas these are the standard techniques in our hospital for no difference in pain scores was seen on D0. There pain management after TKA. was less PONV on D0 and less discomfort urinating Epidural analgesia is a reliable technique, on D7 in the LIA group. Other scores showed no although execution might sometimes be difficult, significant differences between the two groups. may be uncomfortable for the patient and is time consuming. Also, although rare, there is the risk of Discussion complications, such as epidural or spinal hematoma The main findings of this study are that LIA or abscess (16). The incidence of epidural hematoma gives significant better pain and comfort scores is 1 per 4330 to 22189 placements (17, 18). Also (PONV, comfort urinating) than PCEA. postdural puncture headache is an unsettling com-

23-Pennynck.indd 182 12/01/2021 12:25 epidural anesthesia in total knee arthroplasty 183 plication. Another side effect is bladder dys- stay (24). A multidisciplinary consensus, published function, or the need for placement of a bladder in the Acta Orthopaedica in October 2019, confirms catheter with its inherent risks. Epidural opioids LIA as a part of ERAS programs (25). may provoke urine retention and itching. The This study demonstrated a significant difference orthosympathic block may provoke hypotension in pain scores at D7, but it has some limitations. and even hemodynamic instability. Other possible First of all, the PCEA-group and LIA-group were complications are back pain, systemic toxicity, treated by another surgeon. Although they both neural damage and persistent paresthesia. used exactly the same surgical and infiltration Contraindications for epidural analgesia are technique, this may have induced a bias. Secondly, coagulation disorders, local infection, increased patients in the PCEA group received a higher dose intracranial pressure, aortic of mitral valve stenosis of spinal chirocaine, as the epidural gives motor or a refusing patient. Pros and cons should be block anyway. Thirdly, the study only compares weighed in patients with spinal diseases like mul- two different techniques in a single center. A multi- tiple sclerosis (19). center approach would have been more interesting. LIA on the other hand is easy to execute and has few side-effects. A mixture of ropivacaine, ketolorac Conclusion and epinephrine is injected in the knee’s surrounding tissue (10). This mixture can differ, but these three LIA provided lower pain scores on D7 products form the cornerstone of this technique postoperatively, compared to PCEA. Pain scores on (20). It consists of three basic components : a high D2, nausea at D0 and discomfort urinating at D7 volume of diluted and long-acting local anesthetics, were also significantly lower in the LIA group.The non-local analgesics and the possibility of a catheter study demonstrates that LIA is a reliable technique for continuous injection of local anesthetics. The to treat postoperative pain after TKA. concentration of the local anesthetic can be kept low as the small nerve endings do not require a large dose. As the dose remains low, the volume References can be increased providing a long-acting block without exceeding toxicity ranges. The absence of 1. Chan EY, Blyth FM, Nairn L and Fransen M. 2013 Acute large veins or arteries in the infiltrated area gives postoperative pain following hospital discharge after total knee arthroplasty. Osteoarthritis Cartilage. 21(9):1257- even less chance on systemic toxicity. Epinephrine 1263. added to the mixture does not have any analgesic 2. Kehlet H, Jensen TS and Woolf CJ. 2006 Persistent properties, but it gives local vasoconstriction and postsurgical pain: risk factors and prevention. Lancet. 367(9522):1618-1625. therefore a prolonged analgesic effect. 3. Boezaart AP, Davis G and Le-Wendling L. 2012 Recovery In TKA, it is recommended to avoid peri- after orthopedic surgery: techniques to increase duration of articular catheters because of the greater risk pain control. Curr Opin Anaesthesiol. 25(6):665-672. 4. Dillon JP, Brennan L and Mitchell D. 2012 Local infiltration for infection. An alternative could be liposomal analgesia in hip and knee arthroplasty : an emerging tech- bupivacaine, which has a long time of action, but nique. Acta Orthop Belg. 78(2):158-163. it is not yet approved for use and so far it did not 5. Husted H, Hansen HC, Holm G, Bach-Dal C, Rud K, prove to have a superior effect (21). Andersen KL and et al. 2010 What determines length of stay after total hip and knee arthroplasty? A nationwide As the technique for LIA is described before, study in Denmark. Arch Orthop Trauma Surg. 130(2):263- it is interesting to consider that while administering 268. LIA, in some degree an adductor channel block is 6. Labraca NS, Castro-Sanchez AM, Mataran-Penarrocha GA, Arroyo-Morales M, Sanchez-Joya Mdel M and Moreno- given. Since the dose is smaller, there is less risk of Lorenzo C. 2011 Benefits of starting rehabilitation within motor block than a classic adductor channel block. 24 hours of primary total knee arthroplasty: randomized LIA has a prominent place in ERAS (Enhanced clinical trial. Clin Rehabil. 25(6):557-566. 7. Paul JE, Arya A, Hurlburt L, Cheng J, Thabane L, Tidy A Recovery After Surgery) programs. Length of and et al. 2010 Femoral nerve block improves analgesia hospital stay, perioperative pain control, rapidity outcomes after total knee arthroplasty : a meta-analysis of of mobilization and complications are favorably randomized controlled trials. Anesthesiology. 113(5):1144- affected by the use of LIA (22, 23). ERAS has no 1162. 8. Schairer WW, Vail TP and Bozic KJ.2014 What are the impact on readmission rate 30 days after surgery, but rates and causes of hospital readmission after total knee is does affect length of stay (23). ERAS programs arthroplasty? Clin Orthop Relat Res. 472(1):181-187. using LIA have shown to restrain the use of opioids, 9. Husted H, Lunn TH, Troelsen A, Gaarn-Larsen L, Kristensen BB and Kehlet H. 2011 Why still in hospital increase patient satisfaction and decrease the after fast-track hip and knee arthroplasty? Acta orthop. number of complications and duration of hospital 82(6):679-684.

23-Pennynck.indd 183 12/01/2021 12:25 184 s. pennynck et al. 10. Kerr DR and Kohan L. 2008 Local infiltration analgesia: 19. NYSORA. 2020 Epidural Anesthesia and Analgesia a technique for the control of acute postoperative pain 2020. Available from: https://www.nysora.com/regional- following knee and hip surgery: a case study of 325 patients. anesthesia-for-specific-surgical-procedures/abdomen/ Acta Orthop. 79(2):174-183. epidural-anesthesia-analgesia/. 11. Quinn M, Deakin AH, McDonald DA, Cunningham IK, 20. NYSORA. 2020 Intra-articular and Periarticular Infiltration Payne AP and Picard F. 2013 An anatomic study of local of Local Anesthetics 2020. Available from: https://www. infiltration analgesia in total knee arthroplasty. Knee. nysora.com/regional-anesthesia-for-specific-surgical- 20(5):319-323. procedures/lower-extremity-regional-anesthesia-for- 12. Wegrzyn J, Parratte S, Coleman-Wood K, Kaufman KR specific-surgical-procedures/anesthesia-and-analgesia-for- and Pagnano MW. 2013 The John Insall award: no benefit hip-procedures/intra-articular-periarticular-infiltration- of minimally invasive TKA on gait and strength outcomes: local-anesthetics/. a randomized controlled trial. Clin Orthop Relat Res. 21. Bravin LN, Ernest EP, Dietz MJ and Frye BM. 2020 471(1):46-55. Liposomal Bupivacaine Offers No Benefit Over Ropiva- 13. McDonald DA, Deakin AH, Ellis BM, Robb Y, Howe TE, caine for Multimodal Periarticular Injection in Total Knee Kinninmonth AW and et al. 2016 The technique of delivery Arthroplasty. Orthopedics. 43(2):91-96. of peri-operative analgesia does not affect the rehabilitation 22. Jiang HH, Jian XF, Shangguan YF, Qing J and Chen LB. or outcomes following total knee arthroplasty. Bone Joint J. 2019 Effects of Enhanced Recovery After Surgery in Total 98-B(9):1189-1196. Knee Arthroplasty for Patients Older Than 65 Years. Orthop 14. Liem IS, Kammerlander C, Suhm N, Blauth M, Roth Surg. 11(2):229-235. T, Gosch M and et al. 2013 Identifying a standard set of 23. Zhu S, Qian W, Jiang C, Ye C and Chen X. 2017 Enhanced outcome parameters for the evaluation of orthogeriatric co- recovery after surgery for hip and knee arthroplasty: a management for hip fractures. Injury. 44(11):1403-1412. systematic review and meta-analysis. Postgrad Med J. 15. Grosu I, Lavand’homme P and Thienpont E. 2014 Pain 93:736-742. after knee arthroplasty: an unresolved issue. Knee Surg 24. Van Horne A and Van Horne J. 2019 Patient-optimizing Sports Traumatol Arthrosc. 22(8):1744-1758. enhanced recovery pathways for total knee and hip 16. Moen V, Dahlgren N and Irestedt L. 2004 Severe neurological arthroplasty in Medicare patients: implication for transition complications after central neuraxial blockades in Sweden to ambulatory surgery centers. Arthroplast Today. 5(4):497- 1990-1999. Anesthesiology. 101(4):950-959. 502. 17. Bateman BT, Mhyre JM, Ehrenfeld J, Kheterpal S, Abbey 25. Wainwright TW, Gill M, McDonald DA, Middleton RG, KR, Argalious M and et al. 2013 The risk and outcomes Reed M, Sahota O and et al. 2020 Consensus statement for of epidural hematomas after perioperative and obstetric perioperative care in total hip replacement and total knee epidural catheterization : a report from the Multicenter replacement surgery: Enhanced Recovery After Surgery Perioperative Outcomes Group Research Consortium. (ERAS(R)) Society recommendations. Acta Orthop. Anesth Analg. 116(6):1380-1385. 91(1):3-19. 18. Christie IW and McCabe S. 2007 Major complications of epidural analgesia after surgery: results of a six-year survey. Anaesthesia. 62(4):335-341.

23-Pennynck.indd 184 12/01/2021 12:25 (Acta Anaesth. Belg., 2020, 71, 185-196) The effects of intraoperative use of dexmedetomidine on the incidence of postoperative delirium and postoperative cognitive dysfunction : a systematic review

J. Hermans (*), D. Hoogma (*), S. Devroe (*), E. Roofthooft (**), S. Rex (*), M. Van de Velde (*)

Abstract : Background : Postoperative delirium and Introduction cognitive dysfunction contribute to an important increase of complications after major surgery. The use Background of dexmedetomidine in ICU has already established promising results in previous studies regarding its Postoperative Delirium (POD) and postope- delirium-reducing features. However, the results on rative cognitive dysfunction (POCD) have been its intraoperative application remain conflicted. The consistently associated with short term post- goal of this systematic review is to examine whether there is scientific evidence that dexmedetomidine operative complications. The incidence of POD reduces delirium and cognitive dysfunction when it is and POCD has been reported to affect 20% to administered intraoperatively. 40% in the critical ill (1, 2). This even exceeds Methods : This systematic review was fully processed 50% after cardiac surgery (3). The pathogenesis according to the Preferred Reporting Items for Systematic has been studied thoroughly over the last decade. Reviews and Meta-analyses Protocols (PRISMA-P) Nevertheless, the exact pathophysiological mecha- guidelines. The build-up is carried out according to nism has not yet been fully clarified. Many factors the Cochrane Handbook for Systematic reviews of have already been identified in contributing to its Interventions. This review is based on prospective incidence (4-6) but it is clearly a multifactorial recent double-blind Randomized Controlled Trials and condition. Predisposing and precipitating risk retrospective Cohort Studies. factors are summarized in respectively figure 1 Focus : We will focus on the benefits of dexmedetomidine and figure 2 (7). Besides a higher morbidity and towards the prevention of postoperative delirium and cognitive dysfunction. Furthermore, we will make a mortality, POD and POCD also entail a prolonged comparison between cardiac and non-cardiac surgery. in-hospital stay and therefore an increase in hospital Dexmedetomidine infusion was compared with a placebo costs (8-10). Many interventions have been devised and outcomes were assessed by neuropsychological however, due to its multifactorial pathogenesis assessment tests. a straightforward solution remains anything but Results : In total, fourteen articles have been selected. obvious. To date, first-generation antipsychotics, There was no clear evidence concerning a reduction in second-generation antipsychotics, cholinergic en- postoperative cognitive dysfunction. Neither was there a hancers and antiepileptic agents are used to counter reduction in delirium after non-cardiac surgery. On the postoperative delirium but the evidence supporting other hand, after cardiac surgery, a reduction in delirium the efficacy is limited (7). Dexmedetomidine, a was observed in all applicable articles. highly selective and short acting a-2-agonist, has Conclusions : Intraoperative dexmedetomidine most likely prevents postoperative delirium after cardiac surgery. In non-cardiac surgery there was no clear J. Hermans, MD ; D. Hoogma, MD ; S. Devroe, MD ; E. prove for the use of dexmedetomidine in the prevention Roofthooft, MD ; S. Rex, M. Van de Velde, MD. of postoperative delirium. No conclusions could be (*) Department of Cardiovascular Sciences and Department of established regarding postoperative cognitive dysfunction Anaesthesiology, UZ Leuven, Leuven, Belgium. and the use of dexmedetomidine intraoperatively. (**) Department of Anesthesiology, ZNA Middelheim Hospi- tals Antwerp, Belgium. Corresponding author : J. Hermans, Faculteit Farmaceutische Keywords : Dexmedetomidine ; delirium ; cognitive Wetenschappen, UZ Leuven, 3000 Leuven, Belgium. dysfunction ; intraoperative ; clinical trial. E-mail : [email protected]

Paper submitted on Aug 4, 2020 and accepted on Aug 10, 2020. Conflict of interest : none.

24-Hermans.indd 185 12/01/2021 12:26 186 j. hermans et al.

POD POCD disorders and disturbed physical function even up Age > 70 years Age > 65 years to one year after surgery (17). Depression Preoperative cognitive impairment POCD is described as a subtle change in Dementia Preoperative physical impairment mental status which is due to a surgical intervention. Preoperative use of narcotics Cognitive impairment during or benzodiazepines hospitalization The most frequently affected mental properties are Self-reported use of alcohol Delirium verbal learning, working memory, episodic memory Previous history of delirium and processing speed. A preoperative baseline of the Vision impairment patients’ mental state is essential to evaluate POCD. Severe illness Several scales have been developed attempting to BUN/creatinine ratio >18 evaluate a deterioration in mental state. To date the Tobacco use Vascular surgery Mini-mental State Examination scale or MMSE Depressive symptoms scale and the Montreal Cognitive Assessment or Attentional deficits MoCA-scale are considered to be the most valid Fig. 1. — Predisposing factors of postoperative delirium tests to evaluate a deterioration in mental status (POD) and postoperative cognitive dysfunction (POCD) (22). (13, 18). Studies suggest that the MoCA-scale detects a cognitive decay earlier than the MMSE- scale, but the MMSE-scale is the easier one to carry POD: intraoperative precipitating triggers out. The viability of the MMSE-scale still remains Acute pain Use of physical restraints disputable (18). Although its limits, the MMSE- Malnutrition scale is widely acknowledged as the most popular Addition of three or more medications in 24–48 h neuropsychological assessment test regarding Use of a urinary bladder catheter altered mental states. Results however should be Anaemia interpreted with caution. POCD is associated with Electrolyte and fluid abnormalities a prolonged in hospital stay, a faster functional Greater surgical blood loss, greater intraoperative transfusion decline, a diminishment of autonomy and an Fig. 2. — Intraoperative precipitating triggers of postoperative increased mortality (19-21). delirium (POD) (22). It is consequently essential for both entities to be diagnosed in an early stage and predisposing factors should be corrected if possible. A thorough sedative, anxiolytic and analgesic features with preoperative risk stratification should be performed minimal respiratory depression (11, 12). It has been aiming to identify predisposing factors and therefore excessively tested as a sedative in the Intensive predict the patients’ risk for POD and POCD. Care Unit (ICU). Several studies confirmed that Predisposing factors regarding the incidence of postoperative administered dexmedetomidine at POD and POCD are summarized in Figure 1. the ICU contributes to a significant reduction in Eliminating precipitating factors for postoperative the incidence of POD. However, it remains unclear delirium (Fig. 2) showed a drastic reduction in whether the same results could be established when the incidence of delirium (22, 23). Therefore, we dexmedetomidine is administered intraoperatively only can emphasize that a intraoperative reduction because of an absence of randomised controlled in triggering factors and a thorough preoperative trails concerning this subject. screening remains the key-feature in POD/POCD POD is defined by the Diagnostic and Sta- prevention. tistical Manual of Mental Disorders fifth edition Dexmedetomidine is a highly selective a-2 (DSM-V criteria) as an alteration in mental status of adrenoreceptor agonist with no effect on the GABA- the patient which is characterized by a reduction in receptor which generates a state of sedation/sleep awareness and a disturbance in attention (13). Other without any respiratory depression. Although the neuropsychological disorders (for example : changes working mechanism is not fully clarified yet, its in memory deficit, language, visuospatial ability, sedative properties are induced by its affinity for perception or disorientation) can be associated. The the locus coeruleus (an area in the brain regulation key method - and as well the “Golden Standard” – arousal) (24-32). Previous studies demonstrated towards identifying POD, remains the Confusion the anxiolytic and analgesic properties of dexme- Assessment – Method (CAM) (13). Patients who detomidine when used as a sedativum in ICU (24, develop delirium at an ICU are 200% to 400% more 33-38). In a meta-analyse from Flükiger J., et al., likely to prematurely die (14-16). Studies showed (24) – containing 28 double-blind randomized that POD can lead to an increase in cognitive controlled trials (RCTs) – a significant reduction

24-Hermans.indd 186 12/01/2021 12:26

Figure 3: Flow chart of systematic search for randomized or quasi-randomized controlled trials according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses the effects of intraoperative use ofProtocols dexmedetom (PRISMA-P) guidelines.idi ne 187

on the incidence and duration of POD and POCD Records identified through Additional records identified database searching (Pubmed) through other sources (Embase) after major surgery was identified. Dosages of (n = 394) (n = 162)

sedatives (for example Midazolam, Haloperidol, Identification

…) and opioids could be drastically been reduced Records after duplicates removed (n = 370) in the dexmedetomidine group. Similar results were

already described in a study from Pasin L., et al. (39)

Records excluded from 2014 and Xia Z., (40) from 2013. Based on these (n = 353) if Records screened Case reports, opinion results the Food and Drug Administration (FDA) (n = 370) reviews, animal studies,

Screening meta-analyse and

and Therapeutic Goods Administration (TGA) guidelines.

already approved the use of dexmedetomidine as a Full-text articles excluded,

(n = 3) if sedativum in ICU. A study of Sun Y. et al. (41) also Full-text articles assessed for eligibility (RCT’s) Observational study or/and has demonstrated a lower occurrence of emergence Eligibility (n = 17) no results

delirium in children after the use of intraoperative

dexmedetomidine during cardiac surgery. The few

Studies included in similar studies that have been conducted in the adult qualitative synthesis (n = 14) population could not confirm the same assumption Included

as a result of an underpowered sample size.

Fig. 3. — Flow chart of systematic search for randomized or Consequently, using only applicable and recent quasi-randomized controlled trials according to the Preferred randomized controlled trials (RCTs), this review Reporting Items for Systematic Reviews and Meta-analyses will elaborate if dexmedetomidine could achieve Protocols (PRISMA-P) guidelines. 15 similar results when administered intraoperatively.

Methodology evaluated on their usability in the review. For your reference see Tables 1, 2, 3 and 4. Literature search Study characteristics The literature search of this review was guided by the Preferred Reporting Items for Systematic This review will include 12 double-blind Reviews and Meta-Analyses Protocols (PRISMA-P) randomized controlled trials (RCT) and 2 retro- guidelines. Following electronic databases were spective cohort studies published in medical utilized : Pubmed, Embase and Cochrane library. journals between 2012 and 2018. All studies utilized The searched Medical Subject Heads (MeSH) were : dexmedetomidine versus a placebo containing a “dexmedetomidine” and “delirium”. In addition, the saline solution. Dexmedetomidine was started at search was refined using the following keywords : the induction of surgery with or without a loading “intraoperative”, “cognitive dysfunction” and dose and was continued intraoperatively until the “clinical trial”. end of surgery. Of the 14 relevant publications, 13 Eligibility and inclusion criteria were defined articles investigated the incidence of POD or POCD as followed : administration of dexmedetomidine as primary objective. One article investigated this as intraoperative, adult surgical patients (age >18 secondary objective. years), studies in which a placebo is used as a Diagnostic tests control group, surgery under general anaesthesia, POD or POCD as primary or secondary focus, The occurrence of POD or POCD was assessed studies written or translated into English. Exclusion immediately postoperative with a follow-up up criteria included : use of dexmedetomidine as a to 30 days. Both POD and POCD were evaluated sedative added to regional anaesthesia and case by the use of neuropsychological assessment report studies. methods by specialized personnel. The Confusion Assessment Method (CAM) or the Confusion Results Assessment Method in ICU (CAM-ICU) was used in the majority of the studies concerning the A total collection of 370 articles would be evaluation of POD. This test is considered to be relevant for the study. A flow chart of the selection the most relevant and reliable indicator to assess can be found in figure 3. Fourteen publications would POD according to the Diagnostic and statistical comply with the selection criteria for this systematic Manual of Mental disorders (DSM-V) criteria (13). review. All these articles were thoroughly read and This test evaluates the following criteria : 1) Acute

24-Hermans.indd 187 12/01/2021 12:26 188 j. hermans et al.

Limits - Non-representative patients - Non-representative (only ASA I-II) - Small group - Non-representative patients (all ages) - No limitations were found - Small group - Non-standardized methods for POCD evaluation. - No loading dose -Small group method Non-standardized - for POCD evaluation - Small group - Non-representative patients I-II patients). ASA (all ages - Evaluation conducted only 2 hours postoperative indication - Wrong - Small group - No continuous infusion - Non-representative patients I-II patients). ASA (all ages

P P =

Results Significant reduction of POCD (15% versus 27%) after 7 days = 0.05 P No significant reductionafter 30 days. of POCD P= 0.34 Significant reduction of POCD (15% versus 35%) after 5 days < 0.05 .P Significant reduction of POCD (20% versus 42%) after 2 days. < 0.01 P No significant reduction(20%versus 20%) after 1 day. of POCD = 0.99 P No significant reduction(24% versus 20%) after of POCD 7 days. 0.73 Significant reduction(6.3% versus 20%) after 1 to 3 days. < 0.05 P of POCD No significant reductionafter 2 hours. of POCD = 0.02 P Significant reduction on POCD after 1 and 3 days. < 0.009 P

Patients - Total of 122 Total - Age > 60 years - - ASA I-II - Total of 40 Total - Age > 18 years - - ASA I-III - Total of 100 Total - Age > 60 years - - ASA I-III - Total of 50 Total - Age > 60 years - - ASA I-III - Total of 96 Total - Age > 65 years - - ASA I-III - Total of 40 Total - Age > 18 years - - ASA I-II - Total of 80 Total - Age > 18 years - - ASA I-II

Evaluation with aims, methodology, evaluating method, results and limitations with aims, methodology, POCD determined by POCD determined at MMSE evaluation day 7 and 30. POCD determined by POCD determined at MMSE evaluation day 1 and 5. POCD determined by POCD determined at MMSE evaluation day 1 and 2. POCD determined by POCD determined Word Stroop Colour the and Test Interference the MoCa scale at day 1 and day 7. POCD determined by POCD determined scale evaluation MoCA at day 1, 2, 3 and 7. POCD determined by POCD determined after MMSE evaluation 2 hours. POCD determined by POCD determined MMSE on day 1 and 3 postoperative.

Table 1 Table cognitive dysfunction Methodology Dexmedetomidine from Dexmedetomidine induction (loading dose 1.0 infusion continuous and µg/kg the end of until 0.4 µg/kg/h) at surgery. Dexmedetomidine from in- Dexmedetomidine (loading dose 0.8 µg/ duction kg and continuous infusion at 30 minutes until µg/kg/h) 0.3 before the end of surgery. Dexmedetomidine from in- Dexmedetomidine (loading dose 1.0 µg/ duction kg and continuous infusion at 0.4 µg/kg/h) until the end of surgery. Dexmedetomidine from in- Dexmedetomidine infu- in a continuous duction the until 0.4 µg/kg/h sion at end of surgery Dexmedetomidine from in- Dexmedetomidine (loading dose 0.8 µg/ duction kg and continuous infusion at 0.5 µg/kg/h) until the end of surgery Dexmedetomidine from in- Dexmedetomidine (loading dose 1.0 µg/ duction kg and continuous infusion at 0.5 µg/kg/h) until the end of surgery Dexmedetomidine once at Dexmedetomidine dose of in a loading induction 4.0 µg/kg.

Aim Incidence of POCD after laparoscopic cholecystectomy Incidence of POCD after robot-assisted laparoscopic radical cystectomy. Incidence of POCD after laparoscopic cholecystectomy. Incidence of POCD after extended abdominal surgery. Incidence of POCD after laparoscopic ovarium cystectomy. Incidence of POCD after tympanoplasty Incidence of POCD after non- cardiac surgery

study

Type RCT Double-blind Prospective RCT Double-blind Prospective RCT Double-blind Prospective RCT Double-blind Prospective RCT Double-blind Prospective RCT Double-blind Prospective RCT Double-blind Prospective

Title

An overview of articles regarding the postoperative assessment Dexmedetomidine may laparoscopic after function cognitive benefit patients. in elderly cholecystectomy Effects of Dexmedetomidine on of Dexmedetomidine Effects anesthesia recovery period and function postoperative cognitive after robot-assisted of patients laparoscopic radical cystectomy. Effect of Dexmedetomidine on Dexmedetomidine of Effect early postoperative cognitive dysfunction and peri-operative inflammation in elderly laparoscopic undergoing patients cholecystectomy. The effect of Dexmedetomidine The effect of postoperative on the incidence dysfunction in elderly cognitive after prolonged abdominal patients surgery. Effect of Dexmedetomidine-indu- Effect ced anesthesia on the postoperative patients elder of function cognitive cystec- ovarian laparoscopic after tomy. The effect of Dexmedetomidine on Dexmedetomidine of effect The in and anxiety function cognitive middle-ear surgery Effect of Dexmedetomidine Effect Hydrochloride on early cognitive elderly in postoperative function patients

Year 2013 2015 2015 2014 2017 2012 2018

Autor Chen J., et al. Ding L., et al. Li et al. Y., Mohamed Mohamed S. et al. Xu H., et al. F., Yilmaz et al. Zhao W., et al. RCT Randomized Controlled Trial, POCD Postoperative cognitive Dysfunction, MMSE Mini-Mental State Examination, ASA American Society of Anaesthesiologist Score, MoCA Montreal Cognitive Assessment Montreal Cognitive Anaesthesiologist Score, MoCA American Society of ASA POCD Postoperative cognitive Dysfunction, MMSE Mini-Mental State Examination, Trial, Randomized Controlled RCT

24-Hermans.indd 188 12/01/2021 12:26 the effects of intraoperative use of dexmedetomidine 189

Limits

-Cohort study - Non-standardized method for POD evaluation. - No loading dose - No fixed continuous infusion rate. - Stopped early because of futility. - No loading dose -POD evaluation not as primary aim - Cohort study - Non-standardized method for POD evaluation. - No loading dose - No fixed continuous infusion rate. - Non-representative group (> 18 years) - No fixed continuous infusion rate. - Exceptional low incidence of delirium in general. - Small study - Non-representative patients (young and only ASA I-II). - CAM-ICU evaluation conducted only 2 hours postoperative - No loading dose - Small group - Non-representative group (> 18 years) - Dubious study (no power analysis, inconsistencies, …)

Results

: :

Significant reduction of POD after 24 hours. (7% versus 11%) < 0.0001 P No significant reduction of POD after 5 days. (12% versus 11%) = 0.94 P Significant reduction of POD (5.4% versus 7.4%) after 24 hours. = 0.003 P Study 1 Significant reduction of POD (9.5% versus 25%) after 5 days. = 0.04 P Study 2: Significant reduction of POD (18.4% versus 25%) after 5 days. = 0.04 P No significant reduction of POD (4.9% versus 7.7%) after 5 days. = 0.341 P Study 1 Significant reduction of POD (8% versus 24%) after 2 hours. <0.05 P Study 2: Significant reduction of POD (16% versus 24%) after 2 hours. <0.05 P Significant reduction of POD after 7 days.

Patients - Total of 505 Total - Age > 60 years - - ASA I-IV of 390 Total - Age > 68 years - - ASA I-III of 1134 Total - Age > 18 years - - ASA I-IV of 354 Total - Age > 65 years - - ASA I-III of 285 Total - Age > 60 years - - ASA I-IV of 88 Total - - Age 20-44 years - ASA I-II of 60 Total - Age > 18 years - - ASA I-IV

Evaluation POD present with the occurrence of new onset of deviant behaviour. POD daily determined by CAM/CAM- ICU and MMSE evaluation after 5 days. POD present with the occurrence of one of the following: 1) Illusions 2) Confusion 3) Agitation POD determined by CAM evaluation daily up to 5 days. POD determined by CAM/CAM-ICU evaluation daily up to 5 days. POD determined by CAM/CAM- ICU evaluation after 2 hours. POD determined by MMSE evaluation daily up to 7 days.

with aims, methodology, evaluating method, results and limitations with aims, methodology,

Table 2 Table delirium

Methodology

Dexmedetomidine from start CPB as continuous infusion at 0.25-0.6 µg/kg/h until 24 hours postoperative. Dexmedetomidine from induction as continuous infusion at 0.5 µg/kg/h until 2 hours postoperative. Dexmedetomidine from start CPB as continuous infusion at 0.2-0.6 µg/ kg/h until 24 hours postoperative. Study 1: Dexmedetomidine from induction (loading dose 1.0 µg/kg and continuous infusion at 0.2-0.7 µg/ kg/h) until the end of surgery. Study 2: Dexmedetomidine once from induction in a loading dose at 1.0 µg/kg without continuous infusion until the end of surgery. Dexmedetomidine from induction (loading dose 0.6 µg/kg and continuous infusion at 0.4 µg/ kg/h until the end of mechanical ventilation. Study 1: Dexmedetomidine from induction as continuous infusion at 0.4 µg/kg/h until the end of surgery. Study 2: Dexmedetomidine from induction as continuous infusion at 0.2 µg/kg/h until the end of surgery. Dexmedetomidine from induction (loading dose 1.0 μg/kg and continuous infusion at 0.5 μg/kg/h) until the end of surgery.

Aim Incidence of POD after cardiac surgery Incidence of POD after non-cardiac surgery. Incidence of and complications mortality after cardiac surgery Incidence of POD after extensive laparoscopic surgery. Incidence of POD after cardiac surgery. Incidence of POD after non- cardiac surgery Incidence of POD after cardiac surgery

study

Type Cohort study Retrospective RCT Double-blind Prospective Cohort study Retrospective RCT Double-blind Prospective RCT Double-blind Prospective RCT Double-blind Prospective RCT Double-blind Prospective

Title An overview of articles regarding the postoperative assessment

The effect of The effect Dexmedetomidine on in outcomes of cardiac surgery elderly patients Intraoperative infusion of Dex-medetomidine for prevention of postoperative delirium and cognitive dysfunction in elderly patients major elective undergoing a non-cardiac surgery: randomized clinical trial Perioperative Dexmedetomidine improves outcomes of cardiac surgery of the timing and The effect dose of Dexmedetomidine on postoperative delirium in elderly patients after laparoscopic major non- double A cardiac surgery: blind randomized controlled study. Impact of Dexmedetomidine on the incidence of delirium in elderly patients after a randomized cardiac surgery: controlled trial of Dexmedetomidine Effects Combined with Sufentanil on Postoperative Delirium in After General Patients Young Anesthesia Study on cerebral protective of Dexmedetomidine effect during anaesthesia in cardiac valve replacement surgery

Year 2016 2017 2013 2018 2017 2018 2017

Autor Cheng H., et al. Deiner S., et al. Ji F., et al. Lee C., et al. Li X., et al. Liu L., et al. Shu A., et al. RCT Randomized Controlled Trial, POD Postoperative Delirium, MMSE Mini-Mental State Examination, CAM Confusion Assessment Method, CAM-ICU Confusion Assessment Method in ICU, ASA American Society of Anaesthesiologist Score, American Society of ASA Assessment Method in ICU, Assessment Method, CAM-ICU Confusion POD Postoperative Delirium, MMSE Mini-Mental State Examination, CAM Confusion Trial, Randomized Controlled RCT CPB Cardiopulmonary Bypass

24-Hermans.indd 189 12/01/2021 12:26 190 j. hermans et al.

Limits

- Stopped early because of futility. - No loading dose - No fixed continuous infusion rate. . - Small study - Non-representative patients (young and only I-II). ASA - CAM-ICU evaluation conducted only 2 hours postoperative - No loading dose.

Results

: : No significant reduction of POD after 5 days. (12% versus 11%) = 0.94 P Study 1 Significant reduction of POD (9.5% versus 25%) after 5 days. = 0.04 P Study 2: Significant reduction of POD (18.4% versus 25%) after 5 days. = 0.04 P Study 1 Significant reduction of POD (8% versus 24%) after 2 hours. <0.05 P Study 2: Significant reduction of POD (16% versus 24%) after 2 hours. <0.05 P

Patients - Total of 390 Total - Age > 68 years - - ASA I-III of 354 Total - Age > 65 years - - ASA I-III of 88 Total - - Age 20-44 years - ASA I-II

Evaluation POD daily determined by CAM/CAM- ICU and MMSE evaluation after 5 days. POD determined by CAM evaluation daily up to 5 days. POD determined by CAM/CAM- ICU evaluation after 2 hours.

with aims, methodology, evaluating method, results and limitations with aims, methodology,

Table 3 Table

Methodology non-cardiac surgery non-cardiac

Dexmedetomidine from induction as continuous infusion at 0.5 µg/kg/h until 2 hours postoperative. Study 1: Dexmedetomidine from induction (loading dose 1.0 µg/kg and continuous infusion at 0.2-0.7 µg/kg/h) until the end of surgery. Study 2: Dexmedetomidine once from induction in a loading dose at 1.0 µg/kg without continuous infusion until the end of surgery. Study 1: Dexmedetomidine from induction as continuous infusion at 0.4 µg/kg/h until the end of surgery. Study 2: Dexmedetomidine from induction as continuous infusion at 0.2 µg/kg/h until the end of surgery.

Aim Incidence of POD after non-cardiac surgery. Incidence of POD after extensive laparoscopic surgery. Incidence of POD after non- cardiac surgery

study

Type RCT Double-blind Prospective RCT Double-blind Prospective RCT Double-blind Prospective

Title A brief overview of articles regarding the incidence POD after A Intraoperative infusion of Dexmedetomidine for prevention of postoperative delirium and cognitive dysfunction in elderly major elective patients undergoing a randomized non-cardiac surgery: clinical trial of the timing and dose The effect Dexmedetomidine on postoperative delirium in elderly patients after laparoscopic major non-cardiac double blind randomized A surgery: controlled study. of Dexmedetomidine Effects Combined with Sufentanil on Young Postoperative Delirium in After Patients Anesthesia General

Year 2017 2018 2018

Autor Deiner S., et al. Lee C., et al. Liu L., et al. RCT Randomized Controlled Trial, POD Postoperative Delirium, MMSE Mini-Mental State Examination, CAM Confusion Assessment Method, CAM-ICU Confusion Assessment Method in ICU, ASA American Society of Anaesthesiologist Score American Society of ASA Assessment Method in ICU, Assessment Method, CAM-ICU Confusion POD Postoperative Delirium, MMSE Mini-Mental State Examination, CAM Confusion Trial, Randomized Controlled RCT

24-Hermans.indd 190 12/01/2021 12:26 the effects of intraoperative use of dexmedetomidine 191

Limits -Cohort study - Non-standardized method for POD evaluation. - No loading dose - No fixed continuous infusion rate. -POD evaluation not as primary aim - Cohort study - Non-standardized method for POD evaluation. - No loading dose - No fixed continuous infusion rate. - Non-representative group (> 18 years) - Exceptional low incidence of delirium in general. - Small group - Non-representative group (> 18 years) - Dubious study (no power analysis, inconsistencies, …)

Results Significant reduction of POD (7% after 24 hours. versus 11%) < 0.0001 P Significant reduction of POD (5.4% versus 7.4%) after 24 hours. = 0.003 P No significant reduction of POD (4.9% versus 7.7%) after 5 days. = 0.341 P Significant reduction of POD after 7 days.

Patients - Total of 505 Total - Age > 60 years - - ASA I-IV of 1134 Total - Age > 18 years - - ASA I-IV of 285 Total - Age > 60 years - - ASA I-IV of 60 Total - Age > 18 years - - ASA I-IV

Evaluation POD present with the occurrence of new onset deviant behaviour. POD present with the occurrence of one the following: 1) Illusions 2) Confusion 3) Agitation POD determined by CAM/CAM- ICU evaluation daily up to 5 days. POD determined by MMSE evaluation daily up to 7 days. with aims, methodology, evaluating method, results and limitations with aims, methodology,

Table 4 Table

cardiac surgery cardiac

Methodology Dexmedetomidine from start CPB as continuous infusion at 0.25-0.6 µg/kg/h until 24 hours postoperative. Dexmedetomidine from start CPB as continuous infusion at 0.2-0.6 µg/kg/h until 24 hours postoperative. Dexmedetomidine from induction (loading dose 0.6 µg/kg and continuous infusion at 0.4 µg/kg/h until the end of mechanical ventilation. Dexmedetomidine from induction (loading dose 1.0 μg/kg and continuous infusion at 0.5 μg/kg/h) until the end of surgery.

Aim Incidence of POD after cardiac surgery Incidence of complications and mortality after cardiac surgery Incidence of POD after cardiac surgery. Incidence of POD after cardiac surgery

study

Type Cohort study Retrospective Cohort study Retrospective RCT Double-blind Prospective RCT Double-blind Prospective A brief overview of articles regarding the incidence POD after A

Title of Dexmedetomidine The effect in on outcomes of cardiac surgery elderly patients Perioperative Dexmedetomidine improves outcomes of cardiac surgery Impact of Dexmedetomidine on the incidence of delirium in elderly a patients after cardiac surgery: randomized controlled trial of Study on cerebral protective effect during anaesthesia Dexmedetomidine in cardiac valve replacement surgery

Year 2016 2013 2017 2017

RCT Randomized Controlled Trial, POD Postoperative Delirium, MMSE Mini-Mental State Examination, CAM Confusion Assessment Method, CAM-ICU Confusion Assessment Method in ICU, ASA American Society of Anaesthesiologist Score, American Society of ASA Assessment Method in ICU, Assessment Method, CAM-ICU Confusion POD Postoperative Delirium, MMSE Mini-Mental State Examination, CAM Confusion Trial, Randomized Controlled RCT CPB Cardiopulmonary Bypass. Autor Cheng H., et al. Ji F., Ji F., et al. Li X., et al. Shu A., et al.

24-Hermans.indd 191 12/01/2021 12:26 192 j. hermans et al. new onset of delirium ; 2) Inattention ; 3) Altered only ASA I and ASA II patients were included. consciousness ; 4) Disorganized thinking. To Four other separate studies of Li Y. et al. (45), Chen meet the delirium criteria, 1) and 2) are obligatory J. et al. (46), Ding L. et al. (47) and Xu H. et al. and 3) or 4) must be present (22, 42). The Mini- (48) concluded that during laparoscopic procedures Mental State Examination (MMSE) is the best administering dexmedetomidine intraoperative known and most frequently used indicator to assess significantly reduces POCD (respectively P-value POCD (17, 18). This test evaluates an evolution < 0.01, = 0.05, < 0.05 and < 0.05). These studies in the cognitive aspects of mental functions (f. e. displayed only minor limitations. Two other studies orientation, memory, attention, following written by Mohamed S. et al. (49) and Yilmaz F. et al. (50) and verbal commands, …) and is a simple quick could not confirm these results. These studies also way to identify individuals with dementia, cognitive showed some crucial limitations. Besides their impairment and depression (32). A postoperative small sample size (respectively n=50 and n=40), assessment can easily diagnose postoperative they did not apply a standardized diagnostic method cognitive dysfunction when compared to its to identify POCD. Furthermore, the study of Yilmaz preoperative conducted assessment. Another test F. et al. assessed POCD only two hours after minor which investigates POD is the Montreal Cognitive surgery which could explain the lack of significance Assessment (MoCA) criteria. This test examines between the control and study group. short term recall, visuospatial abilities, attention, concentration, working memory and orientation in Postoperative delirium time and space. It is confirmed to be a valid method to evaluate cognitive impairment and is more The other 7 studies investigated exclusively reliable than the MMSE-scale (18) (43). In addition, the incidence of POD. A large prospective study two other articles diagnosed the occurrence of a containing 390 patients by Deiner S. et al. (51) new onset of delirium when an acute emerge of was ceased prematurely because of futility. Interim one of the following four entities occurred : (1) analysis revealed no significant difference in Hallucinations/Illusions, (2) Acute confusion, (3) incidence of POD after 5 days. The author stated Agitation or (4) Altered behaviour. that these non-conclusive results were due to an The RCTs in this systematic review utilized intra-operative underdosing of dexmedetomidine the same study setup and diagnostic methods to as no loading dose was provided. A study written recognize and identify POD/POCD. This enables by Lee C. et al. (52) separated 354 patients into us to make a reliable comparison between the three groups. The first group only received a different study results and create a synopsis loading dose of dexmedetomidine whereas the towards a definitive conclusion for the clinical second group received both a loading dose and practitioner. Eventually this review will present a continuous infusion. Both groups showed a a final recommendation for the administration of substantial decrease in delirium in comparison to intraoperative dexmedetomidine regarding the the control group. Furthermore, the second group evaluation of POCD (Table 1) or POD (Table 2). established an additional significant reduction Furthermore, we tried to differentiate if there was a in POD in comparison to the group with only a difference in results regarding the incidence of POD loading dose. This study showed little limitations in non-cardiac (Table 3) versus cardiac surgery apart from the fact that the continuous infusion (Table 4) based on substantiated scientific evidence. of dexmedetomidine had no fixed rate (0.25-0.6 µg/kg/hour). A large retrospective cohort study Postoperative cognitive dysfunction performed by Ji F. et al. (53) in 2013 included 1134 patients (ASA I-IV, > 18 years) and investigated Of the 14 articles, 7 focused on the occurrence the impact of intraoperative dexmedetomidine on of POCD. A first study from Zhao W. et al (44) postoperative morbidity and mortality as primary investigated the effect of intraoperative administered aim. A secondary aim examined the incidence of dexmedetomidine on its cognitive dysfunction POD. Results confirmed a significant reduction in reducing assets after non-cardiac surgery. Significant POD (5.4% in the study group versus 7.4% in the results (p<0.009) were established nonetheless control group ; P = 0.003). Although the liberal use some limitations were found regarding the setup of of dexmedetomidine (0.2-0.6 µg/kg/hour) and the this study. Only a loading dose of dexmedetomidine limited duration of follow-up (only 24 hours), this was administered after induction and no further study is one of the leading studies because of its continuous infusion was provided. Additionally, large sample size. Limitations however remain the

24-Hermans.indd 192 12/01/2021 12:26 the effects of intraoperative use of dexmedetomidine 193 use of non-standardized assessment methods where into a group investigating the incidence of POD only POD was affirmative in the presence of the new after non-cardiac surgery (Table 3) and a group with onset of one of the following entities : illusions, studies which studied POD after cardiac surgery confusion or agitation. A study performed by Li X (Table 4). Three studies by Deiner S. et al. (51), Lee et al. (54) in 2017 revealed the same results (4.9% C. et al. (52) and Liu L. et al (57) contained a study versus 7.7%). This study however remained not set-up based on non-cardiac surgery. These studies significantly (P = 0.34) due to its smaller sample contained contradictory results as written above. size (n=285) and therefore remained underpowered. However, when we reviewed the 4 other remaining What stands out in this study, is the exceptional articles (Ji F. et al. (53), Li X. et al (54), Cheng H. et low general incidence of delirium as well in the al. (56) and Shu A. et al. (58)), which investigated study as in the control group. The general incidence the incidence of POD after cardiac surgery, we of postoperative delirium amounts to up to 30% discovered consistent results leading to a significant (3) (55) whereas in this study the control group reduction in POD. would knowledge an occurrence of only 8%. Conclusion This exceptional low incidence of POD could be explained by the fact that this study took place in Based on these findings, we can determine a centre in which postoperatively active delirium three different assumptions : preventing measures were taken. The author A reduction on the incidence of POCD by the however emphasized that this study would have use of dexmedetomidine intraoperatively could only been significant if the sample size would have been been be portrayed in 4 out of 7 applicable RCTs. larger. Another retrospective cohort study by Cheng Hence no unambiguous results could be revealed, H. et al. (56) in 2016 concerning 505 elderly ASA no recommendation could be made regarding this I-IV patients yielded significant results (7% versus subject for the clinical practitioner. 11% with a P < 0.0001). However, a few remarks Regarding the incidence of POD after non- concerning this study have to be discussed. First, cardiac surgery, there is no conclusive evidence to the postoperative assessment was conducted after support the use of intraoperative dexmedetomidine. 24 hours and had no further follow-up regarding Out of the 3 bigger RCTs regarding this subject POD. Second, the neuropsychological assessment only 2 of them achieved significant results. Because tests were non-standardized methods for evaluating of contradictory results it is only suspectable that POD. And finally, the dosage of the intraoperative dexmedetomidine reduces POD after non-cardiac administered dexmedetomidine had no fixed rate surgery but to date no exact recommendation can (0.2-0.6 µ g/kg/hour). A study from Liu L. et al be offered. Larger prospective clinical trials are (57), which only admitted young (20-44 years) required to create clarity. Currently there is a large ASA 1 to ASA 2 patients, applied a continuous study going on in China conducted by Wang B. et infusion intraoperatively without any loading dose al. (59) to investigate the incidence of POD after in comparison to a saline infusion. Two different non-cardiac surgery with intraoperative usage of rates where compared to a control group. Eventually dexmedetomidine in a loading and continuous both subgroups would register significant results infusion. Hence the number of the sample size (P<0.05) although its non-representative study will be substantial (n = 620), these findings will group. A smaller study presented by Shu A. et contribute to further decision-making towards the al. (58) in 2017 noted a significant reduction in management of intraoperative dexmedetomidine in postoperative delirium when dexmedetomidine non-cardiac surgery. Results are expected by the fall was applied during cardiac surgery. The validity of 2019. of this study remains very questionable. Their Regarding the reduction of POD after cardiac methodology was not elaborated thoroughly, surgery we can assume that the intraoperative contained inconsistencies and no power analysis application of dexmedetomidine in cardiac surgery was calculated. Significant results were described, is desirable. All the studies involving cardiac surgery nevertheless. portrayed a significant reduction in POD. Although A conclusion regarding the efficacy of some studies showed a few limitations, they all dexmedetomidine on the incidence of POD was not described comparable results. After this review, ambiguous. The majority of these studies portrayed one can conclude that it is advisable to administer a significant reduction in POD whereas other studies intraoperative dexmedetomidine during cardiac could not confirm these results. A sub analysis was surgery. This will result in a substantial reduction conducted regarding POD. We separated the studies of the incidence of POD leading to a noticeably

24-Hermans.indd 193 12/01/2021 12:26 194 j. hermans et al.

reduction in hospital-stay, hospital costs, morbidity in the postoperative setting. Taking into account and mortality. that the effect of postoperative administered dexmedetomidine in the ICU already has proven Discussion its delirium reducing assets, a significant result was obvious. These results eventually were generalised This systematic review provides a scienti- to the perioperative usage of dexmedetomidine. fically based summary about the clinical dis- Therefore, they concluded that the intraoperative cussion making towards the intraoperative use use of dexmedetomidine was advisable with no of dexmedetomidine. It is based on reliable data scientific evidence to support this theory. extracted from several RCTs and retrospective A meta-analysis from Wu M. et al. (61), cohort studies. We experienced however some which was published in the Journal of Clinical limitations. First, there was a big heterogenicity in Anesthesia in 2018 revealed the same remarks. the RCTs between the use of the terms “delirium” This study investigated the reduction of delirium and “cognitive dysfunction”. We used all the data postoperative after cardiac surgery by the use of from these studies as they all described comparable 10 RCTs. This metanalysis confirmed a beneficial results. These results however should be interpreted result towards POD when dexmedetomidine was with caution. Second, the postoperative screening used intraoperatively. However, these results were for POD or POCD was finalised using different based on 7 RCTs where dexmedetomidine was non-standardized neuropsychological tests, thereby administered solemnly postoperative. The eventual triggering the risk for clinical and statistical results remained significant and were generalised heterogeneity. The CAM and CAM-ICU remains for its intraoperative usage. Both meta-analyses the most valid assessment method to detect a new therefore suggested that dexmedetomidine reduces onset of POD. The MMSE-scale on the other hand, POD when used intraoperative. These results is the most frequently used scale to evaluate POCD. could have been biased by the continuation of However, studies suggest this method still contains dexmedetomidine in the postoperative setting on many limitations leading to biased results. Third, ICU. Generalizing these results to solemnly its most of the studies only consisted of a very limited intraoperative usage seems however unreliable. sample size making these studies less reliable. Consequently, we can conclude that further Fourth, the use and start of dexmedetomidine investigation is necessary. A future high-quality was similar in all studies however the continuous large-scaled clinical study would contain of a infusing rate were very inequal. The infusion rate particularly at-risk subpopulation (elderly, ASA I varied substantially (0.2-0.6 µg/kg/h) throughout to IV), a clear-cut methodology (dexmedetomidine the entire surgery. Therefore results could be in a loading dose and a continuous infusion at a biased by inefficient use of the study product. Fifth, fixed rate) and a daily postoperative POD/POCD the studies contained an incoherent population assessment using an official neuropsychological randomisation. Some studies included only ASA assessment test. I-II which would bias the results knowing that the overall incidence of POD or POCD is less likely in References this group. And last, the postoperative assessment was not identical regarding its follow-up. Some 1. Schenning K., Deiner S., et al. 2015. Postoperative delirium studies only evaluated the results once after 24 hours in the geriatric patient., Anesthesiol Clin. 33 : 505-516. 2. Inouye S., Robinson T., Blaum C., Busby-Whitehead J., whereas other studies had a daily check-up the first Boustani M., et al. 2015. Postoperative delirium in older week and did a monthly follow-up afterwards. adults: best practice statement from the American Geriatrics Two meta-analyses have already been published Society. J Am Coll Surg. 220 : 136-148. 3. Groen J., et al. 2012. Treatment of delirium following concerning my research. The first meta-analysis cardiac surgery. J. Card. Surg. 27 :589-593. conducted by Duan X. et al (60) was published in 4. Chan M., et al. 2013. BIS-guided anesthesia decreases the British Journal of Anaesthesia in 2018. This postoperative delirium and cognitive decline., J Neurosurg analysis investigated the perioperative efficacy of Anesthesiol. 25 :33-42. 5. Radtke F., Franck M., Lendner J., Kruger S., Wernecke dexmedetomidine on the incidence of POD. Based on K., Spies C., et al. 2013. Monitoring depth of anaesthesia 18 RCTs they concluded that as well intraoperative in a randomized trial decreases the rate of postoperative as postoperative usage of dexmedetomidine reduces delirium but not postoperative cognitive dysfunction. Br J Anaesth. 110, Suppl. :i98-i105. POD significantly. However, when we take a deeper 6. Leung J., et al. 2013. Does preoperative risk for delirium look into this analysis, we noticed that 14 out of moderate the effects of postoperative pain and opiate use on the 18 RCTs only applicated dexmedetomidine postoperative delirium? Am J GeriatrPsych. 21 :946-956.

24-Hermans.indd 194 12/01/2021 12:26 the effects of intraoperative use of dexmedetomidine 195 7. Campbell N., et al. 2009.Pharmacological Management of 25. Tan J., Ho K., et al. 2010. Use of dexmedetomidine as a Delirium in Hospitalized Adults – A Systematic Evidence sedative and analgesic agent in critically ill adult patients : Review., J Gen Intern Med. 24: 848-853. a meta-analysis. Intensive Care Med. 36:926-39. 8. Moller J., Cluitmans P., Rasmussen L., et al. 1998. Long- 26. Mantz J., Josserand J., Hamada S., et al. 2011. Dexmede- term postoperative cognitive dysfunction in the elderly tomidine: new insights. Eur J Anaesthesiol. 28:3-6. ISPOCD1 study: International Study of Post-Operative 27. Mo Y., Zimmermann A.. et al. 2013. Role of dexmede- Cognitive Dysfunction., Lancet. 351 :857-861. tomidine for the prevention and treatment of delirium in 9. Monk T., Saini V., Weldon B., Sigl J., et al. 2005. Anesthetic intensive care unit patients. Ann Pharmacother. 47:869-76. management and one-year mortality after noncardiac 28. Barr J., Fraser G., Puntillo K., Ely E., Gélinas C., Dasta J., surgery. Anesth Analg. 100 :4-10. et al. 2013. Clinical practice guidelines for the management 10. Brown C., Laflam A., Max L., et al. 2016.The impact of of pain, agitation, and delirium in adult patients in the delirium after cardiac surgical procedures on postoperative intensive care unit. Crit Care Med. 41:263-306. resource use. Ann Thorac Surg. 101 : 1663-1669. 29. Chen K., Lu Z., Xin Y., Cai Y., Chen Y., Pan S., et al. 2015. 11. Mo Y., Zimmermann A., et al. 2013. Role of dexmedeto- Alpha-2 agonists for long-term sedation during mechanical midine for the prevention and treatment of delirium in in- ventilation in critically ill patients. Cochrane Database Syst tensive careunit patients. Ann Pharmacother. 47: 869-765. Rev. 1 12. Paris A., Tonner P., et al. 2005. Dexmedetomidine in 30. Venn R., Bradshaw C., Spencer R., Brealey D., Caudwell anaesthesia. Curr, Opin Anaesthesiol. 18: 412-418. E., Naughton C., et al. 1999. Preliminary UK experience of 13. “The DSM-5 criteria, level of arousal and delirium diag- dexmedetomidine, a novel agent for postoperative sedation nosis: inclusiveness is safer”, BMC Medicine 201412:141, in the intensive care unit. Anaesthesia. 54:1136-42. https://bmcmedicine.biomedcentral.com/ articles/10.1186/ 31. Szumita P., Baroletti S., Anger K., Wechsler M., et al. 2007. s12916-014-0141-2. Sedation and analgesia in the intensive care unit: evaluating 14. Ely E., Shintani A., Truman B., Speroff T. et al. 2004. the role of dexmedetomidine. Am J Health Syst Pharm Delirium as a predictor of mortality in mechanically venti- AJHP. 64:37-44. lated patients in the intensive care unit. JAMA. 291:1753- 32. Folstein M., Folstein S., McHugh P., et al 1975. Mini- 1762. mental state A practical method for grading the cognitive 15. Lin S., Liu C., Wang C., Lin H., et al. 2004.The impact state of patients for the clinician. J. gsychiaf. Res.12: 189- of delirium on the survival of mechanically ventilated 198. patients. Crit Care Med. 32 :2254-2259. 33. Maldonado J.R., Wysong A., Van der Starre P.J., Block 16. Veiga D., Luis C., Parente D., Fernandes V., Botelho M., T., Miller C., Reitz B.A., et al. 2009. Dexmedetomidine Santos P., et al. 2012. Postoperative delirium in intensive and the reduction of postoperative delirium after cardiac care patients: risk factors and outcome. Rev Bras Anestesiol. surgery. Psychosomatics. 50:206-217. 62 :469-83. 34. Geng J., Qian J., Cheng H., Ji F., Liu H., et al. 2016. The 17. Bickel H., Gradinger R., Kochs E., Forstl H. et al 2008. High influence of perioperative dexmedetomidine on patients risk of cognitive and functional decline after postoperative undergoing cardiac surgery: a meta-analysis. PLoS One. 11 delirium. A three-year prospective study. Dement Geriatr 35. Constantin J., Momon A., Mantz J., et al. 2016. Efficacy Cogn Disord. 26:26 -31. and safety of sedation with dexmedetomidine in critical 18. Gluhm S., Goldstein J. et al. 2013. Cognitive Performance care patients: a meta-analysis of randomized controlled on the Mini-Mental State Examination and the Montreal trials. Anaesth Crit Care Pain Med, 35 : 7-15. Cognitive Assessment Across the Healthy Adult Lifespan, 36. Fan H., Zhao Y., Sun Y., Ye J., Chen J., Zhu J., et al. 2017. Cogn Behav Neurol. 26: 1-5. Dexmedetomidine based sedation for post-surgery critically 19. Moller J., Cluitmans P., Rasmussen L. et al. 1998. ill adults: a meta-analysis of randomized controlled trials. ISPOCD Investigators. Long-term postoperative cognitive Iran J Public Health. 46: 1611-1622. dysfunction in the elderly ISPOCD1 study: International 37. Liu X., Xie G., Zhang K., et al. 2017. Dexmedetomidine vs Study of Post-Operative Cognitive Dysfunction. Lancet. propofol sedation reduces delirium in patients after cardiac 351 : 857-861. surgery: a meta-analysis with trial sequential analysis of 20. Monk T., Saini V., Weldon B., Sigl J. et al. 2005. Anesthetic randomized controlled trials. J Crit Care, 38: 190-196. management and one-year mortality after noncardiac 38. Constantin J., Momon A., Mantz J., et al. 2016. Efficacy surgery. Anesth Analg.100:4-10. and safety of sedation with dexmedetomidine in critical 21. Brown C., Laflam A., Max L., et al. 2016. The impact of care patients: a meta-analysis of randomized controlled delirium after cardiac surgical procedures on postoperative trials. Anaesth Crit Care Pain Med. 35:7-15. resource use, Ann Thorac Surg. 101 :1663-1669. 39. Pasin L., Landoni G., Nardelli P., Belletti A., Di Prima 22. Deiner S., Silverstein J., et al. 2009. Postoperative delirium AL., Taddeo D., et al. 2014. Dexmedetomidine reduces and cognitive dysfunction. Br J Anaesth.103 (suppl 1): the risk of delirium, agitation and confusion in critically Ill i41-i46. patients: a meta-analysis of randomized controlled trials. J 23. Chen X., Rui-lin H., Wang C., Wei L., Ma Y., Han Y., et Cardiothorac Vasc Anesth. 28(6):1459-66. al.2013. Effect of Dexmedetomidine on cognitive dys- 40. Xia Z., Chen S., Yao X., Xie C., Wen S., Liu K.. et al. 2013. function after off-pump coronary artery bypass grafting. Clinical benefits of dexmedetomidine versus propofol Chin J Anesthesiol. 33(5): 548-550 in adult intensive care unit patients: a meta-analysis of 24. Flükiger J., Hollinger B., Meier V., Tontisch J., et al. randomized clinical trials. J Surg Res.185:833-843. 2018. Dexmedetomidine in prevention and treatment of 41. Sun Y., et al. 2017. Effects of dexmedetomidine on postoperative and intensive care unit delirium: a systematic emergence delirium in pediatric cardiac surgery., Minerva review and meta-analysis. Ann. Intensive Care 8:92. Pediatr. 69:165-173.

24-Hermans.indd 195 12/01/2021 12:26 196 j. hermans et al. 42. Inouye S., Van Dyck C., Alessi C., Balkin S., Siegal A., 52. Lee C., Lee CH., Lee G., Lee M., Hwang J., et al. 2018. Horwitz R., et al. 1990. Clarifying confusion: the confusion The effect of the timing and dose of dexmedetomidine on assessment method. A new method for detection of postoperative delirium in elderly patients after laparoscopic delirium. Ann Intern Med. 113 : 941-948. major non-cardiac surgery: A double blind randomized 43. Pendlebury S., Bull L., Mehta Z., Rothwell P., et al. 2012. controlled study. J Clin Anesth. 47:27-32. 53. Ji F., Li Z., Nguyen H., et al, Perioperative dexmedetomidine MoCA, ACE-R, and MMSE Versus the National Institute improves outcomes of cardiac surgery. Circulation., of Neurological Disorders and Stroke-Canadian Stroke 2013:127:1576-1584. Network Vascular Cognitive Impairment Harmonization 54. Li X., Yang J., Nie X.L., et al. 2017. Impact of dexmede- Standards Neuropsychological Battery After TIA and tomidine on the incidence of delirium in elderly patients Stroke, Stroke after cardiac surgery: a randomized controlled trial. PLoS 44. Zhao W., Gao C., Cui L., et al. 2018. Effect of Dexmede- One. 12 tomidine Hydrochloride on early cognitive function in 55. Bakker R., Osse R., Tulen J., Kappetein A., et al. 2012. postoperative elderly patients J Adv Med Sci. 1, Issue 02. Preoperative and operative predictors of delirium after 45. Li Y., He R., Chen S., Qu Y., et al. 2015. Effect of cardiac surgery in elderly patients. Eur J Cardiothorac Surg. dexmedetomidine on early postoperative cognitive dys- 41:544-549. function and peri-operative inflammation in elderly patients 56. Cheng H., et al. 2016. The Effect of Dexmedetomidine undergoing laparoscopic cholecystectomy. Exp Ther Med. on Outcomes of Cardiac Surgery in Elderly Patients. J 10(5):1635-1642. Cardiothorac Vasc Anesth. 30 :1502-1508. 46. Chen J., Yan J., Han X., et al. 2013. Dexmedetomidine 57. Liu L., Yuan Q., Wang Y., Gao W., Hou J., Wu Y., Zhao may benefit cognitive function after laparoscopic B., Xia Z., et al. 2018. Effects of Dexmedetomidine cholecystectomy in elderly patients. Exp Ther Med. 5:489- Combined with Sufentanil on Postoperative Delirium in 494. Young Patients After General Anesthesia. Med Sci Monit. 47. Ding L, Zhang H, Mi W, Wang T, He Y, Zhang X, Ma X, 10 :8925-8932. Li H., et al. 2015. ffects of dexmedetomidine on anesthesia 58. Shu A., et al. 2017. Study on the cerebral protective effects of recovery period and postoperative cognitive function dexmedetomidine during anesthesia for surgical correction of patients after robot-assisted laparoscopic radical of congenital heart disease with cardiopulmonary bypass. cystectomy. Int J Clin Exp Med. 8 :11388-11395. Int J Clin Exp Med 10 :16562-16568. 48. Xu G., et al. 2017. Effect of dexmedetomidine-induced 59. Wang B., et al. 2018. Impact of dexmedetomidine infusion anesthesia on the postoperative cognitive function of elder during general anaesthesia on incidence of postoperative patients after laparoscopic ovarian cystectomy. Saudi J Biol delirium in elderly patients after major non-cardiac surgery: Sci. 24: 1771-1775. study protocol of a randomised, double-blinded and 49. Mohamed S., et al. 2014. The effect of Dexmedetomidine placebo-controlled trial. BMJ Open 8 on the incidence of postoperative cognitive dysfunction in 60. Duan X., Coburn M. et al. 2018. Efficacy of perioperative elderly patients after prolonged abdominal surgery. Egypt J dexmedetomidine on postoperative delirium: systematic Anesth. 30 : 331-338. review and meta-analysis with trial sequential analysis of 50. Yılmaz F., Günüşen İ., Çertuğ A., Önen Sertöz Ö., et al. randomised controlled trials. Br J Anaesth. 121:384-397 2012. The effect of dexmedetomidine on cognitive function 61. Wu M., Liang Y. et al. 2018. Perioperative dexmedetomidine and anxiety in middle-ear surgery. Cilt 51, Sayı 1, Sayfa(lar) reduces delirium after cardiac surgery: A meta-analysis 001-008. randomized controlled trials. J Clin Anesth 50 : 33-42. 51. Deiner S., Luo X., Lin H.M., et al. 2017. Intraoperative 62. Higgins J., et al. 2011. Cochrane Handbook for systematic infusion of dexmedetomidine for prevention of post- reviews of interventions : The Cochrane Collaboration, operative delirium and cognitive dysfunction in elderly www.cochrane-handbook.org patients undergoing major elective noncardiac surgery : a randomized clinical trial. JAMA Surg.152.

24-Hermans.indd 196 12/01/2021 12:26 (Acta Anaesth. Belg., 2020, 71, 197-201) Epidural ropivacaine infusion as part of a multimodal postoperative pain treatment following thoracolumbar spinal fusion surgery, a randomised controlled trial

M. Tosi (*), J.T. Bhutia (*), A. Scholliers (*), S.P.M.J. Raeymaeckers (**), P. Forget (****), M. Matic (*), D. Van Schaik (***), S. Vanden Eynde Raeymaeckers (*), T. Opsomer (*), P. Antunes Da Silva (*), J. Poelaert (*)

Abstract : Background : Pain management after Introduction posterior thoracolumbar interbody fusion surgery remains a challenging problem. Epidural analgesia with Many different therapeutic interventions local anesthetics as an adjuvant treatment might prove thought to reduce postoperative pain after posterior beneficial. thoracolumbar spinal fusion surgery have been Objective : To evaluate the effects of continuous epidural investigated (1). A multimodal strategy for pain analgesia with a ropivacaine solution as an adjuvant management can consist of a combination of treatment after posterior thoracolumbar interbody fusion. Study design : A prospective double-blind, randomized, different interventions, but even so pain management placebo-controlled clinical trial. Setting : University after posterior thoracolumbar spinal fusion surgery

Hospital Brussels, a tertiary academic health science remains challenging (2, 3).Better pain management centre. has been associated with less incidence of chronic Methods : Thirty-three patients undergoing spinal fusion back pain, less postoperative complications, better surgery were randomized into two groups. One group surgical outcome, faster time to mobilisation was administered a continuous epidural infusion of a and shorter length of stay, thereby also reducing 0.9% saline solution. The other group was administered financial cost (4, 5, 6). a continuous epidural infusion of a 0.2% ropivacaine Traditionally opioids are the agents of choice solution. The primary outcome measure was the level for pain management (6, 7). The role of opioids has of pain experienced by the patients, which was assessed using the visual analogue scale (VAS). Secondary outcomes were time to mobilisation, total length of stay Maurizio Tosi, MD ; Jigme Tshering Bhutia, MD ; Annelies and opioid consumption on patient demand. Scholliers, MD ; Steven Paul Marc Jan Raeymaeckers, Results : The mean VAS score postoperatively the day MD ; Patrice Forget, MD, PhD ; Milica Matic, MD ; of surgery was scored as 3,5 for the patients receiving David Van Schaik, MD ; Saskia Vanden Eynde, MD ; the ropivacaine infusion. The placebo group scored Tine Opsomer, MD ; Patricia Antunes Da Silva, MD ; Jan Poelaert, MD, PhD 5 on the VAS scale, thus suggesting a beneficial effect (*) Department of Anesthesiology and Perioperative Medicine, of ropivacaine infusion (p-value 0.02). The mean VAS UZ Brussel, B-1090 Jette. score at post-op day 1 of surgery differed between 2 (**) Department of Radiology, UZ Brussel, B-1090 Jette. for the ropivacaine group and 2,8 for the placebo group (***) Department of Orthopedic Surgery and Traumatology, (p-value 0.05). Length of stay was shorter for patients UZ Brussel, B-1090 JETTE (****) Department of Anesthesiology, University of Aberdeen, who had received ropivacaine infusion (mean difference Ireland. 1,8 days ; p-value 0,02). No other significant differences Authors contribution : Maurizio Tosi and Jigme Bhutia were withheld. designed the study. Maurizio Tosi and Jigme Bhutia Conclusion : The use of a continuous epidural recruited the patients and collected the data. Maurizio Tosi ropivacaine infusion after posterior thoracolumbar spinal performed the analyses. Jigme Bhutia interpretated the data. All the authors contributed to scientific discussion, the fusion surgery could improve postoperative VAS scores draft redaction and approved the final version. and may also result in faster hospital discharge rate. Trial Corresponding author : Dr M. Tosi, Department of Anesthe- registration : registered and approved in the EudraCT siology and Perioperative Medicine, UZ Brussel, Avenue du registry (2014-004713-91). Laerbeek 101, B-1090 Jette. E-mail : [email protected]

Keywords : Epidural anesthesia ; spinal surgery ; back Paper submitted on Oct 23, 2019 and accepted on May 26, surgery ; spinal fusion ; pain medicine ; orthopedic sur- 2020 gery ; ropivacaine. Conflict of interest : None

25-Tosi.indd 197 12/01/2021 12:26 198 m. tosy et al. been investigated thoroughly as well as their many included the inability to provide informed consent, different administration routes : oral, subcutaneous, minors (patients under the age of 18 years) and a intramuscular, intravenous, epidural and intrathecal history or suspicion of allergy to local anesthetics. (8). However, use of opioids is associated with Randomization of the subjects into two groups many side effects including respiratory depression, was performed by blocks. A patient-list sorted by date urinary retention, hypotension, nausea and cognitive of scheduled surgery was used to associate subjects impairment, thus limiting their usefulness (6). When a unique number. Those numbers were randomized used via an epidural route, these side effects can be using an online form (www.randomization.org), to partially overcome (8, 9). generate two groups of equal size. Patients in group Epidural analgesia, when compared to other 1 were given placebo (saline), patients in group 2 modalities in postoperative pain control, provides were administered ropivacaine. The randomized list a greater potency and efficacy of the administered was kept confidential by the principal investigator drugs. As a result, drug load can be limited, which (MT). Allocation of the patient to either group diminishes the severity and incidence of side effects was concealed from surgeons, anesthesiologists, (6, 8). participants and all other personnel involved in data The use of epidural local anesthetics is a well- collection. An epidural catheter was placed during established practice, however the advantages of this the procedure, an infusion system was installed to approach in the setting of spinal surgery have only enable the continuous infusion of a clear fluid (NaCl been suggested over the course of the last few years 0.9% for group 1, ropivacaine 0.2% for group 2). (2, 6, 10). The main action of local anesthetics is Both these products were stored in the operating to block neural transmission of pain signals. Used theatre and prepared by the principal investigator via the epidural route local anesthetics may prevent during the surgical intervention. The infusion fluid central pain sensitization, a key factor in chronic was introduced in elastomeric pumps and labelled pain (6). For this study we chose to examine the as study medication. Infusion fluid for bolus effect of ropivacaine, since this drug permits the injection was prepared in syringes, also labelled as preservation of motor function and has limited study medication. In this manner the products were cardiovascular and neurological toxicity (11). handed over to the attending anesthesiologist so as The primary objective of this study is to to ensure full blindness. examine the effect of a continuous epidural All patients received standard preoperative infusion of ropivacaine on postoperative pain after care as per protocol used in our hospital. Patients spinal surgery, assessed by a visual analogue scale were premedicated with transdermal fentanyl (VAS). Secondary objectives include time to first (patches, 25 mcg per hour) and oral administration mobilisation, total length of stay and the use of of gabapentin (300 mg) the evening before the opioids as rescue medication on patient demand. intervention. Patients already on opioids, saw their daily dose added in fentanyl patches. A first Patients and methods VAS score was determined the evening before surgery as reference. On the day of surgery oral A double-blind, randomised, placebo-control- administration of gabapentin (300 mg) was repeated led, clinical trial was set up at the University and acetaminophen (1000 mg) was added. Hospital Brussels (Brussels), a tertiary academic General anesthesia was induced with propofol health science centre. The study protocol was and sufentanyl. Muscle relaxation was achieved approved by the local ethics committee (Medical witch rocuronium, unless the patient presented Ethics Committee UZ Brussel - VUB, Laarbeeklaan chronic kidney or liver dysfunction ; in these cases 101, 1090 Brussels ; head of the ethics committee : cisatracurium was used. The trachea was intubated A. Van Steirteghem ; date of approval 4/12/2014) and ventilation was controlled at the discretion and registered and approved in the EudraCT registry of the anesthesiologist. General anesthesia was (2014-004713-91). Forty patients were included, maintained with sevoflurane. Perioperative pain they all provided written informed consent prior to was managed with ketamine infusion (bolus of 1 mg inclusion. per kg and maintenance of 0.2 mg per kg per hour) Patients who were scheduled to undergo and sufentanyl also at the discretion of the attending thoracic or lumbar posterior interbody fusion surgery anesthesiologist. were eligible for the study, provided they be adults At the end of the procedure, the epidural fitting the American Society of Anesthesiologists catheter (Perifix® Epidural Anesthesia Catheter, B. physical status (ASA) I-III. The exclusion criteria Braun) was introduced by the spine surgeon. The

25-Tosi.indd 198 12/01/2021 12:26 epidural ropivacaine in spinal fusion surgery 199 catheter was placed four to five cm in the epidural space, in the centre of the surgical field so its placement could be confirmed visually. Afterwards a bolus of 10 ml NaCl 0.9% or ropivacaine 0.2% was administered. Continuous infusion was started at the Post Anesthesia Care Unit (PACU) through an elastomeric pump at a basal rate of 7 ml per hour for 72 hours after surgery. At the PACU, pain scores were determined every 10 minutes. Piritramide was administered, 2 mg per bolus, until a VAS score ≤3 was achieved. At the surgical ward, VAS scores were determined every 8 hours and oxycodone was administered on patient demand in case of pain. Standard Fig. 1. — Consort diagram. n = number of patients, group 1 = postoperative pain treatment consisted of 1000 mg placebo group, group 2 = ropivacaine group. acetaminophen every 6 hours, and gabapentin (300 mg) every 12 hours. Side effects were recorded if were excluded because of refusal to participate or any and when they occurred. because they did not meet the inclusion criteria. We aimed to detect a 40% reduction in VAS The included patients were monitored in hospital score in the treatment group as compared with the until they were discharged. Seventeen patients were control group. Assuming a standard deviation of 2, a randomized into group 1 and 16 into group 2. 3 minimum of 16 patients per group would be required. patients from group 1 had to be removed from the We set type I error α = 0.05 (two-sided) and type II error study. Two patients were excluded because they did b = 0.2. For statistical analysis SPSS Statistics® not receive the correct premedication in accordance version 23 was used. The normality of the distribu- with the study protocol. A third patient was excluded tion was assessed using the Kolmogorov-Smirnov because of a dura mater leak that occurred during test. The student T-test was used for analysing the surgery. The leak could not be properly sealed differences between the two groups. The chi-square during surgery so the desired epidural anesthesia test was used for analysis regarding the occurrence could not be guaranteed. In total then, 14 patients of side effects. P values <0.05 were considered to were assigned as treatment group and 16 patients as be statistically significant. No post-hoc correction control group. was used. There were no significant differences be- This report is redacted according to the tween the two groups concerning age, weight, Consolidated Standards of Reporting Trials height, gender, BMI (Body Mass Index), ASA (CONSORT) guidelines (consort-statement.org). physical status, duration of the surgical procedure, preoperative VAS score or preoperative mean daily Results morphine consumption (Table 1). The mean VAS scores are detailed in Table Since our power analysis indicated we needed 2. The VAS scores at day 0 and day 1 were two groups of at least 16 patients, we proposed to significantly lower in group 2 (p-value 0.02 and include 40 patients between December 2014 and p-value 0.05). Intraoperative opioid administration December 2015 in the study (Fig. 1). Seven patients and supplemental opioid consumption on patient

Table 1 Patient characteristics Group 1 Group 2 Difference (95% CI) P Average age, years 62 ±16 55 ± 18 -6.8 (-19.2 to 5.6) NS Weight, kg 70 ± 12.7 79.8 ± 18.6 9.8 (-2.2 to 21.7) NS Height, cm 165.9 ± 7.7 171.7 ± 11.7 5.3 (-2.2 to 12.8) NS BMI, kg/cm2 26.5 ± 3.7 24.9 ± 10.2 1.7 (-1.8 to 5.3) NS ASA I/II/III, n 3/9/2 4/10/2 -0.1 (-0.5 to 0.4) NS Preoperative VAS 3 ± 3.1 3.4 ± 2.8 0.4 (-1.8 to 2.6) NS PMDMa 0 0 / NS a Preoperative mean daily morphine, mg.

25-Tosi.indd 199 12/01/2021 12:26 200 m. tosy et al. Table 2 VAS score (mean ± SD) at each collection time

Time Group 1 Group 2 Difference (95% CI) P First PACU 7.2 ± 3.7 4.5 ± 3.9 2.7 (-1.8 to 2.6) NS Mean PACU 5.7 ± 0.9 3.6 ± 3 1.5 (-1.5 to 5.6) NS Day 0 5 ± 2.9 3.5 ± 1.5 1.6 (-0.1 to 3.6) 0.02 Day 1 2.8 ± 1.5 2 ± 0.7 0.8 (0 to 1.7) 0.05 Day 2 2.1 ± 0.8 1.7 ± 1.2 0.4 (-0.4 to 1.2) NS Day 0 = day of surgery.

Table 3 Supplemental opioid consumption (mean ± SD) at each time

Time Group 1 Group 2 Difference (95% CI) P OR 41.4 ± 14.2 55.5 ± 23.5 14.1 (-1 to 29.1) NS PACU 14.8 ± 10.9 9 ± 7.5 5.7 (-1.2 to 12.6) NS Day 0 10.5 ± 11.4 6.9 ± 7.3 3.6 (-3.4 to 10.6) NS Day 1 11.4 ± 10.5 6.8 ± 7.9 4.5 (-2.3 to 11.4) NS Day 2 11.8 ± 15.9 6 ± 7.5 5.8 (-3.3 to 14.9) NS OR: mg sufentanyl. PACU: mg piritramide. Otherwise: mg oxycodone.

Table 4 Day of mobilisation, length of stay (mean ± SD)

Parameter Group 1 Group 2 Difference (95% CI) P Mobilisation, days 1.5 ± 1 0.8 ± 0.8 0.8 (0 to 1.4) NS Hospitalisation, days 9.6 ± 5.6 6.7 ± 3.9 1.8 (-0.7 to 6.5) 0.02

Table 5 Occurrence of side effects

Group 1 vs. Group 2 PACU Day 0 Day 1 Day 2 Nausea/vomiting, % 14.3 vs. 6.3b 0 vs. 6.3b 0 vs. 6.3b 0 vs. 0b Abdominal discomfort, % 0 vs. 0b 7.1 vs. 0b 7.1 vs. 0b 7.1 vs. 0b Constipation, % 14.3 vs. 6.3b 21.4 vs. 25b 28.6 vs. 21.3b 21.4 vs. 12.5b Pruritus, % 0 vs. 0b 0 vs. 0b 0 vs. 0b 0 vs. 0b Respiratory depression, % 0 vs. 0b 7.1 vs. 0b 0 vs. 0b 0 vs. 0b Paresthesia, % 0 vs. 18.9b 0 vs. 18.9b 0 vs. 12.5b 0 vs. 12.5b Motor blockade, % 0 vs. 0b 0 vs. 0b 0 vs. 0b 0 vs. 0b b P-value non-significant,n = number of patients.

demand is detailed in Table 3. Opioid use is 1 in the ropivacaine group when compared to the expressed in mg sufentanyl during the OR, mg placebo group. Other mean VAS scores did not piritramide at the PACU and mg oxycodone at the differ significantly. The lack of significance for ward. the other VAS scores could be attributed to an The mean time to mobilisation and mean inadequate sample size or by our means of data length of stay are detailed in Table 4. Patients in the collection. Gottschalk et al. (2) also compared ropivacaine group presented a shorter length of stay. epidural ropivacaine 0.1% with placebo and found The occurrence of side effects is detailed in Table 5. significant reduction of more postoperative VAS There were no significant differences between the scores. This study measured VAS scores at fixed two groups concerning side effects. time intervals, which we could not achieve. For our study we used a ropivacaine 0.2% Discussion solution, whereas Blumenthal et al. (12) used a In this trial, significantly lower VAS scores solution of 0.3% ropivacaine. Their choice was were observed postoperatively at day 0 and day based on a single pilot study, which demonstrated

25-Tosi.indd 200 12/01/2021 12:26 epidural ropivacaine in spinal fusion surgery 201 superior pain control with a 0.3% ropivacaine resulted in significantly faster discharge rates in solution compared to a 0.2% solution. However, the ropivacaine group. Based upon these results, a more concentrated solution might also make the a beneficial effect of the use of epidural local patient more prone to motor blockade, which is an anesthetics post thoracolumbar spinal fusion surgery undesirable effect in spinal surgery. can be suggested. In the study of Blumenthal et al. (12) the authors investigated the effect of epidural infusion of ropivacaine in comparison to intravenous morphine References as pain management after scoliosis surgery. The VAS scores in this study were significantly in favour 1. Mathiesen O, Dahl B, Thomsen BA, Kitter B, Sonne N, of the ropivacaine group. Dahl JB and Kehlet H. 2013. A comprehensive multimodal pain treatment reduces opioid consumption after multilevel Opioid consumption did not differ signifi- spine surgery. Eur Spine J. 22(9):2089-96. cantly, although patients in the treatment group 2. Vineyard JC, Toohey JS, Neidre A, Fogel G, Joyner seemed less likely to ask for additional opioid admi- R. 2014. Evaluation of a single-dose, extended-release epidural morphine formulation for pain control after lumbar nistration. This lack of significance is in accordance spine surgery. J Surg Orthop Adv. 23(1):9-12. with a study conducted by Choi et al. (10) who 3. Gottschalk A, Freitag M, Tank S, Burmeister MA, Kreil used a combination of epidural bupivacaine and S and Kothe R, et. al. 2004. Quality of postoperative pain using an intraoperatively placed epidural catheter after hydromorphone administration after lumbar spinal major lumbar spinal surgery. Anesthesiology. 101(1):175- fusion. The authors suggested that the observed wide 80. confidence intervals may fail to detect a significant 4. Shipton EA and Tait B. 2005. Flagging the pain : preventing difference. Studies with larger sample sizes than the burden of chronic pain by identifying and treating risk factors in acute pain. Eur J Anaesthesiol. 22(6):405-12. this study or our own may resolve this issue. 5. Hollmann MW, Strumper D and Durieux ME. 2004. The poor The availability of rescue medication is a man’s epidural : systemic local anesthetics for improving possible confounding variable. Any medication had postoperative outcomes. Med Hypotheses. 63:386-9. 6. Kang H, Jung HJ, Lee JS, Yang JJ, Shin HY and Song to be requested by the patient. Patients may be holding KS. 2013. Early postoperative analgesic effects of a back to ask for rescue medication when needed. single epidural injection of ropivacaine administered The use of patient controlled analgesia (PCA) is a preoperatively in posterior lumbar interbody spinal arthrodesis : a pilot randomized controlled trial. J Bone possible solution for this problem. Gottschalk et al. Joint Surg Am. 95(5):393-9. (2) used this approach and provided their patients 7. Jirarattanaphochai K, Jung S, Thienthong S, Krisana- with an intravenous PCA pump. This way they prakornkit W and Sumananont C. 2007. Peridural Methyl- could demonstrate lower opioid requirements in the prednisolone and Wound Infiltration With Bupivacaine for Postoperative Pain Control After Posterior Lumbar ropivacaine group. Spine Surgery : A Randomized Double-Blinded Placebo- The time to mobilisation was not significantly Controlled Trial. Spine. 32(6):609-16. shorter in the ropivacaine group. This is in 8. Yukawa Y, Kato F, Ito K, Terashima T and Horie Y. 2005. A Prospective Randomized Study of Preemptive Analgesia accordance with other studies in the literature. for Postoperative Pain in the Patients Undergoing Posterior Fisher et al. (13) conducted a study to compare Lumbar Interbody Fusion : Continuous Subcutaneous patient controlled intravenous analgesia (PCIA) to Morphine, Continuous Epidural Morphine, and Diclofenac Sodium. Spine. 30(21):2347-61. PCEA after lumbar spinal fusion. Their epidural 9. Schmidek HH and Cutler SG. 1983. Epidural morphine for solution contained a combination of fentanyl, control of pain after spinal surgery : a preliminary report. bupivacaine and epinephrine. They also found no Neurosurgery. 13(1):37-9. difference in time to mobilisation between groups. 10. Choi S, Rampersaud YR, Chan VW, Persaud O, Koshkin A, Tumber P and Brull R. 2014. The addition of epidural local The major difference with our study is that we use a anesthetic to systemic multimodal analgesia following placebo (epidural infusion of NaCl 0.9%) whereas lumbar spinal fusion : a randomized controlled trial. Can J the control group in the study of Fisher et al uses Anaesth. 61(4):330-9. 11. Zink W and Graf BM. 2004. Benefit-risk assessment of PCIA. Choi et al. (10) failed to detect a difference ropivacaine in the management of postoperative pain. Drug in mobilisation as well. Saf. 27(14):1093-114. In summary, a continuous epidural analgesia 12. Blumenthal S, Min K, Nadig M and Borgeat A. 2005. Double Epidural Catheter with Ropivacaine versus Intravenous infusion with ropivacaine 0.2% after posterior Morphine : A Comparison for Postoperative Analgesia after thoracolumbar spinal fusion resulted in significantly Scoliosis Correction Surgery. Anesthesiology. 102(1):175- lower pain scores at day 0 and day 1. It also 80.

25-Tosi.indd 201 12/01/2021 12:26

(Acta Anaesth. Belg., 2020, 71, 203-210) Comparison of opioid based versus multimodal postoperative pain management in cardiac surgery patients

S. Bogaert (*), H. Peperstraete (**), A. Moerman (***)

Abstract : Cardiac surgery performed by sterno- Introduction tomy is associated with moderate to severe acute postoperative pain. Most common analgesic Cardiac surgery performed by sternotomy schemes for per- and postoperative pain are based is associated with moderate to severe acute on intravenous opioids. Our goal is to compare postoperative pain. Postoperative pain is the standard opioid based regimen to a multimodal pain primary reason for prolonged convalescence and management to determine which therapy provides one of the main concerns of the surgical patient in the most comfort, the fastest extubation time, the the intensive care unit (ICU) department (1). This least pain and the least delirium. pain is multifactorial and multifocal ; it can be Methods : In this prospective randomized double- caused by incision, intraoperative tissue retraction blinded trial, two groups of 25 patients, were and dissection, surgical manipulation of the parietal divided in an opioid group and a multimodal group pleura, posterior rib dislocation or fracture, possible after given written informed consent. Postoperative brachial plexus injury, chest tube insertion and pain was scored at 8-h, 16-h, 24-h, 32-h, 40-h and harvesting of the saphenous vein and internal 48-h by the NRS-scale in the awake patient, or mammarian artery (1-3). the Critical-Care Pain Observation Tool (CPOT) The most common analgesic schemes for in the sedated patient. Delirium is measured by postoperative pain in cardiac surgery are based on the Intensive Care Delirium Screening Checklist intravenous opioids by bolus, with patient- or nurse- (ICDSC)-score. Time to extubation, length of stay controlled delivery systems. Although without (LOS) in ICU, LOS in hospital and rescue pain any doubt, they have a beneficial effect on pain, medication are all registered and compared between opioids are associated with dose-related side effects both groups. including oversedation, ileus, urinary retention, Result : These are preliminary results of an ongoing nausea, vomiting, pruritus, mental confusion and study where we intend to include 100 patients. Forty- respiratory depression, which might lead to a five patients were successfully enrolled. Baseline prolonged extubation time (2-4). characteristics were similar between two groups. In the 1990s anesthetic strategies mainly The multimodal pain scheme was administered in consisted out of high-dose opiates that required 21 patients, 24 patients received an opioid based mechanical ventilation for 12 to 24 hours post- scheme. After linear mixed model analysis no operatively. In the years that followed many centers significant differences were seen in CPOT-score initiated fast-track care pathways. Even though (p=0,057), NRS-score (p=0,919) or consumption in specific approaches differed, these pathways shared rescue pain medication. Also, for the ICDSC-score the common goal of facilitating early extubation no significant difference was found (p=0,595). Time to extubation, LOS at ICU or LOS at the hospital S. Bogaert, MD ; H. Peperstraete, MD ; A. Moerman PhD showed no significant difference between both (*) Department of Anesthesiology, University Hospital Ghent groups. (**) Department of Intensive Care - Cardiac Surgery, University Discussion : This trial shows no significant dif- Hospital Ghent ference in pain, delirium, intubation time, LOS (***) Department of Anesthesiology - Cardiac Anesthesia, University Hospital Ghent on ICU or LOS in hospital between multimodal Corresponding author : Bogaert S. Kloosterstraat 42, 9090 analgesia versus an opioid-based analgetic protocol Melle. in cardiac surgery by sternotomy. E-mail : [email protected] Keywords : Postoperative pain, analgesics, delirium ; Paper submitted on Aug 19, 2020 and accepted on Aug 25, thoracic surgery, sternotomy. 2020. Conflict of interest : none.

26-Bogaert.indd 203 12/01/2021 12:27 204 s. bogaert et al. and limiting ICU resource usage, primarily through who develop delirium after cardiac surgery have minimizing opioid usage. Systematic review and prolonged mechanical ventilation, prolonged analysis of the results of these attempts demonstrated length of hospital stay, high hospital mortality, and that it was a safe approach in low-risk patients and cognitive decline. that duration of intubation and ICU stay was reduced Recent studies show that dexmedetomidine but no change in overall length of hospital stay (5). might be the best agent to reduce the occurrence of However, the significant heterogeneity in the data delirium after cardiac surgery.(8). Dexmedetomidine precludes definitive conclusions on these programs. is an alpha-2 adrenergic receptor agonist that acts In the last decades, much has been written directly on the peripheral nervous system, causing a about multimodal pain protocols and their role in dose-dependent inhibition of C-fibers and Aα-fibers Enhanced Recovery after Surgery (ERAS)-protocols (2, 12). Dexmedetomidine induces sedation while to treat acute postoperative pain in non-cardiac preserving a degree of arousability among patients. surgery.(5). The objective is not only to reduce the Its use resulted in a shorter time to extubation, an dose and side effects of opioids. By working on both increased number of days free from coma or delirium, the central and peripheral pain mechanisms, the aim a reduced incidence of agitated delirium, prevention is to find a holy grail where the patient has minimal of delirium, and lower mortality compared to other suffering and the central neural hyper-excitability agents (8, 11, 13). Hypotension and bradycardia that increases postoperative pain is maximally are common side effects (2, 8, 14-16). The clinical attenuated (3, 6), reducing the risk of transformation efficacy has been proven in non-cardiac surgery by of acute into chronic pain to a minimum. augmenting anesthesia and analgesia, and allowing The prevalence of CPSP after cardiac surgery a reduction in opioid requirements.(2). ranges from 3.8% to 56%. In recent studies evalua- Gabapentin and pregabalin are effective for ting CPSP in non-cardiac surgery a decline in CPSP treatment of various neuropathic pain syndromes. was seen (7). They inhibit central neuronal sensitization and A lot of research has been done on prevention also produce anti-hyperalgesia by decreasing and treatment of delirium after cardiac surgery. excitatory amino acid neurotransmission in the Delirium is characterized by an acute onset and spinal cord through a direct postsynaptic or pre- a fluctuating course of inattention and either synaptic inhibition of Ca² influx. It has been shown disorganized thinking or an altered level of con- that gabapentin reduced pain scores and opioid sciousness. It is known to affect 15% to 70% of requirements in different surgical settings (3, 17- the medical and surgical populations, with the 19). Literature is not conclusive and because of highest prevalence in elderly and critically ill conflicting results the routine use of gabapentin patients.(8). Delirium usually is elicited by an acute and pregabalin to reduce opioid consumption in the noxious insult acting on predisposing factors. Age, cardiac surgical patients is not yet recommended. cerebrovascular disease, preadmission cognitive (17, 19). dysfunction, postoperative pain, use of opioids and Ketamine is not only an anesthetic agent but sedatives, surgical inflammation and depression also has an analgesic effect. The exact mechanism are important risk factors, with general anesthesia, is not yet known but some of the pathways are anesthetic drugs and prolonged mechanical venti- already identified.(20). It binds to the opioid lation being triggers.(9, 10). receptors κ (kappa), δ (delta) and μ (mu) and it has In cardiac surgery the incidence of this been proven that ketamine induces phosphorylation cognitive disorder is as high as 50-70%, despite of mitogen-activated protein kinases by 2-3 improvements in surgical and perioperative care times that of traditional opioid drugs.(14, 21). achieved in the last decades.(10, 11). The use of Another pathway of its analgesic effect is by the cardiopulmonary bypass with continuous flow, muscarinic acetylcholine receptors in the central hypothermia, microembolization, impairment of nervous system.(14). Ketamine also effects other cerebral oxygenation, and impaired autoregulation, ion channels including sodium channels and and eventual occurrence of cardiac dysfunction, voltage sensitive calcium channels leading to local all affect neurologic function and further increases anesthetic and gabapentin like effects.(20). Because the risk of postoperative delirium.(8). Factors of the unique effect of keeping hemodynamic predisposing patients to delirium after cardiac stability during induction, ketamine can be useful surgery and strategies to prevent and treat this in cardiac surgery. The analgesic effect, the absence devastating condition are poorly understood. of respiratory depression and hemodynamic sta- Furthermore, it is very well-known that patients bility make it an excellent drug to use in the

26-Bogaert.indd 204 12/01/2021 12:27 opiod versus multimodal in open cardiac surgery 205 ICU (20). Previous literature demonstrated a Study Participants significant analgesic effect from the intraoperative administration of ketamine, and current reviews All Patients, aged 45 or older, undergoing a continue to cite ketamine as a potential agent for use first sternotomy for coronary artery bypass graft in multimodal analgesia strategies for the enhanced (CABG) surgery with or without valve replacement recovery of cardiac surgery patients. However, it requiring cardiopulmonary bypass were candidates has to be acknowledged that not all studies showed for trial inclusion. In women we actively checked a reduced opioid consumption or lower pain scores for being in menopause to rule out a systematical postoperatively.(22). pregnancy test in women. The perioperative use of intravenous lidocaine Patients with a preoperative left ventricular has many beneficial effects in open procedures, ejection fraction less than 30%, preexisting cognitive such as an earlier return of gastrointestinal tract impairment measured by a preoperative MMSE function, less postoperative opioid consumption, less than 21, Alzheimer disease, Parkinson disease, improvement of postoperative cognitive dysfunction history of recent seizures, or hypersensitivity to and reduced stay in the hospital.(19, 23, 24). The any of the study medications, and patients taking exact working mechanism is not identified but the medications for cognitive decline or under chronic anti-inflammatory effects of lidocaine mediated pain medication therapy were excluded. through interactions with polymorphonuclear cells Participants who did not speak Dutch were and the inhibition of G protein-coupled receptors excluded because of their inability to complete the may play a crucial role (24). cognitive assessments, which have been extensively Magnesium sulfate’s analgesic mechanisms validated in Dutch. are also not fully identified, but it is thought that the NMDA receptor is blocked by calcium regulation Randomization and blinding mechanisms. Because the NMDA receptor plays After informed consent, patients were assigned a role in the transmission of pain, magnesium has randomly (1:1) by computer randomization to either become a subject of interest for potential use in a ‘standard’ pain protocol or a multimodal pain postoperative pain schemes.(25, 26). It has been protocol. The attending anesthesiologist was aware proven that peri-operative intravenous magnesium of the randomization. Patients and study nurses can reduce opioid consumption especially in the were blinded to the pain management. first 24h.(26). The goal of the present study is to compare Procedures a standard pain regimen “Fentanyl – Tradonal – Paracetamol – Oxynorm” to a multimodal pain In the multimodal pain protocol, patients management protocol “Gabapentin – Magnesium received Pregabaline 75mg one hour preoperatively. sulfate – minimal dose of Fentanyl – Ketamine – Once in the operating room, a dexmedetomidine Lidocaine – Dexmedetomidine – Paracetamol” to infusion was started at 0.8µg/kg/h. Postoperatively determine which therapy provides the most comfort, at ICU, the continuous infusion was lowered to the fastest extubation time, the least pain and the 0.5 µg/kg/h and stopped 12 hours postoperatively. lowest incidence of delirium. Induction of anesthesia was performed with Propofol until loss of consciousness and Fentanyl 1-2 µg/kg. Methodology Fentanyl dose was limited to 2.5µg/kg during the peroperative period. If needed, the anesthesiologist Study design was allowed to exceed this dose, as long as it was documented. After induction of anesthesia, a bolus This prospective randomized double-blinded of ketamine 0.5 mg/kg IV was given followed by a (for participant and study staff) study was conducted continuous infusion of 0.3 mg/kg/h. This continuous at Ghent University Hospital from September infusion was stopped postoperatively at ICU 2019 to March 2020. Institutional review board together with the continuous infusion of propofol. approval was obtained at 17/7/2019 from chairman Also, a bolus of Lidocaine 1.5 mg/kg IV was given Prof. dr. D. Matthys of the Committee on Clinical at induction followed by a continuous infusion of Investigations and the Ethical Committee, Corneel 1.3 mg/kg/h. This Lidocaine infusion was stopped Heymanslaan 10, 9000 Gent (eduract number : 12 hours postoperatively. Magnesium was given 2019-000515-84) and all patients provided written 25 mg/kg IV at induction and was repeated after informed consent. weaning of the extracorporal circulation.

26-Bogaert.indd 205 12/01/2021 12:27 206 s. bogaert et al. If a patient was assigned to the standard sedatives during the first 48 hours postoperatively, pain protocol, no pain medication was given pain scores, patient delirium and cognition assess- preoperatively. Induction was also done with ment data, and adverse events. Propofol until loss of consciousness and Fenta- Data were abstracted from the medical record nyl. Fentanyl dose was limited to 15µg/kg to assess secondary endpoints. IV. Postoperatively, a continuous infusion of Statistical analysis was done with IBM SPSS® Remifentanil was started at a dose of 0.05-0.1 µg/ software, version 26. We used descriptive statistics kg/h and stopped 12 h postoperatively. Magnesium and used for parametric data : unpaired t-test, or 2.5 mg/kg was given once at induction. student t-test, for non-parametric data : Wilcoxon’s For both groups, postoperative pain manage- test. For repeated measures of non-independent ment consisted of Paracetamol 1g/6h IV and Tra- data, to test for fixed effects, we used linear mixed madol 100mg/6h IV if needed and as long as the models type III tests. Level of significance was set patient was ventilated. Once a patient was extubated at (p) < 0.05. rescue pain medication was Oxycodone 5mg. No other analgesics were allowed. Results

Study outcomes Study population

The primary endpoint of the study was The flow chart of study inclusion is presented postoperative pain at 8-h, 16-h, 24-h, 32-h, 40-h and in figure 1. This is a preliminary report, halfway of 48-h after cardiac surgery. Pain was measured by our study with 100 patients. A total of 50 patients a dedicated study nurse by means of a NRS-scale provided informed consent after ruling out any of ranging from 0 to 10 if the patient is awake, and the exclusion criteria and were enrolled, of whom 50 by the CPOT ranging from 0 to 8 in the sedated were randomized. Five patients who were eligible patient. Secondary endpoints include delirium in and signed informed consent dropped out because the awake patient measured at 8-h, 16-h, 24-h, 32- the attending anesthesiologist refused to implement h, 40-h and 48-h after surgery, by the ICDSC-score the protocol for this particular patient. No patients ranging from 0 to 8. Also included in the secondary dropped out after the protocol was started. endpoints were : time till extubation in hours, Twenty-one patients received the multimodal length of hospital stay in hours, length of stay at pain therapy and 24 patients received the “standard” the ICU department in hours, Mean arterial blood opioid-based pain therapy. pressure (MAP), vasopressor need (VIS-score), Baseline characteristics are shown in table heart frequency and total consumption of rescue 1. With the exception of length, no significant pain medication registered in mg/kg. Table 1 Sample size calculation The baseline characteristics

Opioid based Following data were used for sample size Characteristics Multimodal (n=21) calculation : Two independent study groups, (n=24) continuous endpoints, an anticipated incidence of Demographics NRS-score in the standard protocol of 4 vs 3 in the Male 17 (81) 19 (79,2) multimodal group, with a standard deviation of 1.5, Female 4 (19) 5 (20,8) based on a previous study by Rafiq S. et al (3), a Age(years) 70,33 68,67 probability of type 1 error of 5% and power of 80%. Length (cm) 173,76* 168,54* Two groups of 35 patients were calculated. Weight (kg) 79,05 80,35 This paper is an intermediary analysis of BMI 26,214 28,154 50 patients divided in 2 groups. This makes this Baseline MMSE 26,9 27,1 analysis underpowered. Surgery Statistical analysis CABG 18 (85,7) 15 (62,5) Valve 2 (9,5) 5 (20,8) Research Electronic Data Capture (REDCap) CABG + valve 1 (4,8) 4 (16,7) was used for data collection and included *Level of significance between both groups p=0.037. First number demographic data, information on analgesics and reported is the absolute number, followed by percentage in parentheses.

26-Bogaert.indd 206 12/01/2021 12:27

Fig1. Flowopiod chart versus of inclusion m ultimodal in open cardiac surgery 207

Assessed for eligibility (n=67) (n = …)

Excluded (n = 17)

- Not meeting inclusion criteria (n = 16) - Not willing to give informed consent (n=1)

Enrollment

Patients Randomized (n = 50)

Allocated to intervention (n = 25)

Received allocated intervention (n = 21) Received allocated intervention (n = 24) Protocol violation during operation (n=4) Protocol violation during operation (n=1)

Patients Did not receive allocated intervention (give Did not receive allocated intervention (give Allocation reasons) (n = …) reasons) (n = …)

up - Lost to follow-up (n =0) Lost to follow-up (n = 0)

Patients Follow Discontinued intervention (give reasons) (n = …) Discontinued intervention (give reasons) (n = …)

Analysed (n = 21) Analysed (n = 24)

Patients Analysis Excluded from analysis (give reasons) (n = …) Excluded from analysis (give reasons) (n = …)

Fig. .1 — Flow chart of inclusion.

differences were found between the two groups.

Since we are not talking about low back pain here, this difference between the two groups will

not be relevant in a possible difference in the pain

experience between the two groups. Coronary artery bypass-grafting was the most frequent performed

intervention (33/45 [73.33%]), followed by single or multiple valve surgery (7/45 [15.55%]) and combined surgery formed the smallest group (5/45 [11.11%]).

Drug Administration Fig. 2. — Mean NRS-scores varying in time. A start value was measured at admission to the hospital, the on admission value All patients in the opioid protocol received was taken at admission on ICU. All patients were sedated at medication as prescribed by the protocol. In the 14admission on ICU. multimodal protocol, the attending anesthesiologist Delirium was allowed to give extra fentanyl when deemed clinically needed. In seven (33%) multimodal There was no significant difference in the cases, the fentanyl dose had to be augmented to ICDSC-scores between groups. Mean difference 5µg/kg. In two (10%) cases, the dose even had to be in the opioid group compared to the multimodal augmented to 7.5 µg/kg. When asked after surgery, group was 0.063 [95% CI -0.171, 0.297] (p=0.595).. all anesthetists reported that they had to augment The ICDSC-score was always scored as 0 when a the dose because of hypertension, especially at the patient is sedated, so there could have been made moment of sternotomy. No other problems were an analyzing error if there was a difference between reported with drug doses in the protocol. both groups in length of intubation or sedation. Analysis of duration of intubation and sedation Pain scores showed no difference between both groups, making sure that this ICDSC comparison was done correctly. No significant differences in pain scores between the two groups were seen. Mean difference Blood pressure in the opioid group compared to the multimodal group was 0.324 [95% CI -0.241, 0.889] (p=0.257) To assess any differences in hemodynamics, and -0.026 [95% CI -0.531, 0.479] (p=0.919) for absolute values of MAP as well as the vasoactive- CPOT and NRS scores, respectively. inotropic score (VIS-score) was compared between The course of the NRS-scores is depicted in both groups. No significant differences were seen in figure 2. MAP (p=0.884) or VIS (p=0.193).

26-Bogaert.indd 207 12/01/2021 12:27 208 s. bogaert et al. Heartrate Table 3 Need for rescue medication in the first 48h *We used median No significant differences in heart rate were time because tested for normality by a Shapiro-Wilk test and seen between both groups (p=0.760). found the data significantly deviate from a normal distribution. Three patients (two in the opioid group and one post-operative in the multimodal group) were in atrial fibrillation Opioid group Multimodal group P-value during the preoperative visit. In two of these three Paracetamol 106 [68 , 170] 106 [68 , 169] 0.909 patients, no atrial fibrillation was diagnosed during (mg/kg) the first 48 hours postoperatively, one patient stayed Tradonal 1.12 [0 , 5] 1.22 [0 , 3.88] 0.913 in atrial fibrillation during the whole first 48 hours (mg/kg) Oxynorm 0.23 [0 , 0.65] 0.35 [0 , 1.67] 0.945 postoperatively. (mg/kg) Only 16 patients were not having a paced Data are presented as median [min , max]. heart rhythm at admission on the intensive care unit (ICU). Twelve (75%) belonged to the opioid group whereas only four (25%) of them belonged day five postoperative and he needed to undergo an to the multimodal group. Thirteen patients (61%) urgent surgery leading to a longer LOS at ICU and a of the multimodal group needed pacing for longer longer LOS at the hospital. than 8 hours. In contrast, only five patients (21%) of the opioid group required pacing for longer than Discussion 8 hours. The decision to start or stop pacing was made by the attending ICU specialist. There was no In this randomized controlled trial, we com- mandatory threshold included in the protocol. pared two different pain management protocols (opioid based vs multimodal). We found no Time to extubation, LOS at ICU and LOS at the differences in postoperative pain at 8-h, 16-h, 24- hospital h, 32-h, 40-h and 48-h. We found no significant When we take a look at table 2, we can see difference in delirium in the awake patient in the that there is no significant difference in time to first 48h post-operative, no significant difference in extubation, LOS at ICU or LOAS at hospital. We time to extubation, LOS in ICU or LOS in hospital. do see there is a trend towards a longer time to In literature conflicting data about multimodal extubation for the multimodal group and a shorter analgesia is very common. This is because there is LOS at hospital. no specific definition of multimodal analgesia and there is no consensus of what drugs and what dose of Need for rescue medication drugs should be included in a multimodal protocol. Our multimodal pain protocol was based on In this trial we could not find any significant protocols used in non-cardiac surgery. It contained difference in need for rescue pain medication pregabalin, ketamine, dexmedetomidine, lidocaine between the two groups (table 3). and low dose opioids. The exact combination of all of our drugs in the multimodal protocol has never Adverse events been studied before in cardiac surgery. Because of this combination of different drugs it is impossible One patient in the multimodal group developed to determine the value of each separate drug. Non- a ruptured abdominal aorta aneurysm during his believers in multimodal analgesia find this its postoperative stay at the ICU. This happened at biggest weakness, believers find this its biggest Table 2 strength (3). Time to extubation and lengths of stay Our multimodal protocol used a fentanyl dose of 2.5µg/kg vs. 15µg/kg in the opioid-based Multimodal Opioid group P value group protocol. During this trial it was found that 2.5µg/ Median* Time to kg of fentanyl in the multimodal protocol does 7 9 0.072 extubation (hours) not always provide an equally stable anesthesia, Median LOS at ICU 22 24 0.284 especially during a strong pain stimulus such as (hours) a sternotomy. In 9 (43%) cases anesthetists had Median LOS at hospital 201.75 99 0.085 (hours) to increase the fentanyl dose up to a maximum of *We used median time because tested for normality by a Shapiro-Wilk 7.5µg/kg, still significantly lower than the 15µg/kg test and found the data significantly deviate from a normal distribution. in the opioid-based protocol.

26-Bogaert.indd 208 12/01/2021 12:27 opiod versus multimodal in open cardiac surgery 209 Compared to literature our single dose of proven that they could have an effect on developing pregabalin 75mg is rather small given that studies postoperative delirium after cardiac surgery (8, 10, were done with doses up to 300mg, all other doses 27). of our tested drugs are similar to those used in The composition and dosing of the multimodal previous studies (5). protocol has not been studied before so it is plausible This study showed no difference between the that other doses or other combinations of drugs will 2groups in experience of postoperative pain within provide different results. the first 48hours. This result is conflicting with the result of Rafiq et al.(3). They found a significant Conclusion difference in postoperative NRS-painscores of 2 vs 3. It is not a big difference and one could argue This trial shows no significant difference in about the clinical significance of this result. pain, delirium, intubation time, LOS on ICU or It has previously been suggested that LOS in hospital and in hemodynamics between postoperative administration of dexmedetomidine multimodal analgesia versus an opioid-based anal- results in a decreased incidence of postoperative getic protocol in adult cardiac surgery patients delirium in cardiac surgery (5, 26). In the present who underwent sternotomy. Although this trial is trial, dexmedetomidine was given both per- and underpowered it suggests that multimodal analgesia postoperative. However, we could not find any in cardiac surgery is feasible and that it might be a statistically significant difference for delirium good alternative for opioid based protocols. nor pain. What we did see was a higher need to postoperative pacing because of bradycardia Acknowledgements which is a known side effect of dexmedetomidine. We do not find this a problem, because every Help with statistical analysis was provided by patient receives temporary epicardial pacing leads (‘cel biostatistiek’ of Ghent University). before closure of the pericardium. Furthermore, as dexmedetomidine is stopped, bradycardia and by References this the pacing need, ends. When we look at time to extubation we 1. Mueller, X.M., et al., 2000, Pain location, distribution, and intensity after cardiac surgery. Chest. 118(2): 391-6. found no significant difference. This is not what 2. Bigeleisen, P.E. and N. Goehner, 2015, Novel approaches we expected because based on what is seen in a in pain management in cardiac surgery. Current opinion in systematic review of Noss et al (5) we would expect anaesthesiology. 28(1): 89-94. 3. Rafiq, S., et al., 2014, Multimodal analgesia versus that a multimodal strategy leads to a shorter time traditional opiate based analgesia after cardiac surgery, leading to extubation. The same counts for LOS on a randomized controlled trial. Journal of cardiothoracic ICU where no significant difference was seen where surgery. 9: 52. we expected that the multimodal protocol would be 4. White, P.F., et al., 2003, Use of a continuous local anesthetic infusion for pain management after median sternotomy. in favor. LOS in hospital showed no difference which Anesthesiology. 99(4): 918-23. is in line of what we see in other studies. Our result 5. Noss, C., et al., 2018, Enhanced Recovery for Cardiac might be compromised, because all patients need Surgery. J Cardiothorac Vasc Anesth. 32(6): 2760-2770. 6. Weinbroum, A.A., 2012, Non-opioid IV adjuvants in the to stay in hospital for at least 7days, independent perioperative period : pharmacological and clinical aspects of their clinical state, since it is a standard practice of ketamine and gabapentinoids. Pharmacol Res. 65(4): in our hospital to perform a postoperative cardiac 411-29. 7. Gjeilo, K.H., et al., 2017, Chronic postsurgical pain in ultrasound check-up on day 7. patients 5 years after cardiac surgery : A prospective cohort study. European journal of pain (London, England). 21(3): Limitations 425-433. 8. Pieri, M., et al., 2019, Trials Focusing on Prevention and Treatment of Delirium After Cardiac Surgery : A systematic This analysis is a halfway analysis and Review of Randomized Evidence. J cardiothor vascul contains only half of the predetermined number anesth. of patients to get enough power, which makes this 9. Atalan, N., et al., 2013, Morphine is a reasonable alternative to haloperidol in the treatment of postoperative study underpowered. hyperactive-type delirium after cardiac surgery. Journal of The 2 groups were found similar without cardiothoracic and vascular anesthesia. 27(5): 933-8. significant differences, but it is impossible to check 10. Subramaniam, B., et al., 2019, Effect of Intravenous Acetaminophen vs Placebo Combined With Propofol or this for every parameter. We did not look at the time Dexmedetomidine on Postoperative Delirium Among Older of aorta clamping or total time on cardiopulmonary Patients Following Cardiac Surgery : The DEXACET bypass. For both of these parameters has been Randomized Clinical Trial. JAMA. 321(7): 686-696.

26-Bogaert.indd 209 12/01/2021 12:27 210 s. bogaert et al. 11. van Eijk, M.M. and A.J. Slooter, 2010, Duration of ICU Analgesics and Techniques : A Review. JAMA Surg. delirium, severity of the underlying disease, and mortality. 152(7): 691-697. American journal of respiratory and critical care medicine. 20. Frenkel, C. and B.W. Urban, 1992, Molecular actions of 181(4): 419-20; author reply 420-1. racemic ketamine on human CNS sodium channels. Br J 12. Tang, C. and Z. Xia, 2017, Dexmedetomidine in peri- Anaesth. 69(3): 292-7. operative acute pain management : a non-opioid adjuvant 21. Gupta, A., L.A. Devi, and I. Gomes, 2011, Potentiation analgesic. J Pain Res. 10: 1899-1904. of mu-opioid receptor-mediated signaling by ketamine. J 13. Tang, C. and Z. Xia, 2017, Dexmedetomidine in Neurochem. 119(2): 294-302. perioperative acute pain management : a non-opioid adju- 22. Cameron, M., et al., 2020, Intraoperative Ketamine for vant analgesic. Journal of pain research. 10: 1899-1904. Analgesia Post-Coronary Artery Bypass Surgery : A 14. Mazzeffi, M., K. Johnson, and C. Paciullo, 2015, Ketamine Randomized, Controlled, Double-Blind Clinical Trial. in adult cardiac surgery and the cardiac surgery Intensive Journal of cardiothoracic and vascular anesthesia. 34(3): Care Unit : an evidence-based clinical review. Ann Card 586-591. Anaesth. 18(2): 202-9. 23. Weibel, S., et al., 2016, Efficacy and safety of intravenous 15. Mu, J.L., A. Lee, and G.M. Joynt, 2015, Pharmacologic lidocaine for postoperative analgesia and recovery after agents for the prevention and treatment of delirium in surgery : a systematic review with trial sequential analysis. patients undergoing cardiac surgery : systematic review Br J Anaesth. 116(6): 770-83. and metaanalysis. Crit Care Med. 43(1): 194-204. 24. Kranke, P., et al., 2015, Continuous intravenous peri- 16. Shehabi, Y., et al., 2019, Early Sedation with Dexmede- operative lidocaine infusion for postoperative pain and tomidine in Critically Ill Patients. The New England journal recovery. Cochrane Database Syst Rev,(7): p. CD009642. of medicine. 380(26): 2506-2517. 25. Castro, J. and M.F. Cooney, 2017, Intravenous Magnesium 17. Maitra, S., et al., 2017, [Perioperative gabapentin and in the Management of Postoperative Pain. J Perianesth pregabalin in cardiac surgery : a systematic review and Nurs. 32(1): 72-76. meta-analysis]. Revista brasileira de anestesiologia. 67(3): 26. Albrecht, E., et al., 2013, Peri-operative intravenous admi- 294-304. nistration of magnesium sulphate and postoperative pain : a 18. Ucak, A., et al., 2011, The effects of gabapentin on acute meta-analysis. Anaesthesia. 68(1): 79-90. and chronic postoperative pain after coronary artery bypass 27. Markham, T., et al., 2019, Assessment of a multimodal graft surgery. Journal of cardiothoracic and vascular analgesia protocol to allow the implementation of enhanced anesthesia. 25(5): 824-9. recovery after cardiac surgery : Retrospective analysis of 19. Wick, E.C., M.C. Grant, and C.L. Wu, 2017, Postoperative patient outcomes. J Clin Anesth. 54: 76-80. Multimodal Analgesia Pain Management With Nonopioid

26-Bogaert.indd 210 12/01/2021 12:27 (Acta Anaesth. Belg., 2020, 71, 211-215) Evaluation of the differential effect of phenylephrine on near-infrared spectroscopy (NIRS)-measured tissue blood volumes : a study protocol

J. Vandermeulen, A.T. Moerman

Abstract : Near-infrared spectroscopy (NIRS) enables fulness lies in the fact that it offers a bedside, noninvasive monitoring of oxygenation and cerebral noninvasive and real-time method to monitor blood volumes. However, concerns have been raised tissue oxygenation (1, 2). It is mostly used for the about extracranial contamination of the signal. We aim evaluation of cerebral oxygen saturation (ScO2), for to assess the occurrence of extracranial disturbance of instance during cardiac surgery or carotid artery the NIRS signal during administration of phenylephrine. surgery. The study will be performed with NIRO 200NX (Hamamatsu Phototonics Corp, Tokyo, Japan) which However, concerns have been reported about employs both the original Modified Beer-Lambert the actual sampling volume and the possible (MBL) method to measure concentration changes in total signal contamination by superficial tissues, such hemoglobin (tHb, expressed in µM), and the Spatially as the skin and subcutaneous adipose layers (3- Resolved Spectroscopy (SRS) method to measure the 6). Presumably the most important source of total hemoglobin content (nTHI, expressed in arbitrary extracranial contamination is interference from the units). Although both parameters generally provide cutaneous circulation. This is a relevant issue as concordant values, SRS has a higher cerebral specificity control of skin blood flow is considerably different and tends not to be affected by extracranial blood flow. from cerebral blood flow and its interference might We hypothesize that extracranial contamination of NIRS significantly alter NIRS-based management. values is plausible when phenylephrine induces more All commercially available NIRS devices pronounced changes in MBL than in SRS. use an identical basic technology to measure ScO2 After ethical committee approval, 20 consenting patients undergoing cardiac surgery with extracorporeal continuously. One particular device, the NIRO circulation will be included. Patients under 18 years of (Hamamatsu Phototonics Corp, Tokyo, Japan), also age and patients who do not need phenylephrine during enables measurement of tissue blood volume, by the procedure will be excluded. Two NIRO sensors will using two different techniques : the technique of be placed bilaterally on the forehead for continuous Spatially Resolved Spectroscopy (SRS) to measure registration of nTHI and tHb. Phenylephrine will be the normalized tissue hemoglobin index (nTHI), administered when clinically indicated. Posthoc, the and the Modified Beer-Lambert (MBL) method to means of 15 values just before and 90 seconds after measure the total hemoglobin concentration (tHb). each administration of phenylephrine will be calculated. Although theoretically blood volumes To adjust for the difference in raw scale units, Z-scores measured with these two techniques should be will be calculated. Distribution of data will be tested for identical, the values have been shown to occasionally normality using the Shapiro-Wilk test. The nTHI and tHb diverge (own observation, not reported). It has been values pre- and post-phenylephrine, and the differences between nTHI and tHb will be analysed with t-test or suggested that SRS has a higher cerebral specificity Wilcoxon Signed Ranks Test, as appropriate. Statistical and tends to be less sensitive to extracranial changes significance will be accepted at p < 0.05 than MBL variables (7-9).

Keywords : NIRS ; tissue blood volume ; Phenylephrine ; J. Vandermeulen, MD ; A.T. Moerman, MD, PhD extracranial contamination. University Hospital Ghent, Corneel Heymanslaan 10, 9000 Gent Corresponding author : Moerman A.T., University Hospital Introduction Ghent, Corneel Heymanslaan 10, 9000 Gent, Belgium. E-mail : [email protected] Background Paper submitted on Aug 19, 2020 and accepted on Aug 25, Near-infrared spectroscopy (NIRS) has been 2020. proven to be a useful tool in clinical care. Its use- Conflict of interest : none.

27-Vandermeulen.indd 211 12/01/2021 12:28 212 j. vandermeulen et al. Therefore, the aim of the present study is to whereas the MBL method measures tHb. It is of evaluate if there is a difference in blood volume great importance to know the difference between measurements between SRS and MBL when those two methods. administering a vasopressor (phenylephrine). Phe-

nylephrine is a selective α1-adrenergic receptor Spatially Resolved Spectroscopy vs. Modified Beer- agonist, commonly used in anesthesia to increase Lambert blood pressure. It binds selectively to α-receptors which cause blood vessels to constrict. Since the Modified Beer-Lambert (MBL) cerebral vasculature lacks significant α-adreno- receptors (10), vasoconstriction mainly occurs in The MBL method is based on the changes in the the extracranial area. amount of detected light induced by concentration We hypothesize that if NIRS measurements change of the substance of interest. The MBL are indeed prone to extracranial contamination, law states that the quantity of light absorbed by a the blood volume values measured with the substance is directly proportional to the absorption MBL method (tHb) will be more affected by the coefficient of the substance, the concentration of the administration of phenylephrine than the values substance, and the path length of the light (2, 11). measured with the SRS method (nTHI). A correction factor is applied for scatter loss due to light backscattered by the tissues. It is important Basic principle of NIRS to mention that with the MBL method there is no distinction between superficial and deep layers. NIRS is an analyzing method that collects By means of the MBL technology, NIRO information about tissue oxygenation and tissue provides a measure of concentration changes of

composition by the use of invisible near infrared O2Hb, HHb and tHb (= O2Hb + HHb) with respect light (1). Light emitted at wavelengths specific to an arbitrary initial value (expressed in µM).

for the peak absorption of the chromophores The biggest advantage of measuring ScO2 with

oxyhemoglobin (O2Hb) (920 mm) and deoxy- this method is its high sensitivity for changes in hemoglobin (HHb) (760mm) penetrates the scalp brain oxygenation. The second advantage of this and skull, and “lights up” brain tissue. Light technique is that it needs only one point for light detectors applied to the forehead measure light detection, enabling miniaturization of the sensor attenuation at a set distance from a NIRS-emitting size.

light source. NIRS then measures the ratio of O2Hb On the other side, there is the theoretical

over total hemoglobin (sum of O2Hb and HHb) at disadvantage of extracranial contamination inter- a depth of around 2 cm below the scalp, assuming fering with the measurement, which is the subject the distance between the light emitter and detector of this study. is 4 cm (9, 11). NIRS reflects the oxygenation of all hemoglobin in the arteries, veins and capillary Spatially Resolved Spectroscopy (SRS) vessels. It is important to notice that NIRS- based cerebral and tissue oximeters do not use SRS measures backscattered light as a function plethysmography. So in contrast to pulse oximetry, of the source-detector distance. The depth of photon NIRS does not differentiate between arterial and penetration is proportional to the source-detector venous blood. distance. Thus analysing the differential signal as Venous blood contributes on average 75% a function of the distance provides assessment of of the volume of total blood in the measured changes occurring in deep tissue layers. This results

brain tissue in the frontal lobe. However ScO2 as in values that are theoretically free from components measured with cerebral oximetry varies between originating in the superficial layers. Therefore it is subjects by approximately 10 %. This is due to the suggested that SRS is not affected by blood flow in fact that different individuals may have a different the superficial layers (12). composition of arteries, veins, capillaries and nonvascular tissue in the frontal lobe. (9, 11) Limitations Currently available NIRS devices use different technologies and computational algorithms to Although NIRS monitoring has greatly generate NIRS values. NIRO employs two different improved in recent years, the technique still has techniques to measure tissue oxygenation and blood limitations and faces issues that might significantly volumes : the technique of SRS measures nTHI, contribute to bias and imprecision in readings.

27-Vandermeulen.indd 212 12/01/2021 12:28 effect of phenylephrine on tissue blood volume 213 One important limitation is the variation tHb (MBL-based variable). The vasoconstriction among individuals. NIRS does not quantify caused by phenylephrine is expected to only affect oxygen molecules but calculates the ratio of light the skin circulation, not the cerebral circulation. absorbencies at predefined wavelengths. The Therefore, we hypothesized that extracranial con- algorithm used to calculate NIRS values assumes tamination of NIRS values is plausible when a fixed distance for light to travel through the phenylephrine induces more pronounced changes sampled area (the optical path length). However, in the superficial method of measurement (MBL, different tissue components produce very different measuring tHb). amounts of photon scattering and absorption. As a result, variations in probe positioning, as well Materials and Methods as interindividual variations in the composition of tissue, may result in variability of the true optical This is a prospective, observational study. path length. Also variations in arterial and venous Approval by the local research ethics committee blood ratio in the sampling area might cause was obtained on 19 March 2019 (reference number reading errors (13). Another possible cause of the 2019/0265). After obtaining written informed variations in NIRS-measured values might be due consent, 20 patients undergoing cardiac surgery to extracranial contamination. The most important with extracorporeal circulation will be included. source of extracranial contamination is likely Patients under 18 years of age and patients who do the interference of cutaneous circulation on the not need phenylephrine during the procedure will be NIRS measurements, which is the subject of our excluded. The standard fasting guidelines according study. Also other possible sources of extracranial to the local fasting protocol are applicable. The contamination are reported, like differences in skin study encompasses no additional benefits nor risks color (1). Darker skin is seen by the cerebral oximeter as patients will be treated by standard of care. as more venous, causing reading bias towards lower saturations. Values were also reported to be affected Study procedure by skull thickness and the area of the cerebrospinal fluid layer (14). It is also known that NIRS reads Two NIRS sensors (NIRO 200NX, Hama- lower in women across the spectrum of saturations. matsu Photonics, Tokyo, Japan) will be placed Whether this is related to lower hemoglobin, thinner respectively on the left and right forehead for cranial bone or other factors in female subjects is continuous registration of nTHI and tHb of not clear (1). the corresponding brain hemisphere. General Another limitation to mention is the presence anesthesia will be induced with fentanyl up to 3 µg of myoglobin in muscle tissue. Hemoglobin and kg-1, propofol up to 2 mg kg-1, and rocuronium up myoglobin share key properties including similar to 1 mg kg-1. Mechanical ventilation with oxygen optical properties. When addressing peripheral enriched air (fractional inspired oxygen 0.6) will be

oxygenation at the thenar or forearm, myoglobin may adapted to keep the end-tidal CO2 around 35 mmHg. become a major contributor to the overall reading. Anesthesia will be maintained with sevoflurane at a Since oxygen affinity of myoglobin is markedly minimum end-tidal concentration of 1.5 %. stronger compared to hemoglobin, myoglobin Phenylephrine will be administered whenever will have a relatively high oxygen saturation and clinically indicated. It will not be administered for can therefore be mistaken for hemoglobin with the purpose of the study. Every administration of overestimating of the saturation. However this is phenylephrine and other vasoactive medication will less an issue when measuring cerebral oxygenation, be marked and posthoc the means of 15 values just since in case of frontal readings only minimal before and 90 seconds after each administration of muscle and thus myoglobin is within the photon phenylephrine will be calculated. pathway. Statistics Aim of the study Statistical analysis will be performed using It has been suggested that SRS-based variables the statistical software SPSS Statistics 23 (SPSS are less sensitive to extracranial changes in blood Inc., Chicago, IL). The raw data will be tested for volume than MBL-based variables. To evaluate normality using the Shapiro–Wilk test. Data will be this hypothesis, we aim to assess the effect of presented as mean ± standard deviation or median phenylephrine on nTHI (SRS-based variable) and [minimum-maximum], as appropriate.

27-Vandermeulen.indd 213 12/01/2021 12:28 214 j. vandermeulen et al. The variable nTHI is expressed in arbitrary A study conducted by Sorensen also examined units, whereas tHb is expressed in μM. To adjust the contribution of extracranial contamination. for the difference in the raw scale units, the raw Two different devices (NIRO-200NX and

unit measures of each data point will be converted INVOS-4100) monitored ScO2 in volunteers during to standard deviation units (Z-scores), according to interventions that are known to affect cerebral the formula : Z-score = (Raw score-mean average) / and skin oxygenation. They observed that INVOS

standard deviation. All data are then normalized to a underestimated ScO2, particularly when skin blood common scale with an average of zero and standard flow was reduced by phenylephrine (4). deviation of 1, allowing the probability estimation Contamination of NIRS values due to that two different categories of data are significantly superficial circulation is not only a problem for different. The nTHI and tHb values pre- and post- cerebral oximetry. Some studies demonstrated that phenylephrine, and the differences between nTHI circulation in the skin can also affect the values and tHb will be analysed with t-test or Wilcoxon when measuring muscle oxygen saturation. Signed Ranks Test, as appropriate. Statistical signi- Mesere et al. conducted a study with the same ficance will be accepted at p < 0.05. setup as ours, except that superficial contamination was evaluated at the muscular level, instead of Discussion the extracranial level (7). They compared SRS and MBL in muscle and skin during selective It has been suggested that changes in cutaneous alterations of blood volumes by thermoregulation. circulation can affect NIRS measurements and SRS appeared to effectively reject interference of might lead to inaccurate values (4-6, 15). However, cutaneous blood flow in both muscle blood volume not all NIRS devices employ the same measurement and oxygenation monitoring. Instead, MBL-based technology. NIRO is the only clinically available variables appeared to be much more sensitive to device that uses both MBL and SRS, all other clinical changes in skin circulation. devices use MBL technology combined with spatial In the present study, we aim to induce resolution (MBL-SR). The aim of spatial resolution selective changes in extracranial circulation by is to correct the obtained MBL measurements for administration of phenylephrine. Phenylephrine contamination. Spatial resolution works by using is a catecholaminergic drug with exclusive α-1 differentially spaced light detectors. The closer activity. Because of the sporadic presence of α-1 receiver will measure the superficial tissue while the receptors in the brain circulation in contrast to the distal optode measures both superficial and deeper abundant existence in peripheral blood vessels (10), tissue. After subtraction of the interference from phenylephrine is extremely fit for measuring the superficial tissues, oxygenation in the deeper tissues influence of extracranial blood volume on NIRS is derived. parameters. However, various studies have suggested The findings in this study have potentially that despite correction by means of the spatial important clinical significance. The problem of resolution technology, NIRS might still be prone to extracranial contamination cannot easily be dis- extracranial contamination. missed. When measuring cerebral oxygenation, The first study which suggested extracranial it is important that changes are not affected by contamination is a study by Davie et al. (6). cutaneous perfusion. NIRS is increasingly used to By inflating a tourniquet around the forehead, guide therapy. If there is inherent contamination, this

significant reductions in ScO2 measured with three might have an influence on detection of deleterious different NIRS devices (INVOS, Foresight and events (such as hypoxia) and may interfere with the Equanox) were observed, suggesting extracranial success of the guided therapy. Therefore, improving contamination. These devices all employ MBL-SR the accuracy of the technology could improve the technology, suggesting insufficient correction of management of such events. If on the other side, the interference of extracranial tissues by the SR NIRS is only used to detect interruptions in the technology. symmetry of cerebral blood flow during carotid Another study examined the effect of the surgery then extracranial contamination is of less position and extracranial contamination on three importance (6). different devices (INVOS-5100C, FORE-SIGHT This study might also provide an answer to ELITE and NIRO). NIRO had the least extracerebral the question of paradoxically cerebral desaturation contamination, particularly in the upright position after phenylephrine reported by Nissen et al. and (9). Moerman et al (16, 17). If our hypothesis would be

27-Vandermeulen.indd 214 12/01/2021 12:28 effect of phenylephrine on tissue blood volume 215 confirmed by our results, cerebral desaturation might troscopy determined cerebral oxygen saturation during be explained as the result of reduction in extracranial administration of norepinephrine. Anesthesiology 117: 263-270 blood volume due to the vasoconstriction caused by 6. Davie S.N. and Grocott H.P. 2012. Impact of extracranial phenylephrine. contamination on regional cerebral oxygen saturation : A major strength of our study design is the a comparison of three cerebral oximetry technologies. Anesthesiology 116: 834-840. fact that the values of MBL and SRS are obtained 7. Messere A. and Roatta S. 2013. Influence of cutaneous and with the same device, at the same sampling site and muscular circulation on spatially resolved versus standard in the same patient. So therefore, no interdevice Beer-Lambert near-infrared spectroscopy. Physiol Rep. 5: 1(7). technologic differences, nor inter- or intraindividual 8. Messere A. and Roatta S. 2015. Local and remote differences can explain discrepancy. thermoregulatory changes affect NIRS measurement in forearm muscles. Eur. J. Appl. Physiol. 115: 2281-2291. 9. Kato S., Yoshitani K., Kubota Y., INamtomi Y. and Ohnishi Conclusion Y. 2017. Effect of posture and extracranial contamination on results of cerebral oximetry by near-infrared spectroscopy. This study aims to evaluate if NIRS measure- J. Anesth. 31: 103-110. ments are affected by extracranial tissue changes. We 10. Strandgaard S. and Sigurdsson S. 2008. Point-Counterpoint : Sympathetic activity does/does not influence cerebral blood hypothesized that if tHb is more affected than nTHI flow. Counterpoint : Sympathetic nerve activity does not when administering phenylephrine, extracranial influence cerebral blood flow. J. Appl. Physiol. 105: 1366- contamination is highly plausible. If this proves to 1367 ; discussion 1367-1368. 11. Steppan J. and Hogue C. 2014. Cerebral and tissue oximetry. be true, it might be safer to adopt the SRS variables Best Practice and Research Clin. Anaesth. 28: 429-439. rather than the MBL variables whenever changes in 12. Canova D., Roatta S., Bosone D. and Micieli G. 2011. skin perfusion cannot be excluded. Inconsistent detection of changes in cerebral blood volume by near infrared spectroscopy in standard clinical tests. J. Appl. Physiol. 110: 1646-1655. References 13. Watzman H., Kurth C., Montenegro L., Rome J., Steven J. and Nicolson S. 2000. Arterial and venous contributions 1. Scheeren TWL., Schrober P. and Schwarte L. 2012. to near-infrared cerebral oximetry. Anesthesiology 93: 947- Monitoring tissue oxygenation by near infrared spectroscopy 953. (NIRS) : background and current applications. J. Clin. 14. Yoshitani K., Kawaguchi M., Miura N., Okuno T., Kanoda Monit. Comput. 26: 279-287. T., Ohnishi Y. and Kuro M. 2007. Effects of hemoglobin 2. Moerman A. and Wouters P. 2010. Near-infrared concentration, skull thickness, and the area of the spectroscopy (NIRS) monitoring in contemporary cerebrospinal fluid layer on near-infrared spectroscopy anesthesia and critical care. Acta Anaesth. Belg. 61: 185- measurements. Anesthesiology 106: 458-462. 194. 15. Germon T., Young A., Manara A. and Nelson R. 1995. 3. Bickler P., Feiner J. and Rollins M. 2013. Factors affecting Extracerebral absorption of near infrared light influences the performance of 5 cerebral oximeters during hypoxia in the detection of increased cerebral oxygenation monitored healthy volunteers. Anesth. Analg. 117: 813-823. by near infrared spectroscopy. J. Neurol. Neurosurg. 4. Sorensen H., Rasmussen P., Siebenmann C., Zaar M., Psychiatry 58: 477-479. Hvidtfeldt M., Ogoh S. and Lundby C. et al. 2015. 16. Nissen P., Brassard P., Jorgensen T. and Secher N.H. Extra-cerebral oxygenation influence on near-infrared 2010. Phenylefrine but not ephedrine reduces frontal lobe spectroscopy-determined frontal lobe oxygenation in oxygenation following anesthesia-induced hypotension. healthy volunteers : a comparison between INVOS-4100 Neurocrit Care 12: 17-23. and NIRO-200NX. Clin. Physiol. Funct. Imaging. 35: 17. Moerman A., Vanbiervliet V., Van Wesemael A., Bouchez 17784. S., Wouters P. and De Hert S. 2015. Assessment of cerebral 5. Sorensen H., Secher N.H., Siebenmann C., Nielsen H.B., autoregulation patterns with near-infrared spectroscopy Kohl-Bareis M., Lundby C. and Rasmussen P. 2012. during pharmacological induced pressure changes. Anes- Cutaneous vasoconstriction affects near-infrared spec- thesiology 123: 327-335.

27-Vandermeulen.indd 215 12/01/2021 12:28

(Acta Anaesth. Belg., 2020, 71, 217-219) Prolonged effect of rocuronium due to predisposing factors, a case report

F. Dhondt, M. Sabbe

Abstract : The duration of neuromuscular blockade objective monitoring for residual NMB is not (NMB) may be prolonged following the administration routinely performed before tracheal extubation. of a neuromuscular blocking agent (NMBA) due to We report a case of an 83-year-old man who different causes. We report a case of an 83-year-old man showed a severe prolongation of NMB following who experienced severe prolongation of NMB following a rapid sequence intubation (RSI) procedure, using the administration of rocuronium. rocuronium as the NMBA, following resuscitation Case presentation : An 83-year-old man, who had been admitted to the geriatrics ward of our hospital two days for a hypercapnic respiratory arrest. prior, was found unresponsive and apneic shortly after the initiation of oxygen supplementation for hypoxemic Case Report Narrative respiratory failure. Basic life support was started by the nursing staff. When the emergency response team arrived, Our patient was an 83-year-old male with the patient was found to have circulatory output, although a prior history of acute myocardial infarction, he remained apneic and unresponsive. The trachea was coronary stenting, pacemaker placement following intubated after the administration of fentanyl, propofol third degree atrioventricular block, atrial fibrillation, and rocuronium using a rapid sequence intubation (RSI) transient ischemic attack, mild chronic renal procedure. When more than two hours following the insufficiency (with an estimated glomerular filtration NMBA administration, the patient failed to show any rate (eGFR) of 50 mL/min/1.73 m2) and bilateral endogenous respiratory drive on a pressure support cataract surgery. At admission he weighed 55 kg ventilation (PSV) setting on the ventilator, a further diagnostic workup was performed. The patient remained for a length of 185 cm. He had been hospitalised paralyzed for a prolonged period of time. Objective two days earlier due to a fall. In the latest months, monitoring of NMB, using a peripheral nerve stimulator the patient became more and more anorexic and and accelerometer, was instrumental in differentiating cachectic. The family noted a decline in cognitive the condition from other underlying causes of apnea. functioning that was yet undifferentiated. Four Conclusion :This case helps to illustrate the benefits days prior to the hospitalisation, a mildly sedating of objective measurement of any residual NMB in the agent (mirtazapine) had been started by his primary emergency department, especially when identifiable risk care physician because of a dysregulation of the factors for prolonged action of an NMBA are present. circadian sleeping pattern. The family reported a steep decline in gait stability and general condition Key words : Neuromuscular block ; curare ; neuromus- since the first intake of mirtazapine. In the ED no cular monitoring. posttraumatic complications were withheld due to the fall. Initial physical examination revealed Introduction a pronounced cachexia and general weakness. At the time of admission, no focal neurologic findings Multiple causes of prolonged duration of action were withheld, nor were there any signs of infection. of neuromuscular blocking agents (NMBA) have Further diagnostic work up revealed a presumed been described in the past. Respiratory acidosis has been proven to prolong the action of vecuronium (1). Fabian Dhondt, MD ; Marc Sabbe, MD, PhD In patients over 75 years of age, NMBA are known Department of Emergency Medicine, UZ Leuven, Leuven, Belgium to be able to have an extended duration of action Corresponding author : Fabian Dhondt, Depart-ment of (2). A combination of factors could severely prolong Emergency Medicine, UZ Leuven, 3000 Leuven, Belgium. the neuromuscular blockade (NMB). The benefits E-mail : [email protected] of objectively monitoring NMB before tracheal Paper submitted on Aug 18, 2020 and accepted on Aug 25, extubation are well known in a postoperative setting 2020. (3, 4). However, in an emergency department (ED) Conflict of interest : none.

28-Dhondt.indd 217 12/01/2021 12:29 218 f. dhondt & m. sabbe chronic and metabolically compensated respiratory bronchopneumonia. A repeated blood biochemistry acidosis. There was no history of obstructive or panel demonstrated an increase of CRP to 80 mg/L. restrictive pulmonary pathology nor were there any A tentative diagnosis was made of a combined clinical signs of current respiratory infection. Further type 1 and type 2 respiratory failure due to bron- laboratory testing demonstrated a mild macrocytic chopneumonia leading to severe hypercapnia, anaemia, no significant signs of inflammation. Any coma and subsequent respiratory arrest. The patient sedating agents were discontinued, and the patient was started on antibiotics appropriate for a lower was subsequently admitted to the geriatric ward for respiratory tract infection. Ventilation was titrated further follow up and diagnostic work up for the using serial ABGs to achieve a neutral pH. At two underlying cause of cachexia and cognitive decline. hours post induction, an attempt was made to switch Our emergency response team (ERT) was over to a pressure support ventilation mode. However, called because of a presumed cardiac arrest. One the patient remained apnoeic. At this point, it was hour prior to the call, the geriatrician had been called unclear whether the cause of apnoea was related to to the patient’s bedside due to a severe decrease in an underlying disease process also responsible for the peripheral oxygen saturation (SaO2) to as low causing the respiratory arrest. A CT angiogram of the as 73%. He was started on oxygen supplementation brain was performed and no intracranial bleeding, via an oxygen mask. During a subsequent check nor any sign of a vessel occlusion was observed. A of the patient’s parameters, the nurse found the peripheral nerve stimulator was brought from the patient to be unresponsive and not breathing. Basic operating room to determine the residual NMB. The life support was initiated. When the ERT arrived, nerve stimulated was the right ulnar nerve. The train the first rhythm check was an organised electrical of four ratio was calculated using an accelerometer activity. A pulse check revealed a pulse to be placed at the thumb. At a current of 40 mA the TOF present. However, the patient remained completely ratio was calculated to be 0. At this point a dose of unresponsive and any spontaneous respiration 1g (>16mg/kg) of sugammadex was administered. remained absent. Pupils appeared equal, miotic and Three minutes later the patient resumed spon- unresponsive to light. An ABG was sent and bag- taneous breathing and following a cessation of valve-mask ventilation with an inspired fraction sedation proved to be alert and cooperative. He was of oxygen (FiO2) of 100% was continued. At that subsequently extubated and continued to maintain a time, the patient’s vitals were as follows : blood free airway and continued to spontaneously breathe pressure (BP) was 68/40 mmHg, heart rate was 60/ initially. He responded adequately to questions. min with broad paced complexes, SaO2 was 100%. Over the course of fifteen minutes, however, he The glycemia was 114 mg/dL. Adrenaline was became more and more somnolent. He ended up titrated intravenously (IV) up to a dose of 50 µg. being reintubated because of recurrence of coma Hereafter, the blood pressure increased to 132/74 due to hypercapnia. The patient remained sedated mmHg and the heart rate increased to a rate of and mechanically ventilated until the next day, around 100/min with an irregular rhythm. An ABG when a new attempt was made to wean the patient of showed a pronounced respiratory acidosis with a mechanical ventilation. He was extubated following pH of 7.05 and a partial pressure of arterial carbon a successful spontaneous breathing trial. He was

dioxide (PaCO2) of 143 mmHg. After consulting able to maintain adequate spontaneous ventilation with the treating physician concerning the presence without support for a prolonged period of time in of any therapeutic restrictions, the decision was the ED and was subsequently readmitted to the made to intubate the trachea and ventilate the geriatrics ward. He was further treated for a lower patient. An RSI procedure was performed using respiratory tract infection. An intensive regimen of 100µg of fentanyl, 40mg of propofol and 65 mg of physiotherapy was started. Further evaluation of the rocuronium (1.2 mg/kg). Subsequently, the patient underlying cause of the weight loss and cognitive was intubated and further adrenaline was titrated decline was initiated. In the following days he IV for hemodynamic stabilisation post induction. continued to improve, and he was eventually Hereafter, the patient was transferred to the ED for discharged from the hospital with further follow up further diagnostic workup and close monitoring. A planned in an outpatient setting. propofol continuous infusion was started at a low dose of 10 mg per hour. Discussion In the ED, a further diagnostic work up was performed. A chest X-ray appeared normal. The patient showed a prolonged duration of However, a CT of the chest was compatible with paralysis after administration of rocuronium. The

28-Dhondt.indd 218 12/01/2021 12:29 prolonged effect of rocuronium due to predisposing factors 219 initial problem requiring intubation was coma after the administration of a NMBA. Objective due to hypercapnia complicated by a respiratory monitoring of the residual neuromuscular blockade arrest. In this case, the objective monitoring of the can help distinguish between a state of prolonged residual NMB objectively helped to distinguish paralysis and other underlying causes of apnea and between a severely prolonged paralysis due to unresponsiveness. The use of objective monitoring rocuronium and some other primary underlying of the residual neuromuscular blockade should be pathology that caused the patient to remain apneic considered in the ED before tracheal extubation and unresponsive. The patient had several risk even at longer time intervals after administration of factors for prolongation of paralysis, namely an an NMBA when known risk factors for a prolonged advanced age (2), a severe respiratory acidosis action are present. (1) and reduced kidney function (5). Since these predisposing factors are not uncommon in the ED and considering the fact that clinical assessment has References been shown to be insufficient to identify clinically 1. Yamauchi, M., Takahashi, H., Iwasaki, H. and Namiki, significant residual NMB (4), it seems prudent to A. 2002. Respiratory acidosis prolongs, while alkalosis also objectively monitor for any significant residual shortens, the duration and recovery time of vecuronium in NMB before attempting tracheal extubation in the humans. J. Clin. Anesth. 14(2):98-101. 2. Stankiewicz-Rudnicki, M. 2016. Neuromuscular blockade in ED even at longer time intervals post administration the elderly. Anaesthesiol Intensive Ther 48(4):257-260. of a NMBA, certainly when any or more of the 3. Naguib, M., Brull, S. J. and Johnson, K. B. 2017. Conceptual known risk factors for a prolonged duration of and technical insights into the basis of neuromuscular monitoring. Anaesthesia. 72 Suppl 1:16-37. action are present. 4. Debaene, B., Plaud, B., Dilly, M. P. and Donati, F. 2003. Residual paralysis in the PACU after a single intubating dose Conclusion of nondepolarizing muscle relaxant with an intermediate duration of action. Anesthesiology. 98(5):1042-8. 5. Khuenl-Brady, K. S. and Sparr, H. 1996. Clinical pharma- There are several identifiable risk factors cokinetics of rocuronium bromide. ClinPharmacokin. that may lead to a prolonged duration of paralysis 31(3):174-83.

28-Dhondt.indd 219 12/01/2021 12:29

(Acta Anaesth. Belg., 2020, 71, 221-229) Systemic clonidine as an adjuvant for postoperative analgesia in ambulatory surgery : a qualitative literature review

N. D’hondt, K. Vermeulen, M. Van de Velde

Abstract : Background : The alfa 2 agonist, clonidine is The role of day case surgery in health care a valuable adjuvant drug during anaesthesia. It has been is becoming increasingly important. Adapted used as sedative and analgesic drug via the parenteral anaesthesia techniques and short-acting products or perineural route. It most likely is also a potent are therefore necessary. It is common knowledge postoperative analgesic drug. that opioid-sparing techniques are a crucial element Objectives : We performed a qualitative systematic in reducing recovery time, which is a major goal review of the literature to evaluate the role and place of in the ambulatory setting of surgery. Clonidine is oral or systemic clonidine for postoperative analgesia during day case surgery. We specifically tried to answer one of the products participating in a multimodal the question if clonidine could reduce postoperative anaesthesia and analgesia policy to partially replace consumption of analgesics or opioids. these opioids. Moreover there is a synergistic Methods : We searched for clinical prospective rando- interaction between clonidine and opioids. mized controlled trials, published between 1991 and Therefore coadministration enhances the effects of 2018, testing systemic clonidine in the surgical population postoperative analgesia by morphine (2). on opioid consumption, postoperative pain intensity with This review focuses on the use of systemic consideration for possible negative effects. Used terms clonidine perioperatively, specifically in day case were “intravenous clonidine postoperative” and “oral surgery. We will try to provide an answer to the clonidine” in Pubmed database and through the online question if perioperative administration leads to research tool Limo from Catholic University of Leuven. a reduction in the postoperative needs of opioids Results : Sixteen randomized controlled trials were or analgesics without compromising on safety. eventually included. After extrapolation of data, seven of the sixteen trials showed a statistically significant Therefore desired effects on analgesia and side favourable effect of the perioperative use of systemic effects will be analyzed. clonidine on postoperative pain score and/or demonstrated a significant reduction in postoperative need of analgesics Methods or opioid consumption. Side effects such as bradycardia and hypotension were often described, but mostly without After clearly defining our research topic and the need for drug intervention. question in June 2019, we searched articles through Conclusion : Systemic clonidine is generally associated Pubmed and Limo (research tool KU Leuven) with lower pain scores and can be safely used in day case according to the PRISMA recommendations. We surgery. Given the heterogeneity of the research groups searched for prospective randomized controlled in the trials, it was impossible to perform a quantitative clinical trials, testing the effect of perioperative use analysis, which makes further research warranted. of clonidine in surgical population on pain intensity and opioid consumption. We looked for trials Keywords : Clonidine ; intravenous ; oral ; postoperative analgesia ; sedation ; side-effects. testing the effect of systemic clonidine administered intravenously or orally. This was a very important

Introduction

Clonidine is an α2 adrenergic receptor ago- Nina D’hondt MD, Kristien Vermeulen MD, Marc Van de Velde nist. It has been used for years to treat several MD, PhD, EDRA medical conditions such as arterial hypertension, Department of Cardiovascular Sciences, Section Anaesthesiology, Katholieke Universiteit Leuven and panic disorders, hyperactivity, pain and to reduce University Hospitals of Leuven, Herestraat 49 3000, Leuven. symptoms of withdrawal (alcohol, drugs, opioids, Corresponding author : D’hondt Nina, Department of benzodiazepines). It is also used as a sedative (1). Anesthesiology, UZ Leuven, Herestraat 49, 3000 Leuven, Belgium. E-mail : [email protected]

29-D'hondt.indd 221 12/01/2021 12:29 222 n. d’hondt et al. criterium, since most trials survey the effect of neuraxial administered clonidine. We searched Pubmed and Limo using terms “intravenous clonidine postoperative”, “oral clonidine”, “postoperative analgesia” and combinations of those. Inclusion criteria were clinical trials, peer- reviewed, English-language, published between 1991 and 2018. However, since clonidine is rather an old product, frequently used in the nineties, many studies date from that period. For this research, we supplemented the clinical trials with more recent articles, such as existing systematic reviews and meta-analyses, by manual reference searching to carry out a literature review.

Results

The search led to the retrieval of ninety studies. Only randomized double-blind controlled trials were retained. Nineteen of them were eligible for further analysis. Three of them were yet excluded because of insufficient quality (jaded score <3), eventually leading us with sixteen trials (Fig 1, table 1). Sixteen randomized controlled trials were analysed. They contained 951 study patients in total Fig. 1. — PRISMA Flowchart of retrieved, excluded and (table 1). After extrapolation of data, seven trials analysed trials. showed a favourable effect of the perioperative use of systemic clonidine compared to control group on postoperative pain score and/or demon- Effect on pain scores strated a reduction in postoperative need of analgesics or opioid consumption (table 2). In seven trials (3-9) systemic clonidine resulted Below, we will discuss the most notable results in a significantly lower pain score compared to the by formulating an answer to relevant questions placebo group (table 2, 3). In four of those seven regarding the use of clonidine in day case surgery. studies clonidine was administered during surgery, so

Table 1. Details Included Randomized Controlled Trials

29-D'hondt.indd 222 12/01/2021 12:29 clonidine as an adjuvant for postoperative analgesia in ambulatory surgery 223

Table 2. Primary, secondary endpoints and adverse effects of analysed trials. plus C = more in clonidine group; min C = less in clonidine group ; ND = no difference of statistical significance ; PL = placebo; x = no registration; PO = postoperatively; * Mean HR significantly lower in clonidine group, but not <60bpm ; ** but higher PCA clonidine doses needed in IV than EPI route ; *** mild bradycardia, mild hypotension, without need of medical intervention ; ° first 6h postoperatively. END POINTS ADVERSE EFFECTS STUDY opioid requirement PO pain score PO satisfaction ; score anxiety PO bradycardia hypotension sedation PO PONV respiratory depression Striebel W.

Table 3. Numerical data on analgesia and opioid consumption of the 7 favourable RCTs. PO = postoperatively ; VAS = visual analogue scale ; VRS = verbal rating scale ; OPS = objective pain scale ; cl = clonidine; pl = placebo; NS = not significant.

intraoperatively. In those four studies (3, 4, 6, 7) the significantly more episodes of hypotension and loading dose of clonidine was at least 3µg kg-1 and in higher level of sedation postoperative. three studies it was followed by a continuous infusion Patients satisfaction, measured by a satis- of clonidine. They all reported a reduction in pain faction score, was a secondary endpoint in five score and in opioid consumption postoperatively. (3, 4, 6, 10, 11) out of the sixteen RCT’s. In three Marinangeli et al published a study (7) where they studies a significantly higher score was found in the compared a placebo group with 3 groups, receiving clonidine group compared to the control group (3, 4, different loading doses of clonidine, respectively 6). Logically, these three studies also demonstrated 2, 3 and 5 µg kg-1, followed by a fixed continuous a lower pain score and lower opioid consumption as infusion. There was a significant difference in VAS a primary endpoint. and opioid consumption as from 3µg kg-1 loading On the other hand, Mayson et al could not prove dose. In the 5µg kg-1 group pain scores were not any significant difference between the placebo and lower than in the 3µg kg-1 group, but there were clonidine group, nor in satisfaction score nor in pain

29-D'hondt.indd 223 12/01/2021 12:29 224 n. d’hondt et al. score and opioid consumption. In the latter case, scores and hemodynamic side effects did not differ oral clonidine was administered at a dose of 3 µg between the three groups. This study suggests that kg-1 90 minutes prior to surgery in adults undergoing not perineural clonidine, but systemic clonidine prostatectomy (11). A possible explanation could be prolongs postoperative analgesia after a peripheral that the dose of clonidine is too low in relation to nerve block. the scope of surgery to obtain a clinically significant – Mikawa et al studied the effect of pre- effect. operatively (105 minutes before surgery) admi- A trial by Bernard et al (12) showed a lower nistered oral clonidine in different doses compared pain score after administration of high dosed i.v. to a placebo group for minor surgery in children clonidine in the postoperative setting, but it was between 5 and 12 years old. They concluded that not a placebo-controlled study. The other group clonidine causes significantly lower pain scores received epidural clonidine, also administered post- postoperatively, directly proportional to preoperative operatively. There was no significant difference degree of sedation, at a dose of 4µg kg-1. Clonidine in pain scores between those two routes of admi- at 2µg kg-1 could not show any significant difference nistration ; but higher PCA doses of clonidine in pain scores. Mild hypotension and bradycardia were needed in intravenous than epidural route. was more frequent in the 4µg kg-1 group, but without This thesis is also confirmed by a study carried out the need for any drug intervention. by M De Kock in 1993 (10) comparing the same dose of epidurally and intravenously administered Opioid consumption clonidine. Opioid consumption perioperatively and postoperatively (during first six hours) was signi- As described above, seven out of sixteen ficantly lower in the epidural compared to the i.v. randomized controlled trials showed statistically group, although VAS did not differ between the two significant lower postoperative pain scores after groups. systemic administration of clonidine in the Striebel et al (13) published a RCT showing no perioperative setting. Five (3-7) of them actually significant difference in postoperative pain score, showed a reduction in opioid consumption. In one neither in opioid consumption, between the placebo study opioid consumption was not included as an group and the group who received intraoperatively endpoint in the study, in the other study no reduction one bolus of clonidine IV. One side note here is that could be demonstrated despite lower pain scores the bolus was a fixed dose of 150 µg of clonidine, (table 2, 3). so not adjusted to body weight. On the other hand, hypotension and bradycardia did occur more PONV frequently in the clonidine group (table 3). Five (8, 9, 11, 14, 15) out of the sixteen Incidence of PONV was included as a studies investigated the effect of preoperatively secondary endpoint in eleven of the sixteen trials administered clonidine. Two (8, 9) of those studies (table 2). Nine of them could not demonstrate any showed a significant decrease in postoperative difference between the clonidine group and the pain scores, but without significantly reduced control group(s). Two trials did : opioid consumption (no endpoint in one study, no – M. De Kock et al (3) published a study significant difference in the other study) : in 1994 where 80 male adults, scheduled for a – Mannion S et al published an RCT con- hemicolectomy, were included and divided into sisting of 36 adult patients undergoing hip surgery, 4 groups. The study was able to demonstrate a who were divided into three study groups : one significant beneficial effect of the intraoperative use placebo-group, one group receiving i.v. cloni- of i.v. clonidine on the incidence and/or severity of dine and one group receiving perineural clonidine PONV (3). (Psoas Compartment Block). The dose of clonidine – In another study dating from 2018, the in both the i.v. and perineural group was 1 µg kg-1 efficacy of premedication with ondansetron was and was administered 45min prior to surgery. The compared with oral clonidine for prevention interval between onset block and first supplementary of postoperative pain and PONV in 30 adults analgesia postoperatively was longer in the i.v. after orthognathic surgery. The study showed group compared to the other two groups. Also, that ondansetron is more effective in preventing pain score at rest was significantly lower 24hours PON and POV than clonidine. On the other hand, postoperatively in the i.v. group, but with no clonidine was able to prolong analgesia time ; but difference in total morphine consumption. Sedation unfortunately there was no significant difference

29-D'hondt.indd 224 12/01/2021 12:29 clonidine as an adjuvant for postoperative analgesia in ambulatory surgery 225 in severe pain incidence between the two groups. There was however a statistically significant However, it should be noted that the study group increase in incidence of hypotension in three trials was rather small (15). (4, 5, 7), considered as a limiting factor for the use of i.v. clonidine. Looking closer on those three Sedation trials, the following remarks can be made : Data on postoperative sedation were reported – Bharti et al (4) compared the use of cloni- in twelve of the sixteen trials. dine in wound infiltration versus i.v. clonidine, In 6 studies (5, 6, 8, 10, 11, 16) there was no administered at the end of surgery (open chole- significant difference in degree of seda-tion between cystectomy). There was also a third placebo-group clonidine and control group. of 20 adults. They concluded that clonidine 3µg In 4 studies (4, 7, 12, 14) there was a signi- kg-1 provided effective postoperative analgesia with ficantly higher sedation score in patients who a reduced morphine requirement when clonidine were administered systemic clonidine. However, was administered intravenously or in the wound in one study (7) the increase in sedation was infiltration compared with placebo. However, the recorded only from a dose of 5µg kg-1 and incidence of postoperative hypotension was signi- higher ; no significant difference was shown ficantly higher in the i.v. clonidine group compared between placebo, 2µg kg-1 and 3µg kg-1 group. with the other two groups. A study by De Kock M et al (3) even proved – Bernard et al (5) published a trial consisting significantly less sedation in the clonidine group. of 50 ASA I patients undergoing spinal surgery. Akin et al published a trial in 2010 (17) including 25 patients received clonidine 5µg kg-1 in bolus 60 children between 2 and 8 years old undergoing followed by an i.v. infusion of 0.3µg kg-1 h-1 in inguinal hernia repair or orchidopexy surgery. the postoperative period, whereas the other 25 Anaesthesia consisted of general anaesthesia patients received placebo (randomly assigned to with caudal block (levobupivacaine 0.25%). The one of the two groups). They recommend the use subjects were randomly divided into 3 groups of i.v. clonidine in the postoperative setting as part 20 patients in a double blinded way : 1 placebo of a multimodal analgesia policy in major surgery group, 1 group received clonidine i.v. 2µg kg-1 at because of its beneficial effect on pain without the start of surgery, and 1 group received clonidine respiratory depression. However, the trial also 2µg kg-1 caudally. The study showed that the time shows a significant decrease in perioperative blood to first analgesia demand was significantly longer pressure, especially if filling pressure is inadequate, after caudal administration of clonidine, where no which may be seen as a convincing limiting factor difference between placebo and i.v. group could be in major surgery. demonstrated. Besides there was no difference in – The third RCT by Marinangeli et al. (7) sedation score between the placebo and i.v. group, showed a higher incidence of hypotension (and whereas sedation after caudal administration was sedation scores) after the administration of 5µg kg-1 significantly higher at certain times (i.e. 30, 60 and clonidine i.v. in bolus at the start of surgery, which 240 min postoperatively). was not demonstrated in the other three groups where respectively 0, 2 and 3µg kg-1 clonidine was Hemodynamics administered at the same moment – and followed by α2 agonists are associated with a bradycardi- a continuous infusion of clonidine during surgery. zing and hypotensive effect, the reason why they Another trial by Striebel et al. (13) showed are used as an antihypertensive agent. Data on intra- lower heart rates in the clonidine group compared and postoperative hemodynamic effects could be to the placebo group. In this population, 150 µg of extracted from fourteen trials (table 2). Two studies clonidine was given over 30 min at the start of the (15, 18) did not mention any information about it. surgery and afterwards 150 µg over the following By hemodynamic instability we mean the 90 min. There was also a significant increase in occur-rence of bradycardia and hypotension that incidence of hypotension ; but no one had to be require a drug intervention. In the majority of the treated for it. They noted that the drop of blood trials the incidence of hypotension or bradycardia pressure and heartrate arised after the continuous in the clonidine group was similar compared to the infusion. control group (table 2). Respiratory depression Data of five studies (5, 7, 9, 13, 14) demonstrated more events of bradycardia, but without Data on respiratory depression were reported the need for drug intervention. in four trials (5, 6, 12, 16). No study could show

29-D'hondt.indd 225 12/01/2021 12:29 226 n. d’hondt et al. a higher incidence of postoperative respiratory the primary afferent neurons in the dorsal horn of depression after the perioperative use of clonidine. the spinal cord. The presynaptic stimulation of α2 receptors effectuates the inhibition of adenylate Discussion cyclase, reducing intracellular cAMP, leading to hyperpolarization of the membrane and finally A large heterogeneity in study design and decreasing the release of norepinephrine in the patient populations was noted between the dif- central nervous system (19, 20). The way clonidine ferent studies included in our analysis. Hence, a acts when added in peripheral nerve blocks is due to quantitative analysis was not possible. Therefore we an increased effect of local anesthetics, decreasing performed a systematic qualitative literature review the conduction on C (pain) and Aα (motor) fibers and analysis. The review will allows us to formulate (21) and through a smaller vasoconstrictor effect recommendations on the use of systemic clonidine. via α1 adrenergic receptor. However, clonidine has Significantly lower pain scores in the post- a selective affinity to α2 receptors 220 times higher operative period could be observed in 7 RCT’s after than to α1 receptors (22). the administration of systemic clonidine compared to the control group. Only in 6 studies did this By acting via other pathways than involved in also appear to be clinically relevant. A significant opioid analgesia, clonidine may be opioid sparing and clinically relevant reduction in postoperative and lead to a reduction of associated side effects morphine consumption could be demonstrated in such as respiratory depression. Besides, clonidine 5 studies, as shown in table 3. Moreover, in three induces a reduction in blood pressure by various studies a higher satisfaction score, translated as mechanisms : a combination of a decrease in vascu- better patient satisfaction, was retained after the lar resistance, a decrease in venous return, heart rate use of clonidine. On the contrary, there was no and myocardial contractility (5). significant reduction in postoperative opioid con- A recent study from 2019 by Flanders et al. sumption in 5 analysed trials, with underdosing as a (18), examined the effect of dexmedetomidine and possible explanation in at least 3 of them. clonidine on the inflammatory response in critical Data on incidence of bradycardia could be illness. extracted from 13 of the 16 RCT’s of which no α2 agonists are often used at intensive care increase in incidence could be demonstrated in units as sedatives and analgesics during critical 8 studies. In 5 studies a higher incidence of mild illness. Flanders et al investigated the potential anti- bradycardia was retained but without clinical inflammatory response of α2 agonists during critical relevance. Regarding the incidence of hypotension, illness by means of a systematic review (22). This 3 studies showed a clinically relevant increased anti-inflammatory response could potentially hypotension following the use of systemic cloni- have a beneficial effect during and after surgery. dine. One study showed that increased haemo- Data from 30 animal studies and 9 human studies dynamic instability occurred only at high doses, were extrapolated and analysed. In animals it was 5 µg kg-1 to be more specific. However, no shown that many mediators of the inflammatory significant nor clinically relevant difference in cascade (such as serum/tissue TFNα, interleukins, incidence of hypotension could be demonstrated in neurofactors, TLR...) were reduced by α2 agonists. 9 studies. 4 studies showed an increased sedation In human studies, dexmedetomidine reduced CRP, postoperatively following the use of clonidine, in TFNα and certain interleukins. All this evidence one study again only as from a dose of 5µg kg-1. supports the thesis that α2 agonists have potential No significant increase in postoperative respiratory anti-inflammatory effects, with possibly an depression was reported in any trial. additional beneficial perioperative effect, but today the clinical relevance of this is questionable. Mechanism Routes of administration Clonidine is a centrally acting imidazoline α2 adrenergic agonist, with centrally analgesic effects Clonidine can be administered through several (19, 20). Its analgesic effect is mediated through routes : intravenously, orally (liquids, transmucosal, brainstem and spinal α2 adrenergic receptors, which capsules), nasal sprays, rectal suspensions. Trans- are associated with descending pain inhibitory dermal route of clonidine also exist by transdermal pathways arising in the locus coeruleus. These patches, but this is not commonly used. Nasal pathways control neurotransmitter release from sprays and rectal suspensions obtain a rapid

29-D'hondt.indd 226 12/01/2021 12:29 clonidine as an adjuvant for postoperative analgesia in ambulatory surgery 227 onset of action but are associated with unreliable At a dose of 4µg kg-1 clonidine reduced the need for absorption. Oral administration is characterised by additional analgesia with minimal side effects. Oral rapid absorption (peak plasma levels within 60-90 clonidine premedication of 2µg kg-1 was found to min), a bioavailability of around 85%, and half-life be insufficient. In addition, clonidine was found to of around 12 hours. Intravenous administration of be superior to midazolam or other benzodiazepines clonidine is the most predictable and most easily and to i.v .fentanyl for prevention of postoperative dose-adjustable. 23 Average elimination half-life of nausea and vomiting and for emergence agitation. clonidine is 12 hours , providing long lasting stable However, studies with larger populations are analgesia (5). warranted (23). The indication of clonidine as a premedication Time of administration for postoperative analgesia is less clear in adults (11, 14, 18. Preoperative use Intraoperative use Given it’s anxiolytic activity, clonidine has In addition to clonidine as a premedication, it often been subject of research on premedica-tion in can be used safely during surgery to reduce the need children undergoing surgery, but it may also help for intra- and postoperative opioids. Intravenous reduce distress in children post-operatively. Oral administration should be at least 3µg kg-1 to provide clonidine should be applied an hour in advance to a postoperative analgesic effect, shown in the trials avoid or treat agitation (23). we investigated. It can be administered as a bolus Only two out of the sixteen trials were charge (over 15 minutes e.g.), and followed by a performed in a child-population. In one clonidine continuous infusion, but in ambulatory setting, the was administered as premedication prior to minor latter seems less appropriate as shown in the trial by surgery. Data from this study (9) showed that Striebel et al (13). They noted that the drop of blood children who were administered oral clonidine 4µg pressure and heartrate arised after the continuous kg-1, had higher preoperative sedation scores, but infusion, which may be an indication to avoid also significant lower pain scores post-operatively continuous infusion perioperatively and limit the and less need for supplementary analgesia. Oral use of clonidine only to a loading dose of 3µg kg-1, clonidine premedication is a good approach in certainly in ambulatory setting. children undergoing minor surgery. This thesis There was no sufficient data to extrapolate is supported by Lambert et al. who published a from our trials about prevention of postoperative review in 2014 investigating the effect of clonidine agitation after intraoperative administration. How- as a premedication in children on postoperative ever, according to the systematic review by Afshari pain. Again, clonidine was recommended as part et al, clonidine at a dose of 3µg kg-1 should be of a multimodal analgesia strategy. Based on this administered 20min before the end of surgery in review, it is likely to be effective and safe providing children older than 1 year to prevent emergence pain relief in children in an adequate dosage, at a agitation. Intraoperative clonidine also prolonged dose of 4µg kg-1 (19). the time to first analgesic administration and In an open label trial by Schmidt et al, the reduced postoperative opioid consumption, while effect of preanesthetic midazolam, clonidine and recovery times were extended by clonidine in this dexmedetomidine on postoperative pain and anxiety child-population. Intraoperative doses of clonidine in children was investigated. Possibly in line with in the literature ranges from 1 to 3µg kg-1 with higher expectations, it was shown that α2 agonists had a doses giving most benefit according to Afshari’s similar anxiolytic effect as midazolam (0.5mg kg- review (23). Although the administration of higher 1), but moreover children given dexmedetomidine doses, like 5 µg kg-1 in the RCT by Marinangeli et or clonidine had less perioperative sympathetic al. (7), also caused significantly more hemodynamic stimulation and consequently less pain postopera- instability. The last trial shows that haemodynamic tively compared to midazolam. In this trial oral effects of clonidine are dose-dependent and can be clonidine was also administered at a dose of 4µg avoided. kg-1 (24). Afshari published a systematic review and Postoperative use meta-analysis in 2019 about the value of clonidine in paediatric anaesthesia. They confirmed again the Only one study 5 from our selection was able previous thesis about oral clonidine premedication. to demonstrate a beneficial effect of postoperative

29-D'hondt.indd 227 12/01/2021 12:29 228 n. d’hondt et al. administration of clonidine on the need of analgesics. 2. Nitta R, Goyagi T and Nishikawa T. 2013. Combination of In this study clonidine was administered at a dose of oral clonidine and intravenous low-dose ketamine reduces -1 the consumption of postoperative patient-controlled anal- 5µg kg , which is quite high. There was a significant gesia morphine after spine surgery. Acta Anaesthesiol difference in postoperative pain score and opioid Taiwanica Off J Taiwan Soc Anesthesiol. 51(1):14-17. consumption compared to the placebo group ; and doi:10.1016/j.aat.2013.03.003 3. de Kock M, Lavandhomme P and Scholtes JL. 1994. this without increased sedation, but unfortunately Intraoperative and PostoperativeAnalgesia Using Intravenous with a higher incidence of hypotension. In minor Opioid, Clonidine and Lignocaine. Anaesth Intensive Care. surgery, such as in the ambulatory setting, we would 22(1):15-21. doi:10.1177/0310057X9402200103 4. Bharti N, Dontukurthy S, Bala I and Singh G. 2013. not recommend to use such high doses of clonidine. Postoperative analgesic effect of intravenous (i.v.) cloni- Besides that we expect less perioperative fluid shifts, dine compared with clonidine administration in wound so we may assume less haemodynamic instability infiltration for open cholecystectomy. Br J Anaesth. 111(4): during the intra- and postoperative period. 656-661. doi:10.1093/bja/aet130 5. Bernard JM, Hommeril JL, Passuti N and Pinaud M. The trial by Salengros et al could not 1991. Postoperative analgesia by intravenous clonidine. demonstrate any benefit from postoperative Anesthesiology. 75(4):577-582. doi:10.1097/00000542- administration. In my opinion this was due to low 199110000-00006 6. De Kock MF, Pichon G and Scholtes JL. 1992. Intra- dosing (16). In the other trials no clear conclusion operative clonidine enhances postoperative morphine could be drawn given too heterogeneous research patient-controlled analgesia. Can J Anaesth J Can Anesth. questions or research groups. 39(6):537-544. doi:10.1007/BF03008314 7. Marinangeli F, Ciccozzi A, Donatelli F et al. 2002. Clonidine According to Afshari’s review clonidine can for treatment of postoperative pain : a dose-finding study. be useful to treat emergence agitation or post- Eur J Pain Lond Engl. 6(1):35-42. doi:10.1053/eujp. operative shivering after paediatric surgery. If no 2001.0270 prior administration of clonidine, give 2-3µg kg-1 8. Mannion S, Hayes I, Loughnane F, Murphy DB and Shorten GD. Intravenous but Not Perineural Clonidine Prolongs intravenously to treat postoperative agitation, with Postoperative Analgesia After Psoas Compartment Block a maximum of 75µg i.v. as initial dose in obese with 0.5% Levobupivacaine for Hip Fracture Surgery: children or adolescents. In case of postoperative Anesth Analg. 2005;100(3):873-878. doi:10.1213/01.ANE. -1 0000143950.13939.7A shivering 1µg kg should suffice (23). 9. Mikawa K, Nishina K, Maekawa N and Obara H. 1996. Oral clonidine premedication reduces postoperative pain Conclusion in children. Anesth Analg. 82(2):225-230. doi:10.1097/ 00000539-199602000-00001 10. De Kock M, Crochet B, Morimont C and Scholtes JL. Based on the present qualitative literature 1993. Intravenous or epidural clonidine for intra- and review, we can conclude that systemic clonidine postoperative analgesia. Anesthesiology. 79(3):525-531. can be a very useful adjunct to multimodal anal- doi:10.1097/00000542-199309000-00016 11. Mayson KV, Gofton EA and Chambers KG. 2002. gesia for postoperative pain management in the Premedication with low dose oral clonidine does not ambulatory surgical setting. Favourable effects on enhance postoperative analgesia of intrathecal morphine. both analgesia and opioid consumption have been Can J Anaesth J Can Anesth. 47(8):752-757. doi:10.1007/ BF03019477 demonstrated, without compromising patient safety. 12. Bernard JM, Kick O and Bonnet F. 1995. Comparison Adequate dosing, 3µg kg-1 in adults and 4µg kg-1 of intravenous and epidural clonidine for postoperative in children, seems necessary to guarantee a robust patient-controlled analgesia. Anesth Analg. 81(4):706-712. doi:10.1097/00000539-199510000-00009 analgesic effect. Higher doses and continuous 13. Striebel WH, Koenigs DI and Kramer JA. 1993. infusions are less suitable in the ambulatory setting, Intravenous clonidine fails to reduce postoperative but could be reserved for major surgery with meperidine requirements. J Clin Anesth. 5(3):221-225. prolonged postoperative monitoring. Preoperative doi:10.1016/0952-8180(93)90019-B 14. Benhamou D, Narchi P, Hamza J, Marx M, Peyrol MT and or intraoperative administration in children and Sembeil F. 1994. Addition of oral clonidine to postoperative intraoperative administration in adults is recom- patient-controlled analgesia with i.v. morphine. Br J mended, however, further study is necessary. Anaesth. 72(5):537-540. doi:10.1093/bja/72.5.537 15. Samieirad S, Sharifian-Attar A, Eshghpour M et al. 2018. Children may benefit the most from this safe and Comparison of Ondansetron versus Clonidine efficacy for slightly sedative analgesic drug, although also in prevention of postoperative pain, nausea and vomiting after adults the opioid sparing effect may be useful. orthognathic surgeries: A triple blind randomized controlled trial. Med Oral Patol Oral Cirugia Bucal. 23(6):e767-e776. doi:10.4317/medoral.22493 References 16. Salengros JC, Hecquet F, Touihri K, Sekkat J, Barvais L and Engelman E. 2011. Low-dose intravenous ketamine 1. Giovannitti JA, Thoms SM and Crawford JJ. 2015. Alpha-2 and clonidine for poor postoperative opioid responsiveness: adrenergic receptor agonists : a review of current clinical a double blind randomized study. Acta Anaesthesiol Belg. applications. Anesth Prog. 62(1):31-39. doi:10.2344/0003- 62(2):65-72. 3006-62.1.31

29-D'hondt.indd 228 12/01/2021 12:29 clonidine as an adjuvant for postoperative analgesia in ambulatory surgery 229 17. Akin A, Ocalan S, Esmaoglu A and Boyaci A. 2010. The 24. Schmidt AP, Valinetti EA, Bandeira D, Bertacchi effects of caudal or intravenous clonidine on postoperative MF, Simões CM and Auler JOC. 2007. Effects of analgesia produced by caudal levobupivacaine in children. preanesthetic administration of midazolam, clonidine, or Pediatr Anesth. 20(4):350-355. doi:10.1111/j.1460-9592. dexmedetomidine on postoperative pain and anxiety in 2010.03259.x children. Paediatr Anaesth. 17(7):667-674. doi:10.1111/j. 18. Holthusen H, Backhaus P, Boeminghaus F, Breulmann 1460-9592.2006.02185.x M and Lipfert P. 2002. Preemptive analgesia: no relevant 23. Khafagy, Hanan F., Reeham S. Ebied, Emad S. Osman, advantage of preoperative compared with postoperative Mohamed Z. Ali and Yasser M. Samhan. 2012. “Peri- intravenous administration of morphine, ketamine, and operative Effects of Various Anesthetic Adjuvants with clonidine in patients undergoing transperitoneal tumor TIVA Guided by Bispectral Index.” Korean Journal of nephrectomy. Reg Anesth Pain Med. 27(3):249-253. Anesthesiology 63 (2): 113-19. doi:10.1053/rapm.2002.30669 24. Nishina Kahoru, Katsuya Mikawa, Makoto Shiga, Yumiko 19. Lambert P, Cyna AM, Knight N and Middleton P. 2014. Takao, Nobuhiro Maekawa and Hidefumi Obara. 2000. Clonidine premedication for postoperative analgesia in “Diclofenac and Flurbiprofen with or without Clonidine for children. Cochrane Database Syst Rev. (1):CD009633. Postoperative Analgesia in Children Undergoing Elective doi:10.1002/14651858.CD009633.pub2 Ophthalmological Surgery.” Paediatric Anaesthesia 10 (6): 20. Sanchez Munoz MC, De Kock M and Forget P. 2017. What 645-51. is the place of clonidine in anesthesia? Systematic review 25. Nitta Rie, Toru Goyagi and Toshiaki Nishikawa. 2013. and meta-analyses of randomized controlled trials. J Clin “Combination of Oral Clonidine and Intravenous Low- Anesth. 38:140-153. doi:10.1016/j.jclinane.2017.02.003 Dose Ketamine Reduces the Consumption of Postoperative 21. Koyyalamudi V, Sen S, Patil S et al. 2017. Adjuvant Agents Patient-Controlled Analgesia Morphine after Spine Sur- in Regional Anesthesia in the Ambulatory Setting. Curr gery.” Acta Anaesthesiologica Taiwanica 51 (1): 14-17. Pain Headache Rep. 21(1):6. doi:10.1007/s11916-017- 26. Blaudszun, Grégoire, Christopher Lysakowski, Nadia 0604-1 Elia and Martin R. Tramèr. 2012. “Effect of Perioperative 22. Flanders CA, Rocke AS, Edwardson SA, Baillie JK and Systemic Α2 Agonists on Postoperative Morphine Con- Walsh TS. 2019. The effect of dexmedetomidine and sumption and Pain Intensity.” Anesthesiology 116 (6): clonidine on the inflammatory response in critical illness: 1312-22. a systematic review of animal and human studies. Crit Care 27. Legrain C, Nkiko G, Lauwick S, Kaba A and Joris J. Lond Engl. 23(1):402. doi:10.1186/s13054-019-2690-4 2010. “Effects of intranvenous clonidine on postoperative 23. Afshari A. 2019. Clonidine in pediatric anesthesia: the analgesia, sedation, and haemodynamic instability in new panacea or a drug still looking for an indication? Curr morbidly obese patients undergoing laparoscopic gastric Opin Anaesthesiol. 32(3):327-333. doi:10.1097/ACO. by-pass.” European Journal of Anaesthesiology. 27 (47): 0000000000000724 203.

29-D'hondt.indd 229 12/01/2021 12:29

(Acta Anaesth. Belg., 2020, 71, 231-237) Cerebral oxygenation monitoring during general anesthesia for arthroscopic and open shoulder surgery in beach chair position : a prospective observational study

Y. Vercammen, I. Myatchin, M. Van de Velde, L. Van Hese, S. Coppens

Abstract : Background and objectives : Patients under- Keywords : Cerebral oxygenation ; near-infrared spec- going shoulder surgery in beach chair position are at risk troscopy ; beach chair position ; shoulder surgery. for severe neurological complications due to cerebral ischemia. Near-infrared spectroscopy is a non-invasive technique that enables continuous monitoring of the Introduction regional cerebral oxygenation. In this investigation we tried to determine the incidence of cerebral desaturation The beach chair position is commonly used events (CDEs) during shoulder surgery in the beach chair for arthroscopic and open shoulder surgery. It has position. benefits from both a surgical and anesthetic point The primary outcome of this study was to evaluate the of view, compared to the lateral decubitus position. relationship between arterial blood pressure and cerebral The sitting position provides better access and oxygenation. The secondary outcome was measuring the smoother manipulations with the shoulder, less risk amount of hypotensive and cerebral desaturation events. of injury of the brachial plexus, less bleeding risk Our aim was to improve the monitoring of patients under and accessibility for conversion to open surgery. It general anesthesia (GA) in beach chair position and allows more anesthesia flexibility to easily convert diminish the risk for neurological complications. Methods : We conducted a prospective monocentric to a supine position and provide better airway access observational study. Data were collected of 52 patients if necessary (1, 2, 3). undergoing arthroscopic or open shoulder surgery under However, there are several reports published GA, whether combined with locoregional anesthesia on severe neurological complications ranging (interscalene block) or not. Arterial blood pressure, heart from cerebral infarction to death (4, 5, 6). It has rate, arterial oxygen saturation and regional cerebral been suggested that these complications were due tissue oxygenation (rScO2) were measured at the baseline to cerebral ischemia. Under GA, the beach chair before positioning and inducing anesthesia, and then position is associated with important hemodynamic continuously until the end of the operation. We registered changes resulting in arterial hypotension and the incidence of cerebral desaturation events, where CDE increased risk of cerebral hypoperfusion. (figure was defined as a 20% decrease of the rScO2 from the 1). The reduction in blood pressure and cardiac baseline. Anesthetic management was standardized in output (CO) is assumed to be mainly the result of all patients. Arterial hypotension was managed using a predetermined protocol. venous pooling of blood in the lower extremities Results : The incidence of intraoperative CDEs was (7). This effect will be more pronounced if a higher 28%. There was an overall significant (p<0.0001) correlation between arterial blood pressure and cerebral tissue oxygenation (rScO2). However, no correlation was Yasmine Vercammen, MD ; Ivan Myatchin, MD, PHD ; Marc shown between predefined CDE and those hypotensive Van de Velde, MD, PHD ; Laura Van Hese, Steve Coppens, events. No correlation was shown between CDE and M.D. length of surgery, age, gender, ASA (American Society Corresponding author : Steve Coppens, University Hospitals of Anesthesiologists) classification or BMI (body mass Leuven, Leuven, Belgium. E-mail : [email protected] index). Conclusion : The CDE rate in our series was in the same The authors declare to have no conflicts of interest with respect range as previously reported rates. There was an overall to the research, authorship, and/or publication of this article. Ethics approval and consent to participate. The study was correlation between arterial blood pressure and cerebral approved by the ethics committee of the University hospital tissue oxygenation (rScO2), suggesting the relevance of Leuven and by the Supervisory Committee on Medical of using NIRS during beach chair position to improve ethics of the “Master in de specialistische geneeskunde” patient safety. programme (Faculty of Medicine of University Leuven). Reference numbers: mp12385 (approval in 2017); mp015427 (reapproval in 2019). Written informed consent was obtained from all patients in this study.

30-Vercammen.indd 231 12/01/2021 12:30 232 s. coppens et al. Any contact with hemoglobin leads to a change in the light spectrum, depending upon the oxygenation status of the hemoglobin molecule (10). Reflected light is detected by the light sensors on the NIRS probes. The cerebral oxygenation is calculated using the Beer-Lambert Law (13). Firstly, in this study, we investigated the relationship between arterial blood pressure and cerebral oxygenation. Although correlation has been proven before, most studies involved small numbers Figure 1. — Hemodynamic and cerebral effects of beach chair and did not consider the possible association between position during GA with kind permission by Clara Lobo, cerebral oxygenation and patient characteristics. Anesthesiology Consultant Centro Hospitalar Trás-Os-Montes Secondly, this study is aimed at obtaining a better e Alto Douro, Vila Real, Portugal. view of the frequency, time lapse and size order of cerebral desaturation events (CDE) to improve the angle of upright position is used. The resulting monitoring of cerebral oxygenation and limit the decline in stroke volume (SV) and cardiac output risk of neurological complications during this type (CO=HRxSV) will result in arterial hypotension of surgery. and impact circulation. Physiological adaptive responses, such as the baroreceptor reflex, will be Material and methods attenuated by anesthetics (8). When blood pressure is below the limit of Study design and participants cerebral autoregulation, cerebral perfusion pressure (CPP) and cerebral blood flow (CBF) is depended We conducted a prospective monocentric ob- on the mean arterial blood pressure (MAP) servational study of 52 patients between 2017 and according to the classical equation CPP=MAP- 2020. We included adult patients who underwent ICP. Intraoperative MAPs should be maintained arthroscopic or open shoulder surgery under GA, above the lower limit of cerebral autoregulation to whether combined with a locoregional anesthesia guarantee adequate cerebral perfusion and avoid (LRA) interscalene block (ISB) or not, at the neurological brain damage (7). In addition, we University Hospital of Leuven. All patients were must consider the gravitational effects of the beach informed of the purpose and the course of the study, chair positioning, where the MAP at the level of and the security policy regarding the data acquired. the brain differs from the MAP at the level of the A written informed consent was obtained before heart. A correction factor for the vertical distance surgery. of the blood column between the arm and the brain should be used to calculate the different MAPs Intra-operative monitoring and treatment (figure 1). It is difficult to determine the individual safe range of MAPs during upright positioning (7, We used routine cardiovascular monitoring

9, 10). On top of that, we also have to consider the (electrocardiography, pulse oximetry (SpO2) and effect of hyperventilation during GA, resulting in invasive arterial blood pressure (ABP) measure- cerebral vasoconstriction and diminished cerebral ment) combined with cerebral oximetry with NIRS

circulation (7). (rScO2) and capnography (etCO2). We opted for As advised before in several reports, the use the Foresight (Casmed, USA) as near-infrared of continuous intraoperative monitoring of cerebral spectroscopy to quantify the regional cerebral

oxygenation to appropriately detect cerebral hypo- tissue oxygenation (rScO2). The NIRS electrodes oxygenation and properly manage it, should be were placed before induction of GA. The first

considered routinely (11). A technique which has measurement of the rScO2-level of the awake patient, gained popularity in this field is cerebral oximetry before administration of hypnotics and opioids, was with near-infrared spectroscopy (NIRS). This is used as the baseline value. Cerebral desaturation a non-invasive transcutaneous technique for the event (CDE) was defined as a 20% decrease of

continuous monitoring of the regional cerebral the rScO2-level from the baseline (12). Parameters

oxygen saturation (rScO2) (5, 12). The NIRS device (ABP, SpO2, heart rate, rScO2) were registered at the emits light from a near infrared spectrum (650- baseline before the positioning and the induction 900nm) through the skull and brain tissue (13). of anesthesia, and then every 5 minutes for the

30-Vercammen.indd 232 12/01/2021 12:30 cerebral oxygenation monitoring during general anesthesia 233

duration of the surgical procedure. Mean arterial Results blood pressure was measured by default at the level of the external acoustic meatus at the University Overview Hospital of Leuven. We implemented an acceptable lower threshold of the MAP of 60mmHg (14). A In the period between 2017 and 2020, 52 decrease of arterial blood pressure of more than 20% patients who underwent shoulder arthroscopy in under the acceptable threshold was conventionally beach chair position under GA, completed the treated with bolus injections of Phenylephrine and/ study. Patients characteristics of included patients or Ephedrine, and/or drips with Phenylephrine. are summarized in table 2. Due to missing baseline

values of rSc02, we adjusted the sample size to 43 Anesthesia protocol for statistical analysis (table 3).

All shoulder procedures were performed at the University Hospital of Leuven’s ambulant care centre (“chirurgisch dag centrum, CDC”) under GA, usually in combination with LRA, unless contra-indicated. Prior to anesthesia, patients had to be in a fasting state for 6 hours. A peripheral intravenous line and a radial arterial catheter (20-gauge) was placed before induction of anesthesia. If no contraindications were noticed and if patient consent was obtained, an ISB was facilitated under ultrasound view and standard ASA (American Society of Anesthesiology) moni- Figure 2. — Plot of blood pressure (systolic blood pressure toring. After a bolus administration of Ropivacaine SBP, diastolic blood pressure DBP and mean arterial pressure MAP) and regional cerebral tissue oxygenation (rScO2) vs. 0.5% (15 to 20cc), a catheter was inserted for time (minutes). postoperative continuous infusion. Standard ASA monitoring and NIRS electrodes were placed Table 1. before induction of general anesthesia. After pre- Implemented definitions of CDE and hypotensive events in oxygenation with FiO2=1, general anesthesia this study was induced with a bolus of sufentanil 0.2mg/kg, rScO2 MAP followed by a bolus of propofol 2mg/kg. Tracheal Cerebral <20% of the baseline

intubation was facilitated by a bolus administration desaturation event rScO2 of rocuronium 0.6mg/kg. General anesthesia was Hypotensive event 1 decline >20% of the maintained using sevoflurane, at MAC ≥ 1.0. baseline Analgesia was achieved with intermittent bolus Hypotensive event 2 <60mmHg injections of sufentanil depending on patients’ reactions, spontaneous movements, sweating and/ Table 2. or sudden increase in heart rate or arterial pressure. Patients characteristics All patients were positioned in beach chair after Age, y 47 ± 16 induction. Male/Female 34/18 BMI, kg/m2 27.2 ± 4.4 Statistical analysis ASA I 20 ASA II 26 ASA III 6 Statistical analysis was performed using IBM® SPSS® Statistics. Data were analyzed for Categorical variables are presented as numbers. Continuous variables are presented as mean ± standard deviation. BMI, body mass index; correlation using the Pearson’s coefficient, r. The ASA, American Society of Anaesthesiologists. level of statistical significance of the observed results was set at p<0.05. Comparisons between Table 3. CDE and non-CDE groups were performed using Adjusted sample size because of missing data the 2-sample t-test for continuous data. Number of Left CDE Right CDE hypotensive events 1, decline of the MAP >20% of the 43 43 baseline hypotensive events 2, MAP<60mmHg 43 43

30-Vercammen.indd 233 12/01/2021 12:30 234 s. coppens et al. Table 4. Table 6. Correlation of blood pressure and left cerebral tissue Cerebral desaturation events (CDE), hypotensive events and oxygenation saturation operative data

Blood pressure, mm Hg Correlation P-value Patients experiencing CDE (n=43) 12 (28%) coefficient, r Number of right CDE (n=43) 1.07 ± 2.55 SBP 0.265 <0.0001 Number of left CDE (n=43) 0.98 ± 2.31 Number of hypotensive events, decline of the MAP DBP 0.303 <0.0001 12.58 ± 6.04 MAP 0.376 <0.0001 >20% of the baseline (n=52) Number of hypotensive events, MAP<60mmHg HR, beats/min 0.376 <0.0001 3.87 ± 3.56 (n=52) Correlation was assessed using the Pearson correlation coefficient, .r Patients experiencing CDE and hypotensive event, 12 (28%) decline of the MAP >20% of the baseline (n=43) Patients experiencing CDE and hypotensive, MAP Table 5. 7 (16%) Correlation of blood pressure and right cerebral tissue <60mmHg (n=43) oxygenation saturation Length of surgery, min (n=52) 85.83 ± 30.07 Categorical variables are presented as numbers (%). Continuous Blood pressure, mm Hg Correlation P-value variables are presented as mean ± standard deviation coefficient, r SBP 0.193 <0.0001 DBP 0.239 <0.0001 Table 7. MAP 0.229 <0.0001 Correlation of the number of cerebral desaturation events and HR, beats/min 0.298 <0.0001 hypotensive events 1, decline of the MAP >20% of the baseline

Correlation was assessed using the Pearson correlation coefficient, .r Correlation coefficient, r P-value Left CDE -0.069 0.660 Blood pressure (SBP, DBP and MAP) and regional Right CDE -0.235 0.129 Correlation was assessed using the Pearson correlation coefficient, .r cerebral oxygen saturation (rScO2)

In our series there was an overall correlation Table 8. of the arterial blood pressure values (SBP, DBP and Correlation of the number of cerebral desaturation events and hypotensive events 2, MAP<60mmHg MAP) and the regional cerebral oxygen saturation Correlation coefficient, r P-value (rScO2), (p<0.0001 for left and right rScO2, table Left CDE -0.348 0.022* 4-5). Figure 2 shows the relationship between these Right CDE -0.205 0.188 parameters during time of surgery. Correlation was assessed using the Pearson correlation coefficient, .r

Table 9. Comparisons between the CDE and non-CDE group

1: CDE, 0: NON-CDE. Comparisons were performed using the independent 2-sample t-test for continuous data. BMI, body mass index; ASA, American Society of Anaesthesiologists.

30-Vercammen.indd 234 12/01/2021 12:30 cerebral oxygenation monitoring during general anesthesia 235 Table 10. Comparison of length of surgery between non-CDE and CDE group

1 : CDE, 0 : NON-CDE. Comparisons were performed using the independent 2-sample t-test for continuous data.

Discussion

The beach chair position is commonly used for arthroscopic and open shoulder surgery. Despite the known surgical and anesthetic benefits (1, 2, 3), we still have to consider the possible catastrophic neurological complications. There are several reports published on severe neurological Figure 3. — Comparison of length of surgery between complications from cerebral infarction to death non-CDE and CDE group. (4, 5, 6). There is an increased risk of cerebral hypoperfusion, after moving an anesthetized patient Heart rate (HR) and regional cerebral oxygen from the supine to the sitting position resulting in

saturation (rScO2) arterial hypotension. From this point of view, we should consider the use of continuous intraoperative A significant correlation was also seen between monitoring to appropriately detect cerebral hypo- heart rate and regional cerebral oxygen saturation oxygenation (11). Cerebral oximetry with near- (p<0.0001 for left and right rScO2, table 4-5). infrared spectroscopy (NIRS) allows in a noninvasive way for the continuous monitoring of the Cerebral desaturation events (CDE) regional cerebral oxygen saturation (rScO2). In our investigation 28% (12 of 43 patients) In this study, we tried to confirm previous experienced intraoperative CDE (table 6). reporting of the correlation between arterial blood In general, no correlation was shown between pressure and cerebral oxygenation. Our results the predefined cerebral desaturation events (CDE) indeed represented an overall significant association and hypotensive events (table 7-8), except for of the blood pressure values (SBP, DBP and MAP)

one significant correlation that was seen between and the regional cerebral oxygen saturation (rScO2) left CDE and hypotensive event 2, defined as (p<0.001). Furthermore, a significant link was seen MAP<60mmHg (p=0.022, table 8). between the heart rate and regional cerebral oxygen

We tried to determine differences between saturation (p<0.0001 for left and right rScO2, table the CDE and non-CDE group. ASA classification, 4-5). gender, length, weight, BMI and age were not We analyzed the frequency, time lapse and size predictive of CDE (table 9). None of these patient order of CDE, predefined as a 20% decrease of the

variables demonstrated a statistically significant rScO2-level from the baseline (12). The CDE rate difference between the CDE group (n=12) and the in our series was in the same range of previously non-CDE group (n=31). There was no difference reported rates, as summarized by Salazar et al. between the non-CDE and the CDE group in terms (15). We tried to determine substantial differences of the length of surgery (p = 0.802, table 10). The between patients with and without CDE events. mean length of surgery in the non-CDE group was In our study, ASA classification, gender, length, 94.35 minutes and 75.58 minutes (range, 75-112 weight, BMI and age turned out not to be predictive minutes) in the CDE group (Fig. 3). of CDE. This can possibly be explained by the

30-Vercammen.indd 235 12/01/2021 12:30 236 s. coppens et al. relatively small sample size. Secondly, all patients hypotensive events seems to be warranted at this came from a relatively healthy population, and, moment. Further research could help us enhance our therefore, received surgery in our one-day care knowledge of cerebral perfusion and avoid potential surgery unit. Interpretations should, therefore, be disastrous outcomes after beach chair positioning. considered with care. Remarkably, the length of surgery in the Acknowledgements non-CDE group was longer than that of the CDE group. However, this did not lead to any statistical The authors sincerely thank all colleagues significance. The trend that overall duration of the of the anesthesiology and orthopedic surgery beach chair position does not negatively influence department of the University hospital of Leuven for CDE’s is of course very interesting and should be their assistance with patient management for this investigated further. A possible explanation could study. be that hemodynamic equilibrium is reached after a certain threshold of surgery duration. There was a significant correlation of arterial blood pressure References and regional cerebral oxygen saturation. However, 1. Rains, D. D., Rooke, G. A., and Wahl, C. J. 2011. the predefined events of cerebral desaturation and Pathomechanisms and complications related to patient arterial hypotension were not correlated. This positioning and anesthesia during shoulder arthroscopy. could lead to the conclusion that cerebral perfusion Arthroscopy, 27 (4), 532-41. 2. Iamsumang, C., and Chernchuijit, B. 2016 The Supine monitoring is a more appropriate and reliable Position for Shoulder Arthroscopy. Arthroscopy Techniques, predictor of CDE’s than invasive blood pressure 5 (5), 1117-1120. monitoring. Nonetheless, previous research did 3. Peruto, C. M., Cicotti, M. G., and Cohen, S. B. 2009. Shoulder arthroscopy positioning : lateral decubitus versus not find a significant clinical impact on cognition beach chair. Arthroscopy, 25 (8), 891-6. of the patient after CDE’s (16). We still found one 4. Aguirre, J. A., Etzensperger, F., Brada, M., Guzzella, S., significant correlation of left CDE hypotensive Saporito, A., and Blumenthal, S., et al. 2019. The beach event 2, defined as MAP<60mmHg (p=0.022). We chair position for shoulder surgery in intravenous general anesthesia and controlled hypotension: Impact on cerebral hypothesize that this is mainly due to data loss and oxygenation, cerebral blood flow and neurobehavioral small sample size, however, further research could outcome. Journal of Clinical Anesthesia, 53, 40-48. be warranted. 5. Salazar, D., Sears, B. W., Andre, J., Tonino, P., and Marra, G. 2013. Cerebral desaturation during shoulder arthroscopy: a prospective observational study. Clinical Orthopaedics and Limitations Related Research, 471, 4027-4034. 6. Papadonikolakis, A., Wiesler, E. R., Olympio, M. A., and Poehling, G. G. 2008. Avoiding catastrophic complications There was no control group in the lateral of stroke and death related to shoulder surgery in the sitting decubitus position, making it impossible to evaluate position. Arthroscopy, 24, 481-482. the hemodynamic and cerebral effects of moving an 7. Murphy, G. S., Greenberg, S. B., and Szokol, J. W. 2019. anesthetized patient into a sitting position. Missing Safety of Beach Chair Position Shoulder Surgery: A Review of the Current Literature. Anesthesia and Analgesia, 129 data resulted in a smaller study population than (1), 101-118. anticipated. Adjusting sample size because of data 8. Soren, L. L., Lyngeraa, T. S., Maschmann, C. P., Van loss could have led to biased conclusions. Lieshout, J. J., and Pott, F. C. 2014. Cardiovascular consequence of reclining vs. sitting beach-chair body The use of NIRS for continuous monitoring position for induction of anesthesia. Frontiers in Physiology, of cerebral oxygenation only evaluates perfusion in 5, 187. the frontal cortex, therefor missing potential other 9. Laflam, A., Joshi, B., Brady, K., Yenokyan, G. , Brown, C., and Everett, A., et al. 2015. Shoulder surgery in the perfusion deficits (13). beach chair position is associated with diminished cerebral autoregulation but no differences in postoperative cognition Conclusion or brain injury biomarker levels compared with supine positioning: the anesthesia patient safety foundation beach chair study. Anesthesia and Analgesia, 120 (1), 176-185. The CDE rate in our series was in the same 10. Miller, R. D., and Stoelting, R. K. 2000. Basics of anesthesia range of previously reported rates. The lack of 5th (Fifth) edition. Cerebral physiology and the effects correlation between the CDE and hypotensive of anesthetics and techniques. New York, NY: Churchill Livingstone. events suggests that both monitors can be beneficial 11. Salazar, D., Sears, B. W., Aghdasi, B., Only, A., Francois, when evaluating the cerebral perfusion status. A., and Tonino, P., et al. 2013. Cerebral desaturation events However, previous research failed to find any during shoulder arthroscopy in the beach chair position: patient risk factors and neurocognitive effects. Journal of clinical implications on cognition after CDE’s. Only Shoulder Elbow Surgery, 22, 1228-1235. strict blood pressure monitoring and prevention of 12. Pant, S., Bokor, D., J., and Low, A. K. 2014. Cerebral

30-Vercammen.indd 236 12/01/2021 12:30 cerebral oxygenation monitoring during general anesthesia 237 oxygenation using near-infrared spectroscopy in the beach- hypotension. Anesthesiology. 119 (3), 507–515. chair position during shoulder arthroscopy under general 15. Salazar, D. H., Davis, W. J., Ziroğlu, N., and Garbis, N. anesthesia. Arthroscopy, 30 (11), 1520-1527. 2019. Cerebral Desaturation Events During Shoulder 13. Ghosh, A., Elwell, C., and Smith, M. 2012. Cerebral Near- Arthroscopy in the Beach Chair Position. Journal of the Infrared Spectroscopy in Adults: A work in progress. American Academy of Orthopaedic Surgeons, 3 (8), e007. Anesthesia and Analgesia, 115 (6), 1373-1383. 16. Cox, R. M., Jamgochian, G. C., Nicholson, K., Wong, J. 14. Walsh, M., Devereaux, P. J., Garg, A. X., Kurz, A., Turan, C., Namdari, S., and Abboud, J. A. 2018. The effectiveness A., and Rodseth, R. N., et al. 2013. Relationship between of cerebral oxygenation monitoring during arthroscopic intraoperative mean arterial pressure and clinical outcomes shoulder surgery in the beach chair position: a randomized after noncardiac surgery: toward an empirical definition of blinded study. Journal of Shoulder Elbow Surgery, 27 (4), 692-700.

30-Vercammen.indd 237 12/01/2021 12:30 ACTA ANÆSTHESIOLOGICA BELGICA 2020 – 71 – Supplement 1

CONTENTS S

Near-infrared spectroscopy for monitoring tissue oxygenation in E. Thys, J. De Puydt, F. Thiessen, 1 breast reconstruction T. Tondu, V. Saldien

Regional anesthesia of the hip joint : review of regional anesthesia C. Coucke, M. Desmet, K. Ver- 7 techniques meylen, M. Van de Velde

Sevoflurane consumption with the Flow-i ventilator in two versions F. De Rydt, A. Kalmar, M. Van de 15 of automatic gas control algorithms and two settings of manually con- Velde trolled anesthesia : an economic and ecological assessment

Case report and narrative review on regional anesthesia for closed J. Kennes, S. Coppens, J. Deprest, 21 shoulder reduction M. Van de Velde

The added value of simulation-based medical education : a quality R. De Meirsman, E. Van Gervenr 29 improvement study

Comparison of the analgesic effects of perineural and intravenous I. Piepers, M. Desmet, M. Van de 37 dexamethasone on peripheral nerve block : a systematic review Veldet

Pain management after open inguinal hernia repair : an updated S. Coppens, J. Gidts, P. Huynen, 45 systematic review and procedure-specific postoperative pain M. Van de Velde, G.P. Joshi, on management (PROSPECT/ESRA) recommendations behalf of the PROSPECT Working Group collaboration : G.P. Joshi, E. Pogatzki-Zahn, M. Van de Velde, S. Schug, H. Kehlet, F. Bonnet, N. Rawal, A. Delbos, P. Lavand’homme, H. Beloeil, J. Raeder, A. Sauter, E. Albrecht, P. Lirk, D. Lobo, S. Freys

The effect of dexamethasone on post-tonsillectomy pain, nausea and L. Beckers, V. Saldien, S. Maes 57 vomiting in children : a review of the evidence

Best access for interscalene plexus block : a prospective observational P. Waelkens, D.S. Dirzu, S. Coppens 67 study

Virtual reality hypnosis for postoperative pain after total knee arthro- L. Depauw, A. Bosteels, S. Maes, L. 73 plasty Sermeus, V. Saldien

Topographic assessment of the medial antebrachial cutaneous nerve at B. De Breuker, N. Mintjens, L. 79 the axilla relative to the axillary artery and vein using ultrasonography Sermeus, M. Breebaart

Efficacy and safety of spontaneous ventilation using dexmede- T. Bongaerts, S. Maes, S. Aerts , D. 87 tomidine-remifentanil in children undergoing direct laryngoscopy for Wildemeersch, V. Saldien surgical procedures : a single cohort observational study

Comparative study on different routes of intraoperative administration B. Clinck, M. Hoffman, M. Jans, 93 of Ropivacaine to reduce postoperative pain in gynecological laparo- H. Coppejans, A. De Bruyn, Y. scopy - a double blinded randomized trial Jacquemyn, K. Wouters, S. Mor- rison, V. Saldien

The systemic inflammatory response to cardiopulmonary bypass : a B. Keunen, M. Van de Velde, L. Al 105 narrative review Tmimi

31-cont-71-A/Suppl.1.indd 1 12/01/2021 12:31 ACTA ANÆSTHESIOLOGICA BELGICA 2020 – 71 – Supplement 1

Rapid sequence induction and intubation in children revisited : a nar- A. Dionys, J. Lauweryns 93113 rative systematic review on current literature with emphasis on airway management

A prospective randomized double-blind trial of the efficacy of bilateral O. De Fré, D. Van Aken, B. 121 lumbar erector spinae plane blocks on the 24h morphine consumption Versyck, L. Sermeus, V. Saldien, after posterior lumbar inter-body fusion surgery : preliminary results M.B. Breebaart

Comparison between the different pharmacokinetic models of Propofol K.W. Abdul-Latif (*), M. Van de 129 target controlled infusion : a narrative review Velde (*,**), L. Al Tmimi

Performance analysis of an ischemic stroke management protocol, a D. De Meirsman, S. De Meirsman, 135 single centre study R. De Meirsman, M. Sabbe, P. Vanelderen, D. Mesotten, S. Van Boxstael

Anesthesia in patients with Duchenne muscular dystrophy : a E. Degryse, N. Goemans, G. 143 retrospective analysis of all Duchenne muscular dystrophy patients Dewinter, M. Verhaegen, M. Van undergoing general and locoregional anesthesia in a tertiary referral de Velde center

Comparison of analogue and digital PCIA, pilot study M. Weyns, V. Saldien 153

Comparing ultrasound-guided retroclavicular versus infraclavicular A. Beaumond, S. Coppens 159 brachial plexus blockade : a retrospective analysis of safety and efficacy for distal upper limb surgery

Measurement of ACT via I-STAT after administration of ad-random B. Vanwijnsberghe, H. De Ruyter, 169 Heparin F. De Somer

Local infiltration of anesthetics gives better results for pain and com- S. Pennynck, A.F. Kalmar, P. 179 fort scores then patient controlled epidural anesthesia in total knee Van Overschelde, P. Byn, H. Van arthroplasty Overschelde

The effects of intraoperative use of dexmedetomidine on the incidence J. Hermans, D. Hoogma, S. Devroe, 185 of postoperative delirium and postoperative cognitive dysfunction : a E. Roofthooft, S. Rex, M. Van de systematic review Velde

Epidural ropivacaine infusion as part of a multimodal postoperative M. Tosi, J.T. Bhutia, A. Scholliers, 197 pain treatment following thoracolumbar spinal fusion surgery, a S.P.M.J. Raeymaeckers, P. Forget, randomised controlled trial M. Matic, D. Van Schaik, S. Vanden Eynde Raeymaeckers, T. Opsomer, P. Antunes Da Silva, J. Poelaert

Comparison of opioid based versus multimodal postoperative pain S. Bogaert, H. Peperstraete, A. 203 management in cardiac surgery patients Moerman

Evaluation of the differential effect of phenylephrine on near-infrared J. Vandermeulen, A.T. Moerman 211 spectroscopy (NIRS)-measured tissue blood volumes : a study protocol

Prolonged effect of rocuronium due to predisposing factors, a case F. Dhondt, M. Sabbe 217 report

Systemic clonidine as an adjuvant for postoperative analgesia in N. D’hondt, K. Vermeulen, M. Van 221 ambulatory surgery : a qualitative literature review de Velde

31-cont-71-B/Suppl.1 .indd 1 12/01/2021 12:32 ACTA ANÆSTHESIOLOGICA BELGICA 2020 – 71 – Supplement 1

Cerebral oxygenation monitoring during general anesthesia for Y. Vercammen, I. Myatchin, M. Van S231 arthroscopic and open shoulder surgery in beach chair position : a de Velde, L. Van Hese, S. Coppens prospective observational study

31-cont-71-C/Suppl.1 .indd 1 12/01/2021 12:33