Regional Anesthesia and Total Knee Arthroplasty Anesthetic and Pharmacological Considerations

PDF hosted at the Radboud Repository of the Radboud University Nijmegen

The following full text is a publisher's version.

For additional information about this publication click this link. http://hdl.handle.net/2066/209723

Please be advised that this information was generated on 2021-09-27 and may be subject to change. Regional anesthesia and total knee arthroplasty Anesthetic and pharmacological considerations

Sietske M.K. Bakker Regional anesthesia and total knee arthroplasty Anesthetic and pharmacological considerations

Sietske M.K. Bakker Regional anesthesia and total knee arthroplasty: Anesthetic and pharmacological considerations.

Proefschrift

ter verkrijging van de graad van doctor aan de Radboud Universiteit Nijmegen op gezag van de rector magnificus prof. dr. J.H.J.M. van Krieken, volgens besluit van het college van decanen in het openbaar te verdedigen op Colofon donderdag 21 november 2019 om 10.30 uur precies Cover design and layout by: burorub grafisch ontwerp, Nijmegen Printing: Koninklijke Van der Most BV, Heerde door ISBN: 978-90-8302808-8 Sietske Marieke Klazien Bakker The studies presented in this thesis were performed at the Department of Anesthesiology, geboren op 21 februari 1990 Sint Maartenskliniek, Nijmegen, The Netherlands. te Apeldoorn

The research described in this thesis was supported entirely by internal funds of the Department of Anesthesiology, Sint Maartenskliniek, Nijmegen, The Netherlands.

The publication of this thesis was financially supported by Sint Maartenskliniek, Nijmegen, The Netherlands.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system of any nature, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the holder of the copyright. Promotor: Table of contents Prof. dr. G.J. Scheffer

Copromotoren: Dr. R. Stienstra Chapter 1 General introduction 7 Dr. P.J.C. Heesterbeek (Sint Maartenskliniek, Nijmegen) Chapter 2 Femoral nerve catheter versus local infiltration analgesia in fast 21 Manuscriptcommissie: track total knee arthroplasty: short-term and long-term outcome Prof. dr. B. W. Schreurs (voorzitter) Prof. dr. C. Kramers Chapter 3 Pharmacokinetics of 400 mg ropivacaine after periarticular local 39 Prof. dr. A. Dahan (LUMC) infiltration analgesia for total knee arthroplasty

Chapter 4 Pharmacokinetics of 400 mg locally infiltrated ropivacaine after 51 total knee arthroplasty without perioperative tourniquet use

Chapter 5 Influence of a tourniquet on opioid consumption after local 63 infiltration analgesia for total knee arthroplasty

Chapter 6 Improving performance by monitoring the success rate of 73 peripheral nerve blocks

Chapter 7 General discussion 83

Chapter 8 Conclusions 95

Summary 99 Nederlandse samenvatting 103 Dankwoord 107 Curriculum vitae 113 Bibliography 117 Research Data Management 121 Theses Sint Maartenskliniek 125 7

Chapter 1

General introduction 8 Introduction 9

General introduction 1 In this thesis several aspects providing postoperative pain relief after orthopedic surgery, with specific interest in total knee arthroplasty, will be discussed. This first chapter starts by laying out the different aspects of anesthesia and analgesia, which can be used to create a condition to perform surgery safely and with a minimum of postoperative pain.

Total knee arthroplasty is a commonly performed surgery to improve chronic refractory joint pain and improve mobility, knee function and quality of life in patients suffering from severe osteoarthrosis.1,2 The history of total knee arthroplasty goes back to the late 19th century when the German surgeon Themistocles Gluck implanted the first hinged knee prosthesis, which was back then made of ivory.3 Over the second half of the 20th century the number of total knee arthroplasties performed annually increased exponentially, and this increase is expected to continue in the next decades.4–7 In this period the component materials varied from acryl to cobalt-chrome and finally to the development of the modern metal-on-polyethylene arthroplasty.8 During the seventies and eighties of the 20th century, several improvements in materials, geometry and fixation were established. Since then, developments have merely focused on durability of the prosthesis, accurate sizing, more natural biomechanics and improving range of motion. Despite all the improvements still 20% of patients undergoing total knee arthroplasty are not satisfied with their clinical outcome. Persistent (chronic) pain, reduced quality of life and lack of functional improvement are main predictors of dissatisfaction.9

Total knee arthroplasty is still associated with severe acute postoperative pain. This is not only uncomfortable for the patient, but has several other detrimental effects. In the short run, pain may hamper early mobilization, recovery and hospital discharge. While in the long run, acute postoperative pain is also a predictor of chronification of pain,10 which is a major setback for both patient and society. Therefore adequate postoperative analgesia is mandatory.11

General versus neuraxial anesthesia General anesthesia is characterized by systemic administration of several types of drug, usually at least an opioid and an anesthetic agent. Opioids obtund the pain stimulus, while the administration of an anesthetic agent provides a state of unconsciousness, thus facilitating all kinds of surgery. Neuraxial anesthesia only anesthetizes the lower part of the body without altering the state of consciousness. Therefore, neuraxial anesthesia can only be used for lower abdominal or lower extremity surgery. Neuraxial anesthesia is performed by placing a needle between the spinous processes of two vertebrae and injecting a local anesthetic, with or without adjuvants, into the subarachnoid (for spinal anesthesia) or epidural (for epidural anesthesia) space. (Fig. 1)

Although spinal12 and epidural13,14 anesthesia were described long before the first total knee arthroplasty was performed, general anesthesia was the dominant form in anesthesia utilized for total knee arthroplasty in its early days. However, since the mid-eighties of the last century, neuraxial anesthesia has been gaining popularity.

Compared to central neuraxis blockade, general anesthesia has been associated with higher rates of postoperative nausea, vomiting, and itching. On the other hand neuraxial anesthesia 10 Introduction Introduction 11

is associated with rare but serious complications such as spinal hematoma, epidural abscess, Multimodal analgesia 1 and nerve injury.15 Furthermore, a patient’s coagulation status should be suitable for neuraxial Considering that severe postoperative pain can delay rehabilitation18,19 and is a prognostic 1 anesthesia, which might be hindered by anticoagulant medication or bleeding disorders. factor for chronic pain,10,25,26 postoperative analgesia is a cornerstone in rehabilitation Finally, even in experienced hands, in a small amount of patients the intrathecal space may after surgery. In the past, intermittent opioid administration was the main cure to relieve not be identified, abolishing the possibility to perform neuraxial anesthesia. postoperative pain. In 1987, the use of a novel electric pump to provide patient controlled analgesia (PCA-pump) was first described in the literature.27 Opioids are strong analgesics which can relieve pain adequately, but are commonly associated with side effects such as nausea, vomiting, drowsiness, respiratory depression, obstipation and itching.

To avoid or diminish these unwanted side effects, postoperative opioid consumption has to be reduced. A multimodal analgesic regimen is therefore standard in fast-track protocols, aiming for adequate pain relief with a minimum of side effects. Paracetamol, in combination with a NSAID, (oral) opioids, gabapentin and a local or regional anesthetic technique are frequently used constituents of fast-track protocols. Paracetamol, a NSAID and gabapentin are usually started preemptively before surgery; dexamethasone can be administered intravenously during surgery with the objective of prolonging the duration of local anesthetics and decrease the incidence of postoperative nausea.28 In general, administration of a COX-2 inhibitor before Figure 1. Needle positioning in spinal and epiural anesthesia. surgery decreases postoperative pain scores, opioid consumption, pruritus, and nausea or vomiting without increasing bleeding complications.29 More specifically for total knee When focusing on total knee arthroplasty, large databases have shown positive results at arthroplasty, the use of etoricoxib leads to a decrease in postoperative opioid use without an several levels when performing neuraxial anesthesia for total knee arthroplasty: spinal increase in side effects.30 anesthesia has been shown to be associated with lower wound infection and blood transfusion rates, and a decrease of overall number of complications, length of hospital stay and 30-day Gabapentin can be prescribed to patients undergoing total knee arthroplasty. Although mortality.16 These beneficial effects of neuraxial anesthesia are particularly pronounced in there is no evidence that gabapentin decreases acute postoperative pain,31 it may have some patients with comorbidities.17 In addition, fast-track or enhanced recovery protocols are protective effect against chronification of postoperative pain.32 Ketamine is another drug that designed to enhance postoperative recovery, reduce morbidity and reduce length of hospital may be part of some fast-track protocols, but because of its psychotomimetic side effects33 its stay.18,19 These protocols usually incorporate the use of spinal anesthesia. Spinal anesthesia use is controversial. is usually performed using a single shot technique. After administration of a local anesthetic into the intrathecal space, sensory block is usually intense enough to perform surgery without Regional anesthetic techniques to reduce opioid use the need for additional sedation. Using a hyperbaric solution offers the possibility to establish In enhanced recovery protocols, patients are encouraged to start mobilizing on the day of a predominantly unilateral block by placing the patient in the lateral decubitus position surgery. This is a challenge for anesthesiologists in treating postoperative pain while side for about 20 minutes following intrathecal injection. With this technique the dose of local effects such as drowsiness or muscle weakness impeding the possibility of safe mobilization, anesthetic can be reduced with approximately 30%,20 thus accelerating the regression of should be minimal or preferably absent. sensory and motor block and facilitating early mobilization. Nerve blocks Epidural anesthesia is nowadays usually performed in combination with spinal anesthesia Peripheral nerve blocks anesthetize a particular area of the body by blocking specific nerves (combined spinal-epidural) or with general anesthesia. A benefit of epidural anesthesia to the part of the body that is to be operated. Compared to epidural analgesia, peripheral compared to spinal anesthesia is the option to prolong the duration of analgesia through nerve blocks allow the possibility of targeted analgesia limited to the operated limb, without an indwelling catheter. Several types of drugs21–24 can be administered through the epidural unwanted, bilateral or sympathetic side effects.34 Femoral nerve blocks (FNB), or previously catheter, but often a mixture of a local anesthetic with an opioid is prescribed. However, the the 3-in-1 blocks, are considered to be the golden standard for optimal analgesia after total continuous administration of a local anesthetic not only provides analgesia by blocking sensory knee arthroplasty.35,36 (Fig. 2) nerve fibers, it can also prolong motor and sympathetic blockade, delaying mobilization and recovery as consequence. Mostly, nerve blocks are performed under ultrasound guidance, alone or accompanied by nerve stimulation. With the introduction and development of ultrasound machines tailored Because of the aforementioned benefits, spinal anesthesia is the preferred and standard to regional anesthesia, the success rate of nerve blocks has improved, the onset time and dose anesthetic technique in the fast-track protocols of many centers around the world. of local anesthetic necessary to establish a block have decreased and the chance of incidental vascular punctures is reduced.37,38 Recent functional developments in catheter systems 12 Introduction Introduction 13

have made it easy to place a perineural catheter and provide a continuous peripheral nerve ropivacaine have a more favourable safety profile and both drugs are used in LIA protocols. Of 1 block. A catheter next to the femoral nerve for example provides the option of continuous or the two, ropivacaine is the preferred drug, since its cardio and neurotoxic potential reduced 1 intermittent bolus administration of local anesthetic to extend postoperative analgesia. In compared to levobupivacaine.45,46 general, with the use of a perineural catheter analgesia can be prolonged for several days; as a result, opioid consumption can be reduced as compared to single shot blocks and adverse Structure and mechanism of action effects related to systemic analgesic medication will be less frequently.39–41 Ropivacaine, like all amide-type local anesthetics, contains a lipophilic benzene ring, a hydrophilic aminogroup and an amide bond connecting these two structures. (Fig. 4)

Figure 2. The femoral nerve Figure 3. Innervation of the knee joint Figure 4. Structure of Ropivacaine When a femoral nerve block is used to provide postoperative analgesia, pain in the popliteal fossa is not covered by this block, since this area is innervated by the tibial and peroneal nerve, Ropivacaine exerts is therapeutic effect at the site of injection by blocking voltage-gated both deriving from the sciatic nerve. (Fig. 3) The majority of patients are not hindered by pain sodium channels in the axons. Only ionized ropivacaine is capable of blocking voltage-gated in the popliteal fossa. However, up to 20% of the patients who undergo total knee arthroplasty, sodium channels, impeding membrane depolarisation, which in turn prevents the conduction experience severe postoperative pain in this area.42 To relieve pain in the popliteal fossa, the of electric impulses through the nerve. sciatic nerve can be blocked or the posterior capsule of the knee joint can be infiltrated with a long acting local anesthetic. Pharmacokinetics Immediately after injection, ropivacaine starts to be absorbed into the central compartment. Local infiltration analgesia (LIA) The vascularity of the infiltrated tissue is one of the main factors determining the resorption Femoral nerve block provides sufficient analgesia in most patients undergoing total knee speed.47 Blood flow is fundamental to systemic absorption; the higher the blood flow, the more arthroplasty, however motor function may also be affected, impeding (independent) rapid the rise in plasma concentration. After systemic absorption, ropivacaine is distributed mobilization if and as long as the motor function is diminished. To provide analgesia after through the whole body and can be taken up by virtually all organs. Microsomal cytochrome total knee arthroplasty and allow for early mobilization without impaired motor function, a P-450 enzymes in the liver metabolize ropivacaine, and subsequently the metabolites are novel technique has been developed. This easy and straightforward technique is called local primarily excreted by the kidneys.48 infiltration analgesia.43,44 Local infiltration analgesia is based on the infiltration of a long acting local anesthetic in the tissues surrounding the knee. Local infiltration analgesia exerts Toxicity its analgesic effect directly at the site of injection, providing analgesia without systemic side Once absorbed into the bloodstream, approximately 95% of ropivacaine is protein-bound effects such as drowsiness or nausea associated with high-dose opioids, or quadriceps muscle in plasma, mainly to 1-glycoprotein.47 Only the unbound, unionized fraction is able to cross weakness as a side effect of femoral nerve block. cell membranes and distribute to other tissues. If the plasma concentration of unbound, unionized ropivacaineα exceeds the toxic threshold, symptoms of systemic toxicity may occur Ropivacaine in the central nervous system (CNS) or the heart. Exceeding the toxic threshold may occur Femoral nerve blocks or local infiltration analgesia for total knee arthroplasty are usually extremely rapid, for instance in case of inadvertent intravascular injection. In such cases, toxic performed with a long-acting local anesthetic such as bupivacaine, levobupivacaine or symptoms typically occur within minutes after injection. Systemic toxicity may also develop ropivacaine. Because of its systemic toxic potential and the high dose of local anesthetic needed gradually after a delay of tens of minutes to hours; this can happen when the total dose of for local infiltration analgesia, bupivacaine is not a logical choice. Both levobupivacaine and local anesthetic is too large, or when elimination is delayed. 14 Introduction Introduction 15

Although the incidence of systemic toxicity after local anesthetic use is very low when a low threshold for perioperative tourniquet use, since orthopedic surgeon experience that 1 standard safety precautions are observed, the consequences can be severe.49 Mild toxic effects perioperative tourniquet use contributes to a drier surgical field, which facilitates the surgical 1 are perioral numbness or metallic taste. However, symptoms can rapidly develop from mild procedure. into serious and potentially disastrous complications, such as seizures, respiratory depression and arrhythmias or cardiac arrest.50 In the central nervous system, local anesthetics affect Clinical outcomes of patients undergoing any kind of treatment are not only determined by the balance between inhibitory and excitatory pathways.51 Although CNS toxicity can present guidelines and protocols. Physicians’ individual performances or choices in treatment may immediately with seizures or CNS depression, a more gradual increase in symptoms is usually also have an influence. To improve one’s behaviour, insight in one’s own performance and the seen. Before seizures develop, patients may experience a variety of other symptoms. Early influence on clinical outcome is essential. Differences in technique between physicians and symptoms can include dizziness, and visual or auditory disturbances such as tinnitus. When the effect on outcome for individual patients may be very subtle. However, monitoring and local anesthetic concentration in the CNS continues to increase, muscle twitching, shivering comparing of individual performances in groups of patients might reveal differences in clinical and tremors can be observed. Ultimately grand-mal seizures and CNS depression including outcome between physicians. If properly managed, this can be used as an instrument for inter- the vasomotor center can occur.52 collegial discussion with the aim of improving both individual and collective performance.

The cardiovascular symptoms are usually less pronounced, since the cardiovascular system Until now, evaluation of individual performance to improve collective outcomes65 and inter- is more resistant to local anesthetic toxicity than the CNS. Higher doses of the same drug are colleague comparison of objective parameters are not widely accepted in our healthcare required to provoke cardiotoxic symptoms as compared to CNS symptoms.53,54 The cardiac system. Moreover, still little is known about comparing individual performance, and what toxicity of local anesthetics is caused by several mechanisms. Interference with sodium and effect, if any, it has on the quality of care. These reservations notwithstanding, tools to improve potassium channels results in rhythm disturbances and may initially be characterized by individual and collective performance in terms of outcome are worth investigating. sympathetic activation presenting with hypertension and tachycardia.55 Next, blocking of the cardiac conduction system will result in brady-arrhythmia and ventricular ectopy, and blocking of calcium channels may cause myocardial depression. Local anesthetics also interfere with Rationale and outline of this thesis mitochondrial respiration, inhibiting the production of ATP by oxidative phosphorylation.56 This is particularly important in high-energy tissues such as the heart, where rapid depletion Recently developed enhanced recovery protocols aim for shorter hospital length of stay of ATP stores may contribute to the loss of contractility and aggravate myocardial depression. and better functional recovery.66 Prompt mobilization, preferably starting on the day of Ultimately, the combination of the cardiac (myocardial depression and brady-arrhythmia) surgery, is one of the main pillars of enhanced recovery protocols. A substantial part of the and CNS (depression of the vasomotor center) toxic effects of local anesthetics may cause patients undergoing total knee arthroplasty experience severe postoperative pain, which cardiovascular collapse.51 may impede early mobilization, slow rehabilitation and predispose to the development of persistent pain.25,26 Therefore adequate postoperative analgesia is a cornerstone in total knee Non-pharmacological reduction of postoperative pain arthroplasty rehabilitation. Physicians, and anesthesiologists in particular, are used to relieve postoperative pain with drugs. Our armoury includes several classes of drugs, which can be administered orally Local infiltration analgesia and femoral nerve block are widely accepted techniques to relieve or parenterally, preemptively or as escape medication. Additionally, we try to decrease pain and reduce opioid consumption after total knee arthroplasty, both with their own the demand for these drugs by applying regional or local anesthetic techniques to relieve benefits and disadvantages. The femoral nerve block has replaced the use of an epidural in postoperative pain. many centers, but this technique is still associated with unilateral motor weakness, which may hamper mobilization. Local infiltration analgesia aims to provide an equal analgesic However, there are more ways to reduce the need for escape medication. Patients can benefit efficacy, while avoiding additional motor impairment.43 In these terms local infiltration from every day, drug related and non-drug related, choices we make in clinical practice. analgesia may be the optimal technique to cover pain after total knee arthroplasty. However, The perioperative use of a tourniquet in total knee arthroplasty has long been subject to the analgesic duration of single shot local infiltration analgesia may be shorter than that of a discussion. For decades, tourniquets were used to facilitate a dry surgical field and reduce continuous femoral nerve block. The use of a catheter in the surgical field to provide continuous intraoperative blood loss. However, the first concerns of an altered risk on thrombotic ropivacaine administration (continuous local infiltration analgesia) is still controversial, since events when a tourniquet is used for knee replacement surgery date back to the nineties it may alter the infection risk.67 of the 20th century.57 Tourniquet use is not only associated with a higher risk of thrombotic events,58 the risk of wound infection is also higher,59,60 the risk for nerve damage is increased The optimal analgesic regimen for total knee arthroplasty has to strike a balance between and postoperative rehabilitation may be hindered.61–63 Paradoxically, when using a tourniquet adequate pain relief and minimal interference with the Fast-Track principles of early total (i.e. both intra- and postoperative) blood loss is not decreased as compared to surgery mobilization. To compare the benefits and disadvantages of a femoral nerve block with the without tourniquet and the fixation of the arthroplasty is not improved.58,64 Furthermore, the LIA-technique we designed the study described in Chapter 2. In this chapter, we compare use of a tourniquet is associated with thigh pain.59,63 Despite these considerations there still is periarticular local infiltration analgesia of the total knee with local infiltration of only the 16 Introduction Introduction 17

posterior knee capsule in combination with a femoral nerve catheter in terms of pain and References 1 functional recovery until one year after surgery in patients undergoing primary total knee 1 arthroplasty. 1. Ethgen, O., Bruyère, O., Richy, F., Dardennes, C. & Reginster, J.-Y. Health-related quality of life in total hip and total knee arthroplasty. A qualitative and systematic review of the literature. The Journal of bone and joint To ensure sufficient analgesia with LIA, a high dose of ropivacaine is infiltrated in the surgery. American volume 86–A, 963–74 (2004). tissues surrounding the knee. Often this dose exceeds the maximum recommended dose 2. Harris, W. H. & Sledge, C. B. Total hip and total knee replacement. The New England journal of medicine 323, of 3-4 mg/kg.68,69 When the LIA-technique was introduced in our institution, its safety track 725–31 (1990). record was still relatively short. After a sudden case of systemic toxicity,70 we carried out the 3. Brand, R. A., Mont, M. A. & Manring, M. M. Biographical sketch: Themistocles Gluck (1853-1942). Clinical pharmacokinetic study described in Chapter 3, to gain more insight in the maximum plasma orthopaedics and related research 469, 1525–7 (2011). concentrations and the timeframe in which patients are most at risk to develop a possible 4. Otten, R., van Roermund, P. M. & Picavet, H. S. J. Trends in the number of knee and hip arthroplasties: systemic toxic reaction. considerably more knee and hip prostheses due to osteoarthritis in 2030. Nederlands tijdschrift voor geneeskunde 154, A1534 (2010). Recently, performing total knee arthroplasty without the use of a tourniquet is gaining 5. Kurtz, S. et al. Prevalence of primary and revision total hip and knee arthroplasty in the United States from popularity. Since local blood flow is one of the main determinants of systemic absorption, the 1990 through 2002. The Journal of bone and joint surgery. American volume 87, 1487–97 (2005). presence or absence of a tourniquet may affect the concentration-time curve. In Chapter 4 6. LROI-rapportage Ingrepen 2010-2015. Available at: http://www.lroi-rapportage.nl/knie-aantallen- we describe the pharmacokinetic profile of the same mixture of ropivacaine and epinephrine ingrepen-2010-2015. (Accessed: 5th March 2019) as used in Chapter 3, however in this cohort no tourniquet was used at the time of infiltration. 7. Patel, A., Pavlou, G., Mújica-Mota, R. E. & Toms, A. D. The epidemiology of revision total knee and hip arthroplasty in England and Wales: a comparative analysis with projections for the United States. A study Subsequently Chapter 5 describes the effect of a tourniquet on concomitant opioid use when using the National Joint Registry dataset. The bone & joint journal 97–B(8), 1076–81 (2015). local infiltration analgesia is used to provide postoperative analgesia. 8. Williams, D. H., Garbuz, D. S. & Masri, B. A. Total knee arthroplasty: Techniques and results. Britisch Columbia Medical Journal 52, 447–454 (2010). Chapter 6 discusses the introduction and development of a monitoring system for the 9. Maratt, J. D., Lee, Y., Lyman, S. & Westrich, G. H. Predictors of Satisfaction Following Total Knee Arthroplasty. collective and individual success rates of peripheral nerve blocks and the effect this monitoring The Journal of Arthroplasty 30, 1142–1145 (2015). system has on postoperative pain scores. 10. Althaus, A. et al. Development of a risk index for the prediction of chronic post-surgical pain. European journal of pain (London, England) 16, 901–10 (2012). In Chapter 7 the main results of this thesis are summarized and discussed together with the 11. Andersen, L. Ø. et al. Subacute pain and function after fast-track hip and knee arthroplasty. Anaesthesia 64, recommendations and directions for future research. 508–513 (2009). 12. Bier, A. Versuche über Cocainisirung des Rückenmarkes. Deutsche Zeitschrift für Chirurgie 51, 361–369 (1899). 13. Cathelin, M. Mode d’action de la cocaine injecte daus l’escape epidural par le procede du canal sacre. Comptes Rendues des Seances de la Societe de Biologie et de ses Filliales 43, 487 (1901). 14. Sicard, J. Les injections medicamenteuse extradurales par voie saracoccygiene. Comptes Renues des Seances de la Societe de Biologie et de ses Filliales 53, 396 (1901). 15. Horlocker, T. T. Complications of Regional Anesthesia and Acute Pain Management. Anesthesiology Clinics 29, 257–278 (2011). 16. Memtsoudis, S. G. et al. Perioperative Comparative Effectiveness of Anesthetic Technique in Orthopedic Patients. Anesthesiology 118, 1046–1058 (2013). 17. Pugely, A. J., Martin, C. T., Gao, Y., Mendoza-Lattes, S. & Callaghan, J. J. Differences in Short-Term Complications Between Spinal and General Anesthesia for Primary Total Knee Arthroplasty. The Journal of Bone and Joint Surgery-American Volume 95, 193–199 (2013). 18. Kehlet, H. & Wilmore, D. W. Evidence-Based Surgical Care and the Evolution of Fast-Track Surgery. Annals of Surgery 248, 189–198 (2008). 19. Husted, H. et al. What determines length of stay after total hip and knee arthroplasty? A nationwide study in Denmark. Archives of orthopaedic and trauma surgery 130, 263–8 (2010). 20. Büttner, B., Mansur, A., Bauer, M., Hinz, J. & Bergmann, I. Einseitige Spinalanästhesie. Der Anaesthesist 65, 847–865 (2016). 18 Introduction Introduction 19

21. Nielsen, P. T., Blom, H. & Nielsen, S. E. Less pain with epidural morphine after knee arthroplasty. Acta 41. Salinas, F. V., Liu, S. S. & Mulroy, M. F. The Effect of Single-Injection Femoral Nerve Block Versus Continuous 1 orthopaedica Scandinavica 60, 447–8 (1989). Femoral Nerve Block After Total Knee Arthroplasty on Hospital Length of Stay and Long-Term Functional 1 22. Pettine, K. A., Wedel, D. J., Cabanela, M. E. & Weeks, J. L. The use of epidural bupivacaine following total knee Recovery Within an Established Clinical Pathway. Anesthesia & Analgesia 102, 1234–1239 (2006). arthroplasty. Orthopaedic review 18, 894–901 (1989). 42. Carli, F. et al. Analgesia and functional outcome after total knee arthroplasty: periarticular infiltration vs 23. Eisenach, J. C., Lysak, S. Z. & Viscomi, C. M. Epidural clonidine analgesia following surgery: phase I. continuous femoral nerve block. British journal of anaesthesia 105, 185–95 (2010). Anesthesiology 71, 640–6 (1989). 43. Kerr, D. R. & Kohan, L. Local infiltration analgesia: a technique for the control of acute postoperative pain 24. Mahoney, O. M., Noble, P. C., Davidson, J. & Tullos, H. S. The effect of continuous epidural analgesia on following knee and hip surgery: a case study of 325 patients. Acta orthopaedica 79, 174–83 (2008). postoperative pain, rehabilitation, and duration of hospitalization in total knee arthroplasty. Clinical 44. Kehlet, H. & Andersen, L. Ø. Local infiltration analgesia in joint replacement: the evidence and orthopaedics and related research 260, 30–7 (1990). recommendations for clinical practice. Acta Anaesthesiologica Scandinavica 55, 778–784 (2011). 25. Puolakka, P. A. E. et al. Persistent pain following knee arthroplasty. European journal of anaesthesiology 27, 45. Zink, W. & Graf, B. M. The toxicity of local anesthetics: the place of ropivacaine and levobupivacaine. Current 455–60 (2010). opinion in anaesthesiology 21, 645–50 (2008). 26. Barrington, J. W., Halaszynski, T. M. & Sinatra, R. S. Perioperative pain management in hip and knee 46. Santos, A. C. & DeArmas, P. I. Systemic toxicity of levobupivacaine, bupivacaine, and ropivacaine during replacement surgery. American journal of orthopedics 43, S1–S16 (2014). continuous intravenous infusion to nonpregnant and pregnant ewes. Anesthesiology 95, 1256–64 (2001). 27. Spetzler, B. & Anderson, L. Patient-controlled analgesia in the total joint arthroplasty patient. Clinical 47. Mather, L. E. & Tucker, G. T. Pharmacokinetics and biotransformation of local anesthetics. International orthopaedics and related research 215, 122–5 (1987). anesthesiology clinics 16, 23–51 (1978). 28. Kirksey, M. A., Haskins, S. C., Cheng, J. & Liu, S. S. Local Anesthetic Peripheral Nerve Block Adjuvants for 48. de Jong, R. H. Biotransformation of local anesthetics: General concepts. International anesthesiology clinics Prolongation of Analgesia: A Systematic Qualitative Review. PloS one 10, e0137312 (2015). 13, 1–20 (1975). 29. Lin, J., Zhang, L. & Yang, H. Perioperative Administration of Selective Cyclooxygenase-2 Inhibitors for 49. Liu, S. S. et al. Cardiac Arrest and Seizures Caused by Local Anesthetic Systemic Toxicity After Peripheral Postoperative Pain Management in Patients After Total Knee Arthroplasty. The Journal of Arthroplasty 28, Nerve Blocks: Should We Still Fear the Reaper? Regional anesthesia and pain medicine 41, 5–21 (2016). 207–213 (2013). 50. Neal, J. M. et al. ASRA Practice Advisory on Local Anesthetic Systemic Toxicity. Regional Anesthesia and Pain 30. Munteanu, A. M., Cionac Florescu, S., Anastase, D. M. & Stoica, C. I. Is there any analgesic benefit from Medicine 35, 152–161 (2010). preoperative vs. postoperative administration of etoricoxib in total knee arthroplasty under spinal 51. Brown, D. L. Local Anesthetic Toxicity. in Complications of Regional Anesthesia (ed. Finucane, B. T.) 61–73 anaesthesia? European Journal of Anaesthesiology 33, 840–845 (2016). (Springer New York, 2007). 31. Hamilton, T. W., Strickland, L. H. & Pandit, H. G. A Meta-Analysis on the Use of Gabapentinoids for the 52. Covino, B. Pharmacology of local anesthetic agents. in Principles and Practice of Anesthesiology (ed. Rogers Treatment of Acute Postoperative Pain Following Total Knee Arthroplasty. The Journal of Bone and Joint M. C., Tinker J. H., Covino B. G., et al.) 1235–1257 (Mosby Year Book, 1993). Surgery 98, 1340–1350 (2016). 53. Conception, M. Acute complications and side effects of regional anesthesia. in Regional Anesthesia and 32. Reddi, D. Preventing chronic postoperative pain. Anaesthesia 71, 64–71 (2016). Analgesia (ed. Brown, D. L.) 446–461 (WB Saunders Company, 1996). 33. Durrani, Z., Winnie, A. P., Zsigmond, E. K. & Burnett, M. L. Ketamine for intravenous regional anesthesia. 54. Covino, B. Toxicity and systemic effects of local anesthetic agents. in Local Anesthetics. Handbook of Anesthesia and analgesia 68, 328–32 (1989). Experimental Pharmacology. vol. 81 (ed. Strichartz, G.) 187–212 (Spinger, 1987). 34. Fowler, S. J., Symons, J., Sabato, S. & Myles, P. S. Epidural analgesia compared with peripheral nerve blockade 55. Blair, M. R. Cardiovascular pharmacology of local anaesthetics. British journal of anaesthesia 47 suppl, 247– after major knee surgery: a systematic review and meta-analysis of randomized trials. British Journal of 52 (1975). Anaesthesia 100, 154–164 (2007). 56. Sztark, F., Malgat, M., Dabadie, P. & Mazat, J. P. Comparison of the effects of bupivacaine and ropivacaine on 35. Hadzic, A., Houle, T. T., Capdevila, X. & Ilfeld, B. M. Femoral Nerve Block for Analgesia in Patients Having Knee heart cell mitochondrial bioenergetics. Anesthesiology 88, 1340–9 (1998). Arthroplasty. Anesthesiology 113, 1014–1015 (2010). 57. Hagan, P. F. & Kaufman, E. E. Vascular complication of knee arthroplasty under tourniquet. A case report. 36. PROSPECT. Procedure specific postoperative pain mangement. Commemendations for Total Knee Clinical orthopaedics and related research 257, 159–61 (1990). Arthroplasty. Available at: http://www.postoppain.org/sections/?root_id=47232§ion=8. (Accessed: 5th 58. Zhang, W. et al. The effects of a tourniquet used in total knee arthroplasty: a meta-analysis. Journal of October 2017) orthopaedic surgery and research 9, 13 (2014). 37. Marhofer, P. et al. Ultrasonographic guidance improves sensory block and onset time of three-in-one blocks. 59. Abdel-Salam, A. & Eyres, K. S. Effects of tourniquet during total knee arthroplasty. A prospective randomised Anesthesia and analgesia 85, 854–7 (1997). study. The Journal of bone and joint surgery. British volume 77, 250–3 (1995). 38. Salinas, F. V. Evidence Basis for Ultrasound Guidance for Lower-Extremity Peripheral Nerve Block. Regional 60. Liu, P. et al. Effects of Unilateral Tourniquet Used in Patients Undergoing Simultaneous Bilateral Total Knee Anesthesia and Pain Medicine 41, 261–274 (2016). Arthroplasty. Orthopaedic Surgery 9, 180–185 (2017). 39. Borgeat, A., Tewes, E., Biasca, N. & Gerber, C. Patient-controlled interscalene analgesia with ropivacaine 61. Vandenbussche, E., Duranthon, L.-D., Couturier, M., Pidhorz, L. & Augereau, B. The effect of tourniquet use in after major shoulder surgery: PCIA vs PCA. British journal of anaesthesia 81, 603–5 (1998). total knee arthroplasty. International orthopaedics 26, 306–9 (2002). 40. Richman, J. M. et al. Does Continuous Peripheral Nerve Block Provide Superior Pain Control to Opioids? A 62. Ledin, H., Aspenberg, P. & Good, L. Tourniquet use in total knee replacement does not improve fixation, but Meta-Analysis. Anesthesia & Analgesia 102, 248–257 (2006). appears to reduce final range of motion. Acta orthopaedica 83, 499–503 (2012). 63. Ejaz, A. et al. Faster recovery without the use of a tourniquet in total knee arthroplasty. Acta Orthopaedica 85, 422–426 (2014). 20 Introduction 21

64. Tai, T.-W. et al. Tourniquet use in total knee arthroplasty: a meta-analysis. Knee Surgery, Sports Traumatology, 1 Arthroscopy 19, 1121–30 (2011). 65. Mohan, N. et al. Walking the tightrope: creation of the physician scorecard at the Rouge Valley Health System. Healthcare Quarterly 8, 83–8 (2005). Chapter 2 66. Ibrahim, M., Alazzawi, S., Nizam, I. & Haddad, F. An evidence-based review of enhanced recovery interventions in knee replacement surgery. The Annals of The Royal College of Surgeons of England 95, 386– 389 (2013). Femoral Nerve Catheter versus 67. Seangleulur, A. et al. The efficacy of local infiltration analgesia in the early postoperative period after total knee arthroplasty. European Journal of Anaesthesiology 33, 816–831 (2016). 68. Ropivacaine 2 mg/ml solution for injection - Summary of Product Characteristics (SPC) - (eMC). (2018). Local Infiltration Analgesia Available at: https://www.medicines.org.uk/emc/product/9686/smpc. (Accessed: 29th March 2019) 69. Rosenberg, P. H., Veering, B. T. & Urmey, W. F. Maximum recommended doses of local anesthetics: a in Fast Track Total Knee multifactorial concept. Regional anesthesia and pain medicine 29, 564–75 (2004). 70. Fenten, M. G. E., Rohrbach, A., Wymenga, A. B. & Stienstra, R. Systemic local anesthetic toxicity after local Arthroplasty: infiltration analgesia following a polyethylene tibial insert exc.hange: a case report. Regional anesthesia and pain medicine 39, 264–5 (2014) Short-Term and Long-Term Outcome

M.G.E. Fenten* and S.M.K. Bakker* G.J. Scheffer A.B. Wymenga R. Stienstra P.J.C. Heesterbeek

Britisch Journal of Anesthesia 121(4), 850-858 (2018)

* Both authors contributed equally to this manuscript 22 FNB versus LIA for total knee arthroplasty 23

Abstract Introduction

Background Total knee arthroplasty (TKA) reduces knee pain and improves knee joint function in patients The aim of this study was to compare the effects on short-term and long-term pain and with knee osteoarthritis.1 TKA may be associated with severe postoperative pain, which in turn functional outcome of periarticular local anesthetic infiltration (LIA) with LIA of the posterior may slow rehabilitation and predispose to the development of persistent pain.2,3 Perioperative 2 knee capsule in combination with a femoral nerve block (FNB) catheter in patients undergoing pain protocols therefore focus on achieving optimal pain relief with a minimum of side effects; total knee arthroplasty. however, these goals may conflict with changing surgical perspectives with emphasis on early mobilization and reduced length of hospital stay. Recently developed fast track protocols Methods (enhanced recovery protocols) aim for shorter hospital length of stay and better functional Eighty patients were randomised to one of two groups: Subjects in group LIA received recovery.4 Fast track protocols that incorporate early mobilization have been shown to improve periarticular LIA with ropivacaine 0.2% for postoperative analgesia; subjects in group FNB functional recovery and patient satisfaction, and are associated with a lower incidence of received LIA of the posterior capsule and a FNB catheter. The primary outcome parameter thromboembolic adverse events.5 was functional capacity of the knee one year after surgery. Secondary parameters included mobility as determined by accelerometer data, pain, satisfaction with the analgesic regimen, FNB provides good analgesia6 and is considered the standard of care by many.7,8 However, hospital length of stay, and use of pain medication three and twelve months after surgery. the use of a FNB has become disputable because like epidural analgesia, it may hamper early mobilization. Recent developments such as local infiltration analgesia (LIA) aim at providing Results adequate analgesia while avoiding motor impairment.9 Several RCTs comparing LIA with FNB There were no differences between groups in long-term functional capacity, patient have been conducted, and three meta-analyses show no differences in the two techniques satisfaction and hospital length of stay. In the first two days, subjects in group FNB had regarding postoperative analgesia and complication rates.10–12 Although LIA may provide slightly lower pain scores and used less opioids, and subjects in group LIA had a higher level acceptable perioperative analgesia, there are no data on long-term functional recovery and of accelerometer activity. Three and twelve months after surgery, subjects in group FNB had persistent pain. lower maximum pain scores and were less likely to use any pain medication one year after surgery. We performed a blinded randomized controlled trial comparing periarticular LIA of the knee with LIA of the posterior knee capsule in combination with a FNB catheter in terms of Conclusions functional outcome and pain in patients undergoing primary TKA. Primary outcome measure Both techniques were similar regarding long-term functional outcome. Subjects in group FNB was knee function, tested with functional tests. Secondary outcomes were perioperative and had slightly lower pain scores and lower opioid consumption after operation, lower maximum long-term knee pain, use of analgesics, length of hospital stay and patient-reported functional pain scores at three and twelve months, and were less likely to use any pain medication one outcome by questionnaires. year after surgery.

Trial Registration Methods NCT01966263 This blinded randomized study was approved by the Independent Review Board Nijmegen (IRBN2013005) and was registered with an international clinical trials registry (www. clinicaltrials.gov, number NCT01966263) before onset of participant enrolment. Patients undergoing primary unilateral TKA were assessed for eligibility during preoperative screening visit. Patients were informed about the study and written informed consent was obtained from all patients. The study was conducted between November 2013 and November 2015 at the Sint Maartenskliniek Nijmegen, The Netherlands, according to the Declaration of Helsinki and later revisions thereof and in accordance with the International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use guidelines for Good Clinical Practice.

Patients Eligible participants were all adults aged 50-80 years with ASA physical health classification I, II or III. Patients presented with non-inflammatory knee osteoarthritis and were scheduled for fast track, primary, unilateral total knee arthroplasty under spinal anesthesia. Exclusion 24 FNB versus LIA for total knee arthroplasty FNB versus LIA for total knee arthroplasty 25

criteria were defined as: any contra-indication for locoregional or spinal anesthesia, 50 mL ropivacaine 0.2% plus epinephrine 1:200.000 and of the subcutaneous tissue with 50 traumatic osteoarthritis or rheumatoid arthritis requiring TKA, an active local or systemic mL ropivacaine 0.2% without epinephrine. Patients in group FNB received three additional infection, known intolerance or contraindication for local anesthetics, paracetamol, NSAIDs boluses of ropivacaine 0.2% via the femoral catheter at 6 hour intervals up to 18 hours after or opioids, body mass index more than 40 kg/m2, inability to walk independently (inability to the initial bolus injection. 2 walk at least 10 meters without a walking aid), scheduled for contra-lateral TKA within a year, 2 or scheduled for another surgery within three months, use of opioids or anti-neuropathic pain Multimodal analgesia medication for more than one year, or physical, emotional or neurological conditions that All patients received oral multimodal analgesia consisting of paracetamol 1000 mg q.i.d., would compromise compliance with postoperative rehabilitation and follow-up. etoricoxib 90 mg once daily, and gabapentin 600 mg b.i.d. or 300 mg if patients were older than 60 years. Study procedure Using a computer-generated sequence of random numbers in eight blocks of 10 and a sealed Breakthrough pain [numeric rating scale (NRS > 3)] in the recovery room was treated with envelope technique, patients were randomized to one of two groups: group FNB or group LIA. intravenous morphine. At the orthopaedic ward, breakthrough pain (NRS > 3) was treated The envelopes were opened just before surgery, when the patient arrived in the anesthetic with oxycodone 5-10 mg ad libitum. room. Patient, anesthesiologist and orthopaedic surgeon were informed about the study allocation. The physical therapists and research assistants who assessed the outcome Rehabilitation and discharge criteria variables were blinded for treatment allocation and the patient was instructed not to discuss Rehabilitation was according to the standard hospital fast track protocol. The protocol the analgesic regimen with anyone. includes preoperative explanation of the protocol and instruction of the patients to ensure maximum involvement, short acting spinal anesthesia, mobilization within 4 hours after Anesthesia and surgical procedure operation, physical therapy twice daily and evaluation of reaching the discharge criteria twice All surgeries were performed according to standard hospital protocol. In the anesthetic room, daily. Weight bearing mobilization was started as soon as spinal anesthesia had worn off standard monitoring (pulse oximeter, non-invasive blood pressure and electrocardiogram) and patients were encouraged to exercise. Patients were discharged when a set of discharge was applied to all patients and intravenous access was established. Before spinal anesthesia, criteria was met (Table 1). patients in group FNB received a femoral catheter (Contiplex, BBraun, Melsungen, Germany) with dual guidance (ultrasound and nerve stimulation). A correct position of the needle Table 1. Hospital discharge criteria and catheter tip was verified by the spread 5-10 mL of NaCl 0.9% injected under ultrasound • Making independent transfers guidance. The catheter was secured with a transparent dressing with an antimicrobial gel Walking independently and safely with walking aid pad. In patients in group LIA a sham femoral catheter was taped to the skin in a similar fashion • as with patients in group FNB. • Climbing stairs independently and safely (if necessary for home situation) • Active knee flexion ≥ 60 degrees, passive knee extension 0 degrees All patients received spinal anesthesia with 10 mg hyperbaric bupivacaine 0.5% in the sitting • Quadriceps muscle force ≥ 3 on Medical Research Council Scale for muscle strength position. Upon completion of the subarachnoid injection patients were turned to the lateral decubitus position, the side of surgery being dependent. The lateral decubitus position was maintained for twenty minutes to achieve a predominantly unilateral block, after which the Outcome measurements patients were turned to the supine horizontal position. Data collection, including conducting all functional tests, was done by blinded physical therapists and research assistants. Functional capacity of the knee was assessed before During surgery, patients received conscious sedation with propofol upon request. A pneumatic operation, at hospital discharge, at three and at twelve months after surgery using the Timed tourniquet was placed around the patient’s thigh and inflated to 50 mmHg above systolic Up and Go Test (TUG), Stair Climbing Test (SCT) and the Six Minute Walking test (6MWT). The blood pressure with a maximum of 250 mmHg. The knee was approached through a medial TUG and SCT measure the time in seconds to perform predefined tests, the 6MWT the distance parapatellar arthrotomy, all patients received a Genesis II posterior-stabilized (PS) TKA (Smith in meters. These tests have been validated for detection of improvement or deterioration in & Nephew, Memphis, TN, USA) with patellar resurfacing. hip or knee function following surgery.13 The functional tests were conducted by three blinded physical therapists, all tests were conducted in the same manner, in the same hall, using the Study interventions same stairs and chair. At three and twelve months the patients performed the functional tests After placement of the prosthetic components and before wound closure, patients in both just before their postsurgical follow up visit to the orthopaedic surgeon at the outpatient groups received local infiltration of the posterior knee capsule with 100 mL ropivacaine clinic. 0.2% plus epinephrine 1:200.000 to cover pain arising from the popliteal fossa. Immediately afterwards, patients in group FNB received a bolus of 20 mL ropivacaine 0.2% via the femoral In addition, knee function was evaluated using the Lower Extremity Functional Scale (LEFS) catheter, while patients in group LIA received local infiltration of the anterior capsule with and the Oxford Knee Score (OKS) preoperatively, at six weeks and at three and twelve months 26 FNB versus LIA for total knee arthroplasty FNB versus LIA for total knee arthroplasty 27

after surgery. At the same time intervals, fear of movement was measured using the Tampa analysis R version 3.4.3 was used as statistical software (The R Foundation for Statistical Scale for Kinesiophobia (TSK), and quality of life was assessed using the EQ5D-3L and a Visual Computing, Vienna, Austria). Analogue Scale (VASQL).

2 During hospital admission, a research assistant blinded for group allocation visited the Results 2 patient twice daily at 8 am and 8 pm to assess postoperative pain by NRS (0 = no pain, 10 = worst possible pain). At each time point, patients were asked to rate their average pain Recruitment and flow of the patients is shown in a Consolidated Standards of Reporting Trials during the previous 12-hour period. In the morning of the first two days after surgery, an (CONSORT) flow diagram (Figure 1). All enrolled subjects received the allocated intervention. accelerometer was attached to the non-operated thigh and in the evening, it was taken Patients in both groups had similar demographic and surgical characteristics. Distribution of off. Accelerometers were used to assess the level of activity: an accelerometer can detect baseline characteristics across the treatment groups is shown in Table 2, adverse events in body movements by measuring orientation and acceleration in three orthogonal planes; Table 3. anteroposterior, mediolateral and vertical. Hereby the intensity of activity over time can be estimated and postures and transfers can be calculated by using the orientation of the meter in relation to the body.14 The patients were instructed to continue their normal routine whilst wearing the accelerometer. Accelerometer data were either classified as ‘active’, which meant the accelerometer was in the upright position and moving; or ‘resting’; which was all other positions.

Twice daily the patients were visited by the physical therapist, who was blinded for group allocation, and who recorded the ability of the patient to mobilize.

At discharge, patients rated their satisfaction with the analgesic regimen on an NRS scale (0 = very dissatisfied, 10 – very satisfied), and range of motion (ROM) was calculated by the sum of knee flexion and extension as measured with a long-arm goniometer.

Pain and the use of analgesic medication for knee pain were assessed before operation, and at three and twelve months after surgery. Patients were asked to rate their average and their maximum pain in the two weeks preceding the contact moment on an 11-point numeric rating scale (NRS), ranging from no pain (0) to worst imaginable pain (10).

Sample size and statistical analysis The null hypothesis of our study was that differences in anesthetic technique for TKA do not affect long-term functional outcome. Of the three tests we used to evaluate functional capacity, the SCT has the largest variance.15,16 To reduce the risk of insufficient power, we Figure 1. CONSORT Flow diagram defined functional capacity of the knee one year after surgery as determined by SCT as the primary outcome parameter. Based on two studies,15,16 the sample size necessary to detect at least a 30% difference with a 90% probability and < 0.05 was 37 patients per group. Allowing Table 2. Baseline characteristics. Displayed as mean (SD) or number of patients. for patient withdrawal during the study period, we included 40 patients per group. Group FNB (n=40) Group LIA (n=40) α Sex, no. M/ no. F 20 / 20 17 / 23 Data were analysed using Stata version 13.1 (Stata Corporation, College Station, TX, USA) and Age (years) 64 (6.9) 66 (6.3) are presented as mean (SD). Knee function outcome scores were analysed using multiple BMI (kg/m2) 30.0 (4.9) 28.4 (4.2) regression models. Linear regression was used to investigate the effect of the analgesic ASA classification, no. I / no. II / no. III 8 / 24 / 8 13 / 15 / 12 technique on the use of pain medication. A multilevel regression model was used to investigate the effect of analgesic technique on pain scores after operation, during the first Side of surgery, no. left / no. right 22 / 18 17 / 23 2 postoperative days. Time, time squared and baseline pain score were used as covariates in Duration of surgery (minutes) 72 (16) 71 (11) this model. The model used a random intercept, to allow for the clustering of the first four Tourniquet time (minutes) 55 (11) 56 (14) pain scores recorded after operation at 8 p.m. and 8 a.m. within each patient. For this specific Hospital length of stay (days) 3.2 (1.1) 3.0 (0.9) 28 FNB versus LIA for total knee arthroplasty FNB versus LIA for total knee arthroplasty 29

Table 3. Adverse events Missing values Description Amount Treatment Sequelae Post-study of AEs interventions Average pain scores during hospitalization Wounds, Bleedings and Infections During the first two days, average pain scores were missed in one to three patients because of 2 Group FNB absence during the regular visit. This concerned different patients every day. Missing average 2 • Infected haematoma 1 Surgical cleaning, antibiotics n.a. Revision of pain scores on the second day after surgery were due to patients being already discharged arthroplasty from the hospital. • Wound infection 1 Wound cleaning, change of No insert during debridement Accelerometer data could not be obtained in all patients due to technical failure. This • Postoperative bleeding 1 Compression bandage and 2 No days immobilization concerned different patients every day. • Wound dehiscence after fall on POD1 1 Surgical cleaning, antibiotics, 2 No days immobilization Functional capacity Group LIA Four patients (two in each group) were discharged before the functional capacity tests could • Wound infection 1 Surgical cleaning, antibiotics No be performed and one patient in group LIA was unable to perform the SCT on the day of immobilization discharge. • Wound dehisce after removing stiches 1 2 weeks immobilization No • Spontaneous haemorrhage in knee 1 Puncture to release blood No At three months, one patient from group FNB was excluded from continued data collection Acute pain due to revision surgery of the knee; one patient from group FNB withdrew from follow up due Group FNB to the development of a malignancy. One patient in group FNB and two in group LIA missed • Severe pain until 10 days after TKA 1 Prolonged hospitalisation for No the appointment at three months. None of these patients had missed the functional capacity analgesic treatment tests at discharge. Limited Range of Motion Group FNB 2 Manipulation under anesthesia Yes Both with good function At twelve months, a third patient from group FNB and three patients from group LIA were after inlay excluded from continued data collection due to TKA of the contralateral knee between three replacement and twelve months. In group LIA, one patient withdrew from follow up at twelve months due 2 Manipulation under anesthesia No to the development of a severe eye disorder requiring intensive treatment. One patient from Group LIA 1 Manipulation under anesthesia No group FNB was unable to perform the 6MWT due to severe backache. These patients were 1 Manipulation under anesthesia Yes other patients than those with missing values at discharge and three months. 1 No treatment Yes Chronic pain Average and maximum pain scores at three and twelve months Group FNB The missing values for average and maximum pain scores at three and twelve months concern • Neuropathic pain 1 Cryoablation No the patients who missed the three-month appointment, withdrew from follow up or were 1 Cryoablation and trigger point Yes Pain resolved excluded from continued data collection for reasons mentioned earlier. injections after resection of lateral facet patella All missing values were classified as completely at random, except for missing pain scores on Group LIA postoperative day 2 due to hospital discharge. • Severe pain of unknown origin 2 Yes Contralateral TKA Group FNB 1 No Group LIA 3 No 30 FNB versus LIA for total knee arthroplasty FNB versus LIA for total knee arthroplasty 31

Long-term outcome variables Pain Average and maximum pain scores decreased over time in both groups. Maximum pain scores, Functional capacity but not average pain scores, were slightly lower in group FNB at three and twelve months after Patients in both groups showed a major improvement in performance of functional knee surgery; p=0.047 and p=0.021 respectively (Table 5 and Figure 2). 2 capacity over time (Table 4). 2 Table 5. NRS Pain scores at 3 and 12 months, displayed as mean (SD) and differences between means Table 4. Functional performances displayed as mean (SD) and adjusted differences for baseline adjusted for baseline pain score. performance. Group FNB N Group LIA N Adjusted difference between Group FNB N Group LIA N Adjusted difference P-value means (95% CI) between means (95% CI) Baseline SCT (sec) NRS average pain 4.7 (2.3) 40 3.9 (2.2) 40 Baseline 21.8 (11.3) 40 17.1 (6.8) 40 NRS maximum pain 7.2 (1.5) 40 6.7 (2.2) 40 Discharge 66.3 (25.9) 38 54.2 (24.7) 37 -8.2(-19.8 to 3.5) 0.166 3 Months 3 months 16.8 (6.4) 37 17.4 (10.4) 38 2.4 (-1.5 to 6.3) 0.222 NRS average pain 2.4 (2.1) 38 2.8 (1.7) 37 0.7(-0.1 to 1.5) 12 months 13.8 (4.7) 37 14.3 (7.1) 36 1.9 (-0.7 to 4.5) 0.153 NRS maximum pain 3.8 (2.8) 38 4.6 (2.2) 37 1.2 (0.0 to 2.4) 12 Months TUG (sec) NRS average pain 1.1 (1.8) 37 1.5 (2.0) 36 0.5 (-0.4 to 1.4) Baseline 10.1 (2.9) 40 9.0 (2.3) 40 NRS maximum pain 1.8 (2.4) 37 3.0 (2.6) 36 1.4 (0.2 to 2.6)

Discharge 21.7 (8.1) 38 19.6 (7.3) 38 -1.6 (-2.1 to 2.0) 0.380 NRS: numeric rating scale, FNB: femoral nerve block, LIA: local infiltration analgesia, CI: confidence interval. 3 months 8.3 (1.5) 37 8.5 (2.4) 38 0.4 (-0.5 to 1.3) 0.356 12 months 7.6 (1.2) 37 7.8 (1.9) 36 0.5 (-0.2 to 1.1) 0.197

6MWT (meters) Baseline 394 (97) 40 432 (99) 40 Discharge 203 (69) 38 219 (66) 38 8 (-22 to 38) 0.603 3 months 440 (81) 37 447 (72) 38 -13 (-44 to 16) 0.368 12 months 505 (84) 36 489 (71) 36 -32 (-64 to -0.4) 0.047

ROM (º) Baseline 107 (17) 40 111 (13) 40 Discharge 74 (15) 38 72 (17) 38 -2 (-9 to 5) 0.569 Figure 2. Box plots for NRS average (A) and maximum (B) pain scores at baseline, 3 and 12 months. Represents outliers outside 1.5x interquartile range. NRS: numeric rating scale, FNB: femoral nerve block, LIA: local 3 months 106 (13) 37 102 (13) 38 -5 (-11 to 1) 0.106 infiltration analgesia. 12 months 112 (17) 37 112 (12) 36 -1 (-8 to 5) 0.580 SCT: stair climbing test (primary outcome measure), TUG: timed up and go test, 6MWT: 6 minute walk test, ROM: range of motion, FNB: femoral nerve block, LIA: local infiltration analgesia, CI: confidence interval. Use of analgesics. Three months after surgery, there was no difference between the groups in the use of Knee function and quality of life analgesics. Twelve months after surgery, patients in group LIA were almost six times more Knee function as measured by LEFS and OKS improved over time in both groups. Knee function likely to use analgesic medication for pain in the operated knee as compared to patients in was comparable between the groups and there were no statistically significant differences at group FNB (Odds ratio 5.9; 95% CI 1.1-31.7; p=0.037) (Table 6). any of the time intervals. The same trend was found regarding quality of life as evaluated by EQ5D-3L and VASQL, as well as fear of movement as measured by TSK. 32 FNB versus LIA for total knee arthroplasty FNB versus LIA for total knee arthroplasty 33

Table 6. Rescue analgesic use during hospitalisation and analgesic use at 3 and 12 months follow up. In FNB vs 27.7 (14.1) minutes in group LIA (adjusted difference between means 7.2, 95% CI 0.5 – hospital oxycodone used was added up per patient for each day. Data are expressed as mean (SD). 13.9). Group FNB N Group LIA N Adjusted difference between means (95% CI) 2 Postoperative, in hospital, oxycodone use (mg) 2 Day of surgery 6.1 (8.9) 40 10.9 (10.3) 40 4.8 (0.5 – 9.0) Day after surgery 15.9 (12.4) 40 28.6 (20.2) 40 12.8 (5.3 – 20.2) Day 2 after surgery 15.1 (16.7) 40 13.0 (15.6) 40 2.1 (-9.3 – 5.1) Analgesic use at 3 months (no / yes) Paracetamol 29/9 38 25/13 38 NSAID 37/1 38 32/6 38 Opioids 35/3 38 33/5 38 Other 37/1 38 37/1 38 Overall use of pain medication at 3 months (no / yes) Total 28/10 38 22/16 38 Analgesic use at 12 months (no / yes) Paracetamol 36 / 1 37 32 / 7 39 NSAID 37 / 0 37 35 / 4 39 Opioids 35 / 2 37 38 / 1 39 Figure 3. Postoperative pain scores during hospitalization. Other 36 / 1 37 39 / 0 39 Postoperative pain scores were assessed twice daily by a blinded research assistant. Data are plotted as means with standard Overall use of pain medication at 12 months (no / yes) deviation (SD). Multilevel analyses on presented pain scores, corrected for baseline pain scores showed a significant pain reduction in group FNB (p=0.001). FNB, femoral nerve block; LIA, local infiltration analgesia; NRS, numeric rating scale. Total 35 / 2 37 30 / 9 39 FNB, femoral nerve block; LIA, local infiltration analgesia, CI: confidence interval. Mobilization Short-term outcome variables There was no difference between the groups in the ability to mobilize. ROM on the day of discharge was equal between groups FNB and LIA (74 (15) degrees and 72 (16) respectively), and Postoperative pain and opioid use there was no difference in hospital length of stay (Table 2). Group FNB showed lower mean pain scores during the first two postoperative days, multilevel analysis revealed a difference of 0.95 (95% CI 0.39 – 1.51, p = 0.001) (Figure 3). Similarly, a Serious adverse events difference in favour of the FNB group was found for maximum pain scores (difference 1.22; None of the patients showed any sign of local anesthetic systemic toxicity (LAST). One falling 95% CI 0.41 – 2.02; p = 0.003). incident was recorded. A patient in the FNB group mobilized unattended shortly after her return to the ward. The effects of spinal anesthesia may not have been fully resolved at this Also, group FNB showed lower opioid consumption on the day of surgery and the day after time and may have contributed to the fall. surgery. Table 6 displays data on postoperative analgesic use. Oxycodone consumption on the day of surgery was 6.1 (8.9) mg in group FNB vs 10.9 (10.3) mg in group LIA (adjusted difference between means 4.8, 95% CI 0.5-9.0). Oxycodone consumption on the day after surgery was 15.9 (12.4) mg in group FNB vs 28.6 (20.2) mg in group LIA (adjusted difference between means 12.8, 95% CI 5.3-20.2). There was no difference in patient satisfaction with the analgesic regimen (8.5 (1.1) in group FNB vs. 8.1 (1.3) in group LIA).

Postoperative activity Group FNB showed lower activity levels on the day of the surgery and the first postoperative day as measured by accelerometry. Time spent active on the day of surgery was 2.3 (2.4) minutes in group FNB vs 4.4 (2.9) minutes in group LIA (adjusted difference between means 2.2, 95% CI 0.9 – 3.4). Time spent active on the day after surgery was 20.5 (14.9) minutes in group 34 FNB versus LIA for total knee arthroplasty FNB versus LIA for total knee arthroplasty 35

Discussion active during the first day was short in both groups, on average 20 and 28 minutes during a twelve-hour interval (8 am until 8 pm). Since there was no difference between the groups on We found no differences in functional recovery of the knee at six weeks, three months and hospital length of stay, the clinical relevance of this difference in mobility is questionable. one year after surgery between the FNB and LIA group. These results agree with several other 2 studies, although differences in methodology and the time of follow up exist.17,18 Because of the potential of FNB to interfere with muscle strength, there is concern that the 2 risk of falling is increased. We observed one falling incident in group FNB. This happened when We found that the maximum pain scores at three months and a year after surgery were the patient, contrary to instructions, mobilized unattended at a time that spinal anesthesia slightly but significantly higher in group LIA, and the odds of taking any pain medication for may not have been fully resolved. We did not observe any other falling incidents. knee pain one year after surgery was almost six times higher in the LIA group. Patients in group FNB also had lower pain scores and less opioid consumption in the immediate postoperative The dose of 400 mg ropivacaine we use for LIA is high, and well above the maximum period. As postoperative pain is a possible risk factor for the development of chronic pain,2,19 recommended dose of 3-4 mg/kg for most patients. Although pharmacokinetic studies the possibility of a causal relation is intriguing and should be kept in mind. However, since involving LIA with 400 mg ropivacaine for TKA found free ropivacaine concentrations to our study was neither powered nor designed to detect the influence of anesthetic technique remain below the toxic threshold27,28 and we observed no signs of LAST in any of our patients, on postoperative pain and chronification of pain after TKA, further study will be necessary to it may be prudent to consider reducing the dose of ropivacaine in patients with a low body elucidate this. weight, or in patients who are otherwise at an increased risk for LAST.

Studies analysing the effect of analgesic technique on long-term recovery after TKA are scarce; Our study has several limitations. We used a combination of ropivacaine and epinephrine most studies comparing FNB and LIA focus mainly on differences in the early postoperative for LIA, but around the world LIA mixtures vary in composition, additives, and dose of local period. Although pain and opioid consumption in the immediate postoperative period are anesthetic; also, we opted for an intermittent bolus technique in the FNB group, whereas others important issues from a perspective of patient comfort and satisfaction, the effect of analgesic may favour a continuous infusion. Our results therefore are not necessarily representative for technique on long-term parameters such as functional recovery and pain is equally essential. different LIA mixtures and different modes of application.

Regarding short-term outcome, we found that patients in group FNB had lower pain scores Although the total dose of ropivacaine in both groups is comparable, the systematic difference and less oxycodone use on the first day and night after operation, as well as on the day after between the two methods of pain relief may favour the FNB group because these patients surgery. The issue of the optimal analgesic regimen for fast track TKA is controversial. Our received three additional boluses with local anesthetic up to 18 hours after surgery, whereas results agree with Carli17 and Kovalak20, who report less opioid use and lower pain scores in the patients in group LIA only received “single shot” infiltration at the end of surgery. To counteract first two days after surgery when comparing femoral nerve catheter combined with LIA of the this difference, an intra-articular catheter allowing similar top-ups would have been posterior part of the knee, compared to LIA of both the posterior and anterior area of the knee. necessary in the LIA group. However, intra-articular catheters are controversial because of However, other studies reported better analgesia with LIA compared to FNB alone.11,21–23 A fear of an increased risk of infection, and for that reason not used in many orthopaedic centres, possible explanation for this difference is the combination of FNB with local infiltration of the including ours. Therefore, despite this systematic difference between the two techniques, a posterior capsule. Different branches of the femoral, obturator and sciatic nerve contribute comparison is still relevant from a clinical perspective. to the innervation of the knee24 and FNB alone does not provide analgesia of the sciatic part of the knee. The addition of a sciatic nerve block has been shown to provide better analgesia For ethical reasons, we refrained from inserting femoral catheters in patients randomized to than FNB alone25 and local infiltration of the posterior capsule likely has a comparable effect. group LIA, taping a sham catheter to the skin instead. Although an in-vivo placebo catheter would have been a better option with respect to blinding, it is our opinion that the risk of an Although the difference between the groups in pain scores and opioid use is statistically invasive sham catheter would not justify the benefits. Thus the subjects were not blinded to significant, the clinical relevance in the immediate postoperative period may be argued. Pain group allocation. The treating anesthetists and surgeons were also not blinded. Given the study scores were low in both groups, and patient satisfaction with the analgesic technique was design, blinding of the anesthetists was not possible. Blinding the operating room personnel, equally high. including the surgeons, would have necessitated sham infiltration of the anterior capsule and subcutaneous tissue with normal saline in the FNB group, causing unnecessary swelling. In Subjects in group FNB showed less activity as measured with the accelerometer on the first addition, it would have made the LIA procedure more complicated with different injectates for two days, but ROM at hospital discharge was similar between groups, as well as length of the anterior part of the knee for the two groups, increasing the risk of unintentional protocol hospital stay and functional test scores. Early mobilization is believed to promote recovery violation. However, because the physical therapists and the research assistants collecting after TKA and is one of the key features of fast track rehabilitation protocols. In our study we the data were blinded and none of the anesthetists, surgeons, or operating room personnel measured the time patients spent active (standing, walking) using accelerometry, an objective were involved in outcome measurements or data collection, we believe these limitations have way to measure physical activity.26 Our data show that patients in group FNB were less active not affected our results. Furthermore, performance on physical tests such as the 6MWT, SCT on the day of surgery and the first postoperative day. However, the total amount of time spent and TUG is not only determined by knee function per se, but also by factors such as age, sex, 36 FNB versus LIA for total knee arthroplasty FNB versus LIA for total knee arthroplasty 37

BMI and possible health issues. Because we used a randomization procedure we assume that References possible confounders were distributed equally among the groups. 1. Carr, A. J. et al. Knee replacement. Lancet 379, 1331–40 (2012). Another limitation is that we asked patients at the three and twelve months visits to rate their 2. Puolakka, P. A. E. et al. Persistent pain following knee arthroplasty. European journal of anaesthesiology 27, 2 average and maximum pain in the two weeks preceding the contact moment, and this is a 455–60 (2010). 2 potential cause of recall bias. A pain diary would have reduced this risk, but would also have 3. Barrington, J. W., Halaszynski, T. M. & Sinatra, R. S. Perioperative pain management in hip and knee been much more time consuming for the patients. Since post-surgery pain was a secondary replacement surgery. American journal of orthopedics 43, S1–S16 (2014). outcome we chose not to use a pain diary but a less burdensome method instead. 4. Ibrahim, M., Alazzawi, S., Nizam, I. & Haddad, F. An evidence-based review of enhanced recovery interventions in knee replacement surgery. The Annals of The Royal College of Surgeons of England 95, 386– In conclusion, FNB and LIA are comparable techniques regarding short-term and long-term 389 (2013). functional outcome after TKA. There were no differences in SCT between the two groups, nor 5. Husted, H. et al. Low risk of thromboembolic complications after fast-track hip and knee arthroplasty. Acta in the other functional capacity tests. We found that FNB compared to LIA provides slightly Orthopaedica 81, 599–605 (2010). better pain relief and less oxycodone use on the first day and night following TKA, as well as on 6. Paul, J. E. et al. Femoral nerve block improves analgesia outcomes after total knee arthroplasty: a meta- the day after surgery. Three and twelve months after surgery, patients in group FNB had lower analysis of randomized controlled trials. Anesthesiology 113, 1144–62 (2010). maximum pain scores and were significantly less likely to use any pain medication one year 7. Hadzic, A., Houle, T. T., Capdevila, X. & Ilfeld, B. M. Femoral Nerve Block for Analgesia in Patients Having Knee after surgery. Further study is needed to determine if this relation is causal or spurious. Arthroplasty. Anesthesiology 113, 1014–1015 (2010). 8. PROSPECT. Procedure specific postoperative pain mangement. Commemendations for Total Knee Acknowledgements Arthroplasty. Available at: http://www.postoppain.org/sections/?root_id=47232§ion=8. (Accessed: 5th We would like to thank Tim Janssen, physical therapist, for his advice and assistance in October 2017) conducting the functional tests and Ewald Bronkhorst, statistical and methodological 9. Kerr, D. R. & Kohan, L. Local infiltration analgesia: a technique for the control of acute postoperative pain consultant, for his advice and assistance in the statistical analyses. Also, we thank Saskia following knee and hip surgery: a case study of 325 patients. Acta orthopaedica 79, 174–83 (2008). Susan, research nurse, and Jolanda Rubrech, research assistant, for managing the logistics of 10. Albrecht, E., Guyen, O., Jacot-Guillarmod, A. & Kirkham, K. R. The analgesic efficacy of local infiltration this study. analgesia vs femoral nerve block after total knee arthroplasty: a systematic review and meta-analysis. British journal of anaesthesia 116, 597–609 (2016). Financial support and sponsorship 11. Fan, L. et al. Comparison of Local Infiltration Analgesia With Femoral Nerve Block for Total Knee Arthroplasty: This study was supported entirely by internal funds of the department of Anesthesiology, A Prospective, Randomized Clinical Trial. The Journal of arthroplasty 31, 1361–5 (2016). Sint Maartenskliniek, Nijmegen, The Netherlands. 12. Mei, S. et al. Analgesia for total knee arthroplasty: a meta-analysis comparing local infiltration and femoral nerve block. Clinics 70, 648–53 (2015). 13. Kennedy, D. M., Stratford, P. W., Wessel, J., Gollish, J. D. & Penney, D. Assessing stability and change of four performance measures: a longitudinal study evaluating outcome following total hip and knee arthroplasty. BMC musculoskeletal disorders 6, 3 (2005). 14. Chen, K. Y. & Bassett, D. R. The technology of accelerometry-based activity monitors: current and future. Medicine and science in sports and exercise 37, S490-500 (2005). 15. Mizner, R. L., Petterson, S. C., Stevens, J. E., Axe, M. J. & Snyder-Mackler, L. Preoperative quadriceps strength predicts functional ability one year after total knee arthroplasty. The Journal of rheumatology 32, 1533–9 (2005). 16. Bruun-Olsen, V., Heiberg, K. E., Wahl, A. K. & Mengshoel, A. M. The immediate and long-term effects of a walking-skill program compared to usual physiotherapy care in patients who have undergone total knee arthroplasty (TKA): a randomized controlled trial. Disability and Rehabilitation 35, 2008–2015 (2013). 17. Carli, F. et al. Analgesia and functional outcome after total knee arthroplasty: periarticular infiltration vs continuous femoral nerve block. British journal of anaesthesia 105, 185–95 (2010). 18. Ng, F.-Y., Ng, J. K.-F., Chiu, K.-Y., Yan, C.-H. & Chan, C.-W. Multimodal periarticular injection vs continuous femoral nerve block after total knee arthroplasty: a prospective, crossover, randomized clinical trial. The Journal of arthroplasty 27, 1234–8 (2012). 19. Macrae, W. A. Chronic post-surgical pain: 10 years on. British Journal of Anaesthesia 101, 77–86 (2008). 38 FNB versus LIA for total knee arthroplasty 39

20. Kovalak, E. et al. A comparison of continuous femoral nerve block and periarticular local infiltration analgesia in the management of early period pain developing after total knee arthroplasty. Acta orthopaedica et traumatologica turcica 49, 260–6 (2015). 21. Li, D., Tan, Z., Kang, P., Shen, B. & Pei, F. Effects of multi-site infiltration analgesia on pain management and Chapter 3 2 early rehabilitation compared with femoral nerve or adductor canal block for patients undergoing total knee arthroplasty: a prospective randomized controlled trial. International orthopaedics 41, 75–83 (2017). 22. Ashraf, A., Raut, V. V., Canty, S. J. & McLauchlan, G. J. Pain control after primary total knee replacement. A Pharmacokinetics of 400 mg prospective randomised controlled trial of local infiltration versus single shot femoral nerve block. The Knee 20, 324–327 (2013). 23. Moghtadaei, M., Farahini, H., Faiz, S. H.-R., Mokarami, F. & Safari, S. Pain Management for Total Knee Ropivacaine after Periarticular Arthroplasty: Single-Injection Femoral Nerve Block versus Local Infiltration Analgesia. Iranian Red Crescent medical journal 16, e13247 (2014). Local Infiltration Analgesia for 24. Horner, G. & Dellon, A. L. Innervation of the human knee joint and implications for surgery. Clinical orthopaedics and related research 221–6 (1994). Total Knee Arthroplasty 25. Wegener, J. T. et al. Value of single-injection or continuous sciatic nerve block in addition to a continuous femoral nerve block in patients undergoing total knee arthroplasty: a prospective, randomized, controlled trial. Regional anesthesia and pain medicine 36, 481–8 (2011). 26. Bussmann, J. B. et al. Measuring daily behavior using ambulatory accelerometry: the Activity Monitor. Behavior research methods, instruments, & computers : a journal of the Psychonomic Society, Inc 33, 349–56 (2001). 27. Brydone, A. S. et al. Ropivacaine plasma levels following high-dose local infiltration analgesia for total knee arthroplasty. Anaesthesia 70, 784–90 (2015). 28. Fenten, M. G. E. et al. Pharmacokinetics of 400 mg ropivacaine after periarticular local infiltration analgesia for total knee arthroplasty. Acta anaesthesiologica Scandinavica 61, 338–345 (2017).

M.G.E. Fenten S.M.K. Bakker D.J. Touw B.J.F. van den Bemt G.J. Scheffer P.J.C. Heesterbeek R. Stienstra

Acta Anaesthesiologica Scandinavica 61(3), 338-345 (2017) 40 Ropivacaine pharmacokinetics with tourniquet use 41

Abstract Introduction

Background Over the past few years enhanced rehabilitation protocols (fast-track surgery protocols) were Although considered safe, no pharmacokinetic data of high-dose, high-volume local infiltration introduced for numerous surgical procedures to improve short- and long-term functional analgesia (LIA) with ropivacaine without the use of a surgical drain or intra-articular catheter recovery, reduce morbidity, decrease length of convalescence and increase patient satisfaction have been described. The purpose of this study is to describe the maximum total and unbound – resulting in reduced hospital costs as a secondary gain.1,2 To facilitate enhanced rehabilitation

ropivacaine concentrations (Cmax, Cu max) and corresponding maximum times (Tmax, Tu max) of a protocols, optimal analgesia with minimal side effects should be provided. Postoperative pain single-shot ropivacaine (200ml 0.2%) and 0.75 mg epinephrine (1000µg/mL) when used for LIA treatment is an integrated part of the fast-track protocol, usually including pre-emptive and 3 in patients for total knee arthroplasty. multimodal analgesia but also locoregional and local infiltration techniques are commonly incorporated. Methods In this prospective cohort study, twenty patients were treated with LIA of the knee for primary In total knee arthroplasty (TKA), local infiltration analgesia (LIA) of the knee allows immediate total knee arthroplasty. Plasma samples were taken at 20, 40, 60, 90, 120, 240, 360 minutes and post-operative mobilization without side effects such as drowsiness (opioids) or impaired at 24 hours after tourniquet release, in which total and unbound ropivacaine concentrations motor function (femoral nerve block) that impedes rehabilitation.3,4 The LIA technique is were determined. based on local infiltration of the soft tissues surrounding the knee with a long-acting local anesthetic (LA). LIA is a straightforward and effective technique, which provides effective Results analgesia in the initial postoperative period after TKA and is adopted by orthopedic surgeons 5 Results are given as median [IQR]. Highest ropivacaine concentration (Cmax) was 1.06 µg/ around the globe.

mL [0.34]; highest unbound ropivacaine concentration (Cu max) was 0.09 µg/mL [0.05]. The corresponding time to reach the maximum concentration for total ropivacaine was 312 Local anesthetics exert their therapeutic effect directly at the site of injection. In a continuous minutes [120] after tourniquet release, and for the unbound fraction 265 [110] minutes after manner, the LA is absorbed from tissue into the central compartment and is then eliminated tourniquet release. from the plasma. Absorption speed and peak plasma concentration vary with the site of injection and type of LA used.6 The commonly used local anesthetic for LIA is ropivacaine, Conclusion chosen for its long-acting profile, reduced cardiotoxicity in comparison to bupivacaine and Although great inter-individual variability was found between the maximum ropivacaine its intrinsic vasoconstrictor properties. In plasma on average 95% of ropivacaine is bound to concentrations, both maximum total and unbound serum concentrations of ropivacaine 1-glycoprotein and approximately 5% of the total plasma concentration of ropivacaine is in remained well below the assumed systemic toxic thresholds of 4.3 and 0.56 µg/mL. the free form. Only unbound, unionized ropivacaine is able to cross the cell membrane of the nerveα cell, exerting its pharmacological effect by blocking voltage-gated sodium channels Trial Registration from inside the nerve cell and preventing depolarization. NL4653 (NTR4796) The plasma levels of ropivacaine are determined by absorption from the injection site, distribution and metabolization by cytochrome P450 enzymes in the liver and subsequent primarily urinary excretion. Plasma levels of unbound ropivacaine are, instead of bound ropivacaine, pharmacological active and determine if, and to what extent, systemic effects of ropivacaine occur. When the plasma concentration of unionized ropivacaine exceeds the toxic threshold in the central nervous system (CNS) or heart, symptoms of systemic toxicity occur.

Doses of ropivacaine used for LIA or peripheral nerve blocks often exceed the recommended maximum dose of 3-4mg/kg.7,8 Concerns have been raised about the high doses of ropivacaine used for LIA with regard to LAST (local anesthetic systemic toxicity).9 Since LIA has only recently gained popularity, its safety track record is relatively short and so far little is known about the pharmacokinetic profile of ropivacaine applied for single-shot LIA of the knee. Thousands of patients have received LIA for TKA and to our knowledge, only one case of LAST after LIA has been described,10 suggesting that the technique is safe. Knowledge about plasma concentrations of ropivacaine and time to reach the highest plasma concentration will help determine safe doses of ropivacaine for LIA and provide a time frame for close monitoring of patients postoperatively. 42 Ropivacaine pharmacokinetics with tourniquet use Ropivacaine pharmacokinetics with tourniquet use 43

The purpose of the present study is to describe the maximum total (Cmax) and unbound (Cu max), After cementing, the knee was infiltrated by the orthopedic surgeon. 300 mg ropivacaine 0.2%

ropivacaine concentration, corresponding maximum times (Tmax resp. Tu max) and proximity (150 mL) was mixed with 0.75 mL epinephrine (1000 µg/mL). Two thirds of this solution was to the toxic threshold, when a solution of 400 mg ropivacaine (200ml, 0.2%) and 0.75 mg injected in the posterior and one third in the anterior capsule. After closure of the parapatellar epinephrine (1000 µg/mL) is used as a single-shot LIA in TKA. arthrotomy another 100 mg of ropivacaine 0.2% (50 mL) without epinephrine was infiltrated in the subcutaneous tissues. Just before tourniquet release, patients received an i.v. bolus of 10 mg/kg tranexamic acid with a maximum of 1000 mg. All patients received a compression Methods bandage before transfer to the recovery room. 3 3 Patients Blood sampling and assays Before onset of participant enrolment, this prospective cohort study was approved by the Pre-operatively, before any intravenous fluids were administered to the patient, a baseline Medical Research Ethics Committee Slotervaart Hospital and Reade and registered at the blood sample was taken. Before start of the surgery, all patients received two peripheral Netherlands Trial Registry (http://www.trialregister.nl, NTR4796). All patients who were intravenous catheters (PIVC). One of the PIVCs, at least 16-gauge in size, was placed in scheduled for enhanced rehabilitation protocol for primary TKA under spinal anesthesia were the antecubital vein or great saphenous vein and was only used for blood sampling. No assessed for eligibility. Eligible patients were those aged between 50-80 years with ASA physical intravenous fluids were administered through this PIVC. Venous blood samples were taken health classification I or II, body mass index less than 40 kg/m2, and hemoglobin 7.5 mmol/L at 20, 40, 60, 90, 120, 240, 360 minutes and at 24 hours after release of the tourniquet. At each or greater. Exclusion criteria were placement of a surgical drain, known hypersensitivity to collection moment, two samples of 5 mL blood were drawn. The first sample was drawn and amide-type local anesthetics, known history of hepatic or renal insufficiency and use of any discarded, as this sample was diluted with NaCL 0.9%. The second (undiluted) 5 mL was kept medications that affect the clearance of ropivacaine.8 in an EDTA tube, centrifuged within one hour after collection and stored at -80ºC until assay of the whole batch. Total ropivacaine concentrations were determined in all serum samples. All eligible patients were informed verbally and in writing about the study and written After determining the sample with the highest total concentration per patient, the unbound informed consent was obtained from all participating patients. The study was conducted ropivacaine levels were determined in ultra filtrates of these samples as well as in the samples at the Sint Maartenskliniek Nijmegen, The Netherlands between January and May 2015 taken immediately before and after this time point. Ropivacaine levels were detected by according to the Declaration of Helsinki and later revisions thereof and in accordance with a validated liquid chromatography – tandem mass spectrometry (LCMSMS) method as the International Conference on Harmonisation of Technical Requirements for Registration of previously described.11 Pharmaceuticals for Human Use guidelines for Good Clinical Practice. Sample size and statistics Anesthetic and Surgical procedure The study was designed to describe the pharmacokinetic profile of 400 mg ropivacaine with All patients were treated according to the standard hospital enhanced rehabilitation protocol. epinephrine, when used for LIA of the knee. Because a sample size calculation is not possible Basic oral pain treatment was started preoperatively at the day of the surgery: paracetamol in a descriptive study, we based the chosen sample size on experience and common sense. In 1000 mg four times daily, etoricoxib 90 mg once daily and gabapentin 300 mg and 600 mg both case of small inter-individual variation and a homogenous study group, a sample size of six once daily or 300 mg twice daily if patients were older than 70 years. Premedication consisted subjects is well accepted for pharmacokinetic studies. Since the study subjects in our study of 10 mg oxazepam orally. Additional pain therapy was started when the postoperative pain are heterogeneic and more than one variable is measured, a larger sample size is required. score (Numerical rating scale, NRS) was ≥4. We assumed that a sample size of 18 subjects would be appropriate, allowing for subjects to withdraw during the study period we chose to include 20 subjects. Analysis was conducted Surgery was performed under spinal anesthesia and upon patient request supplemented using GraphPad Prism 6 software (GraphPad Software Inc., San Diego, CA, USA). Frequency with sedation using propofol. Before placement of spinal anesthesia intravenous access distribution was tested using Kolmogorov-Smirnov test for normality. Normally distributed and routine monitoring (electrocardiogram, non-invasive blood pressure and peripheral data are presented as mean (SD) and non-normal distributed data are presented as median

oxygen saturation) was established in all patients. Spinal anesthesia was performed with [IQR]. Primary end points were peak serum concentration of total ropivacaine in plasma (Cmax)

the patient in the sitting position at the third or fourth lumbar interspace. After obtaining and unbound ropivacaine in ultra-filtrate (Cu max) and the corresponding times to reach peak

a free flow of cerebrospinal fluid, 10 mg hyperbaric bupivacaine was administered and the concentrations in serum (Tmax and Tu max respectively). patient was placed in the lateral decubitus position with the operative side dependent. The lateral decubitus position was maintained for twenty minutes with the aim of obtaining a preferential sensory and motor block on the operative side. A pneumatic tourniquet, inflated to 100 mmHg above normal systolic pressure, with a maximum of 300mmHg, was used on the patient’s thigh to diminish blood loss. Cemented posterior-stabilized total knee replacement with patella resurfacing, was performed according to standard hospital procedure. 44 Ropivacaine pharmacokinetics with tourniquet use Ropivacaine pharmacokinetics with tourniquet use 45

Results

All patients completed the study protocol. None of the patients showed any sign of LAST. Patient characteristics are shown in Table 1.

Table 1. Patient Characteristics Sex (M/F) 12/8 3 Age (years) 58.5 ± 6.7 3 Weight (kg) 89 ± 14 Height (m) 1.73 ± 0.10 BMI (kg/m2) 30 ± 4 Values are proportions or mean ± SD.

Table 2. Pharmacokinetic Parameters. Median [IQR]

Cmax (µg/mL) 1.00 [0.34]

Tmax (min) 240 [120] C (µg/mL) 0.09 [0.05] u max Figure 1. Median total ropivacaine concentrations. Tu max (min) 300 [110] Data are presented as median with first and third quartiles, whiskers represent data in 1.5 interquartile range with outliers plotted as individual points (Tukey boxplot). Cmax = maximum total ropivacaine concentration; Tmax = time to Cmax;

Cu max = maximum free ropivacaine concentration; Tu max = time to Cu max

Pharmacokinetic data A summary of the pharmacokinetic data is given in Table 2. In eight samples the unbound ropivacaine concentration was below the lower limit of quantitation (<0.05 µg/mL). For further calculations these samples were classified as 0.05 µg/mL to prevent underestimation

of mean Cu max.

Median peak plasma concentration of total ropivacaine (Cmax) was 1.06 µg/mL [0.34], median

time to reach maximum plasma concentration (Tmax) was 240 minutes [120]. Figure 1 shows the pharmacokinetic data; Figure 2 shows total ropivacaine plasma concentrations for all

individual patients. Median peak plasma concentration of unbound ropivacaine (Cu max) was

0.09 µg/mL [0.05], median Tu max was 300 minutes [110].

Raw data of total and unbound ropivacaine are presented in Table 3.

Figure 2. Individual datapoints of total ropivacaine plasma concentrations. 46 Ropivacaine pharmacokinetics with tourniquet use Ropivacaine pharmacokinetics with tourniquet use 47

Discussion 7.2 5.9 8.5 9.1 8.6 8.5 7.2 7.8 Na 7.0 8.2 6.8 7.3 8.2 7.5 10.7 Na 8.4 11.5 9.8 % free This is the first study that describes the pharmacokinetic profile of ropivacaine after high- dose, high-volume, single-shot LIA of the knee with epinephrine, without the use of a periarticular catheter or surgical drain. We found a great inter-individual variation in the

plasma concentration of ropivacaine, but even the highest measured Cmax (2.26 µg/mL) and 0.13 0.06 0.11 0.10 0.09 0.09 0.07 0.05 <0.05 0.06 0.08 0.05 0.06 0.08 0.05 0.08 <0.05 0.05 0.07 0.06 Free conc. (mg/mL) Cu max (0.13 µg/mL) are well below the toxic threshold described by Knudsen et al. in 1997, which 3 were 4.3 and 0.56 µg/mL, respectively.12 3 In the literature many different procedures of LIA for knee surgery are described with varying 1.81 1.02 1.29 1.10 1.05 1.06 0.97 0.64 0.41 0.86 0.98 0.73 0.82 0.97 0.67 0.75 0.30 0.59 0.61 0.61 Total conc. Total (mg/mL) doses of ropivacaine and additives such as epinephrine or analgesics. Some authors use an intra- or periarticular catheter for continuous postoperative ropivacaine infusion, and sometimes a surgical drain is left at the operative site. These variations in technique may influence the efficacy of the LIA in terms of pain relief, but they may affect ropivacaine plasma 360 1440 360 360 360 360 60 1440 1440 360 240 360 120 1440 1440 120 1440 360 1440 370* Measurement after peak concentration Time (min) concentrations as well.

The first to describe the pharmacokinetic profile of ropivacaine LIA after TKA was Affas et al. 5.3 8.1 6.4 9.5 Na 7.1 8.9 6.9 9.9 11.3 8.8 6.7 8.4 7.8 7.5 9.7 Na Na 11.8 Na % free in 2012.13 They studied pharmacokinetics of 300 mg ropivacaine, epinephrine and additional ketorolac by taking blood samples at 40 and 60 minutes, and at 2, 4, 6, 12, 24 hours after tourniquet release. Slightly lower mean maximum ropivacaine concentrations (mean: 0.813

µg/mL, range: 0.435-1.735 µg/mL) were found compared to our study. They found the Tmax 0.10 0.12 0.06 0.08 <0.05 0.06 0.05 0.07 0.10 0.09 0.08 0.06 0.07 0.07 0.06 0.07 <0.05 <0.05 0.08 <0.05 Free conc. (mg/mL) between 4-6 hours after tourniquet release and, also similar to our study, some of the patients showed the highest plasma concentrations at 24 hours after tourniquet release. They found slightly lower peak plasma concentrations, which can be explained by the 25% lower dose compared to our study. 1.88 1.47 0.94 0.84 0.67 0.85 0.56 1.02 1.01 0.80 0.91 0.89 0.83 0.90 0.80 0.72 0.57 0.63 0.58 0.39 Total conc. Total (mg/mL) The cause of the great inter-individual variety of plasma concentrations and time to peak plasma concentrations is unclear. Patient characteristics such as protein binding capacity and speed of elimination through cytochrome P450 enzymes may be a factor. Also, the adding 120 240 120 120 120 120 20 240 240 120 90 120 60 240 240 60 240 120 240 120 Measurement before peak concentration Time (min) of epinephrine to the ropivacaine could be of influence, slowing the uptake of ropivacaine by its vasoconstrictive properties, and possibly the injection technique may also contribute to plasma absorption of ropivacaine. 5.3 8.2 8.1 6.5 6.7 7.0 5.6 10.5 8.7 8.9 8.1 5.2 7.6 11.2 5.7 10.8 Na 6.9 16.4 Na % free Brydone et al.14 used 400mg ropivacaine for LIA, a similar dose as in our study, without added epinephrine and they found a more rapid rise in plasma concentration, with a peak concentration at 20 minutes after tourniquet release. A maximum peak concentration of 0.12 0.13 0.11 0.08 0.08 0.08 0.06 0.11 0.09 0.09 0.08 0.05 0.07 0.10 0.05 0.09 <0.05 0.05 0.11 <0.05 Free conc. (mg/mL) 3.093 µg/mL was found, 37% higher than our maximum. In contrast to our protocol, an intra- articular catheter was placed infusing 20 mg ropivacaine per hour, therefore making it unclear whether the rapid rise in plasma concentration and higher peak plasma concentration is due to the intra-articular catheter or to the absence of epinephrine. In their study, potential toxic 2.26 1.58 1.36 1.23 1.19 1.14 1.08 1.05 1.04 1.01 0.99 0.96 0.92 0.89 0.87 0.83 0.80 0.72 0.67 0.61 Total conc. Total (mg/mL) concentrations of total ropivacaine were reached in 2 patients. However, their highest mean free ropivacaine concentration was 0.041µg/mL (SD 0.023) at t=20 minutes, overall sample concentrations varied from 0.001 µg/mL to 0.104 µg/mL and thus stayed far below the toxic 240 360 240 240 240 240 40 360 360 240 120 250* 90 360 360 90 360 240 360 250* Peak concentration Peak Time (min) threshold.

Another study combined 375 mg ropivacaine with epinephrine as LIA with two infusion 15 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 Patient catheters, both infusing 40 mg/hour. Patients reached the maximum unbound ropivacaine Individual ropivacaine concentrations. 3. Individual ropivacaine Table Total ropivacaine serum concentrations (µg/mL) and corresponding unbound ropivacaine ultra-filtrate concentrations (µg/mL). Na = not applicable. * Sample drawn drawn 10 minutes later than (µg/mL). Na = not applicable. * Sample drawn ultra-filtrate concentrations and corresponding unbound ropivacaine (µg/mL) serum concentrations ropivacaine Total scheduled due to logistic reason 48 Ropivacaine pharmacokinetics with tourniquet use Ropivacaine pharmacokinetics with tourniquet use 49

concentration 6 hours after ropivacaine injection, however maximum total plasma This pharmacokinetic study was not powered to investigate the incidence of LAST, however, concentrations were reached 24 hours after the end of surgery. Despite the additional 192 mg based on current data there are no indications that the use of 400mg ropivacaine combined administrated via both catheters, maximum total and unbound ropivacaine concentrations with epinephrine in TKA yields a high risk of LAST. Knudsen et al.12 defined the unbound were slightly lower (0.888 µg/mL (0.539-1689) and 0.050 µg/mL (0.025-0.105), respectively) as ropivacaine toxic threshold in arterial samples to be 0.56 (range: 0.34-0.85) µg/mL. With a compared to our study, which could be explained by the surgical drain. An average volume of maximum unbound ropivacaine concentration of 0.13 µg/mL when ropivacaine is used for LIA, 600 mL shed blood with high ropivacaine concentrations was collected via the surgical drain the peak concentration remains far below the toxic threshold. in the first 6 hours after surgery. Like in our study, free fractions were here highly variable; with 3 a range of 2.7-12.6% and an average free fraction of 4.8%. In conclusion, a single shot of 400 mg ropivacaine with added epinephrine for LIA after TKA 3 appears to be safe, the maximum plasma concentrations of ropivacaine stay far below the A limitation of our study is the scarcity of samples between 6 and 24 hours, just after the peak toxic threshold of 0.56 µg/mL free ropivacaine. However, because of great inter-individual concentration. Based on the literature available when we designed the study, we had expected variety in the absorption of local anesthetics and in the threshold of local anesthetic toxicity, 16 a much faster uptake of ropivacaine. Brydone et al. and Ng et al. found a Tmax of 30 minutes sporadic cases of LAST may still occur. The time to reach Cmax might be several hours after after LIA resp. intra-articular administration of ropivacaine without epinephrine, while tourniquet release, this slow absorption is contributing to the safety profile of high-dose 15 13 Thomassen et al. and Affas et al. found the Tmax 4-6 hours after the administration of LIA with ropivacaine for LIA. added epinephrine. These two latter studies encountered the same problem of ‘missing’ peak plasma concentrations in the sampling scheme. Unfortunately both these studies were yet to Financial support and sponsorship be published when we designed our study and we used the sampling scheme of the two first This study was supported entirely by internal funds of the department of Anesthesiology, studies, expecting an earlier rise in plasma concentrations. Sint Maartenskliniek, Nijmegen, The Netherlands.

One of our patients had unexpected early peak plasma concentrations. Peak concentrations

were found in the first hour after tourniquet release, while Cmax was comparable with other patients (Table 3, patient 3). In this patient the analgesia was already insufficient from 1.5 hour after tourniquet release. This patient was successfully titrated with intravenous morphine and clonidine. The early rise of plasma ropivacaine and insufficient analgesia may be explained by rapid intravenous absorption of ropivacaine, the cause of which is unknown.

In our study, we used 150 mL ropivacaine 0.2% with-, and without epinephrine. The ropivacaine with epinephrine was injected in the periarticular tissue. After closure of the knee capsule, the subcutaneous tissue of the knee was infiltrated with 50 mL of ropivacaine without epinephrine in order to prevent blistering of the skin. Compared with ropivacaine with epinephrine, the portion of ropivacaine without epinephrine will be absorbed more rapidly with an expected 14 Cmax of 20 minutes after tourniquet release. From a theoretical perspective, this portion may be expected to contribute to the rise in the plasma concentration of ropivacaine especially during the first 20-40 minutes. However, since the plasma concentration continued to rise well beyond 40 minutes, we do not believe that the portion of ropivacaine without epinephrine

affected the Cmax and Tmax found in our study.

We cannot exclude the possibility that the true maximum plasma concentration lies between

4 and 6 or even 6 and 24 hours in some patients. This Tmax, which appears to be at least later than four hours after tourniquet release, makes monitoring the patients until beyond this point logistically and financially challenging. In addition, prolonged monitoring would also interfere with the principles of fast track recovery. Although actual maximum plasma concentrations may be missed in our study, maximum plasma levels of unbound ropivacaine remain so far below toxic concentrations that LIA with high-dose ropivacaine and added epinephrine can be considered safe. Prolonged monitoring is in our opinion therefore not necessary.

50 Ropivacaine pharmacokinetics with tourniquet use 51

References

1. Husted, H. et al. What determines length of stay after total hip and knee arthroplasty? A nationwide study in Denmark. Archives of orthopaedic and trauma surgery 130, 263–8 (2010). Chapter 4 2. Perlas, A. et al. The impact of analgesic modality on early ambulation following total knee arthroplasty. Regional anesthesia and pain medicine 38, 334–9 (2013). 3. Kerr, D. R. & Kohan, L. Local infiltration analgesia: a technique for the control of acute postoperative pain Pharmacokinetics of 400 mg 3 following knee and hip surgery: a case study of 325 patients. Acta orthopaedica 79, 174–83 (2008). 4. Andersen, L. Ø., Husted, H., Otte, K. S., Kristensen, B. B. & Kehlet, H. High-volume infiltration analgesia in total knee arthroplasty: a randomized, double-blind, placebo-controlled trial. Acta anaesthesiologica Locally Infiltrated Ropivacaine Scandinavica 52, 1331–5 (2008). 5. Andersen, L. O. & Kehlet, H. Analgesic efficacy of local infiltration analgesia in hip and knee arthroplasty: a after Total Knee Arthroplasty systematic review. British Journal of Anaesthesia 113, 360–374 (2014). 6. Mazoit, J.-X. Local anesthetics and their adjuncts. Paediatric anaesthesia 22, 31–8 (2012). Without Perioperative 7. Ropivacaine 2 mg/ml solution for injection - Summary of Product Characteristics (SPC) - (eMC). (2018). Available at: https://www.medicines.org.uk/emc/product/9686/smpc. (Accessed: 29th March 2019). Tourniquet Use 8. Rosenberg, P. H., Veering, B. T. & Urmey, W. F. Maximum recommended doses of local anesthetics: a multifactorial concept. Regional anesthesia and pain medicine 29, 564–75 (2004). 9. McCartney, C. J. L. & McLeod, G. A. Local infiltration analgesia for total knee arthroplasty. British journal of anaesthesia 107, 487–9 (2011). 10. Fenten, M. G. E., Rohrbach, A., Wymenga, A. B. & Stienstra, R. Systemic local anesthetic toxicity after local infiltration analgesia following a polyethylene tibial insert exchange: a case report. Regional anesthesia and pain medicine 39, 264–5 (2014). 11. Schoenmakers, K. P. W. et al. Pharmacokinetics of 450 mg ropivacaine with and without epinephrine for combined femoral and sciatic nerve block in lower extremity surgery. A pilot study. British journal of clinical pharmacology 75, 1321–7 (2013). 12. Knudsen, K., Beckman Suurküla, M., Blomberg, S., Sjövall, J. & Edvardsson, N. Central nervous and cardiovascular effects of i.v. infusions of ropivacaine, bupivacaine and placebo in volunteers. British journal of anaesthesia 78, 507–14 (1997). 13. Affas, F. et al. A randomized study comparing plasma concentration of ropivacaine after local infiltration analgesia and femoral block in primary total knee arthroplasty. Scandinavian Journal of Pain 3, 46–51 (2012). 14. Brydone, A. S. et al. Ropivacaine plasma levels following high-dose local infiltration analgesia for total knee arthroplasty. Anaesthesia 70, 784–90 (2015). 15. Thomassen, B. J. W., Pool, L., Van Der Flier, R., Stienstra, R. & in ’t Veld, B. A. Safety of retransfusing shed blood after local infiltration analgesia in total knee arthroplasty. Acta orthopaedica Belgica 78, 506–11 (2012). 16. Ng, H.-P. et al. Efficacy of intra-articular bupivacaine, ropivacaine, or a combination of ropivacaine, morphine, and ketorolac on postoperative pain relief after ambulatory arthroscopic knee surgery: a randomized double-blind study. Regional anesthesia and pain medicine 31, 26–33 (2006). S.M.K. Bakker M.G.E. Fenten D.J. Touw B.J.F. van den Bemt P.J.C. Heesterbeek G.J. Scheffer R. Stienstra

Regional Anesthesia and Pain Medicine 43(7), 699-704 (2018) 52 Ropivacaine pharmacokinetics without tourniquet use 53

Abstract Introduction

Background Total knee arthroplasty (TKA) is commonly used to treat severe knee osteoarthritis. During Local infiltration analgesia (LIA) with ropivacaine for Total Knee Arthroplasty (TKA) is the last decades, the prevalence of TKA has been increasing worldwide, and this increase is increasingly used. Despite the high doses of ropivacaine, LIA is considered safe and this expected to continue in the next decades.1 Thus, TKA is at present a frequently performed perception is sustained by pharmacokinetic data demonstrating that maximum concentrations procedure, claiming an increasing part of the orthopedic surgical capacity. In recent years, of ropivacaine stay well below the toxic threshold in plasma. These pharmacokinetic studies research has focused on improving outcome and reducing length of hospital stay by developing all involve TKA procedures with the use of a tourniquet. Recently, performing TKA without fast-track surgery protocols.2,3 the use of a tourniquet is gaining popularity, but no pharmacokinetic data exist when LIA is administered for TKA without the use of a tourniquet. The purpose of this study is to describe TKA is associated with severe postoperative pain that is not only uncomfortable for the patient, the pharmacokinetic profile of a single-shot ropivacaine (200ml 0.2%) and 0.75 mg epinephrine but may also hinder early mobilization, recovery and hospital discharge; adequate treatment (1000µg/mL) when used for LIA in patients for TKA without a tourniquet. of postoperative pain is therefore mandatory. A multimodal pain treatment including local 4 infiltration analgesia (LIA) is often used to provide analgesia. LIA is considered to be a simple Methods and low-invasive technique4 that is characterized by the infiltration of a long acting local In this prospective cohort study twenty patients treated with LIA for TKA without a tourniquet anesthetic in the soft tissues surrounding the knee joint,5,6 providing effective analgesia in were studied. Plasma samples were taken at 20, 40, 60, 90, 120, 240, 360, 480, 600, 720 and the early postoperative period without side effects such as nausea (opioids) or m. quadriceps 1440 minutes after local anesthetic infiltration, in which total and unbound ropivacaine femoris weakness (femoral nerve block). concentrations were determined. LIA is usually performed with ropivacaine in relatively high doses exceeding the maximum Results recommended dose of 3-4 mg/kg,7,8 and therefore local anesthetic systemic toxicity (LAST) is Results are given as median [IQR]. Median peak ropivacaine concentration was 1.16 µg/ a concern. Nevertheless, worldwide thousands of patients have received LIA for knee surgery mL [0.46]; median peak unbound ropivacaine concentration was 0.05 µg/mL [0.02]. The with no apparent signs of systemic toxicity and pharmacokinetic studies9–11 have confirmed corresponding times to reach the maximum concentration for total and unbound ropivacaine the safety of high doses of ropivacaine for LIA, the concentration of unbound ropivacaine in were 360 [240] and 360 [360] minutes respectively. plasma staying well below the toxic threshold. The pharmacokinetic studies mentioned above all concern patients who had surgery under tourniquet-induced exsanguination with Conclusion the tourniquet still inflated at the time of local infiltration. However, because of an increased Although great inter-individual variability in ropivacaine concentration was found, both total awareness of the disadvantages of a tourniquet, total knee arthroplasty without the use of a and unbound maximum serum concentrations remained below the assumed systemic toxic tourniquet is gaining popularity.12–14 thresholds in all samples. It is undetermined if the absence of circulation at the site of infiltration caused by an inflated Trial registration tourniquet and the hyperperfusion following deflation affect the systemic absorption of NL6159 (NTR6306). ropivacaine and to what extent. The aim of the present study is to describe the pharmacokinetic profile of ropivacaine after LIA in TKA surgery without the use of a tourniquet.

Methods

Patients This prospective cohort study was approved by the Medical Research Ethics Committee Slotervaart Hospital and Reade (NL60548.048.17, date of approval February 10th, 2017) before onset of participant recruitment. The study was registered at the Netherlands Trial Registry (http://www.trialregister.nl) with trial ID NTR6306 on January 16th, 2017.

Patients scheduled for primary TKA under spinal anesthesia without the use of a tourniquet were assessed for eligibility. Criteria for inclusion were ASA physical status classification I or II, age 50-80 years, body mass index less than 40kg/m2, and hemoglobin 7.5 mmol/L or greater. Exclusion criteria were the (expected) use of a tourniquet, placement of a surgical drain, 54 Ropivacaine pharmacokinetics without tourniquet use Ropivacaine pharmacokinetics without tourniquet use 55

known hypersensitivity to amide-type local anesthetics, hepatic or renal insufficiency and the last patient had been completed so that all samples could be assayed in one batch. use of medications that are known to affect the clearance of ropivacaine.7 Eligible patients were informed verbally and in writing, and written informed consent was obtained from all The analysis of the ropivacaine concentrations was performed by the Laboratory for participants before the start of the study. Toxicology, Therapeutic Drug Monitoring and Pharmaceutical analysis of the Department of clinical Pharmacy and Pharmacology at the University Medical Center Groningen, Groningen, The study was conducted at the Sint Maartenskliniek Nijmegen, The Netherlands between the Netherlands, using Liquid chromatography-tandem mass spectrometry as described February and May 2017 according to the Declaration of Helsinki and later revisions thereof and previously.11 Total ropivacaine concentrations were measured in all samples. The concentration in accordance with the International Conference on Harmonisation of Technical Requirements of unbound ropivacaine of each patient was assessed in ultrafiltrate of the sample with the for Registration of Pharmaceuticals for Human Use guidelines for Good Clinical Practice. highest total ropivacaine concentration, as well as in those of the preceding and following time point samples. Thus, from each patient three unbound ropivacaine concentrations 4 Anesthetic and Surgical procedure were obtained; the highest value of these three concentrations was taken as the individual 4 The anesthetic and surgical procedures have been described in a previous study conducted maximum unbound ropivacaine concentration (Cu max) and used for calculating the average 11 in our institution and are identical except for the absence of a tourniquet during surgery. Cu max. To calculate the average unbound ropivacaine fraction, we used all free concentrations Briefly, all patients received standard multimodal analgesia with paracetamol 1000 mg q.i.d., and the corresponding total ropivacaine concentrations. etoricoxib 90 mg once daily and gabapentin gabapentin 300 mg and 600 mg both once daily or 300 mg twice daily if patients were older than 70 years. This analgesic regimen was started Sample size and statistics preoperatively on the day of surgery. Postoperative pain was evaluated with a numerical The purpose of this study was to describe the pharmacokinetic profile of ropivacaine 400 mg rating scale (NRS) ranging from 0 (no pain) to 10 (worst pain possible). In case of an NRS score ropivacaine when used for LIA in TKA surgery without the use of a tourniquet. Since this is ≥ 4 patients received oxycodone 5 to 10 mg orally. primarily a descriptive study, the sample size was not calculated but based on choice. Because the methodology of this study is identical to our previous study11 but for the absence of a Surgery was performed under standard monitoring (electrocardiogram, non-invasive blood tourniquet, we decided to study the same number of subjects (n=20). pressure and peripheral oxygen saturation). The anesthetic technique was unilateral spinal anesthesia with 10 mg hyperbaric bupivacaine 0.5%. During surgery patients were either Data were analyzed using Stata version 13.1 (Stata Corporation, College Station, TX, USA). awake or sedated with propofol upon request. Data were tested for normality using the Shapiro-Wilk test. Data are presented as mean (SD) or median [IQR], as appropriate. Peak concentrations of total and unbound ropivacaine

After implanting and cementing a Genesis II Total Knee System with patellar resurfacing (Smith (Cmax and Cu max respectively) and the times to reach these peak concentrations (Tmax and Tu max & Nephew, Memphis, TN, USA), LIA was performed by the orthopedic surgeon using 300 mg respectively) are reported. ropivacaine 0.2% (150 mL) with epinephrine 5 µg/mL for the posterior and anterior capsules; a further 100 mg ropivacaine 0.2% (50 mL) without epinephrine was infiltrated subcutaneously, the total dose of ropivacaine being 400 mg. Upon conclusion of the subcutaneous ropivacaine Results infiltration the time was designated T=0. The study protocol was completed by all patients. The flow of patient recruitment and study Blood sampling and assays participation is shown in Figure 1. No signs or symptoms of LAST were observed in any of the Before anesthesia, all patients received a peripheral intravenous catheter for the patients. Patient characteristics are displayed in Table 1. administration of fluids and medication. Blood samples for the assessment of ropivacaine concentrations were drawn from a separate 16-gauge peripheral intravenous catheter placed Table 1. Baseline characteristics in the antecubital vein of the contralateral arm. No intravenous fluids were administered N = 20 through this sampling catheter, except flushing with 1-2 mL NaCL 0.9% following each Sex (M/F) 9/11 sampling to prevent clotting in the catheter. A baseline sample was taken immediately after Age (years 68.3 ± 5.7 placement of the sampling catheter; consecutive samples were obtained at T = 20, 40, 60, 90, Weight (kg) 88 ± 14 120, 240, 360, 480, 600, 720 and 1440 minutes by one of the investigators. Based on the results Height (m) 1.71 ± 0.08 of our previous study,11 we added the 480-, 600-, and 720-minute samples to the sampling 2 schedule. BMI (kg/m ) 30 ± 5 Values displayed as number of patients or mean ± SD To avoid saline-diluted samples, two samples of 5 mL each were drawn at all time intervals. The first sample was discarded; the second sample was drawn in an EDTA tube, centrifuged and stored at -80ºC within one hour after collection. The samples were kept in storage until 56 Ropivacaine pharmacokinetics without tourniquet use Ropivacaine pharmacokinetics without tourniquet use 57

4 4

Figure 1. Flowchart

Pharmacokinetic data

Median peak plasma concentration of total ropivacaine (Cmax) was 1.16 [0.46] µg/mL; median

time to reach maximum plasma concentration (Tmax) was 360 [240] minutes. Median peak

unbound ropivacaine concentration (Cu max) was 0.05 [0.02] µg/mL, median Tu max was 360 [360] Figure 2. Individual datapoints of total ropivacaine plasma concentrations.

minutes. The highest individual Cu max was 0.082 µg/mL. Mean unbound ropivacaine fraction was 3.7% (1.3).

Figure 2 shows total ropivacaine plasma concentrations for all individual patients. In Table 2 and Figure 3 we have combined the results of the present study with those of our previous study.11 The raw data of total and unbound ropivacaine concentrations are presented in Table 3.

Table 2. Pharmacokinetic data. No Tourniquet cohort (n=20) Tourniquet cohort (n=20)

Cmax (µg/mL) 1.16 [0.46] (0.78; 2.34) 1.00 [0.34] (0.61; 2.26)

Tmax (min) 360 [240] (240; 600) 240 [120] (40; 360)

Cu max (µg/mL) 0.05 [0.02] (0.014; 0.093) 0.09 [0.05] (0.05; 0.13)

Tu max (min) 360 [360] (240; 720) 300 [110] (60; 370)

Data displayed as median [IQR] (minimum value; maximum value). Cmax = maximum total ropivacaine concentration;

Tmax = time to Cmax; Cu max = maximum free ropivacaine concentration; Tu max = time to Cu max

Figure 3. Combination of the median total ropivacaine concentrations of the present (without tourniquet) and our previous (with tourniquet) study11. Data are presented as median with first and third quartiles. whiskers represent data in 1.5 interquartile range with outliers plotted as individual points (Tukey boxplot). 58 Ropivacaine pharmacokinetics without tourniquet use Ropivacaine pharmacokinetics without tourniquet use 59

Discussion 2.8% 5.2% 2.2% 4.7% 2.8% 4.0% 2.8% 2.6% 1.4% 3.7% 3.0% 3.4% 3.0% 4.8% 6.2% 3.4% 3.2% 6.2% 6.0% 5.9% % free This study describes the pharmacokinetic profile of ropivacaine after a single-shot, high-dose LIA in TKA without the use of a tourniquet. Although the variation in ropivacaine concentrations is large, all total and unbound concentrations are well below the toxic threshold as described by Knudsen (4.3 µg/mL for total ropivacaine and 0.56 µg/mL for unbound ropivacaine).15 0.029 0.093 0.017 0.053 0.028 0.059 0.025 0.03 0.011 0.031 0.042 0.074 0.058 0.046 0.043 0.041 0.031 0.053 0.047 0.068 Free conc. (mg/mL)

Many different procedures of LIA for knee surgery have been published with varying administration techniques, post-operative drain use, additives, and differences in doses of ropivacaine.9–11,16 Although these technical differences may affect the pharmacokinetic 1.04 1.79 0.77 1.13 1.00 1.49 0.9 1.15 0.8 0.83 1.41 2.15 1.91 0.95 0.69 1.22 0.98 0.85 0.78 1.15 Total conc. Total (mg/mL) profile, and differences in ropivacaine dose make a comparison of maximum concentrations 4 problematic, the available pharmacokinetic data have shown that ropivacaine concentrations 4 do not reach toxic levels. 720 480 360 480 480 360 360 360 480 360 720 360 360 360 480 720 360 600 600 720 Measurement after peak concentration Time (min) 11 We found a Tmax of 360 minutes in the present study and 240 minutes in our previous study.

This agrees with the results of Affas et al. who reported a Tmax of 240 – 360 min in a study using 9 a LIA mixture of 300 mg ropivacaine with epinephrine and ketorolac for TKA , and a Tmax of 6.0% 4.5% 1.7% 2.8% 4.5% 3.9% 2.4% 3.0% 1.7% 5.4% 2.6% 3.4% 2.8% 4.5% 4.1% 3.2% 2.8% 3.9% 4.7% 4.6% % free 240 minutes using a LIA mixture of 200 mg ropivacaine with epinephrine and ketorolac for 17 total hip arthroplasty. This Tmax of 240 – 360 minutes seen with LIA is much later than the Tmax associated with peripheral nerve blocks. In a study using 450 mg ropivacaine with epinephrine 18 5 µg/mL for combined sciatic/femoral nerve block the mean Tmax was 100 minutes , and others 0.067 0.086 0.013 0.031 0.044 0.039 0.023 0.023 0.014 0.039 0.038 0.059 0.046 0.032 0.029 0.039 0.021 0.035 0.039 0.055 Free conc. (mg/mL) have reported a mean Tmax of 80 minutes for lumbar plexus block and 38 minutes for combined 19 lumbar plexus/sciatic nerve block using ropivacaine with epinephrine 2.5 µg/mL. Tmax is largely determined by absorption from the site of injection and elimination; since differences in the elimination of ropivacaine between LIA and peripheral nerve blocks are unlikely, the most likely explanation for the difference in T is that absorption with LIA is considerably 1.12 1.89 0.77 1.11 0.97 1.01 0.96 0.76 0.82 0.72 1.46 1.73 1.63 0.71 0.71 1.24 0.74 0.91 0.83 1.19 Total conc. Total (mg/mL) max slower than with peripheral nerve block. A possible explanation is that in case of LIA local blood flow may be reduced by postoperative swelling. 480 240 120 240 240 120 120 120 240 120 480 120 120 120 240 480 120 360 360 500* Measurement before peak concentration Time (min) Tourniquets are used to create a bloodless surgical field, with the purpose of facilitating the surgical procedure and limiting intra-operative blood loss. However, these advantages come with a price; the use of a tourniquet may be associated with complications such as venous 12,14 6.3% 2.8% 5.0% 2.8% 2.4% 3.3% 2.3% 3.2% 1.6% 4.4% 2.5% 3.5% 2.5% 5.1% 4.6% 4.2% 2.2% 3.9% 5.0% 4.5% % free thrombosis, wound infections and delayed functional recovery or uncomfortable thigh pain at the level of the tourniquet. Because it was demonstrated that the use of a tourniquet might reduce intraoperative blood loss but has little, if any, effect on total (intraoperative plus postoperative) blood loss,12 awareness of the disadvantages of a tourniquet has increased and 0.073 0.059 0.048 0.032 0.024 0.054 0.024 0.037 0.014 0.039 0.037 0.082 0.058 0.062 0.036 0.056 0.023 0.037 0.046 0.056 Free conc. (mg/mL) caused a shift in the balance towards surgery without a tourniquet.

The purpose of our present study was to investigate if the absence of a tourniquet affects the pharmacokinetics of ropivacaine. With Bier block, it has been shown that after tourniquet 1.17 2.08 0.96 1.15 1.00 1.66 1.04 1.16 0.85 0.89 1.49 2.34 2.31 1.21 0.78 1.34 1.03 0.96 0.92 1.24 Total conc. Total (mg/mL) release a significant portion of the local anesthetic remains in the operated limb for a prolonged time.20 When performing surgery with a tourniquet there is a 10- to 15-minute timeframe between ropivacaine infiltration and restoration of the blood flow. During this 600 360 240 360 360 240 240 240 360 240 600 240 240 240 360 600 240 480 480 600 Peak concentration Peak Time (min) period, systemic absorption is absent and postponed until the time of deflation. In analogy with the observed pharmacokinetics with Bier block, it is conceivable that this delay in absorption affects the pharmacokinetic profile of ropivacaine compared with the situation 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 Patient without the use of a tourniquet, when systemic absorption starts simultaneously with local Individual ropivacaine concentrations. 3. Individual ropivacaine Table Total ropivacaine serum concentrations (µg/mL) and corresponding unbound ropivacaine ultra-filtrate concentrations (µg/mL). Na = not applicable. * Sample drawn at t 500 due to logistic reasons ultra-filtrate concentrations and corresponding unbound ropivacaine (µg/mL) serum concentrations ropivacaine Total 60 Ropivacaine pharmacokinetics without tourniquet use Ropivacaine pharmacokinetics without tourniquet use 61

anesthetic infiltration. Specifically, we wanted to know if the absence of a tourniquet would References

result in a more rapid uptake of local anesthetic and a higher Cmax, with possible implications for patient safety in terms of the risk of LAST. Comparing the data of the present study with 1. Patel, A., Pavlou, G., Mújica-Mota, R. E. & Toms, A. D. The epidemiology of revision total knee and hip those of our previous study,11 it is obvious that the presence or absence of a tourniquet during arthroplasty in England and Wales: a comparative analysis with projections for the United States. A study infiltration has no major impact on the pharmacokinetic profile of ropivacaine. Maximum using the National Joint Registry dataset. The bone & joint journal 97–B(8), 1076–81 (2015).

concentrations are comparable, and the difference in Tmax is marginal with considerable 2. Husted, H., Holm, G. & Jacobsen, S. Predictors of length of stay and patient satisfaction after hip and knee overlap and most likely explained by the absence of samples between 360 and 1440 minutes in replacement surgery: Fast-track experience in 712 patients. Acta Orthopaedica 79, 168–173 (2008). the study with tourniquet (Table 3 and Figure 3). To what extent the intrinsic vasoconstrictive 3. Husted, H. et al. What determines length of stay after total hip and knee arthroplasty? A nationwide study properties of ropivacaine itself, the addition of epinephrine to the local anesthetic, or the phase in Denmark. Archives of orthopaedic and trauma surgery 130, 263–8 (2010). of hyperaemia following tourniquet deflation affect systemic absorption remains speculative. 4. Kopp, S. L. et al. Anesthesia and Analgesia Practice Pathway Options for Total Knee Arthroplasty. Regional 4 Anesthesia and Pain Medicine 42, 683–697 (2017). 4 In the present study, the unbound fraction of ropivacaine was 3.7%, whereas in our 5. Kerr, D. R. & Kohan, L. Local infiltration analgesia: a technique for the control of acute postoperative pain previous study,11 this fraction was considerably higher (8.3% (SD 2.0), calculated from all following knee and hip surgery: a case study of 325 patients. Acta orthopaedica 79, 174–83 (2008). free concentrations and corresponding total concentrations). In different pharmacokinetic 6. Kehlet, H. & Andersen, L. Ø. Local infiltration analgesia in joint replacement: the evidence and studies, the unbound fraction of ropivacaine varies widely, the two main reasons being recommendations for clinical practice. Acta Anaesthesiologica Scandinavica 55, 778–784 (2011). differences in analytical methods, or differences in alpha-1-acid glycoprotein.21 Because we 7. Rosenberg, P. H., Veering, B. T. & Urmey, W. F. Maximum recommended doses of local anesthetics: a used the same analytical method and ultrafiltration technique in both studies and differences multifactorial concept. Regional anesthesia and pain medicine 29, 564–75 (2004). in alpha-1-acid glycoprotein between the two study populations seem unlikely, we have no 8. Ropivacaine 2 mg/ml solution for injection - Summary of Product Characteristics (SPC) - (eMC). (2018). explanation for this difference in the unbound fraction of ropivacaine. Available at: https://www.medicines.org.uk/emc/product/9686/smpc. (Accessed: 29th March 2019) 9. Affas, F. et al. A randomized study comparing plasma concentration of ropivacaine after local infiltration A limitation of our previous study11 was the scarcity of samples between 6 and 24 hours. Based analgesia and femoral block in primary total knee arthroplasty. Scandinavian Journal of Pain 3, 46–51 (2012).

on the literature available at that time, we expected Tmax in the range between 1 and 6 hours. 10. Brydone, A. S. et al. Ropivacaine plasma levels following high-dose local infiltration analgesia for total knee To avoid missing actual peak concentrations beyond six hours, we extended the sampling arthroplasty. Anaesthesia 70, 784–90 (2015). schedule in the present study with additional samples at 8, 10 and 12 hours after ropivacaine 11. Fenten, M. G. E. et al. Pharmacokinetics of 400 mg ropivacaine after periarticular local infiltration analgesia

infiltration. Indeed, in the present study we found maxT to be later than 6 hours in 35 % of the for total knee arthroplasty. Acta anaesthesiologica Scandinavica 61, 338–345 (2017). patients. 12. Tai, T.-W. et al. Tourniquet use in total knee arthroplasty: a meta-analysis. Knee Surgery, Sports Traumatology, Arthroscopy 19, 1121–30 (2011).

The finding that with or without tourniquet maxT occurs at 4-6 hours after infiltration may be 13. Alcelik, I. et al. A comparison of outcomes with and without a tourniquet in total knee arthroplasty: a disturbing from a safety perspective; patients have been discharged from the recovery and systematic review and meta-analysis of randomized controlled trials. The Journal of arthroplasty 27, 331–40 are no longer frequently monitored for signs of LAST at the time ropivacaine concentrations (2012).

reach their maximum. On the other hand, the Cmax and Cu max reported by others and by us are 14. Zhang, W. et al. The effects of a tourniquet used in total knee arthroplasty: a meta-analysis. Journal of far below the toxic threshold,9–11 and worldwide, large numbers of patients have been treated orthopaedic surgery and research 9, 13 (2014). with LIA without signs of LAST. Based on these observations the conclusion that LIA with or 15. Knudsen, K., Beckman Suurküla, M., Blomberg, S., Sjövall, J. & Edvardsson, N. Central nervous and without tourniquet is a safe technique seems warranted. However, it should always be kept in cardiovascular effects of i.v. infusions of ropivacaine, bupivacaine and placebo in volunteers. British journal mind that local anesthetic concentration per se is not the only determinant of LAST: Individual of anaesthesia 78, 507–14 (1997). patient factors such as comorbidities, co-medication, blood flow at the site of infiltration, and 16. Thomassen, B. J. W., Pool, L., Van Der Flier, R., Stienstra, R. & in ’t Veld, B. A. Safety of retransfusing shed blood general condition play an equally important role.22, 23 after local infiltration analgesia in total knee arthroplasty. Acta orthopaedica Belgica 78, 506–11 (2012).

In conclusion, this study demonstrates that the absence of a tourniquet does not affect the maximum plasma concentrations or the time to reach maximum plasma concentrations of ropivacaine when used as LIA for TKA. Maximum plasma concentrations of unbound ropivacaine stay far below the toxic threshold of 0.56 µg/mL.

Financial support and sponsorship This study was supported entirely by internal funds of the department of Anesthesiology, Sint Maartenskliniek, Nijmegen, The Netherlands. 63

Chapter 5

Influence of a Tourniquet on Opioid Consumption after Local Infiltration Analgesia for Total Knee Arthroplasty

S.M.K. Bakker N.M. Kosse S. Crnic G.J. Scheffer R. Stienstra

Turkisch Journal of Anaesthesiology and Reanimation 47(2), 107-11 (2019) 64 Influence of a tourniquet on opioid consumption 65

Abstract Introduction

Background Total knee arthroplasty (TKA) is usually performed in the setting of an enhanced recovery To provide postoperative analgesia after total knee arthroplasty (TKA), local infiltration protocol. TKA may be associated with severe postoperative pain that may hamper early analgesia (LIA) with ropivacaine is increasingly used. TKA may be performed with or without mobilization. Adequate pain relief is therefore an essential part of enhanced recovery. In the use of a tourniquet. The absence of local blood flow when infiltrating local anesthetic many centers, local infiltration analgesia1,2 (LIA) in combination with oral analgesics has below an inflated tourniquet may affect the rate of systemic absorption, and this may have been introduced as the preferred method of postoperative analgesia because it combines an effect on the duration and intensity of analgesia as compared to LIA without the use of a adequate analgesia with a minimum of side effects. LIA protocols may vary in terms of dose tourniquet. The aim of this study was to investigate the influence of tourniquet use during and additives, but are characterized by local infiltration of the tissues surrounding the knee surgery on the time to first request (TTFR) of opioids, and opioid consumption. joint with a long acting local anesthetic (LA) such as ropivacaine.

Methods TKA can be performed with or without the use of a tourniquet. When the LIA mixture is Two historical time-based cohorts (one with and one without tourniquet during surgery) of infiltrated in the absence of a tourniquet, systemic absorption will commence immediately. 300 patients underwent primary TKA under spinal anesthesia and received LIA to provide By contrast, in case of an inflated tourniquet systemic absorption will be delayed until the postoperative analgesia. The cohorts were compared for TTFR of opioids and opioid tourniquet is deflated and local blood flow restored. On the other hand, after deflation of 5 consumption. a tourniquet there will be a period of hyperemia with a concomitant increase in systemic absorption. It is not known if these differences in systemic absorption with or without Results tourniquet alter the pharmacokinetic profile of LIA sufficiently to affect the duration of TTFR did not significantly differ between the tourniquet and non-tourniquet group with a analgesia. median [25th-75th percentile] of respectively 240 [102-651] and 282 [100-720] minutes. Median [25th-75th percentile] oxycodone use was higher in the tourniquet group with 50 [20-90] vs. 40 Until recently, TKA in our hospital was performed with the use of a tourniquet and LIA for [10-77.5] mg (p=0.01). postoperative pain relief. Perceived advantages of a tourniquet are a dry surgical field and reduced intraoperative blood loss. However, the use of a tourniquet has become subject to Conclusion discussion. In two meta-analyses, Zhang3 and Tai4 showed that total blood loss is not affected There was no difference in the time to first opioid consumption, suggesting that the presence by tourniquet use, but it does increase the risk of thromboembolic events and might even of an inflated tourniquet during local anesthetic injection does not alter systemic absorption hinder early postoperative rehabilitation. Based on these results, our standard practice has sufficiently to affect the duration of analgesia. However, the use of a tourniquet was associated recently changed to performing TKA without the use of a tourniquet. The purpose of this with a higher opioid consumption, which is most likely caused by pain resulting from the cohort study was to investigate if the presence or absence of an intraoperative tourniquet tourniquet itself. affects the duration of LIA.

Methods

The protocol of this study was approved by the hospital’s investigational review board. The Medical research and Ethics Committee ‘Slotervaart Hospital and Reade, Amsterdam, The Netherlands’ reviewed the study on June 7th 2017 and determined, based on the Dutch Medical Research Involving Human Subjects Act (Dutch acronym: WMO), that the research activities described meet the requirements for exemption from Ethical Committee review.

Patient demographics, procedure characteristics, opioids consumption and use of analgesics from 600 patients were extracted from the patient data management system. Medio 2016 our standard surgical procedure for primary TKA changed from perioperative tourniquet use, to surgery without the use of a tourniquet. The tourniquet cohort (Group T) comprised the last 300 patients who underwent primary TKA with perioperative tourniquet use before protocol change. The non-tourniquet cohort (Group NT) included the first 300 patients who underwent primary TKA after protocol change. Patients who were not treated according to the standard surgical procedure and perioperative protocol for the time-period in which they underwent surgery, were excluded. 66 Influence of a tourniquet on opioid consumption Influence of a tourniquet on opioid consumption 67

Procedure oxycodone,5 0.7 for intravenous piritramide6, and 0.05 for oral tramadol5. An average NRS pain In our hospital, TKA is performed according to a standard protocol, detailing the anesthetic and score at rest and during activity was calculated for each patient from all NRS scores recorded surgical procedure, perioperative pain treatment and rehabilitation, and discharge criteria. during hospitalization; average pain scores per group are based on these individual scores. All patients receive spinal anesthesia with 10 mg hyperbaric bupivacaine 0.5% in the sitting position. After bupivacaine injection patients are turned into the lateral decubitus position for Sample size calculation and statistical analysis 20 minutes with the side of surgery dependent to achieve a predominantly unilateral block. Our null hypothesis was that the presence or absence of a tourniquet would not result in Upon request patients receive conscious sedation with propofol (1-4 mg/kg/h) during surgery. differences in the TTFR. A difference of 180 min was chosen as clinically relevant. Data gathered In Group T, a pneumatic tourniquet is placed on the patient’s thigh and automatically inflated in our hospital for quality monitoring of primary TKA without the use of a tourniquet showed a to 50 mmHg above systolic blood pressure; in group NT, no tourniquet is used. After placement standard deviation of 561 minutes in the TTFR. Based on these data, the sample size required of the implants and before wound closure, all patients receive LIA: The posterior and anterior to identify a difference in TTFR of at least 180 min with a power of 80% was 300 patients per knee capsules are infiltrated with 100 mL and 50 mL ropivacaine 0.2% + epinephrine 5 µg/ group (two-sided, level of significance 0.05). Data analysis was performed using Stata version mL respectively; the subcutaneous tissue is infiltrated with 50 mL ropivacaine 0.2% without 13.1 (Stata Corporation, College Station, TX, USA). Data were tested for normality using the epinephrine. All patients receive an i.v. bolus of 10 mg/kg tranexamic acid with a maximum Shapiro-Wilk test. Descriptive statistics of patient and operation characteristics are presented of 1000 mg. This dose is administered just before tourniquet release in group T; in group NT, as percentage (95% confidence interval), mean ± SD or median and IQR [25th-75th percentile], 5 tranexamic acid is administered in the anesthetic room before the start of surgery. After as appropriate. Differences between the cohorts were tested for significance using the Mann- 5 wound closure, all patients receive a compression bandage before transfer to the recovery Whitney and Chi-square test for numerical and categorical variables respectively. A P-value room. <0.05 was considered statistically significant.

Standard oral multimodal analgesia includes paracetamol 1000 mg q.i.d., etoricoxib 90 mg once daily and gabapentin 600 mg b.i.d. or 300 mg if age is > 60 years. Pain is evaluated with Results a numerical rating scale (NRS) ranging from 0 (no pain) to 10 (worst pain imaginable). On the orthopedic ward, NRS scores are recorded by a nurse at least once during each 8-hour shift. All patients underwent surgery between December 2015 and March 2017. In this period 108 In addition, patients are instructed to contact the nurses when pain exceeds NRS 3 and these patients underwent TKA under general anesthesia. These patients are not included in the scores are recorded as well. In the recovery room, NRS scores are recorded and in case of an cohorts. Patients in both cohorts had similar demographic and surgical characteristics (Table NRS > 3, pain relief is titrated with intravenous increments of morphine 1-2 mg or piritramide 2). In the tourniquet and non-tourniquet cohort respectively 31 and 27 patients did not use 1.5-3 mg until the NRS is ≤ 3. At the orthopedic ward, breakthrough pain (NRS > 3) is treated a NSAID because of an allergy or kidney insufficiency. All patients received paracetamol and with oxycodone 5-10 mg ad libitum or i.m./s.c. opioids if oxycodone alone is not sufficient. gabapentin. There was no statistically significant difference in the TTFR between the two Sporadically, tramadol is used when other opioids are not well tolerated. The distribution of groups with a median [IQR] of 240 [102-651] minutes for Group T versus 282 [100-720] minutes opioids on the recovery and orthopedic ward is standardized: Opioids are checked, signed and for Group NT (p=0.482). registered in the patient data management system by two nurses at the time of distribution. According to our standard TKA protocol, patients are encouraged to mobilize starting on the There was no difference in blood transfusion requirements between the groups (Table 2). day of surgery. Patients must fulfill functional criteria before discharge (Table 1). Total opioid consumption (median [IQR]) was significantly higher in group T: 37.5 [17-67] mg, as compared to group NT 27 [10-60] mg (p=0.014). This difference was caused by a higher Table 1. Functional discharge criteria oxycodone consumption of patients in group T. There were no differences in parenteral opioid Active knee flexion ≥ 60 degrees, passive knee extension 0 degrees and tramadol consumption, nor were there differences in the percentage of patients using Quadriceps muscle force ≥ 3 on Medical Research Council Scale for muscle strength opioids. Data on opioid consumption are summarized in Tables 3A and B.

Making independent transfers There were no differences between the groups in the average NRS pain scores at rest (median Walking independently and safely with walking aid [IQR] 1.9 [1.4-2.5] for group T and 1.9 [1.3-2.5] for group NT) and during activity (median [IQR] 3.1 Climbing stairs independently and safely (if necessary for home situation) [2.4-3.7] for group T and 2.9 [2.3-3.5] for group NT).

Outcome measurements There was no difference between Group T and Group NT in median hospital length of stay The primary outcome parameter was the time to first request (TTFR) of postoperative pain (median [IQR] 2 [2-3] and 2 [2-3] days respectively). relief, defined as the time (in minutes) between ropivacaine infiltration and the first gift of any opioid. Secondary outcomes included hospital length of stay, transfusion requirements and the amount of opioids administered during hospitalization. Total opioid consumption was converted to intravenous morphine equivalent, using conversion factors of 0.67 for oral 68 Influence of a tourniquet on opioid consumption Influence of a tourniquet on opioid consumption 69

Table 2. Group characteristics Discussion Group T Group NT (n=300) (n=300) The purpose of this study was to investigate the influence of an intra-operative tourniquet on Sex, no. M/ no. F 107 / 193 125 / 175 the duration of LIA as determined by the TTFR of opioids. There was no difference in the TTFR between the two cohorts and thus our null hypothesis is retained. We found that the total Age (yr) 65 (6.7) 65 (9.3) opioid consumption was higher in patients in the tourniquet group. There was no difference Number of patients with age > 60 years 212/300 221/300 in hospital length of stay. Number of patients not receiving etoricoxib* 31/300 27/300 Weight (kg) 86 (17.3) 87 (17.7) The rationale for the use of a pneumatic tourniquet during TKA is twofold: To limit Height (cm) 171 (10) 170 (13) intraoperative blood loss and to facilitate a bloodless surgical field. However, tourniquets are BMI (kg m-2) 28.9 (5.4) 29.2 (5.2) associated with several complications such as deep venous thrombosis, wound infections, Duration of surgery, minutes 67 (15) 71 (13) and delayed functional recovery caused by tissue ischemia underneath and distal to the Tourniquet Time 50 [44-60] n/a tourniquet.3,4 Tourniquet use is also associated with increased postoperative pain: The first Time from LIA to tourniquet deflation 13 [10-16] n/a report describing the influence of a tourniquet on postoperative pain after TKA dates from 5 7 5 Number of patients requiring RBC transfusion 3 4 1995. Several RCTs followed after this first study, most of them reaching the same conclusion: Use of a tourniquet increases postoperative pain after TKA.8–12 Total units of RBC transfused 5 7 Data are displayed as number of patients, mean (SD), median [25th-75th percentile], n/a = not applicable. *Due to allergy or kidney insufficiency. RBC = red blood cells. When using LIA, it stands to reason that the duration of analgesia will be affected by absorption of the local anesthetic from the site of injection. Local blood flow is one of the main determinants of systemic absorption.13 Studies investigating lidocaine plasma levels Table 3A. Opioid consumption I after intravenous regional anesthesia demonstrate that prolonging the tourniquet time after 14 Group T (n=300) Group NT (n=300) P-value injection slows down peak systemic absorption and lowers peak concentration. Likewise, TTFR 240 [102-651] 282 [100-720] 0.482 compared to performing LIA without a tourniquet, the absence of circulation at the site of Total opioid consumption (mg in se) 38 [17-67] 27 [10-60] 0.014 injection when performing LIA with an inflated tourniquet will delay the initial absorption. Whether this delay alters the pharmacokinetic profile sufficiently to exert an effect on the Oxycodone consumption (mg in se) 33.5 [13.3-60.3] 26.8 [6.7-51.9] duration of LIA is speculative; a delay in systemic absorption might result in an increased Oxycodone consumption (mg) 50 [20-90] 40 [10-77.5] 0.012 presence of local anesthetic at the site of injection, which might lengthen the duration of Data are displayed as median [25th-75th percentile]. Total opioid consumption is the sum of oxycodone and tramadol consumption, and i.v., i.m. and s.c. opioids administration converted to intravenous morphine standard equivalent (se). analgesia. However, in the absence of pharmacokinetic data these contemplations remain Oxycodone consumption is presented as actual consumption (mg) and consumption converted to intravenous morphine hypothetical. We were unable to demonstrate an increased duration of analgesia in group T standard equivalent (mg in se). as determined by a difference in the TTFR and under the conditions of this study we conclude that the presence or absence of a tourniquet does not affect the duration of LIA. Table 3B. Opioid consumption II Group T (n=300) Group NT (n=300) P-value Although the average NRS scores were similar, total opioid consumption was significantly Use of any opioid (%) 91.0 (87.2 to 93.7) 90.0 (86.0 to 92.9) 0.676 higher in group T. Since there was no difference in the TTFR, it seems that in group T Oxycodone use (%) 90.0 (86.0 to 92.9) 88.3 (84.1 to 91.5) 0.511 postoperative pain was more intense, requiring more opioids after the effect of LIA had worn off. The most likely explanation for this observation is pain resulting from the tourniquet itself. Use of parenteral opioids total (%) 24.0 (19.5 to 29.2) 23.3 (18.9 to 28.5) 0.848 Tramadol use (%) 3.0 (1.6 to 5.7) 3.3 (1.8 to 6.1) 0.816 Our findings are in agreement with the study from Ejaz’s et al.8 in which patients underwent TKA Data are displayed as percentage (95% CI) of patients using postoperative opioids. without LIA. In that study, patients in the tourniquet group had a higher opioid consumption as compared to the non-tourniquet group. Similar results were described by Abdel et al.7 who reported an extended interval between opioid injections in the first 24 hours after surgery in patients who underwent surgery without the use of a tourniquet.

Although a femoral nerve block is still considered the golden standard by many,15,16 LIA is increasingly used for postoperative analgesia after TKA. LIA was introduced in 2008 as a multimodal opioid-sparing technique for knee and hip surgery with the aim to achieve adequate analgesia without motor impairment and with reduced opioid-related side effects, 70 Influence of a tourniquet on opioid consumption Influence of a tourniquet on opioid consumption 71

resulting in rapid recovery and reduced length of hospital stay.2 Although in different studies References mixtures used for LIA vary in composition and in volume, two systematic reviews and meta- analyses have concluded that there are no significant differences in postoperative pain scores 1. Kehlet, H. & Andersen, L. Ø. Local infiltration analgesia in joint replacement: the evidence and at rest and opioid consumption between LIA and femoral nerve block.17,18 Thus, in terms of recommendations for clinical practice. Acta Anaesthesiologica Scandinavica 55, 778–784 (2011). postoperative pain relief, LIA appears to be an acceptable alternative for femoral nerve block. 2. Kerr, D. R. & Kohan, L. Local infiltration analgesia: a technique for the control of acute postoperative pain following knee and hip surgery: a case study of 325 patients. Acta orthopaedica 79, 174–83 (2008). Our study has several limitations. Although measuring postoperative pain scores is part of 3. Zhang, W. et al. The effects of a tourniquet used in total knee arthroplasty: a meta-analysis. Journal of our perioperative protocol and NRS scores are recorded regularly, the exact timing of taking orthopaedic surgery and research 9, 13 (2014). these scores is not standardized, and the number of pain scores recorded per patient may vary. 4. Tai, T.-W. et al. Tourniquet use in total knee arthroplasty: a meta-analysis. Knee Surgery, Sports Traumatology, The average pain scores per group are less solid when compared to a prospective analysis and Arthroscopy 19, 1121–30 (2011). should therefore be interpreted with caution. However, because our patients are instructed 5. Pain Management : Opioid Dose Converter. Available at: http://www.paindata.org/calculator.php. to ask for pain relief and our perioperative protocol ensures that patients are treated with (Accessed: 5th January 2018) opioids when the NRS pain score exceeds 3, we believe that the difference in total opioid 6. Information, N. C. for B. Piritramide - Pubchem. Open chemistry database. Available at: https://pubchem. consumption combined with similar average pain scores reliably reflects a difference in the ncbi.nlm.nih.gov/compound/Piritramide. (Accessed: 4th January 2018) 5 intensity of postoperative pain between the two groups and that patients in group T need 7. Abdel-Salam, A. & Eyres, K. S. Effects of tourniquet during total knee arthroplasty. A prospective randomised 5 more opioids to establish adequate pain relief. study. The Journal of bone and joint surgery. British volume 77, 250–3 (1995). 8. Ejaz, A. et al. Faster recovery without the use of a tourniquet in total knee arthroplasty. Acta Orthopaedica A second limitation of our study is that we did not evaluate the influence of a tourniquet on 85, 422–426 (2014). possible differences in short-term functional recovery between the two groups. However, 9. Ledin, H., Aspenberg, P. & Good, L. Tourniquet use in total knee replacement does not improve fixation, but since patients must meet a defined set of functional criteria before discharge and we found no appears to reduce final range of motion. Acta orthopaedica 83, 499–503 (2012). difference in hospital length of stay, we conclude that under the conditions of this retrospective 10. Vandenbussche, E., Duranthon, L.-D., Couturier, M., Pidhorz, L. & Augereau, B. The effect of tourniquet use in analysis, the use of a tourniquet does not interfere with short-term functional recovery. total knee arthroplasty. International orthopaedics 26, 306–9 (2002). 11. Tai, T.-W., Chang, C.-W., Lai, K.-A., Lin, C.-J. & Yang, C.-Y. Effects of Tourniquet Use on Blood Loss and Soft-Tissue Conclusions Damage in Total Knee Arthroplasty. The Journal of Bone and Joint Surgery-American Volume 94, 2209–2215 We found no difference in the TTFR after LIA for TKA with or without tourniquet, suggesting (2012). that the effect of an inflated tourniquet during local anesthetic infiltration does not alter 12. Kumar, N. et al. Evaluation of pain in bilateral total knee replacement with and without tourniquet; a systemic absorption sufficiently to affect the duration of analgesia. The use of a tourniquet prospective randomized control trial. Journal of clinical orthopaedics and trauma 6, 85–8 (2015). was associated with a higher total opioid consumption, which is most likely caused by pain 13. Mather, L. E. & Tucker, G. T. Pharmacokinetics and biotransformation of local anesthetics. International resulting from the tourniquet itself. anesthesiology clinics 16, 23–51 (1978). 14. Tucker, G. T. & Boas, R. A. Pharmacokinetic aspects of intravenous regional anesthesia. Anesthesiology 34, Financial support and sponsorship 538–49 (1971). This study was supported entirely by internal funds of the department of Anesthesiology, 15. Fischer, H. B. J. et al. A procedure-specific systematic review and consensus recommendations for Sint Maartenskliniek, Nijmegen, The Netherlands. postoperative analgesia following total knee arthroplasty. Anaesthesia 63, 1105–1123 (2008). 16. Paul, J. E. et al. Femoral nerve block improves analgesia outcomes after total knee arthroplasty: a meta- analysis of randomized controlled trials. Anesthesiology 113, 1144–62 (2010). 17. Mei, S. et al. Analgesia for total knee arthroplasty: a meta-analysis comparing local infiltration and femoral nerve block. Clinics 70, 648–53 (2015). 18. Albrecht, E., Guyen, O., Jacot-Guillarmod, A. & Kirkham, K. R. The analgesic efficacy of local infiltration analgesia vs femoral nerve block after total knee arthroplasty: a systematic review and meta-analysis. British journal of anaesthesia 116, 597–609 (2016). 73

Chapter 6

Improving Performance by Monitoring the Success Rate of Peripheral Nerve Blocks

S.M.K. Bakker R. Stienstra

Anesthesia & Analgesia 126(2), 644-647 (2018) 74 Individual performance monitoring 75

Abstract Introduction

In our hospital, we introduced a system to measure the collective and individual efficacy of Increasing attention is being payed to multimodal performance evaluation of medical brachial plexus and popliteal nerve blocks with the objective to create transparency as an professionals and benchmarking health care outcomes. Both objective criteria such as clinical instrument for monitoring and improvement. Initially, individual results were anonymous, outcomes and more subjective aspects such as evaluation of patient satisfaction may be but after one year anonymity was lifted within the team of anesthesiologists and results are part of a physician’s appraisal. To guarantee a certain level of education, continuous medical now discussed quarterly. education (CME) is mandatory for board-certified specialists.

Collective performance of interscalene, supraclavicular and popliteal blocks improved Attending refresher courses or congresses enhances knowledge and familiarity with recent significantly over time. Sharing and discussing collective and individual performance has developments. However, the effect size is considered to be small for physicians’ performances resulted in critical self-appraisal and increased willingness to learn from each other, and and patient outcome,1 and the true impact of CME is usually not examined. strengthened the team’s ambition for further improvement. Recertification programs suggest a certain level of knowledge, but as far as we know, none of the recertification programs require proof of technical skills. This is in sharp contrast with e.g. aviation, where all pilots – regardless of their experience- are trained and examined several times a year on knowledge, crew resource management and technical skills.

We believe that besides CME and training stressful situations, “continuous performance 6 evaluation” may contribute to maintain and improve both cognitive and technical skills. A critical condition for improvement is transparency. We decided to evaluate our efficacy regarding peripheral nerve blocks (PNBs) with the objective to establish a benchmark for our collective performance and transparency in individual performance, as well as a monitoring system for future changes in policy.

Methods

In accordance with Dutch law (Dutch Medical Research Involving Human Subjects Act[ WMO]), this survey meets the requirements for exemption from Research Ethics Committee review under the WMO. A statement to this effect was obtained from the Research Ethics Committee Slotervaart Hospital and Reade, Amsterdam, The Netherlands. All anesthesiologists involved were asked for permission to publish these data.

Assessing postoperative pain intensity in the recovery or post anesthesia care unit with the Numerical Rating Scale (NRS) is standard procedure in our practice; pain relief is considered adequate with NRS scores ≤ 3. Accordingly, we defined a PNB as successful when the first postoperative NRS score was ≤ 3 in the surgical area. All pain scores are obtained by qualified recovery room nurses, who are instructed to wait with the first measurement until the patient is awake and fully responsive.

For practical reasons, we limited this survey to patients receiving an upper extremity block (interscalene, supraclavicular, infraclavicular or axillary) or a popliteal block. To cover pain at the medial part of the lower leg and discomfort from a tourniquet, popliteal blocks in our practice are always combined with either a saphenous nerve block or a block of the femoral and lateral cutaneous femoral nerve. All blocks are placed with dual guidance (ultrasound and nerve stimulation) using ropivacaine 0.75% or mepivacaine 1.5%, without additives. All procedures are performed with the ultrasound probe in short axis view. An in-plane approach 76 Individual performance monitoring Individual performance monitoring 77

is used for all upper extremity blocks and for the vast majority of popliteal blocks. Depending Table 1. Overview of the PNB’s Performed, Average Success Rates per Six-Months’ Period on the kind of surgery and the patient’s and surgeon’s preference surgery is performed either Type of PNB jul-14 jan-15 jul-15 jan-16 jul-16 P-value awake, with additional sedation or general anesthesia. Sedation is provided with propofol jan-15 jul-15 jan-16 jul-16 jan-17 alone or in combination with remifentanil, general anesthesia is performed with propofol/ Interscalene (N) 359 383 323 400 348 remifentanil. Success Rate 88.0% 87.2% 92.9% 93.0% 94.3% 0.0001* Supraclavicular (N) 43 57 33 29 11 From July 1st, 2014 we started extracting data from the Patient Data Monitoring System (Chipsoft, Amsterdam, the Netherlands) into Excel spreadsheets to calculate individual Success Rate 81.4% 91.2% 93.9% 96.6% 100% 0.0127* and collective success rates for each type of block. After the first six months, the individual Infraclavicular (N) 34 28 50 47 37 performance was discussed in one-on-one dialogs between the individual anesthesiologist Success Rate 94.1% 96.4% 96.0% 97.9% 94.6% 0.78 and the department head, at this time individual data were not shared within the department’s Axillary (N) 89 74 61 85 84 staff. The collective performance was published in the hospital’s Quality & Safety department’s Success Rate 98.9% 100% 98.4% 100% 97.6% 0.50 three-monthly performance report. Popliteal (N) 415 360 383 344 455 Success Rate 93.3% 93.3% 93.5% 96.5% 96.9% 0.0027* One year after the introduction, lifting the anonymity of the individual performance scores The columns show six-month average success rates; data are numbers or percentages. P-value: 2 test for trend of success was discussed and unanimously agreed on. From that moment on, individual scores are rates. * = Statistically significant. χ shared internally between the anesthesiology staff and the results are discussed every three 6 months. Anonymity is maintained in external communications. 6

We did not distinguish between patients receiving a single shot or a continuous PNB. The collective data are presented as six-month averages and include all blocks, i.e. blocks placed by staff, residents and fellows. Analysis was conducted using Prism 7 software (GraphPad Software Inc., San Diego, CA, USA). Statistical analysis of the collective performance over time used the 2 test for trend, Pearson’s r was calculated to assess a possible correlation between the individual number of blocks and success rate; P<0.05 was considered significant. χ

Results

All anesthesiologists involved agreed to publication of these data and have given written consent.

The data from 10 patients were excluded from evaluation because of lacking NRS scores. The number of blocks and collective performance per six-month period are summarized in Table 1. Figure 1 shows a graphical representation of the upper extremity blocks. Figure 1. Collective performance of upper extremity blocks. The markers show the collective success rates (%). “A” marks the baseline results, i.e. the first time participants learned their personal and the collective average scores. During the time interval AB, participants were only aware of their own personal scores, as well as the collective average scores. During the interval BC, anonymity was lifted and participants became aware of each other’s scores as well. “C” marks the introduction of three-monthly evaluations of the results.

Trend analysis shows a significant increase of the success rate over time for interscalene (P = 0.0001), supraclavicular (P = 0.0127) and popliteal blocks (P = 0.0027), whereas the changes in infraclavicular and axillary blocks are not significant.

78 Individual performance monitoring Individual performance monitoring 79

We found no correlation between the individual number of blocks and performance: Pearson’s not be successful at all from the perspective of the patient. Placement of a PNB may be r = 0.08, 95 % CI = -0.39 – 0.5 (Figure 2). technically correct, but if surgery and consequently postoperative pain extends outside the blocked area, the block will be scored as unsuccessful. Furthermore, our monitoring system does not correlate the different surgical interventions with individual performance. Some surgical interventions may be associated with lower success rates of PNB, whereas others may be highly successful; unequal division of the different surgical interventions may skew individual results. These limitations notwithstanding, we believe that our evaluation of block performance is a valuable first step in self-assessment and skills monitoring.

Although evaluation of individual performance can improve collective outcomes,2 inter- colleague equation of objective parameters is not yet widely accepted, and still little is known about comparing individual performance and the effect on the quality of care. Introducing a monitoring system that reveals personal performance is a challenge when people are not used to it, because it can be perceived as external or punitive. A key element for successful implementation is mutual trust. To overcome initial anxiety and avoid skepticism, the matter was first discussed extensively within the team of anesthesiologists. Emphasis was placed on nonjudgmental registration, a safe environment, and anonymity of the individual scores. It 6 Figure 2. Absence of Correlation between Individual Number of Blocks and Success Rate between July 1st, was emphasized that only the collective performance data would be shared externally, and 6 2014 and January 1st, 2017. Pearson’s r = 0.08, 95 % CI = -0.39 – 0.5. that sharing personal scores with fellow anesthesiologists was an individual choice. Individual scores and position on the ranking would be shared only face-to-face between the individual and the head of the department with the guarantee that they solely would be used as personal Discussion evaluation tool and there would be no disciplinary consequences.

One goal of our monitoring system was to provide transparency in individual and Lifting the anonymity obviously poses a new challenge for the safe environment. It was agreed collective success rates of PNBs as a starting point for continuous performance evaluation beforehand that the decision to lift the anonymity had to be unanimous and that if one or and improvement. The results show that the collective performance of interscalene, more staff members would not agree, anonymity would be maintained. Each staff member supraclavicular and popliteal blocks improved significantly over time; the success rates of was approached individually by mail and asked for his opinion with emphasis on privacy and axillary and infraclavicular blocks did not change significantly. A possible explanation is that the right to oppose without explanation. The entire staff turned out to be in favor, and thus these latter blocks already had a high success rate to start with. anonymity was lifted.

We have been able to set a benchmark that can be used to monitor personal fluctuations. In A notable but welcome spin-off from our registration system was the development of an addition, the system can be used to monitor the effect of changes in e.g. dose and concentration open atmosphere in which it is possible to evaluate techniques and personal opinions and of local anesthetic, adjuvants, concomitant medication and/or equipment. exchange experiences without judgment, but with the purpose of improving performance. Not unexpectedly, awareness of personal performance works as an incentive for self- Our study has several limitations. Most importantly, we did not attempt to control for possible appraisal and, when appropriate, for improvement. While anesthesiologists usually perform confounders, and we have no control group with which to compare our data. Therefore, our their PNBs without fellow anesthesiologists watching, we have observed an increase in asking results are subject to confounder bias. While there has been no change in patient characteristics a second opinion from a colleague, especially in situations where the ultrasound picture is or surgical techniques, we cannot exclude the possibility of other factors influencing the ambiguous. We have seen that the introduction of the performance monitoring system has outcome and therefore the statistical results of our study should be interpreted with caution. resulted in a reduced threshold to ask each other’s opinion or to seek each other’s advice, and The number of blocks was not equally divided between the staff anesthesiologists, residents we have noticed individual adaptations in technique, in particular local anesthetic volume. and fellows. However, we found no correlation between the individual number of the different After the internal anonymity was lifted we started with three monthly meetings where both blocks performed, and the corresponding success rates (Figure 2); we therefore believe that collective and individual performance are analyzed and discussed with the purpose of further differences in the individual number of performed blocks is not a factor affecting the individual improvement in terms of increasing collective scores and reducing inter-individual variation. or collective results. We defined a block as being successful when the first postoperative NRS When discussing individual performance, it is paramount to keep reminding each other in the awake patient in the recovery room was ≤ 3. This is an arbitrary definition that does not that the focus should not be on individuals, but rather on the performance as a group take into account that “unsuccessful” blocks may still contribute significantly to postoperative and on identifying possible causes for unsuccessful blocks and developing strategies for pain relief. Also, an initially successful block may wear off more rapidly than expected and improvement. To emphasize this we have recently changed the format of the meeting, 80 Individual performance monitoring Individual performance monitoring 81

sending each participant a personal list of his own unsuccessful blocks during the previous References three months with a request to scrutinize these blocks for possible explanations, which are then discussed during the meeting. 1. Mansouri, M. & Lockyer, J. A meta-analysis of continuing medical education effectiveness. The Journal of continuing education in the health professions 27, 6–15 (2007). Our monitoring system has been in place for 2 ½ years, and it seems that after an initial period 2. Mohan, N. et al. Walking the tightrope: creation of the physician scorecard at the Rouge Valley Health of increase in success rates, the results are leveling off (the most recent rise in the success System. Healthcare Quarterly 8, 83–8 (2005). rate of supraclavicular blocks should be interpreted with caution, since it is based on only 11 procedures). We believe that given all the variables, a 100% success rate for all procedures is unattainable and that future efforts should focus on maintaining the present high success rates and minimizing inter-individual variation.

In conclusion, we have introduced a monitoring system for the collective and individual success rates of PNBs of the upper extremity and for popliteal blocks. Despite its limitations, it has provided us with an objective measure of our performance, resulted in an increase in the success rates of interscalene, supraclavicular, and popliteal blocks, and has opened the possibility of monitoring the effect of changes in the department’s policy regarding the 6 different aspects of PNB. 6 Financial support and sponsorship This study was supported entirely by internal funds of the department of Anesthesiology, Sint Maartenskliniek, Nijmegen, The Netherlands. 83

Chapter 7

General discussion 84 General discussion, clinical implications and future perspectives 85

General discussion

Total knee arthroplasty (TKA) is a frequently performed surgery to improve chronic refractory joint pain and improve mobility, in patients suffering from severe osteoarthrosis.1,2 Although TKA is performed to improve pain and function in the long term, the first days after surgery are notorious for the risk of severe postoperative pain. Severe postoperative pain is not only uncomfortable for the patient, it can delay rehabilitation and is assumed to be a prognostic factor for the development of chronic pain.3–5 Therefore adequate postoperative analgesia, preferably without impeding mobilization, is a cornerstone in rehabilitation after TKA.

Different strategies exist to alleviate postoperative pain after TKA. Postoperative regimens including continuous epidural infusion of local anesthetic provide excellent analgesia, but have become less attractive because of the concomitant motor weakness, hampering early mobilization. Opioid analgesia is associated with systematic side effects such as nausea and drowsiness.

Local infiltration analgesia (LIA) and femoral nerve blocks (FNB) are two other methods to reduce postoperative pain after TKA while diminishing opioid requirements. To gain more insight and knowledge on these two methods, we performed several studies that make up the core of this thesis. This concluding chapter summarizes and discusses the main findings in relation to the current literature, the clinical implications arising from these studies and 7 proposes future directions. Finally, the conclusions of this thesis are presented.

Discussion of main findings

The study described in Chapter 2 was designed to compare periarticular LIA of the knee with LIA of the posterior knee capsule in combination with a FNB catheter in terms of pain and functional outcome in patients undergoing primary TKA. In this randomized controlled trial, short and long-term outcome was assessed in terms of pain, activity levels and functional recovery.

In this study, Group FNB had lower pain scores and lower opioid consumption until the first postoperative day. This benefit of continuous FNB on pain and opioid consumption was not reported in two meta-analyses.6,7 However, in these analyses no distinction was made between single-shot and continuous femoral nerve block. Furthermore, in these meta-analyses the effect of popliteal fossa infiltration or sciatic nerve block as an addition to FNB was not included in the assessment. The sciatic nerve plays a substantial role in the innervation of the knee joint8 and blocking the sciatic nerve contributes to better analgesia after TKA.9 Popliteal fossa infiltration likely has a comparable effect. Omitting a sciatic nerve block or posterior fossa infiltration as variable in models in these meta-analysis, may have underestimated the positive effect of a continuous FNB compared to local infiltration of the anterior capsule of the knee.

Pain and opioid consumption in the early postoperative period may be important issues from the patient’s perspective, however optimal functional recovery in the long term is the main goal of TKA. Studies focussing on the influence of analgesic technique on functional recovery 86 General discussion, clinical implications and future perspectives General discussion, clinical implications and future perspectives 87

and long-term rehabilitation after TKA are scarce. investigating lidocaine plasma levels after intravenous administration demonstrate that prolonging the tourniquet time after infiltration slows down peak systemic absorption and Femoral nerve block is associated with the possibility of quadriceps motor weakness. Since limits peak concentration.21 Therefore the early peaks found in a number of patients in our walking time is associated with a better funtional outcome,10 the LIA technique11,12 is used in tourniquet study are surprising. In general, the hyperaemic period after tourniquet release order to provide postoperative pain relief without affecting motor function. Indeed, in our might contribute to this phenomenon. Immediately after tourniquet release the blood flow study group FNB showed a clinically small, but statistically significant lower activity level until through the ischemic limb is restored and can increase to almost 5 times above baseline.22 the first postoperative day. However, since there was no difference in the functional recovery An increased blood flow increases the concentration gradient governing passive diffusion one year after TKA, the clinical relevance of this diminished postoperative mobility in group of ropivacaine into the systemic circulation. Other possible explanations for very early FNB is questionable. The absence of differences in functional recovery one year after surgery is peaks (40 minutes) found in the study described in Chapter 3 are an accidental omission of in agreement with another study13 with a crossover design. Nevertheless, walking time during adrenaline in the LIA mixture, or difficulty to identify the veins and arteries when a tourniquet hospitalisation and functional recovery six weeks after surgery were slightly more favourable is used. An inflated tourniquet causes the collapse of arteries and veins and may hamper the in group FNB in the study by Carli.10 identification of the blood vessels when aspirating before infiltrating ropivacaine. Possibly, the risk of accidental intravascular ropivacaine injection is increased when a tourniquet Finally, a remarkable finding is the higher maximum pain score and analgesics consumption disables the ability to identify the blood vessels. in group LIA one year after total knee arthroplasty. Although the pathophysiology of persistent postoperative pain is not completely understood, significant lower pain scores until a month In Chapter 5 we investigated the influence of perioperative tourniquet use on the time to first after surgery are also reported by patients treated with continuous regional anesthetic request (TTFR) of opioids and total opioid consumption when local infiltration analgesia is techniques in a prospective observational registry.14 In summary, the fear that FNB hampers used to reduce postoperative pain. We found no statistically significant difference in median functional recovery after TKA is not supported by clinical evidence; FNB may have a beneficial TTFR between the two groups, however, there was a difference in total opioid consumption. effect in preventing chronic postoperative pain. Patients in the Tourniquet group consumed significantly more opioids as compared to 7 patients who underwent total knee arthroplasty without perioperative tourniquet use. The 7 Chapter 3 is the first study describing the pharmacokinetic profile of ropivacaine after observation that patients in the tourniquet group used more opioids is most likely caused high-dose, high-volume, single-shot LIA of the knee with epinephrine, without a draining by the pain resulting from the tourniquet itself, be it due to compression, ischemia, or both. system such as a periarticular catheter or surgical drain. A phenomenon often seen in both This is in accordance with other studies investigating tourniquet influence on pain or opioid pharmacodynamic and pharmacokinetic studies is a high inter-individual variability.15 consumption after total knee arthroplasty.19,23–26 Likewise, we also found a large variation in the plasma concentrations, however even the

maximum total (Cmax 2.26 µg/ml) and unbound (Cu max (0.13 µg/ml) ropivacaine concentration Combining the findings described in Chapter 4 and Chapter 5 we conclude that the use of a in our study remained well below the assumed toxic thresholds of 4.3 and 0.56 µg/ml, tourniquet does not delay systemic absorption of ropivacaine to the extent that it affects the respectively.16 duration or intensity of analgesia after local infiltration analgesia. On the contrary, looking at the data on total opioid consumption it seems that postoperative pain is reduced when no During TKA a tourniquet is often used to create a dry surgical field and reduce intraoperative tourniquet is used. blood loss. However, the use of a tourniquet is associated with an increased risk of thromboembolic events, wound infections and prolonged thigh pain.17–19 These findings have In Chapter 6 we discussed a monitoring system to evaluate the collective and individual raised important questions about the necessity of perioperative tourniquet use. Therefore the performance with regard to the efficacy of peripheral nerve blocks. In healthcare increasing standard practice in several centers has recently been changed to performing TKA without attention is being payed to multimodal performance evaluation of medical professionals and the use of a tourniquet. Chapter 4 describes the pharmacokinetic profile of a high-dose, benchmarking of health care outcomes. However, in contrast to another high-risk branch, single-shot LIA when no tourniquet was used at the moment of infiltration. In analogy with e.g. aviation, there are hardly any objective measurements to guarantee a physician’s skills.

our study in which a tourniquet was used (Chapter 3), the maximum total (Cu max 2.34 µg/mL) The tool we introduced in this study helped us to evaluate and improve our performance

and unbound (Cu max 0.093 µg/mL) remained below the toxic threshold. Comparing these two regarding peripheral nerve blocks. We found an improvement of maximum 6.3 % in success

pharmacokinetic studies, we did not observe a clinically relevant difference in Cmax or Tmax rate of peripheral nerve blocks, without any other intervention except discussing the different between ropivacaine infiltration with or without tourniquet. techniques in a group of anesthesiologists.

In our study with tourniquet a number of patients showed an early Cmax (Tmax ≤ 240 minutes There is not much known about comparing individual performance and the effect on the 27 in 11/20 patients), whereas in the study without tourniquet none of the patients had a Tmax < quality of care. A paper by St. Jacques discusses a monitoring system to evaluate and 240 minutes. This is in contrast with the theory underlying the pharmacokinetic principles stimulate the punctuality of individual anesthesiologists and its effect on maximizing the of locally infiltrated medication. Vascularity is a main determinant of systemic absorption productivity of the group of anesthesiologists as a whole. To stimulate the change in behaviour speed.20 When using a tourniquet an artificial decrease in vascularity is created. Studies the anesthesiologists were publicly ranked and a financial reward was offered, depending 88 General discussion, clinical implications and future perspectives General discussion, clinical implications and future perspectives 89

on the physician’s performance in relation to the others. At the end of the study period it Potential limitations and future perspectives turned out that the combination of public profiling and a financial incentive can change a physician’s behaviour and improve his productivity. However, some serious objections may be Despite careful designing of the studies described in this thesis, several limitations and raised against this method to improve productivity. Firstly, only productivity was measured directions for future research can be designated. without any attention to actual skills or quality of care. Healthcare institutions have to be economical, but this should never subrogate their main goal: providing high quality health The study described in Chapter 2 has some small but unavoidable limitations. Firstly, patients care. Furthermore, it is not unlikely that public ranking and financial rewarding for the high participating in this study were not blinded for ethical reasons. However, we believe this performers will enhance rivalry between colleagues. limitation has not affected our results, since the physical therapists assessing the ability to mobilize and the research assistants collecting the pain scores were blinded. Furthermore, Another study28 described a system focused on cost-effectiveness, evolving from a personal performance on physical tests such as the 6MWT, SCT and TUG is not only determined by scoring system to a well accepted benchmark system, similar to our monitoring system. knee function, but also by factors such as age, sex, body mass index; also, possible health In accordance with our experience, this report suggests that with a performance and issues might adversely affect the performance on these tasks. We assume that by using benchmarking system it is possible to change specialist behaviour without rewarding. randomisation these confounding factors were equally divided between the two groups, and therefore did not exert a significant bias. Since individual performance monitoring is still unfamiliar in the healthcare sector, mutual trust is of great importance before introducing individual monitoring systems. A notable finding of this study is the slightly but significantly higher maximum pain scores in group LIA. The number of patients reporting high mean pain scores (Numeric rating scale Our performance monitoring focused on upper extremity and popliteal nerve blocks and not [NRS] >3) one year after surgery was similar in both groups (10%). However, remarkable is the on the regional anesthetic techniques for TKA such as LIA or femoral nerve blocks. Although high rate of patients in group LIA (42%) reporting high maximum pain scores (NRS >3) one year our monitoring system was thus not aimed at improving postoperative pain scores for TKA after surgery compared to FNB (26%) and to the literature (30%).29 Additionally, the odds of 7 patients, we expect the improved success rate demonstrated for upper extremity and popliteal using analgesics for knee pain a year after total knee arthroplasty was almost 6 times greater 7 nerve blocks to apply equally for other block procedures. in the LIA group. This relation may be spurious and clinical relevance of this difference can be argued; still, patients in group FNB also had lower pain scores and less opioid consumption in the immediate postoperative period. Since early postoperative pain is a possible risk factor for Clinical implications the development of chronic pain,3,30 the possibility of a causal relation is intriguing. Further studies, which should include a larger number of patients, will be necessary to elucidate this. In our pharmacokinetic studies we demonstrated that after the administration of high dose locally infiltrated ropivacaine, total and unbound ropivacaine plasma concentrations remain Chronic postoperative pain or persistent postsurgical pain (PPSP) is defined as pain that lasts far below the assumed toxic threshold of 4.3 and 0.56 µg/mL. The results of our studies suggest beyond the healing of injured tissue and the related inflammatory processes.31 PPSP is hard that in routine surgery, total knee arthroplasty can be safely performed without the use of to treat, therefore is has serious consequences for both society and individual patients, since a tourniquet, reducing postoperative opioid consumption without increasing the risk of chronic pain affects all dimensions of quality of life.32 Younger age, female gender, obesity, systemic toxicity. psychological distress, preoperative and acute postoperative pain have so far been identified as predictors of chronic postsurgical pain after total knee arthroplasty.3 However, effective A femoral nerve catheter provides better analgesia in the early postoperative period than preoperative and individual identification of patients prone to develop chronic postoperative LIA of the anterior aspect of the knee, with both postoperative pain scores and opioid pain is, yet, not possible. Future research should focus on factors predicting chronic post- consumption being lower. Also considering the lower maximum pain scores and lower surgical pain and strategies to reduce the impact of these factors when possible. analgesic consumption in group FNB one year after surgery, these results suggest that there might be an association between the type of regional anesthesia and the risk to develop Recently, it has been hypothesized that central sensitisation may be predictive for the chronic pain. However, it should be kept in mind that the success rate of a peripheral nerve development of persistent postoperative pain. Quantitative sensory testing (QST) might be an block also depends on the individual skills of the performer. instrument to assess pain processing and the degree of central sensitisation prior to surgery. QST contains a broad spectrum of test batteries to assess sensitisation. However, preoperative Finally, we consider individual performance monitoring to be a valuable first step in self- QST does not show consistent results.33 Although several studies34–36 have identified possible assessment and skills monitoring, which can contribute to the quality of care. promising predictors for chronic pain, more exploration in this discipline is necessary to determine the possible role of QST in clinical practice. Because of the huge impact of chronic pain on both the individual patient as well as society as a whole, the possibility to identify patients at risk preoperatively and strategies to reduce or prevent the development of chronic pain would be a great achievement. 90 General discussion, clinical implications and future perspectives General discussion, clinical implications and future perspectives 91

Femoral nerve blocks may be associated with weakness of the quadriceps muscle, which may References hamper early mobilization. Although there is no evidence that impaired motor function, as a result of femoral nerve blocks, adversely affects recovery after TKA, there might be a 1. Ethgen, O., Bruyère, O., Richy, F., Dardennes, C. & Reginster, J.-Y. Health-related quality of life in total hip and future for motor function sparing techniques. Recent studies have shown that a single shot total knee arthroplasty. A qualitative and systematic review of the literature. The Journal of bone and joint adductor canal block provides similar analgesia but limited quadriceps weakness compared surgery. American volume 86-A, 963–74 (2004). to femoral nerve block, allowing for easier mobilization.37,38 However, this benefit is less 2. Harris, W. H. & Sledge, C. B. Total hip and total knee replacement. The New England journal of medicine 323, pronounced or absent when comparing continuous femoral with continuous adductor canal 725–31 (1990). blocks.39–41 Future research should determine which role is provided for adductor canal blocks 3. Puolakka, P. A. E. et al. Persistent pain following knee arthroplasty. European journal of anaesthesiology 27, in enhanced recovery protocols after TKA. 455–60 (2010). 4. Barrington, J. W., Halaszynski, T. M. & Sinatra, R. S. Perioperative pain management in hip and knee The study discussed in Chapter 3 has one main limitation; the scarcity of samples between 6 replacement surgery. American journal of orthopedics 43, S1–S16 (2014).

and 24 hours after ropivacaine infiltration, made it impossible to determine the maxC and Tmax 5. Feizerfan, A. & Sheh, G. Transition from acute to chronic pain. Continuing Education in Anaesthesia, Critical in some of the patients. To overcome this hiatus, we extended the sampling schedule until 12 Care & Pain 15, 98–102 (2015). hours after ropivacaine infiltration in the subsequent study Chapter( 4). 6. Albrecht, E., Guyen, O., Jacot-Guillarmod, A. & Kirkham, K. R. The analgesic efficacy of local infiltration analgesia vs femoral nerve block after total knee arthroplasty: a systematic review and meta-analysis. Future research should also focus on prolonging the analgesic duration of locally infiltrated British journal of anaesthesia 116, 597–609 (2016). ropivacaine. There may be an role for multivesicular liposomes containing a local anesthetic, 7. Zhang, L., Ma, J., Kuang, M. & Ma, X. Comparison of Periarticular Local Infiltration Analgesia With Femoral however proof of these multivesicular liposomes extending the duration of a single-shot local Nerve Block for Total Knee Arthroplasty: a Meta-Analysis of Randomized Controlled Trials. The Journal of anesthetic is not yet irrefutable.42 Arthroplasty 33, 1972–1978 (2018). 8. Horner, G. & Dellon, A. L. Innervation of the human knee joint and implications for surgery. Clinical 7 Lastly, when extracting data or monitoring healthcare quality via the hospital information orthopaedics and related research 221–6 (1994). 7 system, a certain amount of bias will slip into a dataset. However, this bias represents the daily 9. Wegener, J. T. et al. Value of single-injection or continuous sciatic nerve block in addition to a continuous clinical practice and still provides us with valuable information. With current progressions in femoral nerve block in patients undergoing total knee arthroplasty: a prospective, randomized, controlled data warehousing and efforts and attention to develop reliable monitoring systems, the daily trial. Regional anesthesia and pain medicine 36, 481–8 (2011). practice probably can become a valuable source to monitor and improve quality of care. 10. Carli, F. et al. Analgesia and functional outcome after total knee arthroplasty: periarticular infiltration vs continuous femoral nerve block. British journal of anaesthesia 105, 185–95 (2010). 11. Kerr, D. R. & Kohan, L. Local infiltration analgesia: a technique for the control of acute postoperative pain following knee and hip surgery: a case study of 325 patients. Acta orthopaedica 79, 174–83 (2008). 12. Kehlet, H. & Andersen, L. Ø. Local infiltration analgesia in joint replacement: the evidence and recommendations for clinical practice. Acta Anaesthesiologica Scandinavica 55, 778–784 (2011). 13. Ng, F.-Y., Ng, J. K.-F., Chiu, K.-Y., Yan, C.-H. & Chan, C.-W. Multimodal periarticular injection vs continuous femoral nerve block after total knee arthroplasty: a prospective, crossover, randomized clinical trial. The Journal of arthroplasty 27, 1234–8 (2012). 14. Bugada, D. et al. Effects of anaesthesia and analgesia on long-term outcome after total knee replacement: A prospective, observational, multicentre study. European journal of anaesthesiology 34, 665–672 (2017). 15. Turner, R. M., Park, B. K. & Pirmohamed, M. Parsing interindividual drug variability: an emerging role for systems pharmacology. Wiley interdisciplinary reviews. Systems biology and medicine 7, 221–41 (2015). 16. Knudsen, K., Beckman Suurküla, M., Blomberg, S., Sjövall, J. & Edvardsson, N. Central nervous and cardiovascular effects of i.v. infusions of ropivacaine, bupivacaine and placebo in volunteers. British journal of anaesthesia 78, 507–14 (1997). 17. Zhang, W. et al. The effects of a tourniquet used in total knee arthroplasty: a meta-analysis. Journal of orthopaedic surgery and research 9, 13 (2014). 18. Tai, T.-W. et al. Tourniquet use in total knee arthroplasty: a meta-analysis. Knee Surgery, Sports Traumatology, Arthroscopy 19, 1121–30 (2011). 19. Kumar, N. et al. Evaluation of pain in bilateral total knee replacement with and without tourniquet; a prospective randomized control trial. Journal of clinical orthopaedics and trauma 6, 85–8 (2015). 92 General discussion, clinical implications and future perspectives General discussion, clinical implications and future perspectives 93

20. Mather, L. E. & Tucker, G. T. Pharmacokinetics and biotransformation of local anesthetics. International 41. Jæger, P. et al. Adductor Canal Block Versus Femoral Nerve Block for Analgesia After Total Knee Arthroplasty. anesthesiology clinics 16, 23–51 (1978). Regional Anesthesia and Pain Medicine 38, 526–532 (2013). 21. Tucker, G. T. & Boas, R. A. Pharmacokinetic aspects of intravenous regional anesthesia. Anesthesiology 34, 42. Zhao, B., Ma, X., Zhang, J., Ma, J. & Cao, Q. The efficacy of local liposomal bupivacaine infiltration on pain and 538–49 (1971). recovery after Total Joint Arthroplasty: A systematic review and meta-analysis of randomized controlled 22. Dedichen, H. & Myhre, H. O. Reactive hyperaemia of the human lower limb. A comparison between the effect trials. Medicine 98, e14092 (2019). of tourniquet occlusion, selective arterial occlusion and intra-arterial papaverine injection. Acta chirurgica Scandinavica 141, 517–21 (1975). 23. Ejaz, A. et al. Faster recovery without the use of a tourniquet in total knee arthroplasty. Acta Orthopaedica 85, 422–426 (2014). 24. Ledin, H., Aspenberg, P. & Good, L. Tourniquet use in total knee replacement does not improve fixation, but appears to reduce final range of motion. Acta orthopaedica 83, 499–503 (2012). 25. Vandenbussche, E., Duranthon, L.-D., Couturier, M., Pidhorz, L. & Augereau, B. The effect of tourniquet use in total knee arthroplasty. International orthopaedics 26, 306–9 (2002). 26. Tai, T.-W., Chang, C.-W., Lai, K.-A., Lin, C.-J. & Yang, C.-Y. Effects of Tourniquet Use on Blood Loss and Soft-Tissue Damage in Total Knee Arthroplasty. The Journal of Bone and Joint Surgery-American Volume 94, 2209–2215 (2012). 27. St Jacques, P. J., Patel, N. & Higgins, M. S. Improving anesthesiologist performance through profiling and incentives. Journal of clinical anesthesia 16, 523–8 (2004). 28. Mohan, N. et al. Walking the tightrope: creation of the physician scorecard at the Rouge Valley Health System. Healthcare Quarterly 8, 83–8 (2005). 7 29. Beswick, A. D., Wylde, V. & Gooberman-Hill, R. Interventions for the prediction and management of chronic 7 postsurgical pain after total knee replacement: systematic review of randomised controlled trials. BMJ open 5, e007387 (2015). 30. Macrae, W. A. Chronic post-surgical pain: 10 years on. British Journal of Anaesthesia 101, 77–86 (2008). 31. Chapman, C. R. & Vierck, C. J. The Transition of Acute Postoperative Pain to Chronic Pain: An Integrative Overview of Research on Mechanisms. The Journal of Pain 18, 359.e1-359.e38 (2017). 32. Wylde, V. et al. Chronic pain after total knee arthroplasty. EFORT open reviews 3, 461–470 (2018). 33. Sangesland, A., Støren, C. & Vaegter, H. B. Are preoperative experimental pain assessments correlated with clinical pain outcomes after surgery? A systematic review. Scandinavian Journal of Pain 15, 44–52 (2017). 34. Petersen, K. K., Graven-Nielsen, T., Simonsen, O., Laursen, M. B. & Arendt-Nielsen, L. Preoperative pain mechanisms assessed by cuff algometry are associated with chronic postoperative pain relief after total knee replacement. PAIN 157, 1400–1406 (2016). 35. Wylde, V. et al. Preoperative widespread pain sensitization and chronic pain after hip and knee replacement: a cohort analysis. Pain 156, 47–54 (2015). 36. Lunn, T. H., Gaarn-Larsen, L. & Kehlet, H. Prediction of postoperative pain by preoperative pain response to heat stimulation in total knee arthroplasty. Pain 154, 1878–85 (2013). 37. Grevstad, U. et al. Effect of adductor canal block versus femoral nerve block on quadriceps strength, mobilization, and pain after total knee arthroplasty: a randomized, blinded study. Regional anesthesia and pain medicine 40, 3–10 (2015). 38. Kim, D. H. et al. Adductor Canal Block versus Femoral Nerve Block for Total Knee Arthroplasty. Anesthesiology 120, 540–550 (2014). 39. Machi, A. T. et al. Discharge Readiness after Tricompartment Knee Arthroplasty: Adductor Canal versus Femoral Continuous Nerve Blocks-A Dual-center, Randomized Trial. Anesthesiology 123, 444–56 (2015). 40. Sztain, J. F. et al. Continuous Adductor Canal Versus Continuous Femoral Nerve Blocks: Relative Effects on Discharge Readiness Following Unicompartment Knee Arthroplasty. Regional anesthesia and pain medicine 40, 559–67 (2015). 95

Chapter 8

Conclusions 96 Conclusions 97

Conclusions

The main conclusions that can be drawn from the work described in this thesis are:

• The choice of anesthetic technique, femoral nerve block or local infiltration analgesia, does not influence functional recovery after total knee arthroplasty. Chapter( 2)

• Patients treated with a femoral nerve block have lower postoperative pain scores and opioid consumption compared to patients treated with local infiltration analgesia. Chapter( 2)

• Patients treated with local infiltration have higher maximum pain scores at 3 and 12 months after total knee arthroplasty and are almost six times more likely to use analgesic medication 1 year after surgery. (Chapter 2)

• Total and free ropivacaine concentrations after a high-dose, single-shot ropivacaine with epinephrine remain below the assumed toxic threshold for local anesthetic systemic toxicity. (Chapters 3 and 4)

• Perioperative tourniquet use does not substantially decrease systemic absorption of locally infiltrated ropivacaine. Chapters( 3 and 4)

• Perioperative tourniquet use does increase postoperative opioid consumption. (Chapter 5)

• Individual and collective performance monitoring have the ability to contribute to 8 improving physicians’ practical skills and quality of health care. (Chapter 6) 99

Summary 100 Summary 101

Summary

In patients suffering from severe osteoarthrosis, total knee arthroplasty is a frequently performed surgery to improve chronic refractory joint pain and improve mobility. Total knee arthroplasty is well known for its chance on severe postoperative pain. Chapter 1 provides a general introduction to the different kinds of anesthesia for total knee arthroplasty, including regional anesthesia techniques. Multimodal analgesia as well as several aspects of the local anesthetic ropivacaine are discussed and an outline of this thesis is described. In the subsequent chapters the results of our studies investigating these methods and factors are further presented and discussed.

Methods to provide postoperative analgesia after total knee arthroplasty Two currently popular methods to provide postoperative analgesia after total knee arthroplasty are a continuous femoral nerve block and local infiltration analgesia. Nerve blocks usually offer sufficient analgesia in most of the patients, however motor function is also affected. This might impede (independent) mobilisation as long as motor function is diminished. Because local infiltration analgesia provides analgesia without affecting motor function, virtually immediate postoperative mobilization is possible and it is hypothesized that this might enhance rehabilitation after total knee arthroplasty. In Chapter 2 the influence of these techniques on early and late functional recovery and pain relief are extensively discussed. Apart from functional outcome and medication use one year after surgery, the quality of short-term postoperative pain relief is discussed. It was found that patients receiving a femoral nerve block had slightly lower pain scores and reduced analgesic consumption, both in the immediate postoperative period as well as 12 months after surgery. Clinically relevant differences in knee function were not found between the groups.

Pharmacokinetics of high dose locally infiltrated ropivacaine Local infiltration analgesia is an easy and straightforward technique to provide postoperative pain relief after total knee arthroplasty. Since this technique has only been introduced a little more than a decade ago its safety track is still relatively short. To gain more insight in the pharmacokinetics of the high dose of ropivacaine used for local infiltration analgesia,Chapters 3 and 4 describe the pharmacokinetic profile of ropivacaine when used for local infiltration analgesia in total knee arthroplasty with (Chapter 3) or without (Chapter 4) tourniquet. Similar to most pharmacokinetic studies, we found a high inter-individual variation in the

plasma concentrations in both studies, but even the highest measured ropivacaine Cmax (2.34

µg/ml) and Cu max (0.13 µg/ml) remained well below the assumed toxic thresholds.

Factors affecting postoperative pain and opioid consumption Since postoperative pain may have an influence on the development of chronic pain, identifying non pharmacological factors to decrease postoperative pain and opioids consumption are another field worth exploring.

Chapter 5 describes the influence of perioperative tourniquet use on the time to first request (TTFR) of opioids and total opioid consumption when local infiltration analgesia is used to reduce postoperative pain. We found no statistically significant difference in median TTFR between the Tourniquet and non-Tourniquet group. However, opioid consumption was higher in the Tourniquet group 37.5 (17-67) mg as compared to patients who underwent total knee 102 Summary 103

arthroplasty without perioperative tourniquet use, 27 (10-60) mg. Possible explanations for the effect of a tourniquet on local anesthetic uptake and on postoperative pain are discussed.

Chapter 6 describes the influence of monitoring immediate postoperative pain on the success Nederlandse samenvatting rate of peripheral nerve blocks. Improvement of individual and collective performance is demonstrated and discussed, as well as the strategy and risks of introducing a monitoring system in anesthesia practice.

In Chapter 7 the main finding of this thesis of this discussed in a broader prespective, addressing the clinical implications, potential limitations and future perspectives.

Finally, Chapter 8 sums up the conclusions that can be drawn from this thesis. 104 Samenvatting 105

Samenvatting

Patiënten die lijden aan ernstige slijtage van het kniegewricht (gonartrose) kunnen gebaat zijn bij een totale knieprothese (totale knie arthroplastiek). Een knieprothese wordt geplaatst met het doel de mobiliteit te verbeteren en om op den duur pijnklachten te verminderen. Echter, de eerste dagen na totale knie vervangende chirurgie zijn geassocieerd met ernstige postoperatieve pijn.

Hoofdstuk 1 omvat een algemene introductie op dit proefschrift waarin een achtergrond en aanleiding tot de onderzoeken worden geschetst. Eveneens worden verschillende vormen van anesthesie (verdoving) en analgesie (pijnbestrijding), welke routinematig voor totale knievervangende chirurgie toegepast worden, uiteengezet. In de hierop volgende hoofdstukken worden de resultaten van de uitgevoerde onderzoeken gepresenteerd en besproken.

Methoden van postoperatieve pijnbestrijding na totale knie prothese chirurgie Lokale infiltratie analgesie en het nervus femoralis blok zijn momenteel populaire postoperatieve pijnbestrijdingstechnieken. Zenuwblokkades bieden meestal voldoende pijnstilling, maar gaan gepaard met motorische blokkade. De spierzwakte die hiervan het gevolg kan zijn kan (onafhankelijke) mobilisatie belemmeren zolang de motorische functie is verminderd. Lokale infiltratie analgesie, waarbij een lokaal anestheticum direct in het weefsel rondom de knie wordt geïnfiltreerd, verschaft pijnstilling zonder de motorisch functie te beïnvloeden, hierdoor is vrijwel onmiddellijk postoperatieve mobilisatie mogelijk. Er wordt verondersteld dat directe postoperatieve mobilisatie de revalidatie na een totale knie prothese kan verbeteren. In Hoofdstuk 2 wordt ingegaan op de invloed van deze twee technieken op vroeg en laat functioneel herstel, pijnverlichting en medicijngebruik. Patiënten die een blokkade van de nervus femoralis kregen hadden iets lagere pijnscores en een verminderde inname van pijnstillende medicatie, zowel in de onmiddellijke postoperatieve periode als 12 maanden na de operatie. Er werden geen klinisch relevante verschillen in kniefunctie gevonden tussen de groepen.

Farmacokinetiek van hoog gedoseerde lokaal geïnfiltreerde ropivacaïne Lokale infiltratie analgesie is een eenvoudige, pragmatische en doelgerichte methode om in pijnstilling na een totale knie prothese te voorzien.

Aangezien deze techniek iets meer dan een decennium geleden is geïntroduceerd, zijn de gegevens over veilige toepassing nog relatief schaars. Om meer inzicht te krijgen in de farmacokinetiek van de hoge dosis ropivacaïne die wordt gebruikt voor lokale infiltratie analgesie, wordt in de hoofdstukken 3 en 4 het farmacokinetisch profiel van ropivacaïne, toegepast als lokale infiltratie analgesie bij totale knieartroplastiek, met Hoofdstuk( 3) of zonder (Hoofdstuk 4) tourniquet op het moment van infiltratie, beschreven. Beide studies toonden, in overeenstemming met de meeste andere farmacokinetische studies, een hoge interindividuele variatie in de plasmaconcentraties, maar zelfs de hoogst gemeten ropivacaïne

Cmax (2,34 μg / ml) en Cu max (0,13 μg / ml) bleven ruim onder de veronderstelde toxische drempels.

Niet medicamenteuze factoren die van invloed zijn op postoperatieve pijn en opiaatconsumptie Omdat postoperatieve pijn een rol kan spelen bij de chronificatie van pijn, is de identificatie 106 Samenvatting 107

van niet-farmacologische factoren die postoperatieve pijn en opioïd consumptie kunnen verminderen de moeite waard.

Hoofdstuk 5 beschrijft de invloed van een perioperatieve tourniquet op de TTFR (Time To Dankwoord First Request, het tijdstip waarop voor het eerst om pijnstilling wordt gevraagd) en de totale opioïd consumptie wanneer lokale infiltratie analgesie wordt toegepast als methode van postoperatieve pijnbestrijding bij patiënten die een totale knie arthroplastiek ondergingen. Er was geen statistisch significant verschil in de mediane TTFR tussen de groep met- en de groep zonder tourniquet. De totale opioïd consumptie was in de Tourniquet groep echter hoger dan in de groep waarbij geen tourniquet werd gebruikt (37.5 (17-67) mg respectievelijk 27 (10-60) mg). Mogelijke verklaringen voor de invloed van een tourniquet op de absorptie van lokaal anestheticum en op de mate van postoperatieve pijn worden besproken.

Hoofdstuk 6 beschrijft de invloed van routinematig monitoren van postoperatieve pijn op het succespercentage van perifere zenuw blokkades. Een individuele en collectieve verbetering van succespercentages wordt getoond en besproken.

Eveneens wordt verder ingegaan op de strategie en risico’s die gepaard gaan met het introduceren van een individueel monitoringsysteem.

In Hoofstuk 7 worden de belangrijkste bevindingen van dit proefschrift in breder perspectief getrokken. Verder worden de klinische implicaties, beperkingen en mogelijke toekomstperspectieven besproken.

Tenslotte presenteert Hoofstuk 8 de conclusies van dit proefschrift. 108 Dankwoord 109

Dankwoord

Zoals promovendi die mij zijn voorgegaan weet ik inmiddels ook: “Promoveren doe je niet alleen (binnen kantoortijden)”. Daarom wil ik graag een ieder bedanken voor zijn bijdrage aan het tot stand komen van dit proefschrift, alsmede ook hen die een bijzondere invulling hebben gegeven aan mijn schaarse vrije tijd afgelopen jaren. Een bijzonder woord van dank gaat uit naar:

Prof. dr. Scheffer (promotor), beste Gert Jan, hartelijk dank voor jouw steun, enthousiasme, vertrouwen en de grenzeloze vrijheid bij het vormgeven van mijn promotietraject.

Dr. R. Stienstra (co-promotor), beste Ruud, je wordt gekenmerkt door een aanstekelijke gedrevenheid en enorme passie voor ons vak. Het wekelijkse vrijdagmiddagoverleg in de Maartenskliniek begon altijd als zakelijke voortgangsbespreking, maar eindigde steevast in satirische analyse van mijn adrenaline opwekkende hobby’s of de (on)mogelijkheid tot het consumeren van een vleesloze maaltijd. Hoewel jij je reeds in de nadagen van jouw carrière als anesthesioloog bevond toen ik aan mijn promotietraject begon, ben je jouw tijd ver vooruit met allerlei gadgets en begrijp je dat jonge onderzoekers goede koffie nodig hebben. Dit manuscript is een afsluiting van onze samenwerking, maar geen einde van ons weerzien. Ik ben dankbaar dat ik jou als wetenschapper, arts en mens heb mogen leren kennen.

Dr. P.J.C. Heesterbeek (co-promotor), beste Petra, zonder jouw fiat was dit proefschrift er nooit gekomen. Met Maaike en jou had ik mijn eerste sollicitatie gesprek voor dit promotietraject, jullie waren de eerste horde op weg naar deze thesis. Na deze horde die ik alleen moest nemen, volgden er voor ons samen nog velen. De METC, reviewers en natuurlijk niet te vergeten de stormbaan in Deventer. Bedankt voor jouw hulp en kritische noten.

Leden van de manuscriptcommissie, prof. dr. B.W. Schreurs, prof. dr. A. Dahan en prof. dr. C. Kramers hartelijk dank voor de beoordeling van dit manuscript.

Veel dank gaat uit naar al mijn medeauteurs, jullie specifieke kennis tilde de artikelen naar een hoger niveau of maakte het mogelijk om de data van honderden patiënten binnen no time te extraheren en filteren. Sakib, zonder jouw hulp was ons artikel er überhaupt niet geweest.

Het Orthoresearch team dat zich over de anesthesiologie promovendi ontfermt, Miranda, Berbke, Katrijn en Menno. Nienke, jij hebt me echt door de laatste loodjes heen geholpen. De andere Orthoresearch dames van één hoog achter: Jolanda, Saskia en Malou. Mijn vaste adres voor een theekransje en dé vakantieoppas van mijn planten. Alle overige research collega’s wil ik bedanken voor de gezelligheid en lunchwandelingen. Miranda, Noël, alle inval- hardlooptrainers en mede-hardlopers, jullie waren mijn stok achter de deur om woensdag om half 6 tijd voor mezelf in te ruimen. Clara en Richarda, managementassistentes zoals jullie zijn onmisbaar voor de afdeling en dus ook voor mijn proefschrift. Bedankt voor jullie strakke organisatie en onuitputtelijke spraakzaamheid.

Collega promovendi Karin en Maaike. Karin, jij was al bijna klaar toen ik begon, toch ik heb genoten van jouw gezellige aanwezigheid op de momenten dat je naar de Maartenskliniek kwam om jouw proefschrift af te maken. 110 Dankwoord Dankwoord 111

Lieve Maaike, ook naar jou gaat veel dank uit. Je vertrouwde mij jouw lopende onderzoeken Bijzondere zweefvliegers: Max, destijds bracht jij me de beginselen van het zweefvliegen en toe en bracht mij de fijne kneepjes van het congresleven bij. Met veel plezier kijk ik terug op later ook van het skiën bij. Bedankt voor jouw interesse in mijn bezigheden. Lieve Joop en Rikkie, onze samenwerking en avonturen. all-inclusive mentoren tijdens mijn instructieopleiding, maar ook daarna. Jan Deelen, ook tot jou wil ik mij kort richten. Met meer dan 60 jaar leeftijdsverschil ben jij mijn oudste vriend. Dit proefschrift was er nooit geweest zonder alle studiedeelnemers of zonder de Jouw interesse in de mensen om je heen, jouw oprechte, maar onterechte bescheidenheid en anesthesiologen van de Sint Maartenskliniek. Andrea, Natasja, Thecla, Ivar, Roel, Jorrit, Edgar ruimdenkende kijk op het hedendaagse leven, maken diepe indruk op mij. en Thomas, jullie wil ik in het bijzonder bedanken aangezien jullie in beide hiervoor genoemde categorieën vallen. Ivar, jij volgde Ruud op als medisch manager, bedankt voor jouw flexibiliteit Elisabeth de Ruiter en Dennis Boon, bedankt voor jullie lovende woorden in de aanbevelingsbrief in deze periode. die jullie, vanuit jullie rol als huisarts en als ouders van Daan, Suus en Kiki, geschreven hebben toen ik deelnam aan de decentrale selectie. Het was het begin van de weg naar dit proefschrift. Verder wil ik graag alle doktersassistentes, anesthesiemedewerkers, pijn- en recoveryverpleegkundigen bedanken voor de prettige samenwerking, bezorgdheid over mijn Lieve Oma, zonder jouw trots zou promoveren slechts half zo leuk zijn. werktijden, helpende handen en gezelligheid op ieder denkbaar tijdstip van de dag. Pap en Mam, een stevige thuisbasis is belangrijk voor geluk in het leven, ik ben dan ook dankbaar De ondersteuning bij het afronden van dit proefschrift blijft niet beperkt tot medewerkers dat jullie support niet gestopt is toen ik het huis uit ging. Lieve Pap, jij timmert in Nijmegen nu van de Sint Maartenskliniek. Alle huidige collega’s in het Radboud en Bernhoven wil ik graag letterlijk aan mijn nieuwe thuis en lieve Mam, jouw onuitputtelijke zorgzaamheid maakt dat bedanken voor jullie interesse en aanmoedigingen. Een bijzondere dank gaat uit naar de alles, altijd en voor iedereen, tot in de puntjes geregeld is. Dit maakt ieders huis een thuis. opleiders uit het Radboudumc en Bernhoven ziekenhuis. Christiaan Keijzer-Broeders, Laura Blok en Hester Hoogenboom, jullie hebben voor mij de voorwaarden geschept om mijn Geert, de rol van broer vervul jij met verve. Wat laat je mij lachen met jouw gekke uitspraken. promotie vlot af te kunnen maken naast de opleiding, jullie oplossingsgezindheid en flexibele Al heb jij zelf een andere benaming voor jouw rol, ik ben erg blij dat jij samen met Julia als insteek waardeer ik enorm. paranimf naast mij wil staan.

En tenslotte, Derek. Traditiegetrouw ben jij de laatste die genoemd wordt. Jij hebt vaak gezegd dat je geen bijdrage hebt geleverd aan mijn proefschrift en daarom niet in het dankwoord Naast de mensen die in de letterlijke zin aan mijn proefschrift hebben bijgedragen, wil ik graag genoemd hoeft te worden. Toch noem ik jou hier vol overtuiging, want niets is minder nog een aantal bijzondere mensen bedanken die de afgelopen jaren minstens zo belangrijk waar. Je hebt engelengeduld met mij gehad en met één blik leg je mij haarfijn uit welke zijn geweest als zij die hiervoor genoemd zijn. wetenschappelijke tradities en medisch abracadabra helemaal niet zo vanzelfsprekend zijn. Door onze asynchrone werktijden hebben we vaak maar weinig tijd samen, dat maakte de Mijn fantastische vrienden en familie. Wat heb ik veel bijzondere mensen om mij heen. Zonder afgelopen jaren behoorlijk complex. Met het afronden van mijn proefschrift komt er hopelijk iemand te kort te doen wil ik mij tot aan aantal van jullie in het bijzonder richten: iets meer lucht in ons leven. Laten we verder vooral blijven genieten van het leven zoals we al doen. Josta, ik vind je fantastisch als mens en als dokter.

Michelle, jij bent er al zo lang dat ik me onze eerste ontmoeting niet meer kan herinneren. Jij maakte als 3-jarige net zo lang stampij bij jouw ouders, totdat we naar dezelfde basisschool zouden gaan. De basis van onze vriendschap. Behalve dankbaar voor onze vriendschap, ben ik ook ontzettend trots op jouw doorzettingsvermogen.

Kyra, samen wonen, vliegen, sporten en wildkamperen. Samen kunnen wij alles aan.

Julia, door de jaren heen hebben we veel beleefd en raken nooit uitgepraat. We hebben vaak dezelfde smaak of ideeën, maar deinzen niet terug voor elkaars kritisch inzichten. Vandaar dan ook dat de rol van paranimf jou op het lijf geschreven is. 113

Curriculum vitae 114 Curriculum vitae 115

Curriculum vitae

Sietske Marieke Klazien was born on the 21st of Februari, 1990 in Apeldoorn.

After graduating from Gymnasium Apeldoorn in 2008, she started with the Bachelor of Medicine at VUmc School of Medical Sciences (Amsterdam). Between 2010 and 2014 she worked as a research assistent in the Deparmtent of Hematology, VUmc (Head: Prof. Dr. P.C. Huijgens, later Prof. Dr. S. Zweegman).

She reveiced her medical degree in 2014 after which she started working at the Sint Maartenskliniek (Nijmegen) as a research fellow under supervesion of Dr. R. Stientra.

The research on local and regional analgesia for total knee arthroplasty became the foundation of this thesis. Sietske is BROK® certified since September 2015 (BROK® ref. 15851). In 2018 she started her residency in anesthesiology at the Radboud University Medical Center (dr. Keijzer- Broeders). 117

Bibliography 118 Bibliography 119

Bibliography

Okuyama N, Sperr WR, Kadar K, Bakker S.M.K., Szombath G, Handa H, Tamura H, Kondo A, Valent P, Várkonyi J, van de Loosdrecht A, Ogata K. Prognosis of acute myeloid leukemia transformed from myelodysplastic syndromes: a multicenter retrospective study Leukemia research 2013 Aug;37(8):862–7.

Fenten M.G.E., Bakker S.M.K., Touw D.J., van den Bemt B.J., Scheffer G.J., Heesterbeek P.J., Stienstra R. Pharmacokinetics of 400 mg ropivacaine after periarticular local infiltration analgesia for total knee arthroplasty Acta Anaesthesiologica Scandinavica. 2017 Mar;61(3):338-345.

Bakker S.M.K., Stienstra R Improving Performance by Monitoring the Success Rate of Peripheral Nerve Blocks Anesthesia and Analgesia. 2018 Feb;126(2):644-647. This article is awarded with the “Mathieu Gielen Award”.

Bakker S.M.K., Fenten M.G.E, Touw D.J., van den Bemt B.J.F., Heesterbeek P.J.C., Scheffer G.J., Stienstra R. Pharmacokinetics of 400 mg Locally Infiltrated Ropivacaine After Total Knee Arthroplasty Without Perioperative Tourniquet Use Regional Anesthesia and Pain Medicine. 2018 Oct;43(7):699-704

Fenten M.G.E, Bakker S.M.K., Scheffer G.J., Wymenga A.B., Stienstra R., Heesterbeek P.J.C. Femoral nerve catheter vs local infiltration for analgesia in fast track total knee arthroplasty: short-term and long-term outcomes Britisch Journal of Anaesthesia. 2018 Oct;121(4):850-858. This article is awarded with the ”Prof. dr. Ritsema van Eck-award”

Bakker S.M.K., Kosse N.M., Crnic S., Scheffer G.J., Stienstra R. Influence of a Tourniquet on Opioid Consumption After Local Infiltration Analgesia for Total Knee Arthroplasty Turkisch Journal of Anaesthesiology and Reanimimation 2019 Apr;47(2):107-11. 121

Research Data Management 122 Research Data Management 123

Research Data Management

This thesis is based on the results of human studies, which were conducted in accordance with the principles of the Declaration of Helsinki. The medical and ethical review board Committee on Research Involving Human Subjects Region Arnhem Nijmegen, Nijmegen, the Netherlands or Medical Research Ethics Committee Slotervaart Hospital and Reade, Amsterdam, the Netherlands has given approval to conduct of these studies.

Individual projects are stored at Sint Maarktenskliniek research server (V:\) under research_ archief (Chapters 2, 3, 4, and 5) or anesthesiology server H:\OK\ANAEST\Performance (Chapter 6). Hardcopies of the projects are stored at the Sint Maartenskliniek archive.

Privacy of the participants in Chapter 2, 3 and 4 is warranted by use of encrypted and unique individual subject codes. Informed consents and decryption keys of these chapters are stored and locked away separately form the anonymized questionaires and CRFs.

Data of Chapters 5 and 6 is anonymously extracted from the patient data management system. On first arrival in the Sint Maartenskliniek all patients are informed about the use of anonymous patient data and the ‘opt-out’ arrangement to object to the use of anonymized data. None of the patients in these cohorts objected to the use of anonymous data.

Data management was performed in secured and locked Excel files. Before analysis a second investigator monitored all data. Data was converted from excel to STATA, Prism or R to perform statistical analysis.

The data will be saved for 15 years after termination of the individual studies. Using patient data in future research is only possible after a renewed permission by the patient as recorded in the informed consent. Datasets are available from the corresponding author on reasonable request. 125

Theses Sint Maartenskliniek 126 Theses Sint Maartenskliniek 127

Theses Sint Maartenskliniek

Fenten, M. (2019). Optimizing locoregional anesthesia for fast track orthopedic surgery. Radboud University Nijmegen, Nijmegen. The Netherlands.

Claassen, A. (2019). Strategies for patient education in rheumatic diseases. Radboud University Nijmegen, Nijmegen. The Netherlands.

Minten, M. (2019). On the role of inflammation and the vallue of low dose radiation therapy in osteoarthritis. Radboud University Nijmegen, Nijmegen. The Netherlands.

Bekker, C. (2018). Sustainable use of medication. Medication waiste and feasibility of redispensing. Utrecht University, The Netherlands.

Bikker, I.(2018). Organizing timely treatment in multi-disciplinary care. University of Twente, The Netherlands.

Bouman, C. (2018). Dose optimisation of biologic DMARDs in rheumatoid arthritis - long-term effects and possible predictors. Radboud University Nijmegen, The Netherlands.

Mahler, E. (2018). Contributors to the management of osteoarthritis. Utrecht University, The Netherlands.

Tweehuysen, L. (2018). Optimising biological treatment in inflammatory rheumatic diseases. Predicting, tapering and transitioning. Radboud University Nijmegen, Nijmegen, The Netherlands.

Geerdink, Y. (2017). Getting a grip on hand use in unilateral cerebral palsy. Radboud University, Nijmegen, The Netherlands.

Remijn, L. (2017). Mastication in children with cerebral palsy. Radboud University, Nijmegen, The Netherlands.

Selten, E. (2017). Beliefs underlying treatment choices in osteoarthritis. Radboud University, Nijmegen, The Netherlands.

Van Hooff, M. (2017). Towards a paradigm shift in chronic low back pain? Identification of patient profiles to guide treatment. VU University Amsterdam, Amsterdam, The Netherlands.

Lesuis, N. (2016). Quality of care in rheumatology. Translating evidence into practice. Radboud University, Nijmegen, The Netherlands.

Luites, J. (2016). Innovations in femoral tunnel positioning for anatomical ACL reconstruction. Radboud University, Nijmegen, The Netherlands.

Pakvis, D. (2016). Survival, primary stability adn bone remodeling assessment of cementless sockets. An appraisal of Wolff's law in the acetabulum. Radboud University, Nijmegen, The Netherlands. 128 Theses Sint Maartenskliniek Theses Sint Maartenskliniek 129

Schoenmakers, K. (2016). Prolongation of regional anesthesia. Determinants of peripheral nerve Van Kessel, M. (2014). Nothing left? How to keep on the right track. Spatial and non-spatial block duration. Radboud University, Nijmegen, The Netherlands. attention processes in neglect after stroke. Radboud University, Nijmegen, The Netherlands.

Altmann, V. (2015). Impact of trunk impairment on activity limitation with a focus on Brinkman, M. (2013). Fixation stability and new surgical concepts of osteotomies around the wheelchair rugby. Leuven University, Leuven, Belgium. knee. Utrecht University, Utrecht, The Netherlands.

Bevers, K. (2015). Pathophysiologic and prognostic value of ultrasonography in knee Kwakkenbos, L. (2013). Psychological well-being in systemic sclerosis: Moving forward in osteoartrhitis. Utrecht University, Utrecht, The Netherlands. assessment and treatment. Radboud University, Nijmegen, The Netherlands.

Cuperus, N. (2015). Strategies to improve non-pharmacological care in generalized Severens, M. (2013). Towards clinical BCI applications: assistive technology and gait osteoarthritis. Radboud University, Nijmegen, The Netherlands. rehabilitation. Radboud University, Nijmegen, The Netherlands.

Kilkens, A. (2015). De ontwikkeling en evaluatie van het Communicatie Assessment & Interventie Stukstette, M. (2013). Understanding and treating hand osteoarthritis: a challenge. Systeem (CAIS) voor het aanleren van (proto-)imperatief gedrag aan kinderen met complexe Utrecht University, Utrecht, The Netherlands. ontwikkelingsproblemen. Radboud University, Nijmegen, The Netherlands. Van der Maas, A. (2013). Dose reduction of TNF blockers in Rheumatoid Arthritis: clinical and Penning, L. (2015). The effectiveness of injections in cuffdisorders and improvement of pharmacological aspects. Radboud University, Nijmegen, The Netherlands. diagnostics. Maastricht University, Maastricht, The Netherlands. Zedlitz, A. (2013). Brittle brain power. Post-stroke fatigue, explorations into assessment and Stegeman, M. (2015). Fusion of the tarsal joints: outcome, diagnostics and management of treatment. Radboud University, Nijmegen, The Netherlands. patient expectations. Utrecht University, Utrecht, The Netherlands. Beijer, L. (2012). E-learning based speech therapy (EST). Exploring the potentials of E-health for Van Herwaarden, N. (2015). Individualised biological treatment in rheumatoid arthritis. dysarthric speakers. Radboud University, Nijmegen, The Netherlands. Utrecht University, Utrecht, The Netherlands. Hoogeboom, T. (2012). Tailoring conservative care in osteoarthritis. Maastricht University, Wiegant, K. (2015). Uitstel kunstknie door kniedistractie. Utrecht University, Utrecht, Maastricht, The Netherlands. The Netherlands. Boelen, D. (2011). Order out of chaos? Assessment and treatment of executive disorders in brain- Willems, L. (2015). Non-pharmacological care for patients with systemic sclerosis. Radboud injured patients. Radboud University, Nijmegen, The Netherlands. University, Nijmegen, The Netherlands. Heesterbeek, P. (2011). Mind the gaps! Clinical and technical aspects of PCL-retaining total knee Witteveen, A. (2015). The conservative treatment of ankle osteoarthritis. University of replacement with the balanced gap technique. Radboud University, Nijmegen, Amsterdam, Amsterdam, The Netherlands. The Netherlands.

Zwikker, H. (2015). All about beliefs. Exploring and intervening on beliefs about medication to Hegeman, J. (2011). Fall risk and medication. New methods for the assessment of risk factors in improve adherence in patients with rheumatoid arthritis. Radboud University, Nijmegen, commonly used medicines. Radboud University, Nijmegen, The Netherlands. The Netherlands. Smulders, E. (2011). Falls in rheumatic diseases. Risk factors and preventive strategies in Koenraadt, K. (2014). Shedding light on cortical control of movement. Radboud University, osteoporosis and rheumatoid arthritis. Radboud University, Nijmegen, The Netherlands. Nijmegen, The Netherlands. Snijders, G. (2011). Improving conservative treatment of knee and hip osteoarthritis. Smink, A. (2014). Beating Osteoarthritis. Implementation of a stepped care strategy to manage Radboud University, Nijmegen, The Netherlands. hip or knee osteoarthritis in clinical practice. VU University Amsterdam, Amsterdam, The Netherlands. Vriezekolk, J. (2011). Targeting distress in rheumatic diseases. Utrecht University, Utrecht, The Netherlands. Stolwijk, N. (2014). Feet 4 feet. Plantar pressure and kinematics of the healthy and painful foot. Radboud University, Nijmegen, The Netherlands. 130 Theses Sint Maartenskliniek Theses Sint Maartenskliniek 131

Willems, P. (2011). Decision making in surgical treatment of chronic low back pain. Jongerius, P. (2004). Botulinum toxin type-A to treat drooling. A study in children with cerebral The performance of prognostic tests to select patients for lumbar spinal fusion. Maastricht palsy. Radboud University, Nijmegen, The Netherlands. University, Maastricht, The Netherlands. Van de Crommert, H. (2004). Sensory control of gait and its relation to locomotion after a spinal Aarts, P. (2010). Modified constraint-induced movement therapy for children with unilateral cord injury. Radboud University, Nijmegen, The Netherlands. spastic cerebral palsy: the Pirate group intervention. Radboud University, Nijmegen, The Netherlands. Van der Linde, H. (2004). Prosthetic prescription in lower limb amputation. Development of a clinical guideline in the Netherlands. Groningen University, Groningen, The Netherlands. Groen, B. (2010). Martial arts techniques to reduce fall severity. Radboud University, Nijmegen, The Netherlands. Hendricks, H. (2003). Motor evoked potentials in predicting motor and functional outcome after stroke. University of Nijmegen, Nijmegen, The Netherlands. Van Koulil, S. (2010). Tailored cognitive behavioral therapy in fibromyalgia. Radboud University, Nijmegen, The Netherlands. Hosman, A. J. F. (2003). Idiopathic thoracic spinal deformities and compensatory mechanisms. University of Nijmegen, Nijmegen, The Netherlands. Van den Bemt, B. (2009). Optimizing pharmacotherapy in patients with rheumatoid arthritis: an individualized appraoch. Radboud University, Nijmegen, The Netherlands. Donker, S. (2002). Flexibility of human walking: a study on interlimb coordination. Groningen University, Groningen, The Netherlands. Van Nes, I. (2009). Balance recovery after supratentorial stroke. Influence of hemineglect and the effects of somatosensory stimulation. Radboud University, Nijmegen, The Netherlands. Hochstenbach, J. (1999). The cognitive, emotional, and behavioural consequenses of stroke. University of Nijmegen, The Netherlands. Ruiter, M. (2008). Speaking in ellipses. The effect of a compensatory style of speech on functional communication in chronic agrammatism. Radboud University, Nijmegen, The Netherlands. De Kleuver, M. (1998). Triple osteotomy of the pelvis.An anatomical, biomechanical and clinical study. University of Nijmegen, Nijmegen, The Netherlands. Baken, B. (2007). Reflexion on reflexes. Modulation during gait. Radboud University, Nijmegen, The Netherlands. Van Balen, H. (1997). A disability-oriented approach to long-term sequelae following traumatic brain injury. Neuropsychological assessment for post-acute rehabilitation. University of Gaasbeek, R. (2007). High tibial osteotomy. Treatment of varus osteoarthritis of the knee. Nijmegen, Nijmegen, The Netherlands. Radboud University, Nijmegen, The Netherlands. Tromp, E. (1995). Neglect in action: a neuropsychological exploration of some behavioural Koëter, S. (2007). Patellar instability. Diagnosis and treatment. Radboud University, Nijmegen, aspects of neglect. University of Nijmegen, Nijmegen, The Netherlands. The Netherlands. Van Lankveld, W. (1993). Coping with chronic stressors of rheumatoid arthritis. University of Langeloo, D. (2007). Monitoring the spinal cord during corrective spinal surgery: a clinical study Nijmegen, Nijmegen, The Netherlands. of TES-MEP. Radboud University, Nijmegen, The Netherlands. Geurts, A. (1992). Central adaptation of postural organization to peripheral sensorimotor De Haart, M. (2005). Recovery of standing balance in patients with a supratentorial stroke. impairments. University of Nijmegen, Nijmegen, The Netherlands. Radboud University, Nijmegen, The Netherlands. De Rooij, D. (1988). Clinical and serological studies in the connective tissue diseases. University Den Otter, R. (2005). The control of gait after stroke: an electromyographic approach to of Nijmegen, Nijmegen, The Netherlands. functional recovery. Groningen University, Groningen, The Netherlands.

Spruit, M. (2005). Surgical treatment of degenerative disc conditions of the lumbar spine. Biomechanical, clinical and radiological aspects. University Utrecht, Utrecht, The Netherlands.

Weerdesteyn, V. (2005). From the mechanisms of obstacle avoidance towards the prevention of falls. Radboud University, Nijmegen, The Netherlands. 234 5 6