Volume 29 - No 1 - January 2021

Netherlands Journal of Critical Care

Bi-monthly journal of the Dutch Society of Intensive Care

REVIEW REVIEW ORIGINAL ARTICLE Value of electroencephalography for prognosis Late development of Takotsubo syndrome Validation of the Nursing Activities Score (NAS) and treatment of comatose patients after following intensive care unit discharge using time-and-motion measurements in circulatory arrest V. Visvanathan, A. Kucia, B. Reddi, J. Horowitz Dutch intensive care units J. Hofmeijer, M.J.A.M. van Putten C.C. Margadant, M.E. Hoogendoorn, R.J. Bosman, J.J. Spijkstra, S. Brinkman, N.F. de Keizer

Netherlands Journal of Critical Care

NETHERLANDS JOURNAL OF CRITICAL CARE

EXECUTIVE EDITORIAL BOARD D.W. Donker, editor in chief CONTENTS I. van Stijn, managing editor H. Dupuis, language editor EDITORIAL D. van Dijk, associate editor M.M.J. van Eijk, associate editor 4 EEG in patients admitted after cardiac arrest: daily clinical practice N. Kusadasi, associate editor J. Horn C.L. Meuwese, associate editor [email protected] REVIEW 6 Value of electroencephalography for prognosis and treatment of comatose patients COPYRIGHT after circulatory arrest Netherlands Journal of Critical Care ISSN: 1569-3511 J. Hofmeijer, M.J.A.M. van Putten NVIC p/a Domus Medica P.O. Box 2124, 3500 GC Utrecht T: +31-(0)30 - 760 74 44 REVIEW 14 Late development of Takotsubo syndrome following intensive care unit discharge © 2021 NVIC. All rights reserved. Except as V. Visvanathan, A. Kucia, B. Reddi, J. Horowitz outlined below, no part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, ORIGINAL ARTICLE electronic, mechanical, photocopying, recording 22 Validation of the Nursing Activities Score (NAS) using time-and-motion measurements or otherwise, without prior written permission of the publisher. Permission may be sought directly in Dutch intensive care units from NVIC. C.C. Margadant, M.E. Hoogendoorn, R.J. Bosman, J.J. Spijkstra, S. Brinkman, N.F. de Keizer

DERIVATIVE WORKS Subscribers may reproduce tables of contents CASE REPORT or prepare lists of articles including abstracts 28 Jaundice due to autoimmune haemolytic anaemia in a patient with Gram-negative for internal circulation within their institutions. septic shock Permission of the publisher is required for resale or distribution outside the institution. Permission M.A. Lantinga, F.N. Polderman, D.E. Issa, C.P.C. de Jager of the publisher is also required for all other derivative works, including compilations and CASE REPORT translations. 31 Leptospira-induced acute disseminated encephalomyelitis ELECTRONIC STORAGE Y. Özbay, P. Brussee, J.G. van der Hoeven Permission of the publisher is required to store or use electronically any material contained in this journal, including any article or part of an article. CASE REPORT 36 Treacher Collins syndrome: a case report SUBSCRIPTIONS R. Ibrahim, D. Hejazi Albasha, H. Daood An annual subscription to the Netherlands Journal of Critical Care consists of 6 issues. Issues within Europe are sent by standard mail and CASE REPORT outside Europe by air delivery. Cancellations 41 A case of cerebral and coronary air embolism after flushing of a pleural drain should be made, in writing, at least two months before the end of the year. J.W. Smithuis, M.C.A. Müller, F.F. Smithuis, R.P. Weenink The annual subscription fee for the Netherlands is 5 175, for Europe 5 295, for the rest of the world 5 390. Subscriptions are accepted on a prepaid ABSTRACTS & CASE REPORTS INTENSIVISTENDAGEN 2021 basis only and are entered on a calendar year basis. 44 Please transfer the fee to ABNAMRO NL55 ABNA 0524 5618 93, swift-code: ABNANL2A. Do not forget to mention the complete address for 68 NVIC conference and course agenda delivery of the Journal. 69 Editorial board 69 International advisory board CLAIMS Claims for missing issues should be made within 70 Information for authors two months of the date of dispatch. Missing issues will be mailed without charge. Issues claimed beyond the two-month limit must be prepaid at back copy rates.

ADVERTISING/ BUSINESS CONTACTS For advertising, please contact Eventex.

Eventex Sint Bernulphusstraat 13D 6861GS Oosterbeek T: +31 (0)26 – 333 75 77 E: [email protected]

Netherlands Journal of Critical Care is indexed in:

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 3 Netherlands Journal of Critical Care Submitted November 2020; Accepted November 2020

EDITORIAL

EEG in patients admitted after cardiac arrest: daily clinical practice

J. Horn Department of Intensive Care, Amsterdam UMC, University of Amsterdam, Amsterdam, the Netherlands

Correspondence [email protected]

Keywords - electroencephalography; cardiac arrest; coma; prognosis

The role of electroencephalography (EEG) in patients admitted EEG recording will be started. However, the neurophysiology to your intensive care unit (ICU) after cardiac arrest has department is busy, but they manage to come to the ICU at changed substantially in the last few years. In the old days, an 16.00 hrs. Applying all the electrodes takes about 30 minutes, so EEG was advised in patients who did not wake up after clearance the registration starts about 6.5 hours after cardiac arrest. If you of sedative drugs following target temperature management. manage to do this, you have done very well: this is very quick! Then, first a median nerve somatosensory evoked potential Even in centres where continuous EEG registration is well (SSEP) test was done and if no cortical N20 response was found, organised, it takes a mean of 11 hours to start EEG recording.[4] an EEG could help further.[1,2] The review on EEG registrations At 12 hours after cardiac arrest, 50% of the patients included in after cardiac arrest by Jeannette Hofmeijer and Michel van the study were hooked onto the EEG registration. Putten, in this issue of the journal, shows that the EEG is much Only 40% of all cardiac arrests occur during office hours, so an more valuable in the first 12-24 hours after cardiac arrest.[3] early start to EEG registration in all cardiac arrest patients would EEG registration early after cardiac arrest is easier said require a neurophysiologist to be available 24/7 on the ICU.[5] In than done in daily clinical practice. Two factors are crucial: the Netherlands, this is usually not the case. And is this really optimal recording and reliable assessment. For both parts, the necessary? Would you come to decisions on continuation or Department of Clinical Neurophysiology is necessary as a true withdrawal of care within the first 24 hours after cardiac arrest? ‘partner in crime’. Qualitatively good EEG recording on the In the current Dutch guideline on prognostication of postanoxic ICU is technically more demanding than in the lab. The many coma, you are advised to start the prognostication process electrical devices on the ICU and involuntary muscle activity when a patient does not wake up after clearance of sedative of the patient can lead to significant noise in the registration. drugs.[6] The exact moment when you expect that the sedative Assessment of an EEG may not be real rocket science, but when medication administered is no longer having an effect depends decisions about continuation of supportive care are based on on the type of drug and dosage used, target temperature used the results, this has to be a careful process performed by an (lower body temperatures lead to slowed metabolism) and experienced neurologist or neurophysiologist who is aware of organ failure delaying clearance. The team of treating physicians the pitfalls of ICU EEG registrations. So, how can you organise at the bedside should decide on this moment, but this is usually this in daily clinical care? later than 24 hours after cardiac arrest. When a patient is then Imagine, it is Tuesday morning 10.00 hrs and you are called by still in a coma, the EEG registration already recorded can be the cardiologist because they are expecting a patient who had a checked for the pattern seen at 24 hours. Use of continuous cardiac arrest in the local supermarket. The presenting rhythm EEG recording makes it easy to look back, for example, 3AM. was ventricular fibrillation and cardiac ischaemia is seen on the Apart from this possibility of looking back, is a continuous EEG ECG. So, after stabilisation in the emergency room, the patient recording really needed? Not necessarily. Studies from Lausanne is taken to the cath lab. An occlusion of the LAD is successfully have shown that a standard 20-30 minute EEG registration treated and around noon the patient arrives on your ICU. As it performed at 24 hours gives you similar information.[7] If you took 10 minutes to achieve return of spontaneous circulation, do this registration later than 24 hours, then the specificity for you decide to treat this patient with target temperature poor outcome prediction remains very high but the sensitivity management and start sedative medication to enable diminishes[4] This is caused by the recovery of the EEG to a less temperature control. Ideally, that same afternoon, continuous malign pattern in a majority of patients with a poor outcome

4 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care EEG in patients admitted after cardiac arrest

in the end.[8] Still, for EEG registration on the ICU early after assisted analyses for assessment opens the way to continuous cardiac arrest, the cooperation of the neurophysiology team EEG recording on the ICU for many more neuro-ICU patients. is necessary. In the Amsterdam UMC, we have arranged that the technicians from the neurophysiology department start continuous EEG recording during office hours (08.00-17.00), References seven days a week. This makes it possible to start registration 1. Wijdicks EF, Hijdra A, Young GB, Bassetti CL, Wiebe S. Practice parameter: within 24 hours in almost all patients. Registration is continued prediction of outcome in comatose survivors after cardiopulmonary resuscitation (an evidence-based review): report of the Quality Standards Subcommittee of until the patient wakes up or until a maximum of 72 hours after the American Academy of Neurology. Neurology. 2006;67:203-10. cardiac arrest. The neurophysiology technician checks the EEG 2. Sandroni C, Cariou A, Cavallaro F, et al. Prognostication in comatose survivors of cardiac arrest: an advisory statement from the European Resuscitation Council and electrodes daily and applies them again if necessary. Such a European Society of Intensive Care Medicine. Resuscitation. 2014;85:1779-89. 3. Hofmeijer J, van Putten MJAM. Value of electroencephalography for prognosis continuous EEG recording for a prolonged period can help to and treatment of comatose patients after circulatory arrest. Neth J Crit Care. identify epileptic activity developing over time. 2021;29:6-12. 4. Ruijter BJ, Tjepkema-Cloostermans MC, Tromp SC, et al. Early electroencephalography Several developments that will facilitate EEG recording on the for outcome prediction of postanoxic coma: A prospective cohort study. Ann ICU are ongoing. Easy-to-use EEG electrodes, straightforward Neurol. 2019;86:203-14. 5. Bagai A, McNally BF, Al-Khatib SM, et al. Temporal differences in out-of-hospital enough to allow application by the ICU nurse, are currently cardiac arrest incidence and survival. Circulation. 2013;128:2595-602. being developed.[9] They consist of a limited number of 6. Federatie Medisch Specialisten. Richtlijn: Prognose van postanoxisch coma. Available from https://richtlijnendatabase.nl/richtlijn/prognose_van_ electrodes on the forehead and will give less information than postanoxisch_coma/startpagina.html the classical full-head EEG registration. However, brain injury 7. Fatuzzo D, Beuchat I, Alvarez V, Novy J, Oddo M, Rossetti AO. Does continuous EEG influence prognosis in patients after cardiac arrest? Resuscitation. 2018;132:29-32. after cardiac arrest is a global type of injury affecting the whole 8. Oh SH, Park KN, Shon YM, et al. Continuous Amplitude-Integrated brain so a more limited recording might suffice in most cardiac Electroencephalographic Monitoring Is a Useful Prognostic Tool for Hypothermia- Treated Cardiac Arrest Patients. Circulation. 2015;132:1094-103. [10] arrest patients. 9. Muraja-Murro A, Mervaala E, Westeren-Punnonen S, et al. Forehead EEG electrode set versus full-head scalp EEG in 100 patients with altered mental state. Epilepsy Interpretation of the EEG recording can be supported by using Behav. 2015;49:245-9. basic training programs for ICU team members.[11] Furthermore, 10. Rittenberger JC, Weissman A, Baldwin M, et al. Preliminary experience with point- of-care EEG in post-cardiac arrest patients. Resuscitation. 2019;135:98-102. translation of the EEG recording using computer assisted 11. Citerio G, Patruno A, Beretta S, et al. Implementation of continuous qEEG in two analyses into scores interpretable by everyone will be the neurointensive care units by intensivists: a feasibility study. Intensive Care Med. 2017;43:1067-8. next step. Combining easy recording methods and computer-

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 5 Netherlands Journal of Critical Care Submitted May 2020; Accepted September 2020

REVIEW

Value of electroencephalography for prognosis and treatment of comatose patients after circulatory arrest

J. Hofmeijer1,2, M.J.A.M. van Putten1,3 1Clinical Neurophysiology, Technical Medical Centre, University of Twente, Enschede, the Netherlands 2Department of Neurology, Rijnstate Hospital, Arnhem, the Netherlands 3Department of Clinical Neurophysiology, Medisch Spectrum Twente, Enschede, the Netherlands

Correspondence J. Hofmeijer - [email protected]

Keywords - EEG, postanoxic encephalopathy, postanoxic coma, outcome prediction, electrographic status epilepticus

Abstract synaptic activity is very sensitive to the effects of ischaemia, All comatose patients after circulatory arrest initially have a the EEG is sensitive to detection of ischaemia-induced cerebral severely abnormal disturbed electroencephalogram. The speed malfunctioning.[3] However, the specificity of pathological EEG of normalisation is a robust contributor to prediction of outcome. activity for reliable prediction of poor or good outcome has long Differences between patients with poor and good outcome are been uncertain.[4] Over the past decade, various specific EEG largest <24 hours after the arrest. Lasting suppression at ≥12 patterns have been associated with poor or good outcome. It has hours or synchronised patterns with >50% suppression at ≥24 become clear that the EEG can contribute to reliable outcome hours are invariably associated with poor outcome. This includes prediction if EEG patterns are classified in relation to the burst suppression with identical bursts and generalised periodic time since circulatory arrest. Here, we review the evidence of discharges on a suppressed background. Recovery towards reliability of EEG-based outcome prediction, discuss treatment continuous patterns within 12 hours is strongly associated with of epileptiform patterns and provide future perspectives. a good outcome. Predictive values are highest at <24 hours despite the use of mild therapeutic hypothermia or sedative Dynamics of brain activity after circulatory arrest medication. Additional value of electroencephalography Within 10 to 40 seconds after circulatory arrest the EEG becomes reactivity for the prediction of poor outcome is negligible. iso-electric.[5] Just as deep coma in the first hours after the arrest does Computer-assisted analysis is equally reliable and may facilitate not preclude full functional recovery, recovery of brain functioning is the use of the electroencephalogram at the bedside on intensive possible with iso-electricity on early EEG. In such cases improvement care units. Whether or not treatment of electrographic status of EEG activity within 12 to 24 hours is vital.[6-8] Absence of relevant epilepticus improves outcome is being studied in the Dutch improvement within that time window is invariably associated multicentre randomised TELSTAR trial (NCT02056236). with a poor outcome.[9-11] On the other hand, with recovery towards continuous, physiological rhythms within 12 hours, Introduction neurological prognosis is very good (figure 1).[9,10] Comatose patients after circulatory arrest have an uncertain prognosis. Despite treatment on intensive care units, the EEG background pattern is at least as reliable as SSEP for outcome is poor in approximately half of all patients with prediction of outcome out-of-hospital cardiac arrest as a result of severe postanoxic Studies on the association between the EEG background pattern encephalopathy.[1] Early recognition of patients without chances and outcome unrelated to timing of the EEG reported moderate of recovery of brain function may prevent continuation of futile predictive values.[12-17] High predictive values have been found by treatment and contribute to communication between doctors EEG classification in relation to the time since circulatory arrest. and patients. Seven prospective cohort studies report on the value of ongoing The electroencephalogram (EEG) measures electrical potential suppression at 12 to 24 hours after circulatory arrest for prediction of differences between pairs of scalp electrodes. These primarily poor outcome. Six studies partly overlap and together consist of 864 result from the sum of post-synaptic potentials, so EEG activity patients from five Dutch hospitals.[9,10,18-21] The seventh included mainly reflects cortical synaptic activity.[2] Since cortical 100 patients from Yale University Hospital.[11] In addition, there

6 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care EEG after circulatory arrest

Figure 1. Upper panel: EEG pattern evolving from iso-electric to continuous, physiological activity within 12 hours. This evolution is favourable and the patient had a good outcome. Lower panel: EEG pattern evolving from nearly iso-electric to burst suppression and beyond 24 hours back to low voltage activity. This evolution is 100% unfavourable and the patient had a poor outcome is a retrospective cohort study in 211 patients from Italy.[22] In all comatose or death]. In the American study, the best achieved these studies, consecutive, unselected comatose patients after score on the Glasgow Outcome Scale during admission was cardiac arrest were included. Continuous EEG measurements used [4 or 5 = good, 1, 2 or 3 = poor]. started within 12 to 24 hours and continued for at least three The eight studies together included 1175 patients. The days, or until the patient died or recovered. Twenty-one proportion of patients with a poor outcome varied from 52- electrodes were used according to the international 10-20 54% in the Dutch to 71% in the American studies. An iso- system. Patients were treated according to standard protocols electric, suppressed (<10 μV) or low voltage (<20 μV) EEG at for comatose patients after circulatory arrest. This indicated ≥24 hours after cardiac arrest was invariably associated with a targeted temperature management [TTM] at 33°C with the poor outcome. Lasting suppression or synchronised patterns necessary sedation [propofol, midazolam] in approximately with >50% suppression at ≥12 hours after cardiac arrest were also three quarters, and TTM at 36°C in about one quarter of all invariably associated with a poor outcome.[10,11,20] This included included patients. Withdrawal of treatment was considered burst suppression with identical bursts.[24] The sensitivity of after ≥48-72 hours, during normothermia, and off sedation. these patterns together in identifying patients with a poor Decisions were based on international guidelines including outcome varied between 28 and 84%. With no false positives incomplete return of brainstem reflexes, treatment-resistant in a total of 1175 patients, these EEG measures are at least as myoclonus, and bilateral absence of somatosensory evoked reliable as absent SSEP for prediction of poor outcome, since potentials [SSEPs].[23] The EEG in the first 2 hours was not SSEP guidelines are based on cohorts that included a total of 678 taken into account. EEG analyses were performed offline, after patients, and four false positives were reported.[4] In addition, registration. Evaluators were blinded to the time of the epoch a continuous EEG pattern at 12 hours is strongly associated since the arrest, treatment, and patient outcome. In the Dutch with a good neurological outcome.[11,20,22] If patients with such studies, outcome at six months was classified as good [cerebral a beneficial evolution of the EEG died, it was generally from performance category [CPC] 1 or 2 indicating no or moderate failure of other organs than the brain, mostly the heart. disability] or poor [CPC 3, 4 or 5, indicating severe disability, At least six other cohort studies, together including 1587 patients,

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 7 Netherlands Journal of Critical Care EEG after circulatory arrest

confirmed the reliability of EEG measures for prediction of poor Highest predictive value within 24 hours, despite outcome with false-positive rates of <2%,[25-30] but terminology medication varies. Some researchers use the term ‘highly malignant’ EEG Intuitively, analogous to the clinical course, the value of the EEG patterns. This is ill defined. However, suppressed patterns to predict patient outcome should increase with time elapsing and synchronous patterns with >50% suppression were since circulatory arrest.[4] However, based on the data, the always included in definitions of ‘highly malignant’ patterns opposite turns out to be the case. Differences between patients and invariably associated with a poor outcome. Reliability of with and without chances of recovery, as well as predictive values burst suppression with identical bursts has been confirmed by for good and poor outcome, are the largest within the first 24 visual EEG analysis in 522 patients with no false positives.[31] A hours after arrest.[38] An important cause is the evolution towards small cohort study suggests that repeated routine recordings aspecific EEG activity beyond 24 hours in many patients who are possibly as reliable as continuous EEG.[32,33] In the group eventually have a poor outcome.[10] Whether or not such activity of patients with indeterminate outcome perspectives, EEG still includes qualitative or quantitative predictive characteristics characteristics hold potential to predict the chance of permanent warrants further study.[34] Furthermore, it is generally considered eurological deficits after late awakening, but this needs further that the EEG is not useful as a predictor during treatment with research.[34] hypothermia or sedative medication.[4] This is a misapprehension, not supported by data.[10,11,39] Although ion channel kinetics and EEG background pattern contributes to multimodal neurotransmitter release are temperature dependent, effects prediction of poor outcome of few degrees are small and mild therapeutic hypothermia to In at least four cohorts, EEG background pattern data were 32ºC affects the EEG only mildly.[40] Furthermore, propofol- combined with clinical, biochemical, or SSEP data.[11,21,25,35] The induced EEG changes are well known. With the dosages that previously established high predictive values of absent pupillary are mostly used during targeted temperature management, light or SSEP responses at 48-72 hours for prediction of poor patterns remain continuous with anteriorisation of the ‘alpha’ outcome were confirmed. Additionally, EEG parameters were rhythm, and iso-electricity will never be induced.[41] If burst found to be complementary to these conventional predictors. suppression is observed, bursts are heterogeneous and appear ‘Highly malignant’ EEG patterns are not always associated and disappear gradually.[42] This is a physiological response with absent SSEP,[36] and in a substantial proportion of patients of a relatively healthy brain to sedation and contrasts sharply only one or two predictors of poor outcome were present. This with the observed pathological burst suppression patterns indicates that with all tests together, more patients with a poor with identical bursts, with flat interburst intervals and abrupt outcome could be identified reliably than with a single modality. transitions between suppression and burst activity (figure 2).[24] Only in patients with a continuous EEG pattern with a dominant Moreover, mean doses of sedative medication were lower in frequency of ≥8 Hz from 12 hours after cardiac arrest the SSEP patients with unfavourable EEG patterns than in those with was always present and this test therefore may be withheld.[21,37] favourable patterns.[9,10,20]

Figure 2. Two examples of burst suppression with identical bursts. The interval between the two epochs is approximately 24s in the left and 15s in the right panel. Note the similarity of shapes of subsequent bursts and the iso-electricity of interburst intervals. Burst suppression with identical bursts has only been observed after severe hypoxic or ischaemic brain damage and was invariably associated with a poor outcome[24]

8 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care EEG after circulatory arrest

Figure 3. EEG fragments with generalised periodic discharges [GPDs]. In the example on the left, GPDs evolved from a background pattern with continuous activity. Between GPDs, there appears to be continuous background activity and there is no complete bilateral synchronicity and GPD shapes and inter-GPD intervals vary. This is a favourable evolution of GPDs and the patient had a good outcome. In the example on the right, GPDs evolved from an iso-electric pattern. There is no background activity between the GPDs and GPDs appear bilaterally synchronous. GPD shapes and inter-GPD intervals are identical. This is an unfavourable evolution of GPDs and the patient had a poor outcome.[66] Because of the apparent heterogeneity of GPD patterns, only well-described subgroups have relevant prognostic value

Burst suppression and status epilepticus Burst suppression and status epilepticus are classically classified as ‘burst suppression with identical bursts’ (figure 2): considered to be ‘unfavourable’ EEG patterns in patients subsequent bursts in a particular channel are almost ‘photographic’ with a postanoxic coma.[11,12,15,17,43-46] However, specificity for copies. Burst suppression with identical bursts was not observed in predictions of poor outcome based on unselected groups of a series of 240 EEGs during anaesthesia or traumatic brain injury. burst suppression or status epilepticus EEGs is moderate.[4,47] Otherwise, this pathological EEG pattern may be seen in up to 20% This is because such patterns are also observed in a considerable of patients with postanoxic encephalopathy and a poor outcome, proportion of patients who eventually have a good outcome.[18] mostly on the first or second day.[24] Burst suppression with identical Only specific, well-defined subgroups of burst suppression or bursts indicates severe encephalopathy and is invariably associated status epilepticus reliably predict a poor outcome. with a poor outcome.[9-11,24]

Burst suppression Status epilepticus Burst suppression can be defined as an EEG with high amplitude The reported incidence of electrographic status epilepticus in activity of at least four phases and a duration of at least 500 ms comatose patients after cardiopulmonary resuscitation varies [bursts], alternated by periods of low [<10μV] or absent activity from 10 to 35% and depends on diagnostic criteria.[15,39,56-59] [suppressions] for more than 50% of the time.[48] Such patterns Distinct epileptiform patterns, with evolving seizures, are can be physiological, for instance during early development, or rare.[60,61] Other rhythmic activity, such as generalised periodic pathological, for example in almost half of all comatose patients discharges or rhythmic delta activity, is more common.[39,60,62,63] within the first 48 hours after cardiac arrest.[18] Also, burst It is unclear whether these various patterns all reflect true suppression can be induced by anaesthetics.[49] The mechanisms epileptiform activity, with the possibility to return to normal, involved in burst suppression are divergent, and range from or rather are a direct expression of severe encephalopathy, reversible changes in synaptic functioning and Ca2+ homeostasis in which treatment with antiepileptic drugs would be futile. to selective neural death.[50-52] [64,65] On the EEG, potential reversibility of status epilepticus in Characteristics to classify burst suppression patterns into postanoxic coma is associated with evolution from patterns subgroups with differences in clinical significance include the with continuous background activity, as opposed to evolution duration of the bursts and interburst intervals, maximum peak-to- from a discontinuous background pattern.[66] Furthermore, as peak voltage, area under the curve, the ratio of power in high versus compared with epileptiform patterns of patients with a poor low frequencies,[53] and combinations with other pathological outcome, in patients who eventually recovered, such patterns had patterns, such as generalised periodic discharges.[54,55] For example, a higher background continuity, higher discharge frequency [0.90 longer suppressions are associated with poorer recovery in patients vs.1.63 Hz], lower relative discharge power, and lower discharge with postanoxic coma.[13] Extreme similarity of burst shape is a periodicity (figure 3).[11,66,67] distinct feature of some burst suppression patterns, which are

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 9 Netherlands Journal of Critical Care EEG after circulatory arrest

Figure 4. Case 1 A-C: three EEG epochs at 5, 12 and 20 hours after cardiac arrest [CA], showing a favourable evolution towards a continuous EEG pattern within 24 hours. This is strongly associated with a good outcome. Case 2 A-C: Three EEG epochs showing an indeterminate evolution. At t=27 hours after arrest, the EEG still shows significant suppressions intermixed with delta and theta activity. Outcome is uncertain. Lower panels: quantitative analysis of the EEG patterns above with the Cerebral Recovery Index [CRI]. Corresponding epochs are indicated with A-C. Case 1 shows an increase towards CRI >0.5 within 24 hours, with a final CRI=0.9. This is strongly associated with good outcome. For Case 2, CRI ≤0.5 at all points in time and CRI=0.2 at 24 hours. This is strongly associated with a poor outcome. Note that in both patients the EEG is nearly iso- electric in the first hours of the recording. In case 2, prognosis remains uncertain with visual analysis of the EEG, but can be classified as poor with use of the CRI

Treatment of status epilepticus even more likely in patients with myoclonia, since the risk of a poor Apart from classification, the usefulness of treatment of outcome is larger and neuronal necrosis more common.[1,4,67,73,74] electrographic status epilepticus after circulatory arrest is In a retrospective cohort study of 139 patients, non-standardised, unclear.[68-70] Ambivalence in this respect is reflected by the moderately intensive treatment with antiepileptic drugs did not way these patterns are treated by Dutch and American epilepsy improve outcome of electrographic status epilepticus after cardiac experts: approximately two thirds give antiepileptic drugs, but arrest.[62] Effects of intensive treatment according to status epilepticus only one third treats as aggressively as in clinically overt status guidelines is currently being studied in the randomised, multicentre epilepticus.[71,72] For most neurologists the threshold to treat patients Treatment of Electroencephalographic STatus epilepticus After with overt myoclonia is lower than for patients with non-convulsive cardiopulmonary Resuscitation [TELSTAR] trial [NCT02056236; electrographic seizures. However, irreversible damage is probably www.TELSTARtrial.nl].[58]

10 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care EEG after circulatory arrest

EEG reactivity the importance of evolution in time: in both groups there is EEG reactivity can be defined as any change in frequency or improvement of the mean EEG pattern. However, in patients amplitude of the EEG background pattern resulting from with a good outcome, mean improvement is twice as fast as in application of an external stimulus.[75,76] However, consensus patients with a poor outcome. about the characteristics of changes in a responsive EEG has long been lacking.[77] External stimulation typically consists Conclusion of auditory [shouting or clapping], somatosensory [painful In comatose patients after circulatory arrest, the EEG background pressure to the nail bed or supraorbital nerve], or visual [passive pattern in the first 24 hours provides reliable information on eye opening] input.[78] the severity of encephalopathy and enables reliable prediction Absent reactivity to external stimulation of the EEG background of outcome in 40-50% of patients, despite treatment with pattern is much studied as a potential predictor of poor outcome hypothermia or sedative medication. For poor outcome of comatose patients after circulatory arrest. Two prospective prediction, the EEG is as reliable as and complementary to the and one retrospective cohort studies report strong associations SSEP. The EEG is the first modality to also allow prediction of between absent EEG reactivity and poor outcome.[16] However, a good outcome. Computer-assisted interpretation of the EEG these results could not be replicated with a recent systematic may assist in outcome prediction and facilitate bedside use multicentre study, and the additional predictive value of absent at intensive care units. Epileptiform patterns are of unknown EEG reactivity testing, in addition to the EEG background significance and effects of treatment with antiepileptic drugs pattern, was futile.[79] Otherwise, adequate EEG reactivity to are uncertain. Whether or not treatment of electrographic stimuli within the first 48 hours was strongly associated with status epilepticus improves outcome is being studied in the good recovery.[44,79-82] randomised multicentre Treatment of Electroencephalographic STatus epilepticus After cardiopulmonary Resuscitation Computer-assisted analysis [TELSTAR] trial [NCT02056236]. Application of the EEG on the intensive care unit is limited by the complexity of the signal, which typically cannot be interpreted Acknowledgement by general intensive care nurses or staff. Computer-assisted Parts of this article have been published previously.[93] The editor analysis may help.[83] Techniques to assist in the interpretation of Clinical Neurophysiology provided permission for publication of continuous EEG background patterns include time frequency of this updated version in the Netherlands Journal of Critical Care. trend curves,[84,85] quantification of hemispheric asymmetry,[86] and an explicit classification of the EEG in common categories Disclosure [e.g. iso-electricity, burst suppression or diffusely slowed Michel J.A.M. van Putten is co-founder of Clinical Science patterns].[87] Systems, Leiden (www.clinicalscience.systems). Jeannette A few articles present techniques specifically aiming at outcome Hofmeijer has no conflicts of interest. prediction in patients with a postanoxic encephalopathy. One of the earliest studies is on the use of amplitude-integrated EEG References [88] [aEEG]. In a prospective cohort of 34 patients, all 20 patients 1. Zandbergen EG, de Haan RJ, Stoutenbeek CP, Koelman JH, Hijdra A. Systematic with a continuous aEEG pattern at normothermia regained review of early prediction of poor outcome in anoxic-ischaemic coma. Lancet. 1998;352:1808-12. consciousness. All 14 patients with flat patterns, burst suppression, 2. van Putten M. Fysiologie van het EEG. In: Leerboek Klinische Neurofysiologie. M. or status epilepticus aEEG patterns died in hospital.[43,88] Other Zwarts, G. van Dijk, M. van Putten, W. Mess [editors]. Houten: Bohn Stafleu van Loghum; 2014. p. 123-30. quantitative EEG features studied include the burst suppression 3. Hofmeijer J, van Putten MJAM. Ischemic cerebral damage: an appraisal of synaptic failure. Stroke J Cereb Circ. 2012;43:607-15. ratio and entropy measures, with differences between patients 4. Sandroni C, Cariou A, Cavallaro F, et al. Prognostication in comatose survivors of with good and poor outcome on a group level, but limited cardiac arrest: An advisory statement from the European Resuscitation Council [13,89] and the European Society of Intensive Care Medicine. Intensive Care Med. predictive value for individual patients. 2014;40:1816-31. 5. van Dijk JG, Thijs RD, van Zwet E, et al. The semiology of tilt-induced reflex syncope in relation to electroencephalographic changes. Brain J Neurol. The Cerebral Recovery Index [CRI] was introduced in 2013 and 2014;137[Pt 2]:576-85. is based on a combination of features, including amplitude and 6. Jøogensen EO, Malchow-Møller A. Natural history of global and critical brain ischaemia. Part III: cerebral prognostic signs after cardiopulmonary resuscitation. [90,91] continuity, derived from an 18-channel EEG recording. The Cerebral recovery course and rate during the first year after global and critical CRI is normalised in the range [0-1], with 0 indicating severe ischaemia monitored and predicted by EEG and neurological signs. Resuscitation. 1981;9:175-88. encephalopathy and 1 indicating normal brain functioning. In 7. Jørgensen EO, Malchow-Møller A. Natural history of global and critical brain ischaemia. Part I: EEG and neurological signs during the first year after independent training and test sets using deep learning, CRI at cardiopulmonary resuscitation in patients subsequently regaining consciousness. 12 and 24 hours after cardiac arrest predicted poor outcome Resuscitation. 1981;9:133-53. 8. Jørgensen EO, Malchow-Møller A. Natural history of global and critical brain without false positives at 58% sensitivity and good outcome at ischaemia. Part II: EEG and neurological signs in patients remaining unconscious a specificity of 95% and a sensitivity of 48% (figure 4).[92] Note after cardiopulmonary resuscitation. Resuscitation. 1981;9:155-74.

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 11 Netherlands Journal of Critical Care EEG after circulatory arrest

9. Tjepkema-Cloostermans MC, Hofmeijer J, Trof RJ, Blans MJ, Beishuizen A, van 35. Kim JH, Kim MJ, You JS, et al. Multimodal approach for neurologic prognostication Putten MJAM. Electroencephalogram Predicts Outcome in Patients With of out-of-hospital cardiac arrest patients undergoing targeted temperature Postanoxic Coma During Mild Therapeutic Hypothermia. Crit Care Med. management. Resuscitation. 2019;134:33-40. 2015;43:159-67. 36. Beuchat I, Solari D, Novy J, Oddo M, Rossetti AO. Standardized EEG interpretation 10. Hofmeijer J, Beernink TMJ, Bosch FH, Beishuizen A, Tjepkema-Cloostermans MC, in patients after cardiac arrest: Correlation with other prognostic predictors. van Putten MJAM. Early EEG contributes to multimodal outcome prediction of Resuscitation. 2018;126:143-6. postanoxic coma. Neurology. 2015;85:137-43. 37. Fredland A, Backman S, Westhall E. Stratifying comatose postanoxic patients for 11. Sivaraju A, Gilmore EJ, Wira CR, et al. Prognostication of post-cardiac arrest coma: somatosensory evoked potentials using routine EEG. Resuscitation. 2019;143:17-21. early clinical and electroencephalographic predictors of outcome. Intensive Care 38. Hofmeijer J, Tjepkema-Cloostermans MC, van Putten MJAM. Outcome prediction Med. 2015;41:1264-72. in postanoxic coma with electroencephalography: The sooner the better. 12. Zandbergen EGJ, Hijdra A, Koelman JHTM, et al. Prediction of poor outcome Resuscitation. 2015;91:e1-2. within the first 3 days of postanoxic coma. Neurology. 2006;66:62-8. 39. Rossetti AO, Logroscino G, Liaudet L, et al. Status epilepticus: an independent 13. Wennervirta JE, Ermes MJ, Tiainen SM, et al. Hypothermia-treated cardiac arrest outcome predictor after cerebral anoxia. Neurology. 2007;69:255-60. patients with good neurological outcome differ early in quantitative variables of 40. Schomer DL, Lopes da Silva F. Niedermeyer’s Electroencephalography: Basic EEG suppression and epileptiform activity. Crit Care Med. 2009;37:2427-35. Principles, Clinical Applications, and Related Fields. 6th ed. Philadelphia: 14. Kawai M, Thapalia U, Verma A. Outcome from therapeutic hypothermia and EEG. Lippincott Williams & Wilkins; 2011. J Clin Neurophysiol Off Publ Am Electroencephalogr Soc. 2011;28:483-8. 41. Hindriks R, van Putten MJAM. Meanfield modeling of propofol-induced changes 15. Rittenberger JC, Popescu A, Brenner RP, Guyette FX, Callaway CW. Frequency in spontaneous EEG rhythms. NeuroImage. 2012;60:2323-34. and timing of nonconvulsive status epilepticus in comatose post-cardiac arrest 42. Reddy RV, Moorthy SS, Mattice T, Dierdorf SF, Deitch RD. An subjects treated with hypothermia. Neurocrit Care. 2012;16:114-22. electroencephalographic comparison of effects of propofol and methohexital. 16. Rossetti AO, Carrera E, Oddo M. Early EEG correlates of neuronal injury after brain Electroencephalogr Clin Neurophysiol. 1992;83:162-8. anoxia. Neurology. 2012;78:796-802. 43. Rundgren M, Westhall E, Cronberg T, Rosén I, Friberg H. Continuous amplitude- 17. Legriel S, Hilly-Ginoux J, Resche-Rigon M, et al. Prognostic value of electrographic integrated electroencephalogram predicts outcome in hypothermia-treated postanoxic status epilepticus in comatose cardiac-arrest survivors in the cardiac arrest patients. Crit Care Med. 2010;38:1838-44. therapeutic hypothermia era. Resuscitation. 2013;84:343-50. 44. Thenayan EAL, Savard M, Sharpe MD, Norton L, Young B. Electroencephalogram 18. Cloostermans MC, van Meulen FB, Eertman CJ, Hom HW, van Putten MJAM. for prognosis after cardiac arrest. J Crit Care. 2010;25:300-4. Continuous electroencephalography monitoring for early prediction of 45. Oh SH, Park KN, Kim YM, et al. The prognostic value of continuous amplitude- neurological outcome in postanoxic patients after cardiac arrest: a prospective integrated electroencephalogram applied immediately after return of cohort study. Crit Care Med. 2012;40:2867-75. spontaneous circulation in therapeutic hypothermia-treated cardiac arrest 19. Sondag L, Ruijter BJ, Tjepkema-Cloostermans MC, et al. Early EEG for outcome patients. Resuscitation. 2013;84:200-5. prediction of postanoxic coma: prospective cohort study with cost-minimization 46. Sadaka F, Doerr D, Hindia J, Lee KP, Logan W. Continuous Electroencephalogram analysis. Crit Care. 2017;21:111. in Comatose Postcardiac Arrest Syndrome Patients Treated With Therapeutic 20. Ruijter BJ, Tjepkema-Cloostermans MC, Tromp SC, et al. Early Hypothermia: Outcome Prediction Study. J Intensive Care Med. 2015;30:292-6. electroencephalography for outcome prediction of postanoxic coma: A 47. Westhall E, Rundgren M, Lilja G, Friberg H, Cronberg T. Postanoxic status prospective cohort study. Ann Neurol. 2019;86:203-14. epilepticus can be identified and treatment guided successfully by continuous 21. Glimmerveen AB, Ruijter BJ, Keijzer HM, Tjepkema-Cloostermans MC, van Putten electroencephalography. Ther Hypothermia Temp Manag. 2013;3:84-7. MJAM, Hofmeijer J. Association between somatosensory evoked potentials and 48. Hirsch LJ, Brenner RP, Drislane FW, et al. The ACNS subcommittee on research EEG in comatose patients after cardiac arrest. Clin Neurophysiol. 2019;130:2026-31. terminology for continuous EEG monitoring: proposed standardized terminology 22. Spalletti M, Carrai R, Scarpino M, et al. Single electroencephalographic patterns as for rhythmic and periodic EEG patterns encountered in critically ill patients. J Clin specific and time-dependent indicators of good and poor outcome after cardiac Neurophysiol. 2005;22:128-35. arrest. Clin Neurophysiol. 2016;127:2610-7. 49. Yoon JR, Kim YS, Kim TK. Thiopental-induced burst suppression measured by 23. Wijdicks EFM, Hijdra A, Young GB, Bassetti CL, Wiebe S, Quality Standards the bispectral index is extended during propofol administration compared with Subcommittee of the American Academy of Neurology. Practice parameter: sevoflurane. J Neurosurg Anesthesiol. 2012;24:146-51. prediction of outcome in comatose survivors after cardiopulmonary resuscitation 50. van Putten MJAM, van Putten MHPM. Uncommon EEG burst-suppression in [an evidence-based review]: report of the Quality Standards Subcommittee of severe postanoxic encephalopathy. Clin Neurophysiol. 2010;121:1213-9. the American Academy of Neurology. Neurology. 2006;67:203-10. 51. Liley DTJ, Walsh M. The Mesoscopic Modeling of Burst Suppression during 24. Hofmeijer J, Tjepkema-Cloostermans MC, van Putten MJAM. Burst-suppression Anesthesia. Front Comput Neurosci. 2013;7:46. with identical bursts: a distinct EEG pattern with poor outcome in postanoxic 52. Brandon Westover M, Ching S, Kumaraswamy VM, et al. The human burst coma. Clin Neurophysiol. 2014;125:947-54. suppression electroencephalogram of deep hypothermia. Clin Neurophysiol. 25. Bongiovanni F, Romagnosi F, Barbella G, et al. Standardized EEG analysis to reduce 2015;126:1901-14. the uncertainty of outcome prognostication after cardiac arrest. Intensive Care 53. Akrawi WP, Drummond JC, Kalkman CJ, Patel PM. A comparison of the Med. 2020;46:963-72. electrophysiologic characteristics of EEG burst-suppression as produced by 26. Nakstad ER, Stær-Jensen H, Wimmer H, et al. Late awakening, prognostic isoflurane, thiopental, etomidate, and propofol. J Neurosurg Anesthesiol. factors and long-term outcome in out-of-hospital cardiac arrest - results 1996;8:40-6. of the prospective Norwegian Cardio-Respiratory Arrest Study [NORCAST]. 54. Fugate JE, Wijdicks EFM, Mandrekar J, et al. Predictors of neurologic outcome in Resuscitation. 2020;149:170-9. hypothermia after cardiac arrest. Ann Neurol. 2010;68:907-14. 27. Scarpino M, Carrai R, Lolli F, et al. Neurophysiology for predicting good and 55. Søholm H, Kjær TW, Kjaergaard J, et al. Prognostic value of electroencephalography poor neurological outcome at 12 and 72 h after cardiac arrest: The ProNeCA [EEG] after out-of-hospital cardiac arrest in successfully resuscitated patients multicentre prospective study. Resuscitation. 2020;147:95-103. used in daily clinical practice. Resuscitation. 2014;85:1580-5. 28. Backman S, Cronberg T, Friberg H, et al. Highly malignant routine EEG predicts 56. Snyder BD, Hauser WA, Loewenson RB, Leppik IE, Ramirez-Lassepas M, Gumnit RJ. poor prognosis after cardiac arrest in the Target Temperature Management trial. Neurologic prognosis after cardiopulmonary arrest: III. Seizure activity. Neurology. Resuscitation. 2018;131:24-8. 1980;30:1292-7. 29. Westhall E, Rossetti AO, van Rootselaar A-F, et al. Standardized EEG interpretation 57. Mani R, Schmitt SE, Mazer M, Putt ME, Gaieski DF. The frequency and timing accurately predicts prognosis after cardiac arrest. Neurology. 2016;86:1482-90. of epileptiform activity on continuous electroencephalogram in comatose 30. Bevers MB, Scirica BM, Avery KR, Henderson GV, Lin AP, Lee JW. Combination of post-cardiac arrest syndrome patients treated with therapeutic hypothermia. Clinical Exam, MRI and EEG to Predict Outcome Following Cardiac Arrest and Resuscitation. 2012;83:840-7. Targeted Temperature Management. Neurocrit Care. 2018;29:396-403. 58. Ruijter BJ, van Putten MJ, Horn J, et al. Treatment of electroencephalographic 31. Barbella G, Novy J, Marques-Vidal P, Oddo M, Rossetti AO. Prognostic role of status epilepticus after cardiopulmonary resuscitation [TELSTAR]: study protocol EEG identical bursts in patients after cardiac arrest: Multimodal correlation. for a randomized controlled trial. Trials. 2014;15:433. Resuscitation. 2020;148:140-4. 59. Seder DB, Sunde K, Rubertsson S, et al. Neurologic outcomes and 32. Alvarez V, Sierra-Marcos A, Oddo M, Rossetti AO. Yield of intermittent versus postresuscitation care of patients with myoclonus following cardiac arrest. Crit continuous EEG in comatose survivors of cardiac arrest treated with hypothermia. Care Med. 2015;43:965-72. Crit Care. 2013;17:R190. 60. Hirsch LJ, LaRoche SM, Gaspard N, et al. American Clinical Neurophysiology 33. Crepeau AZ, Fugate JE, Mandrekar J, et al. Value analysis of continuous EEG in Society’s Standardized Critical Care EEG Terminology: 2012 version. J Clin patients during therapeutic hypothermia after cardiac arrest. Resuscitation. Neurophysiol. 2013;30:1-27. 2014;85:785-9. 61. Knight WA, Hart KW, Adeoye OM, et al. The incidence of seizures in patients 34. Rey A, Rossetti AO, Miroz J-P, Eckert P, Oddo M. Late Awakening in Survivors of undergoing therapeutic hypothermia after resuscitation from cardiac arrest. Postanoxic Coma: Early Neurophysiologic Predictors and Association With ICU Epilepsy Res. 2013;106:396-402. and Long-Term Neurologic Recovery. Crit Care Med. 2019;47:85-92.

12 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care EEG after circulatory arrest

62. Hofmeijer J, Tjepkema-Cloostermans MC, Blans MJ, Beishuizen A, van Putten 78. Admiraal MM, van Rootselaar AF, Horn J. International consensus on EEG reactivity MJAM. Unstandardized treatment of electroencephalographic status epilepticus testing after cardiac arrest: Towards standardization. Resuscitation. 2018;131:36-41. does not improve outcome of comatose patients after cardiac arrest. Front 79. Admiraal MM, van Rootselaar A-F, Hofmeijer J, et al. Electroencephalographic Neurol. 2014;5:39. reactivity as predictor of neurological outcome in postanoxic coma: A multicenter 63. Milani P, Malissin I, Tran-Dinh YR, et al. Prognostic EEG patterns in patients prospective cohort study. Ann Neurol. 2019;86:17-27. resuscitated from cardiac arrest with particular focus on Generalized Periodic 80. Rossetti AO, Oddo M, Logroscino G, Kaplan PW. Prognostication after cardiac Epileptiform Discharges [GPEDs]. Neurophysiol Clin. 2014;44:153-64. arrest and hypothermia: a prospective study. Ann Neurol. 2010;67:301-7. 64. Young GB, Claassen J. Nonconvulsive status epilepticus and brain damage: 81. Crepeau AZ, Rabinstein AA, Fugate JE, et al. Continuous EEG in therapeutic further evidence, more questions. Neurology. 2010;75:760-1. hypothermia after cardiac arrest: prognostic and clinical value. Neurology. 65. Tjepkema-Cloostermans MC, Hindriks R, Hofmeijer J, van Putten MJAM. 2013;80:339-44. Generalized periodic discharges after acute cerebral ischemia: Reflection of 82. Tsetsou S, Oddo M, Rossetti AO. Clinical outcome after a reactive hypothermic selective synaptic failure? Clin Neurophysiol. 2013;125:255-62. EEG following cardiac arrest. Neurocrit Care. 2013;19:283-6. 66. Ruijter BJ, van Putten MJAM, Hofmeijer J. Generalized epileptiform discharges 83. Friberg H, Westhall E, Rosén I, Rundgren M, Nielsen N, Cronberg T. Clinical review: in postanoxic encephalopathy: Quantitative characterization in relation to Continuous and simplified electroencephalography to monitor brain recovery outcome. Epilepsia. 2015;56:1845-54. after cardiac arrest. Crit Care. 2013;17:233. 67. Rossetti AO, Oddo M, Liaudet L, Kaplan PW. Predictors of awakening from 84. Friedman D, Claassen J, Hirsch LJ. Continuous electroencephalogram monitoring postanoxic status epilepticus after therapeutic hypothermia. Neurology. in the intensive care unit. Anesth Analg. 2009;109:506-23. 2009;72:744-9. 85. Oddo M, Villa F, Citerio G. Brain multimodality monitoring: an update. Curr Opin 68. Chong DJ, Hirsch LJ. Which EEG patterns warrant treatment in the critically ill? Crit Care. 2012;18:111-8. Reviewing the evidence for treatment of periodic epileptiform discharges and 86. van Putten MJAM. Extended BSI for continuous EEG monitoring in carotid related patterns. J Clin Neurophysiol. 2005;22:79-91. endarterectomy. Clin Neurophysiol. 2006;117:2661-6. 69. Cronberg T. Should Postanoxic Status Epilepticus Be Treated Aggressively? Yes! J 87. Cloostermans MC, de Vos CC, van Putten MJAM. A novel approach for computer Clin Neurophysiol. 2015;32:449-51. assisted EEG monitoring in the adult ICU. Clin Neurophysiol. 2011;122:2100-9. 70. Rossetti AO. Should Postanoxic Status Epilepticus be Treated Aggressively?-No! J 88. Rundgren M, Rosén I, Friberg H. Amplitude-integrated EEG [aEEG] predicts Clin Neurophysiol. 2015;32:447-8. outcome after cardiac arrest and induced hypothermia. Intensive Care Med. 71. Abend NS, Dlugos DJ, Hahn CD, Hirsch LJ, Herman ST. Use of EEG monitoring 2006;32:836-42. and management of non-convulsive seizures in critically ill patients: a survey of 89. Noirhomme Q, Lehembre R, Lugo ZDR, et al. Automated analysis of background neurologists. Neurocrit Care. 2010;12:382-9. EEG and reactivity during therapeutic hypothermia in comatose patients after 72. Bouwes A, Kuiper MA, Hijdra A, Horn J. Induced hypothermia and determination cardiac arrest. Clin EEG Neurosci. 2014;45:6-13. of neurological outcome after CPR in ICUs in the Netherlands: results of a survey. 90. Tjepkema-Cloostermans MC, van Meulen FB, Meinsma G, van Putten MJ. A Resuscitation. 2010;81:393-7. Cerebral Recovery Index [CRI] for early prognosis in patients after cardiac arrest. 73. Krumholz A, Stern BJ, Weiss HD. Outcome from coma after cardiopulmonary Crit Care. 2013;17:R252. resuscitation: relation to seizures and myoclonus. Neurology. 1988;38:401-5. 91. Tjepkema-Cloostermans MC, Hofmeijer J, Beishuizen A, et al. Cerebral Recovery 74. Young GB, Gilbert JJ, Zochodne DW. The significance of myoclonic status Index: Reliable Help for Prediction of Neurologic Outcome After Cardiac Arrest. epilepticus in postanoxic coma. Neurology. 1990;40:1843-8. Crit Care Med. 2017;45:e789-e797. 75. Young GB. The EEG in coma. J Clin Neurophysiol. 2000;17:473-85. 92. Tjepkema-Cloostermans MC, da Silva Lourenço C, Ruijter BJ, et al. Outcome Prediction 76. Horn J, Cronberg T, Taccone FS. Prognostication after cardiac arrest. Curr Opin Crit in Postanoxic Coma With Deep Learning. Crit Care Med. 2019;47:1424-32. Care. 2014;20:280-6. 93. Hofmeijer J, van Putten MJAM. EEG in postanoxic coma: Prognostic and 77. Admiraal MM, van Rootselaar A-F, Horn J. Electroencephalographic reactivity diagnostic value. Clin Neurophysiol. 2016;127:2047-55. testing in unconscious patients: a systematic review of methods and definitions. Eur J Neurol. 2017;24:245-54.

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 13 Netherlands Journal of Critical Care Submitted June 2020; Accepted October 2020

REVIEW

Late development of Takotsubo syndrome following intensive care unit discharge

V. Visvanathan1,2, A. Kucia1,3, B. Reddi1,2,4, J. Horowitz1,4,5 1University of South Australia, Adelaide, Australia 2Royal Adelaide Hospital, Adelaide, Australia 3Lyell McEwin Hospital, Adelaide, Australia 4University of Adelaide, Adelaide, Australia 5The Queen Elizabeth Hospital, Adelaide, Australia

Correspondence V. Visvanathan - [email protected]

Keywords - Takotsubo syndrome, acute heart failure, hemodynamic instability, respiratory deterioration

Abstract O2 by nasal cannula and a temperature of 39.5°C. The admission Takotsubo syndrome (TTS), previously called Takotsubo electrocardiogram (ECG) showed sinus rhythm with first- cardiomyopathy and stress cardiomyopathy, is an acute form degree heart block, low voltages consistent with the patient’s of regional heart failure that usually mimics acute myocardial obesity, a heart rate of 89 beats/min, and mild widespread P-R infarction. TTS is often precipitated by catecholamine release depression. The QTc interval was normal and there were no associated with an acute emotional and/or physical stressor abnormalities of the ST segment or T wave. Electrolytes were such as critical illness. We describe a case of TTS following within the normal limits. Abnormal laboratory results included discharge from the ICU to a general ward. The medical creatinine 241 µmol/l and urea 10.9 mmol/l indicating acute emergency response team was called twice to the patient while kidney injury; arterial blood gas showed signs of metabolic on the general ward before she was readmitted to the ICU. This acidosis with bicarbonate (HCO3) 17 mmol/l, pH 7.15, and case indicates that the risk of developing TTS is not confined pCO2 49 mmHg; C-reactive protein was 540 mg/l; white cell to the acute phase of a critical illness. Given that TTS may count 13.5 x 109/l. Haemoglobin was 88 g/l. An echocardiogram be associated with serious haemodynamic instability and/or was performed on day 9 following admission to assess valvular respiratory deterioration, clinicians need to be aware that such function as there was some concern for endocarditis secondary deterioration, even following transfer from the ICU, may be due to bacteraemia. The echocardiogram showed normal left to TTS. ventricular (LV) systolic and valvular function. No wall motion abnormalities were reported although the test was suboptimal Case study due to patient obesity and positioning. A 58-year-old female presented to the emergency department The patient's index ICU admission was prolonged and of a rural hospital with pyrexia, right hip pain, limited mobility complicated by multiple medical problems including septic and hypotension two days after a fall. Investigations revealed shock (managed with antibiotics and noradrenaline infusion a fractured right acetabulum with minimal displacement titrated between 0.05 μg/kg/min to 0.1 μg/kg/min to maintain and Klebsiella urosepsis. Past medical history included the mean arterial pressure above 65 mmHg), respiratory failure previous cigarette smoking, hypertension, osteoarthritis, (treated with non-invasive positive pressure ventilator support hyperlipidaemia, obesity (BMI 33.02 kg/m2), bilateral total hip and diuretics), sepsis (treated with multiple antibiotics), replacements and chronic back pain. Medication at presentation anaemia requiring blood transfusion, acute renal failure consisted of aspirin, perindopril, gemfibrozil, meloxicam, requiring haemodialysis and functional bowel obstruction. pregabalin, paroxetine, carbamazepine, sustained-release She was found to have an infected right hip requiring multiple morphine and paracetamol. She was transferred to the intensive debridements, removal of the hip prosthesis and eventually care unit (ICU) of a metropolitan hospital where her initial femoral head ostectomy. observations were blood pressure 88/49 mmHg, respiratory The patient was discharged from the ICU to an orthopaedic rate 16 breaths/min, oxygen saturation of 97% on 3 litres/min ward on day 28 following admission and subsequently the

14 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Takotsubo syndrome after ICU discharge

Figure 1a. 12-lead ECG (day 1 following ICU readmission)

Figure 1b. ECG showing widespread T-wave inversion and QT prolongation (day 3 following ICU readmission) medical emergency response (MER) team was called twice in R-wave progression in precordial leads V2 and V3 (figure 1a) the 48 hours after ICU discharge (day 30 after admission) for which may raise suspicion of anterior myocardial infarction, acute respiratory deterioration, hypertension, tachycardia and but can also transiently occur in TTS. ECGs following chest pain. Table 1 shows details of precipitants of the MER resolution of wall motion abnormalities showed normal R-wave calls including signs and symptoms of clinical deterioration, progression. Serial chest X-rays showed the development of clinical findings and MER team management. The ECG upper lobe venous congestion, increased cardiac silhouette following readmission to the ICU showed sinus tachycardia size, linear interstitial opacities and 'bats wing' distribution of (130 beats/min) with first-degree heart block and reduced airspace opacification that developed over a few hours, strongly

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 15 Netherlands Journal of Critical Care Takotsubo syndrome after ICU discharge

suggestive of cardiogenic pulmonary oedema. Due to worsening with potential TTS during ICU admission and following ICU respiratory failure, the patient was transferred to ICU for non- discharge in order to increase awareness of this potentially invasive ventilatory support. On day 1 following readmission lethal inflammatory disorder. to the ICU the troponin T was elevated at 396 ng/l and NT- TTS is a cause of acute heart failure that is associated with proBNP was 35,000 ng/l. The ECG 3 days after readmission characteristic left ventricular (LV) regional wall motion showed widespread T-wave inversion and QTc prolongation abnormalities (RWMAs) in the absence of causative occlusive (figure 1b). Four days after readmission, a transthoracic coronary artery disease.[2] The vast majority of patients echocardiogram (TTE) was undertaken using contrast showing diagnosed with TTS are older women, but cases of TTS have moderate to severe LV systolic impairment, LV ballooning been identified in both genders and in all age groups.[2,3] In involving mid segments (figure 2) and apex without LV outflow most cases TTS is precipitated by endogenous (associated with tract obstruction (LVOTO), some hypokinesis in the distal right emotional or physical stressors), or exogenous (pharmacological) ventricular free wall and mild to moderate mitral regurgitation catecholamine surges.[4,5] TTS may be triggered by other with an ejection fraction (EF) of 35%. The echo was repeated two diseases and is denoted as 'secondary TTS'.[1,6] Myocardial weeks later, showing a low normal LV ejection fraction (LVEF) injury is associated with diffuse inflammatory activation[7] and (50-55%), mild to moderate mitral regurgitation and a dilated oedema throughout the left ventricle: this persists for at least left atrium. No wall motion abnormalities were noted. On day 81 three months, together with evidence of impaired myocardial following first admission, the patient was transferred to a facility energetics[8] and eventual diffuse, but variable fibrous scarring. for ongoing rehabilitation and survived to discharge home. Takotsubo syndrome diagnosis Introduction TTS signs and symptoms at onset usually mimic those of We have described the case of a 58-year-old woman who acute coronary syndrome including chest pain and dyspnoea developed Takotsubo syndrome (TTS) secondary to a fall and in some cases severe hypotension or cardiac arrest.[9] In and fracture of the right acetabulum, which was complicated the past, TTS has largely been diagnosed ‘by accident’ during by urosepsis, renal failure and hypotension necessitating ICU emergency cardiac catheterisation for management of patients admission. Consistent with the relatively poor prognosis of with presumptive acute myocardial infarction. Initially, the ‘secondary’ TTS[1] this patient had a long stay in hospital but diagnosis was essentially one of exclusion: there was either no survived to discharge. This case is also noteworthy for the late coronary stenosis, despite the advanced age of most patients, or development of TTS (post discharge from ICU), the delay to the stenosis which was present did not account for the pattern diagnosis (based on ECG, biomarkers and echocardiographic of regional systolic dysfunction.[9] However, the problem with appearances alone), and the lack of adjustment of therapy for this method was that it relied entirely on cardiac catheterisation the occurrence of TTS. We will review aspects of this case findings: it is theoretically possible to have a myocardial infarct which are particularly relevant to the management of patients without major coronary stenosis, for example on the basis of coronary artery spasm, embolic infarct in atrial fibrillation, and spontaneous coronary artery dissection. Furthermore, there are many patients, such as those who were severely ill in the ICU, who could not readily undergo immediate cardiac catheterisation because of clinical circumstances.[10] The patient in the case reported here falls into this category. In response to this dilemma, there has been progressive development of a set of criteria for presumptive (preliminary) diagnosis of TTS, based on ECG appearances, biomarkers, and echocardiographic changes.[11] The ECG in TTS can initially show elevated, normal or non- specifically altered ST segments, left bundle brunch block and/or T-wave abnormalities.[12,13] Widespread deep T-wave inversion and QT-interval prolongation are commonly seen around 24-36 hours following onset of TTS symptoms.[12,14] Often, ECG changes in TTS are ‘multi-regional,’ suggesting Figure 2. Image of left ventricular apical ballooning abnormalities extending beyond the distribution of a single Reproduced from Chandrasegaram MD, Celermajer DS, Wilson MK. Apical coronary artery.[12] The most diagnostically useful biomarkers ballooning syndrome complicated by acute severe mitral regurgitation with left ventricular outflow obstruction–case report. J Cardiothoracic Surg. 2007;2:1-4. for TTS are the combination of a small elevation in cardiac Available via license: Creative Commons Attribution 2.0 Generic troponin concentrations, together with a very large elevation

16 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Takotsubo syndrome after ICU discharge

Table 1. Patient’s vital signs during medical emergency response (MER) calls transthoracic echocardiography (TTE) can provide a presumptive diagnosis of TTS. The key echocardiographic feature in MER call 1 MER call 2 (day 2 post-ICU (day 3 post-ICU distinguishing TTS from acute coronary syndrome is the large area transfer) transfer) of myocardial systolic dysfunction that extends beyond the territory Symptoms Dyspnoea Dyspnoea of a single coronary artery and symmetrical regional abnormalities Chest tightness Chest pain [2] Feeling lightheaded in TTS. TTE is also used to detect potential TTS complications Reasons for MER call Tachycardia Tachycardia including left ventricular outflow obstruction (LVOTO), mitral Tachypnoea Tachypnoea regurgitation, right ventricular involvement, apical thrombus Hypertension Hypertensive Hypoxia Hypoxia formation and cardiac rupture, as well as monitoring the recovery Chest pain Chest pain of LV function.[11,17] Heart rate (beats/min) 100 140 In the last decade, cardiac magnetic resonance imaging (CMR) Respiratory rate (breaths/min) 26 32 has become the key diagnostic modality for patients with Systolic BP (mmHg) 166 173 suspected TTS.[18] TTS is associated with extensive myocardial Diastolic BP ((mmHg) 110 101 oedema (a result of both myocardial inflammation and increased Temperature 38.2oC 38.1oC vascular permeability) as well as the absence of late gadolinium [7,8,18] SpO2 80%; 4 l/min 79%; 6 l/min enhancement without evidence of acute myocardial infarction. Nasal catheter Simple O mask 2 In addition to visualisation of regional wall motion abnormalities, Glasgow Coma Scale 14 14 CMR allows for precise quantification of right and left ventricular Blood glucose (mmol/l) 5.5 7.6 function, exclusion of myocarditis and the assessment of other 12-lead ECG Sinus tachycardia Ventricular bigeminy abnormalities such as pericardial and pleural effusion and thrombi in the left and right ventricle.[18] Performing CMR may MER management 02 10 l/min NRB Transfer to ICU for GTN patch BiPAP ventilation be extremely challenging, particularly because of severe illness, IV frusemide ABG; CXR; ECG obesity and/or claustrophobia, as in the case reported here, but where possible, a patient with suspected TTS should have the BP = blood pressure; SpO2 = peripheral capillary oxygen saturation; ECG = electrocardiogram; NRB = non-rebreather mask; GTN = glyceryl trinitrate; IV = diagnosis established definitively via CMR. intravenous; ABG = arterial blood gas CXR = chest X-ray; BiPAP = bilevel positive airway pressure Significance of Takotsubo syndrome When first described in 1990, TTS was thought to be a relatively in NT-proBNP and BNP concentrations[15] as was seen in the benign condition with complete recovery within a few weeks case we present here. The high ratio of natriuretic peptide to in most cases. However, this concept of TTS being a transient troponin may be useful in discriminating between TTS and disorder is incorrect.[8,19,20] With improved recognition of TTS, acute coronary syndrome.[15] It is likely that the disproportionate it has been found that in-hospital mortality is similar to that rise in NT-proBNP reflects inflammation rather than heart of patients with acute coronary syndrome (4-5%),[2,17] although failure. The performance of NT-proBNP estimation is highly attributable mortality remains unclear. A physical trigger for recommended when evaluating the cause of any cardiac injury TTS (such as acute illness or medical procedures) has been found in the ICU setting. to be an independent predictor for in-hospital complications, with TTS secondary to neurological disease having the worst Coronary angiography with left ventriculography has in the short- and long-term prognosis.[17] Several early complications past been considered to be the gold standard diagnostic tool for may develop from TTS. A syndrome resembling cardiogenic diagnosis of TTS,[3] but this is essentially a diagnosis of exclusion. shock occurs in up to 10% of TTS patients and is associated Coronary angiography in TTS typically shows normal coronary with a significant increase in 28-day mortality compared with arteries or mild to moderate coronary artery disease, while patients without cardiogenic shock.[13] LVOTO, transient ventriculography will show regional wall motion abnormalities mitral regurgitation, and right ventricular involvement may characteristic of TTS.[2] Most cases of TTS involve apical and theoretically contribute to shock,[17] but hypotension and shock mid-ventricular regional wall motion abnormalities (termed in TTS are likely multifactorial and not purely due to impaired apical ballooning), often with hyperkinetic basal segments, cardiac output.[9,21,22] Nguyen et al.[9] have demonstrated that although TTS may occasionally demonstrate basal (inverted Syndecan-1, a component of glycocalyx, is released into plasma TTS), localised, or even global wall motion abnormalities. during the acute stages of TTS. TTS patients often exhibit Approximately one third of patients will have concomitant wall features of pulmonary oedema in the setting of pulmonary motion abnormalities of the right ventricle.[16] capillary wedge pressures that are not substantially elevated. For patients in whom emergency coronary angiography is In these cases, pulmonary oedema is not haemodynamically not a practical option, including critically ill ICU patients, determined and likely results from fluid extravasation into the

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 17 Netherlands Journal of Critical Care Takotsubo syndrome after ICU discharge

pulmonary interstitium due to inflammatory changes resulting TTS resulting from an acute illness (secondary TTS) may be from ‘shedding’ of the endothelial glycocalyx[9] that in many common.[34,35] Estimates of incidence in ICU vary between ways is analogous to glycocalyx shedding in the setting of 1.5%[55] to 28%.[56] Studies using retrospective data report a lower septic shock.[23] Pleural or small pericardial effusions may also incidence and likely missed cases,[1,34] whereas most prospective be seen. Hypotension may result from a combination of fluid studies using TTE to screen all ICU patients report a higher extravasation plus increased vascular sensitivity to the dilator incidence.[34,57-60] A recent prospective single-centre study of 280 effects of nitric oxide.[24,25] For these patients, fluid replacement ICU patients who had TTE, ECG and troponin measurement may be considered, but for those who develop hypoxia, performed on admission, at 24 and 48 hours following ventilatory assistance may be required for the first 2-3 days, admission, at discharge and in the case of clinical deterioration, until glycocalyx shedding ceases, and the overall inflammatory as well as measuring BNP in all patients on admission and in crisis which underlies TTS is attenuated. the case of LV failure, found a TTS incidence of 4.6% (7.6% Intraventricular thrombosis is reported to occur in around among females only).[34] It is uncertain whether there are times 3% of patients in the acute phase of TTS[26] with the attendant when the risk of developing TTS in ICU patients is greater: risk of embolic complications, including stroke. Ventricular for example, at ICU admission, peak severity of the primary tachyarrhythmias (monomorphic and polymorphic ventricular illness, when complications (such as sepsis) occur, during tachycardia and ventricular fibrillation) occur in an estimated catecholamine administration, during invasive procedures, or 10% of patients with TTS and are linked to worsened survival other events. In the case that we have presented, TTS occurred in both acute and convalescent phases of TTS.[27] New onset of following resolution of acute illness and discharge from ICU to atrial arrhythmia (atrial fibrillation and atrial flutter) occurs in general wards suggesting that the risk of developing TTS may up to 25% of patients with TTS and is associated with a worsened not be confined to the most severe phase of illness. short-term[28,29] and long-term prognosis.[28-30] Atrioventricular Chest pain and dyspnoea are the usual presenting symptoms for block has less frequently been described in the setting of TTS TTS, but ICU patients rarely exhibit or communicate symptoms and may lead to torsade de pointes.[31] Cardiac arrest at the time of chest pain or breathlessness that may be associated with of TTS presentation or during the acute phase of the syndrome TTS onset due to intubation, sedation, analgesia, and the is associated with increased in-hospital and long-term mortality primary underlying illness that may obscure these symptoms compared with TTS patients without cardiac arrest.[32] in critically ill patients. The first indications of TTS may be At long-term follow-up in the International Takotsubo the presence of associated complications that include acute Registry (InterTAK), the rate of major adverse cardiac and heart failure, pulmonary oedema, stroke, cardiogenic shock, cerebrovascular events was 9.9% per patient-year and death arrhythmia, or cardiac arrest manifested by symptoms of (all-cause) was 5.6% per patient-year.[13] Longer-term mortality impaired consciousness, neurologic complications, and sudden appears to occur predominantly in the first four years after TTS respiratory or haemodynamic deterioration.[10,35,59] A reasonable diagnosis and is partially related to concomitant non‐cardiac approach to TTS detection includes a baseline ECG and illnesses including malignancy.[2,33] troponin on admission to ICU, and daily ECGs throughout ICU admission to observe for pathognomonic features of TTS. In the Takotsubo syndrome in critically ill patients case of ECG abnormalities or clinical deterioration suggestive Acute illness is a potential trigger for TTS, particularly in of TTS, troponin and BNP should be measured and TTE conditions with high sympathetic activity. TTS has been performed if coronary angiogram or CMR are not possible. If identified in patients with respiratory insufficiency, sepsis, clinicians do not recognise TTS as the cause of the associated neurological disease, major surgery,[1,34,35] and numerous other cardiopulmonary or neurological deterioration, appropriate disease processes. Indeed, neurogenic stress cardiomyopathy, management may be delayed. a form of acute reversible myocardial dysfunction associated ICU patients with diagnosed TTS are more likely to experience with a variety of neurological conditions that include shock, arrhythmias, and require ventilation more frequently than subarachnoid haemorrhage,[36,37] intracerebral haemorrhage,[38] ICU patients with causes of LV dysfunction other than TTS.[35] To acute ischaemic stroke,[39] traumatic brain injury,[40] seizure,[41] date, there does not appear to be convincing evidence of higher transient global amnesia,[42] electroconvulsive therapy,[43] mortality for patients with TTS compared with patients with hydrocephalus,[44,45] encephalitis,[46] multiple sclerosis,[47-50] similar comorbidities without TTS,[1,34,36] but ICU patients with myasthenia gravis[51] and other less common neurological TTS have a more complex course of hospitalisation and incur conditions, has been recognised for decades. There have been higher healthcare costs compared with ICU patients without conflicting views on whether neurogenic stress cardiomyopathy TTS.[1] is in fact TTS,[52] but there now seems to be general acceptance In the absence of evidence from randomised clinical trials to that neurogenic stress cardiomyopathy and TTS are the same support therapeutic strategies in the TTS population, current entity.[53,54] management is based on clinical experience and expert

18 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Takotsubo syndrome after ICU discharge

consensus,[2,17] and consists of supportive care and prevention ICU to a general ward with less nursing presence and monitoring of complications. Treatment algorithms have been proposed, can be psychologically distressing for patients.[65,66] Indeed, post- notably the expert consensus document from the International traumatic stress disorder has been reported in 20-43% of Takotsubo (InterTAK) Registry[17] and the position statement patients after ICU discharge.[67,68] This patient was noted to be from the Heart Failure Association of the European Society of extremely anxious following ICU transfer. It is plausible that Cardiology.[2,17] In the absence of contraindications, angiotensin- psychological stress may have contributed to the development converting enzyme inhibitors or angiotensin receptor blockers of TTS in this patient, and the propensity for psychological are recommended for use in patients with normal or high stress to cause TTS must not be underestimated. systemic vascular resistance and an LVEF <45% and beta The exact timing of TTS onset in this patient is uncertain as blockers in patients with LVEF <45%.[2,17] It may be harmful to TTS was not considered as a cause of the clinical deterioration, administer either catecholamines[61] or organic nitrates during resulting in a delay in the appropriate investigations. An the acute stages of TTS. Inflammatory changes found within echocardiogram was not undertaken until four days after the myocardium in TTS patients appear to stem from abnormal readmission, but the onset of chest discomfort and dyspnoea at post-receptor signalling in response to β-adrenoceptor agonist the time of the MER calls, along with respiratory deterioration, binding[5,62] coupled with activation of nitric oxide synthase may have been indicators of TTS onset. Troponin and NT- and release of the hypotensive autacoid nitric oxide, which proBNP were taken on day 1 after readmission to the ICU may contribute to hypotension directly or via the formation and both were elevated. A definite diagnosis of TTS cannot of peroxynitrite.[25] Given the putative role of catecholamine- be made without information on the status of the coronary induced myocardial damage in the aetiology of TTS, there is arteries and even when the coronary anatomy is known, general consensus that mechanical circulatory support should be differentiation is difficult between post-ischaemic myocardial considered in preference to these agents.[52-54,56] An intra-aortic stunning or aborted myocardial infarction. Some of the clinical balloon pump is the most commonly used device, although it findings, such as acute heart failure and troponin and NT- is not recommended in patients with LVOTO.[2,17] There are proBNP elevation, could have resulted from ischaemia, but in limited reports of veno-arterial extracorporeal membrane the context of the pathognomonic patterns of LV dysfunction oxygenation[63] and Impella mechanical circulatory support[64] found on echocardiography and the ECG as well as temporal being used successfully in TTS but these are expensive and not resolution of LV function and echocardiographic abnormalities, always available. The Ca2þ-sensitiser levosimendan has been used TTS is the most likely diagnosis. safely and effectively as an alternative inotrope to catecholamine agents in TTS,[60,61] but is not recommended for patients with Conclusion LVOTO.[17] The role of prophylactic anticoagulation in higher Increasing reports of TTS in the ICU suggest that it may not be risk TTS cases has not yet been determined but given the risk uncommon in critically ill patients, particularly in conditions of stroke or arterial embolism may be considered in those at associated with high sympathetic activity. TTS is associated higher risk of developing thrombus in akinetic apical TTS.[2,17] with serious complications including arrhythmias, shock and emboli. ICU patients with TTS have a more complex course of Discussion hospitalisation and incur higher healthcare costs compared with The case study presented here illustrates some of the challenges those without TTS. Recognising TTS in ICU patients can be in recognising and managing a patient with TTS in the ICU. The challenging. Clinicians should retain a high index of suspicion of patient in this case experienced an index ICU admission that TTS if patients show unexpected haemodynamic or respiratory was prolonged and complicated by multiple medical problems deterioration. Increased awareness and understanding of TTS including sepsis, respiratory failure, anaemia, acute renal failure and depth of knowledge of this condition may reduce early and functional bowel obstruction. She was readmitted to the ICU diagnostic errors, minimise delays in onset of appropriate two days later following two MER calls due to acute respiratory treatment and prevent further complications. TTS may not be deterioration associated with chest pain, hypertension and limited to the ICU admission and risk following ICU discharge. tachycardia, and subsequently diagnosed with TTS. The patient in this case was exposed to emotional and physical Disclosures stressors associated with critical illness, including invasive All authors declare no conflicts of interest. No funding or procedures and the use of catecholamines in ICU. It is not clear financial support was received. why TTS developed following transfer from the ICU when the This study was approved by the Royal Adelaide Hospital Human patient was purportedly in the process of recovery rather than Research Ethics Committee and was performed in accordance during the ICU admission when physical stresses were likely with the ethical standards. HREC reference number: HREC/16/ greater. Discharge from the ICU represents a positive step in terms RAH/302; CALHN Reference number: R20160719; SSA/17/ of physical recovery, but the transition from high acuity care in the RAH/37.

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 19 Netherlands Journal of Critical Care Takotsubo syndrome after ICU discharge

References 28. Möller C, Eitel C, Thiele H, Eitel I, Stiermaier T. Ventricular arrhythmias in patients with Takotsubo syndrome. J Arrhythmia. 2018; 34:369-75. 29. El-Battrawy I, Lang S, Ansari U, et al. Impact of concomitant atrial fibrillation on 1. Oras J, Lundgren J, Redfors B, et al. Takotsubo syndrome in hemodynamically the prognosis of Takotsubo cardiomyopathy. Europace. 2017;19:1288-92. unstable patients admitted to the intensive care unit–a retrospective study. Acta 30. Berthon C, Jesel L. Atrial arrhythmias in Takotsubo cardiomyopathy: Incidence, Anaesthesiol Scand. 2017; 61:914-24. predictive factors and prognosis. Arch Cardiovasc Dis Suppl. 2018;10:100. 2. Lyon AR, Bossone E, Schneider B, et al. Current state of knowledge on Takotsubo 31. Oshima T, Ikutomi M, Ishiwata J, et al. Takotsubo cardiomyopathy associated syndrome: a Position Statement from the Taskforce on Takotsubo Syndrome of with complete atrioventricular block and torsades de pointes. Int J Cardiol. the Heart Failure Association of the European Society of Cardiology. Eur J Heart 2015;181:357-9. Fail. 2016;18:8-27. 32. Gili S, Cammann VL, Schlossbauer SA, et al. Cardiac arrest in takotsubo syndrome: 3. Ghadri JR, Wittstein IS, Prasad A, et al. International expert consensus document results from the InterTAK Registry. Eur Heart J. 2019;40:2142-51. on Takotsubo syndrome (part I): clinical characteristics, diagnostic criteria, and 33. Nguyen TH, Stansborough J, Ong GJ, Surikow S, Price TJ, Horowitz JD. Antecedent pathophysiology. Eur Heart J. 2018;39:2032-46. cancer in Takotsubo syndrome predicts both cardiovascular and long-term 4. Ong GJ, Nguyen TH, Stansborough J, Surikow SY, Horowitz JD. Incremental mortality. Cardiooncology. 2019;5:20. ‘Therapeutic’ Myocardial Exposure to Catecholamines: Incidence and Impact in 34. Doyen D, Moschietto S, Squara F, et al. Incidence, clinical features and outcome Takotsubo Syndrome. Cardiovasc Drugs Ther. 2020;34:95-100. of Takotsubo syndrome in the intensive care unit. Arch Cardiovasc Dis. 5. Lyon AR, Rees PS, Prasad S, Poole-Wilson PA, Harding SE. Stress (Takotsubo) 2020;113:176-88. cardiomyopathy—a novel pathophysiological hypothesis to explain 35. Muratsu A, Muroya T, Kuwagata Y. Takotsubo cardiomyopathy in the intensive catecholamine-induced acute myocardial stunning. Nat Clin Pract Cardiovasc care unit. Acute Med Surg. 2019;6:152-7. Med. 2008;5:22-9. 36. Milaković B, Nastasović T, Lepić M, et al. Takotsubo cardiomyopathy in aneurysmal 6. Núñez-Gil IJ, Almendro-Delia M, Andrés M, et al. Secondary forms of Takotsubo subarachnoid hemorrhage: A case report. Vojnosanitetski Pregled. 2019;76:1087-93. cardiomyopathy: a whole different prognosis. Eur Heart J. Acute Cardiovasc Care. 37. Piliponis L, Neverauskaitė-Piliponienė G, Kazlauskaitė M, et al. Neurogenic stress 2016;5:308-16. cardiomyopathy following aneurysmal subarachnoid haemorrhage: a literature 7. Neil C, Nguyen TH, Kucia A, et al. Slowly resolving global myocardial inflammation/ review. Sem Cardiovasc Med. 2019;25:44-52. oedema in Tako-Tsubo cardiomyopathy: evidence from T2-weighted cardiac MRI. 38. Shiromoto T, Shibazaki K, Kimura K, Sakai K. Case of cerebellar hemorrhage Heart. 2012;98:1278-84. complicated with takotsubo cardiomyopathy-usefulness of plasma brain 8. Scally C, Rudd A, Mezincescu A, et al. Persistent long-term structural, functional, natriuretic peptide measurement for the diagnosis. Rinsho Shinkeigaku. and metabolic changes after stress-induced (Takotsubo) cardiomyopathy. 2012;52:778-81. Circulation. 2018;137:1039-48. 39. Jung JM, Kim JG, Kim JB, et al. Takotsubo-like myocardial dysfunction in ischemic 9. Nguyen TH, Liu S, Ong GJ, Stafford I, Frenneaux MP, Horowitz JD. Glycocalyx stroke: a hospital-based registry and systematic literature review. Stroke. shedding is markedly increased during the acute phase of Takotsubo 2016;47:2729-36. cardiomyopathy. Int J Cardiol. 2017;243:296-9. 40. Cheah CF, Kofler M, Schiefecker AJ, et al. Takotsubo cardiomyopathy in traumatic 10. Chockalingam A. Stress cardiomyopathy of the critically ill: Spectrum of secondary, brain injury. Neurocrit Care. 2017;26:284-91. global, probable and subclinical forms. Indian Heart J. 2018 ;70:177-84. 41. Stöllberger C, Wegner C, Finsterer J. Seizure associated Takotsubo cardiomyopathy. 11. Izumo M, Akashi YJ. Role of echocardiography for takotsubo cardiomyopathy: Epilepsia. 2011;52:e160-7. clinical and prognostic implications. Cardiovascular Diagn Ther. 2018 ;8:90. 42. Finsterer J, Stöllberger C. Transient global amnesia: The cerebral Takotsubo?. 12. Frangieh AH, Obeid S, Ghadri JR, et al., on behalf of the InterTAK Collaborators. Journal of the Neurological Sciences. 2017;376:196-7. ECG Criteria to Differentiate Between Takotsubo (Stress) Cardiomyopathy and 43. Narayanan A, Russell MD, Sundararaman S, Shankar KK, Artman B. Takotsubo Myocardial Infarction. J Am Heart Assoc. 2016;5:1-13. cardiomyopathy following electroconvulsive therapy: an increasingly recognised 13. Templin C, Ghadri JR, Diekmann J, et al. Clinical features and outcomes of phenomenon. Case Reports. 2014 Nov 25;2014:bcr2014206816. takotsubo (stress) cardiomyopathy. New Engl J Med. 2015;373:929-38. 44. Johnson J, Ragheb J, Garg R, Patten W, Sandberg DI, Bhatia S. Neurogenic 14. Pant S, Deshmukh A, Mehta K, et al. Burden of arrhythmias in patients with stunned myocardium after acute hydrocephalus: report of 2 cases. J Neurosurg Takotsubo cardiomyopathy (apical ballooning syndrome). Int J Cardiol. Pediatr. 2010;5:428-33. 2013;170:64-8. 45. Chiriboga AL, Yoon JW, Freeman WD, Odunukan OW, Cheshire WP. Takotsubo 15. Nguyen TH, Neil CJ, Sverdlov AL, et al. N-terminal pro-brain natriuretic protein cardiomyopathy in the setting of acute hydrocephalus secondary to levels in takotsubo cardiomyopathy. Am J Cardiol. 2011;108:1316-21. neurocysticercosis. Clinical Auton Res. 2016;26:235-41. 16. Citro R, Bossone E, Parodi G, et al. Clinical profile and in-hospital outcome of 46. Venkatraman A, Bajaj NS, Khawaja A, Meador W. Cardiogenic shock from atypical Caucasian patients with takotsubo syndrome and right ventricular involvement. Takotsubo cardiomyopathy attributed to acute disseminated encephalomyelitis Int J Cardiol. 2016;219:455-61. lesion involving the medulla. Clin Auton Res. 2016;26:149-51. 17. Ghadri JR, Wittstein IS, Prasad A, et al. International Expert Consensus Document 47. Biesbroek PS, Nijveldt R, de Jong BA, van Rossum AC, van Loon RB. Reversed on Takotsubo Syndrome (Part II): Diagnostic Workup, Outcome, and Management. takotsubo cardiomyopathy in a patient with new-onset multiple sclerosis: Cause, Eur Heart J. 2018;39:2047-62. link, or just coincidence? Int J Cardiol. 2016;207:217-9. 18. Gunasekara MY, Mezincescu AM, Dawson DK. An update on cardiac magnetic 48. Prestipino E, Squitieri M, Razzolini L, Pastò L, Forleo P, Amato MP. A case of resonance imaging in Takotsubo cardiomyopathy. Curr Cardiovasc Imaging Rep. Takotsubo syndrome during a multiple sclerosis brainstem relapse. Mult Scler 2020;13: https://doi.org/10.1007/s12410-020-09536-0. Relat Disord. 2018;24:1-2. 19. Schwarz K, Ahearn T, Srinivasan J, et al. Alterations in cardiac deformation, timing 49. Kozu K, Suzuki H, Nishiyama S, et al. Multiple sclerosis lesion in the medulla of contraction and relaxation, and early myocardial fibrosis accompany the oblongata in a patient with takotsubo cardiomyopathy. Int J Cardiol. apparent recovery of acute stress-induced (takotsubo) cardiomyopathy: an end 2016;222:980-1. to the concept of transience. J Am Soc Echocardiography. 2017;30:745-55. 50. Peller M, Balsam P, Budnik M, Marchel M, Opolski G. (2016). Reverse Takotsubo 20. Nowak R, Fijalkowska M, Gilis-Malinowska N, et al. Left ventricular function after syndrome in a patient with diagnosed multiple sclerosis. Kardiologia Polska. takotsubo is not fully recovered in long-term follow-up: A speckle tracking 2016;74:1029. echocardiography study. Cardiol J. 2017;24:57-64. 51. Shivamurthy P, Parker MW. Cardiac manifestations of myasthenia gravis: a 21. Stiermaier T, Santoro F, Eitel C, et al. Prevalence and prognostic relevance of atrial systematic review. IJC Metab Endocr. 2014;5:3-6. fibrillation in patients with Takotsubo syndrome. Int J Cardiol. 2017;245:156-61. 52. Tavazzi G, Zanierato M, Via G, Iotti GA, Procaccio F. Are Neurogenic Stress 22. Chong CR, Neil CJ, Nguyen TH, et al. Dissociation between severity of takotsubo Cardiomyopathy and Takotsubo Different Syndromes With Common Pathways?: cardiomyopathy and presentation with shock or hypotension. Clin Cardiol. Etiopathological Insights on Dysfunctional Hearts. JACC Heart Failure. 2017;5:940-2. 2013;36:401-6. 53. Guglin M, Novotorova I. Neurogenic stunned myocardium and takotsubo 23. Singh K, Neil CJ, Nguyen TH, et al. Dissociation of early shock in takotsubo cardiomyopathy are the same syndrome: a pooled analysis. Congest Heart Fail. cardiomyopathy from either right or left ventricular systolic dysfunction. Heart, 2011;17:127-32. Lung Circ. 2014;23:1141-8. 54. Win CM, Pathak A, Guglin M. Not Takotsubo: A Different Form of Stress-Induced 24. Uchimido R, Schmidt EP, Shapiro NI. The glycocalyx: a novel diagnostic and Cardiomyopathy—A Case Series. Congest Heart Fail. 2011;17:38-41. therapeutic target in sepsis. Crit Care. 2019;23:16. 55. Champion S, Belcour D, Vandroux D, et al. Stress (Tako-tsubo) cardiomyopathy in 25. Nguyen TH, Neil CJ, Sverdlov AL, et al. Enhanced NO signaling in patients with critically-ill patients. Eur Heart J Acute Cardiovasc Care. 2015;4:189-96. Takotsubo cardiomyopathy: short-term pain, long-term gain? Cardiovasc Drugs 56. Park JH, Kang SJ, Song JK, et al. Left ventricular apical ballooning due to severe Ther. 2013;27:541-7. physical stress in patients admitted to the medical ICU. Chest. 2005;128:296-302. 26. Surikow SY, Nguyen TH, Stafford I, et al. Nitrosative stress as a modulator of 57. Cinotti R, Piriou N, Launey Y, et al. Speckle tracking analysis allows sensitive inflammatory change in a model of Takotsubo syndrome. JACC Basic Transl Sci. detection of stress cardiomyopathy in severe aneurysmal subarachnoid 2018;3:213-26. hemorrhage patients. Intensive Care Med. 2016;42:173-82. 27. Ding KJ, Cammann VL, Szawan KA, et al. Intraventricular Thrombus Formation and 58. Yang HS, Kim HJ, Shim HJ, et al. Soluble ST2 and troponin I combination: useful Embolism in Takotsubo Syndrome: Insights From the International Takotsubo biomarker for predicting development of stress cardiomyopathy in patients Registry. Arter Thromb Vasc Biol. 2020;40:279-87. admitted to the medical intensive care unit. Heart Lung. 2015;44:282-8.

20 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Takotsubo syndrome after ICU discharge

59. Pelliccia F, Kaski JC, Crea F, Camici PG. Pathophysiology of Takotsubo syndrome. 64. Napp LC, Møller J, Pappalardo F, et al. TCT-636 International experience with Circulation. 2017;135:2426-41. Impella mechanical circulatory support for Takotsubo syndrome with shock. J 60. Singh K, Akashi YJ, Horowitz J. Takotsubo syndrome therapy: current status and Am Coll Cardiol. 2018;72(13 Suppl):B254. future directions. International Cardiovasc Forum J. 2016; 5:23-9. 65. Chaboyer W, Thalib L, Alcorn K, Foster M. The effect of an ICU liaison nurse on 61. Padayachee L. Levosimendan: the inotrope of choice in cardiogenic shock patients and family's anxiety prior to transfer to the ward: an intervention study. secondary to takotsubo cardiomyopathy. Heart Lung Circ. 2007;16:S65-70. Intensive Crit Care Nurs. 2007 ;23:362-9. 62. Paur H, Wright PT, Sikkel MB, et al. High Levels of Circulating Epinephrine Trigger 66. Forsberg A, Lindgren E, Engström Å. Being transferred from an intensive care unit to Apical Cardiodepression in a β2-Adrenergic Receptor/Gi– Dependent Manner: A a ward: Searching for the known in the unknown. Int J Nurs Pract. 2011;17:110-6. New Model of Takotsubo Cardiomyopathy. Circulation. 2012;126:697-706. 67. Parker AM, Sricharoenchai T, Raparla S, Schneck KW, Bienvenu OJ, Needham DM. 63. Rojas-Marte G, John J, Sadiq A, Moskovits N, Saunders P, Shani J. Medication- Posttraumatic stress disorder in critical illness survivors: a meta-analysis. Crit Care induced Takotsubo Cardiomyopathy presenting with cardiogenic shock—utility Med. 2015;43:1121-9. of extracorporeal membrane oxygenation (ECMO): case report and review of the 68. Ramnarain D, Gnirrip I, Schapendonk W, Rutten A, Van der Nat G, Van der Lely N. literature. Cardiovasc Revasc Med. 2015;16:47-51. Post-traumatic stress disorder after ICU discharge: results of a post-ICU aftercare program. Crit Care. 2015;19:553.

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 21 Netherlands Journal of Critical Care Submitted June 2020; Accepted October 2020

ORIGINAL ARTICLE

Validation of the Nursing Activities Score (NAS) using time- and-motion measurements in Dutch intensive care units

C.C. Margadant1,5, M.E. Hoogendoorn2,5, R.J. Bosman3,5, J.J. Spijkstra4,5, S. Brinkman1,5, N.F. de Keizer1,5 1Department of Medical Informatics, Amsterdam Public Health Research Institute, Amsterdam UMC, Vrije Universiteit Amsterdam, Amsterdam, the Netherlands 2Department of Anaesthesiology and Intensive Care, Isala, Zwolle, the Netherlands 3Department of Intensive Care, OLVG, Amsterdam, the Netherlands 4Department of Intensive Care, Amsterdam UMC, Vrije Universiteit Amsterdam, Amsterdam, the Netherlands 5National Intensive Care Evaluation (NICE) Foundation, Amsterdam, the Netherlands

Correspondence C.C. Margadant - [email protected]

Keywords - Nursing Activities Score (NAS), time-and-motion techniques, validation, nursing workload, patient acuity

Abstract Introduction Background The Nursing Activities Score (NAS) is widely used There are concerns regarding excessively high nursing workload, for measuring the workload of intensive care unit (ICU) nurses. both in general and ICU wards.[1] An excessively high nursing However, the performance of the NAS to measure actual workload can lead to burnout and job dissatisfaction among nursing time has not been comprehensively and externally nurses[2] and have a deleterious effect on patients.[3] Workload validated. The aim of this study is to validate the NAS using has risen due to an increased turnover of patients, increased time-and-motion measurements in Dutch ICUs. complexity of patients, together with nursing shortages.[4] All Methods We measured nursing time for patients admitted this makes planning of nursing capacity important. In the last to seven Dutch ICUs, between November 2016 and October 30 years different instruments have been developed to measure 2017. The patient(s) that were under the care of a chosen the nursing workload to give insight into the nursing staff nurse were followed by the observers during the entire needed per shift and provide much needed input for capacity shift and measurements were performed using an in-house planning.[5] developed web application. To validate the reliability of To assess nursing workload in the ICU, Cullen et al.[6] created the NAS, we first converted NAS points per activity into the Therapeutic Intervention Scoring System (TISS). The TISS minutes. Next, we compared the converted time per NAS was originally developed to classify nursing workload in relation item and the converted total nursing time per patient with to the severity of illness of ICU patients. The TISS consists of the actual observed time. We used Wilcoxon signed-rank 76 therapeutic interventions that receive 1-4 points based on tests at nursing activity level and Pearson’s R and R2 at patient the severity of illness. It appeared that nursing workload is only level for these comparisons. partly related to severity of illness, since less severely ill patients Results A Pearson’s correlation of R=0.59 (R2=0.35) was found could also generate a high nursing workload. For instance, between the total converted NAS time and the total observed a patient recovering from a serious illness with agitated time per patient. The median converted NAS time per patient delirium would not score high in severity of illness, but could (202.6 minutes) was higher compared with the observed time demand very intensive nursing care, up to continuous bedside per patient (114.3 minutes). At NAS item level, we found care throughout the day. This made the TISS less effective in significant differences between the converted NAS time and the assessing nursing workload. Therefore, the Nursing Activities observed time for all separate NAS items. Score (NAS) was developed by Miranda et al. in 2003.[7] The Conclusions The NAS overestimates the nursing time needed NAS describes activities that largely represent the work actually for patients in Dutch ICUs. Therefore, we advise revisions of performed by nurses at the bedside in caring for patients and was the time weighting assigned to each NAS item to obtain better developed to measure the nursing workload for each individual insight into the true nursing workload so that this information patient. The points assigned to nursing activities provide an can be used for more effective nursing capacity planning. average time consumption in caring for the patients instead

22 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Validation of the NAS Score

of representing the severity of illness. The NAS was created Time-and-motion by using the work-sampling approach: at random moments The study involved time-and-motion measurements for per shift the nurse was asked what he or she was doing at that patients admitted to the ICU. We included different types of specific moment. The researchers applied a weighting for each hospitals (academic, teaching and non-teaching hospitals) and activity. The total NAS for an individual patient is the sum of different shifts (day, evening and night). We performed an equal NAS points for all activities, varying between 0 to 177 points. A number of measurements in all types of hospitals and shifts. At score of 100 NAS points is equivalent to the amount of care that the start of a shift, one nurse was chosen by the observer. The can be provided by one full-time equivalent nurse during either patient(s) that were under the responsibility of this nurse were one shift or one day. A score above 100 points indicates that the followed by the observer during the entire shift. A longer-term care needed can only be provided by more than one nurse.[7] patient could theoretically be observed on different dates during The NAS is considered a valuable tool and is widely used for different shifts and therefore could possibly be followed during workload measurement in ICUs.[8,9] However, the performance more than one measured shift. The measurements took place of the NAS has not been comprehensively validated. One on different days of ICU admission (e.g. first ICU admission day study showed that the NAS might either underestimate or through to last ICU admission day) and with different types of overestimate the actual nursing time required by patients and nurses (registered and student nurses). We randomly selected therefore recommended revision of the original NAS because nurses who took care of patients that were expected to stay of inadequate measurement of nursing activities.[4] The study by during the whole shift in order to measure as many nursing Stafseth et al. suggests that the reliability and validity of the NAS activities as possible. are good. However, this study strongly suggests more research Observers were researchers CM and MH and ten student nurses. in other countries and larger groups of patients.[10] The students were trained in performing time-and-motion measurements by oral and written instructions and one day of Furthermore, research has demonstrated that the work- measuring together with one of the researchers. The observers sampling approach, as used for the development of the NAS, used an in-house developed web application to record start and does not lead to an accurate representation of the true nursing stop times of each performed nursing activity. The application workload. This is due to the fact that the weighting of nursing included all activities occurring in the NAS (table 1). If two activities is based on the probability that a particular nursing nurses were simultaneously performing nursing activities for activity occurred.[11] The total amount of time in a shift is divided the same patient, this was also registered by pressing the ‘two over the nursing activities that were carried out. When nursing nurses button’ and multiplying this time by two in the analysis. activities frequently occur or take a lot of time, they would When two different activities were carried out by two nurses, also occur more frequently in the work-sampling approach. these activities could be measured simultaneously. However, this approach will not lead to precise measurements, Measurements were conducted between 1 November 2016 but will only approximate the time of the different activities. and 1 October 2017. Participation of the hospitals was on a Thus, in contrast to time-and-motion techniques in which voluntary basis. Seven hospitals were willing to participate. every minute of a nursing shift is measured, the work-sampling Data were processed anonymously. approach does not measure the real amount of time spent on nursing activities, which could lead to less accurate results.[12] Ethical approval Therefore, the time-and-motion technique is considered the The Institutional Research Board of the Amsterdam University best technique for time measurement.[13] Medical Centre reviewed the research proposal and waived the The aim of this study is to validate the NAS in the Dutch ICU need for informed consent (IRB protocol W17_366). setting using the time-and-motion technique, and to identify which nursing activities are underestimated or overestimated in the NAS. Data analysis Nursing activities that occurred less than ten times in the total Methods dataset were excluded from the analysis. Most NAS items have Setting a fixed number of NAS points but some items have different We conducted an observational study. All 82 Dutch ICUs categories corresponding to different numbers of NAS points participate in the National Intensive Care Evaluation (NICE) depending on the duration of that activity (e.g. bedside with quality registry. Fifteen of these ICUs are participating in the hourly vital signs, bedside for two hours or more, or four newly implemented voluntary nursing capacity module[14], hours or more). For these duration-dependent activities, we seven of which voluntarily took part in this study. Data on first used the measured time for that activity to assign the characteristics of the ICUs (such as number of ICU beds) and correct number of points. For example, a nurse performed data on patient characteristics (such as age, BMI, admission hygiene procedures on a patient for 1.2 hours during a shift, type, and mortality) were extracted from the NICE registry. according to our time measurements. This NAS item has three

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 23 Netherlands Journal of Critical Care Validation of the NAS Score

Table 1. NAS activities with their points according to Miranda et al. (2003),[7] and the median converted NAS times and observed times per NAS item

NAS item NAS points Median Median observed Difference in minutes, per activity converted NAS time (minutes) median [IQR] time (minutes) [IQR]

1a Present at bedside and continuous observation or active for <2 hours 4.5 21.6 14.22 [7.26-26.17] 7.38 [-4.57-14.35]* 1b Present at bedside and continuous observation or active for ≥2 hours 12.1 NA NA NA 1c Present at bedside and continuous observation or active for ≥4 hours 19.6 NA NA NA 2 Laboratory, biochemical and microbiological investigations 4.3 20.64 5.45 [3.13-8.81] 15.19 [11.83-17.51]* 3 Medication, vasoactive drugs excluded 5.6 26.88 2.24 [0.90-4.91] 24.64 [21.97-25.98]* 4a Performing hygiene procedures ≤2 hours 4.1 19.68 11.58 [3.95-27.8] 8.1 [-8.12-15.73]* 4b Performing hygiene procedures >2 hours 16.5 NA NA NA 4c Performing hygiene procedures >4 hours 20.0 NA NA NA 5 Care of drains 1.8 8.64 2.41 [0.92-4.64] 6.23 [4.0-7.72]* 6a Mobilisation and positioning, performing procedure(s) up to 3 times per 24 hours 5.5 26.4 2.46 [0.91-4.88] 23.94 [21.52-25.49]* 6b Mobilisation and positioning, performing procedure(s) >3 times per 24 hours, or 12.4 59.52 4.82 [2.17-9.33] 54.69 [50.19-59.49]* with two nurses 6c Mobilisation and positioning, performing procedure(s) with 3 nurses 17.0 81.6 2.4 [0.89-6.16] 79.2 [75.44-80.71]* 7a Support or care of patient or relatives for about 1 hour 4.0 19.2 2.4 [0.89-6.16] 16.8 [13.2- 22.89]* 7b Support or care of patient or relatives for about 3 hours 32.0 NA NA NA 8a Administrative or managerial tasks for <2 hours 4.2 20.16 40.91 [28.53-60.33] -20.74 [-40.17- -8.37]* 8b Administrative or managerial tasks for about 2 hours 23.2 111.4 130.0 [126.3-157.4] -18.67 [-46.02- -14.92]* 8c Administrative or managerial tasks for about 4 hours 30.0 NA NA NA 9 Respiratory support 1.4 6.72 2.99 [1.42-5.9] 3.73 [0.82-5.30]* 10 Care of artificial airways 1.8 8.64 1.43 [0.5-4.77] 7.21 [3.87-8.14]* 11 Treatment for improving lung function 4.4 21.12 1.32 [0.64-2.79] 19.80 [18.33-20.48]* 12 Vasoactive medication 1.2 5.76 1.99 [0.95-4.99] 3.78 [-0.77-4.81]* 13 Intravenous replacement of large fluid losses 2.5 NA NA NA 14 Left atrial monitoring 1.7 NA NA NA 15 Cardiopulmonary resuscitation after arrest 7.1 NA NA NA 16 Haemofiltration techniques 7.7 36.96 18.76 [7.83-36.66] 18.20 [-1.67-28.78]* 17 Qualitative urine output measurement 7.0 33.6 1.35 [0.66-2.45] 32.25 [31.15-32.96]* 18 Measurement of intracranial pressure 1.6 7.68 0.91 [0.28-2.62] 6.77 [5.07-7.4]* 19 Treatment of complicated metabolic acidosis 1.3 NA NA NA 20 Intravenous hyperalimentation 2.8 13.44 2.64 [0.79-4.1] 10.80 [9.41-12.65]* 21 Enteral feeding through gastric tube 1.3 6.24 1.87 [0.81-4.64] 4.37 [1.6-5.43]* 22 Specific interventions in the ICU 2.8 NA NA NA 23 Specific interventions outside the ICU 1.9 9.12 18.18 [5.69-27.46] -9.06 [-18.34- 3.43]* Total per patient - 202.56 98.52 84.7 [155.04-241.2] [71.86-127.72] [50.31-127.72]*

N = 371 patients and 46,319 measured nursing activities. * Indicates a significant P-value of <0.05 (Wilcoxon signed-rank test); NA: not measured during measurement. This is a shortened version of the NAS; the full version can be found on [link] categories: performing hygiene procedures for less than two one NAS point corresponds to 3.84 minutes of nursing care hours, for more than two hours, or for more than four hours. during an 8-hour shift ((8 hours * 60 mins)/100)*0.8).[7,15] With In the above-mentioned example, the activity took 1.2 hours this information, we were able to convert the NAS scores into and would therefore be assigned to the category for less than an estimated nursing time per patient and per nursing activity two hours, which corresponds to 4.1 NAS points. To validate (from now on referred to as ‘converted NAS time’. Next we the NAS, we first converted the originally assigned NAS points compared the time per NAS item and the total nursing time per activity into time. Based on Miranda et al.[7] 100 NAS points per patient, based on NAS scores according to the model, with correspond to 100% of care time provided by one nurse during observed times from the time-and-motion measurements. For a shift and hence 1 NAS point corresponds to 1% of care time the observed time, we took the sum of the times of all performed provided by one nurse. As stated by the author of the NAS, a nursing activities per patient per shift in minutes (from now on nurse is productive in 80% of the 8-hour shift; this means that referred to as ‘observed time’.

24 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Validation of the NAS Score

The median and interquartile ranges (IQR) of the converted NAS times and the observed times were calculated. First, the difference between the total converted NAS times and the total observed times per patient were visualised by scatterplots. Second, the correlation between the total converted NAS times and the total observed times per patient were assessed with the Pearson’s correlation test. In addition, we also assessed the R2, a measure for the proportion of the variance. For each nursing activity separately, medians and interquartile ranges (IQR) of the converted NAS times and observed times were calculated and differences were tested with the Wilcoxon signed-rank test. All statistical analyses were performed using R statistical software, version 3.3.2.[16]

Results Baseline results Table 2 shows the ICU characteristics of the seven included ICUs compared with all Dutch ICUs; no significant differences were found between the included ICUs and all Dutch ICUs. During Figure 1. The correlation between the total converted NAS time in minutes and our study, a total of 287 unique patients were observed during the total observed time in minutes per patient. A full nursing shift is 480 minutes. 371 different shifts with time-and-motion measurements. In Blue diagonal shows equal converted and observed time per patient these patients, 46,319 nursing activities were measured. In 45% of the measurements, nurses took care of two patients losses and treatment of metabolic acidosis/alkalosis, since these per shift. In 15% nurses took care of three patients per shift. two nursing activities usually fall under the category bedside For the remaining 40%, nurses cared for one patient per shift. activities. The patients in our study had a significantly higher in-hospital mortality rate (22.3% versus 13.0%) and length of ICU stay (3.2 Total patient time and times per NAS item days versus 1.0 day) compared with all Dutch patients in the The median converted NAS time per patient (202.6 minutes; same period (table 3). Furthermore, acute renal failure, chronic IQR 155.0-241.2 minutes) was significantly higher (p<0.001) respiratory insufficiency, and cirrhosis differed between the compared with the observed time per patient (144.3 minutes; groups, with a higher percentage in the patients in our study. IQR 81.3 – 168.4 minutes), see figure 1. A Pearson’s correlation For the other patient characteristics, the included patients and of R=0.59 (R2=0.35) was found between the total converted all Dutch ICU patients in this period were comparable. NAS time and the total observed time per patient (table 1). Table 2. ICU characteristics At the NAS item level, we found significant differences between Variable Included ICUs All Dutch ICUs the converted NAS times and observed times for all items. These (n=7) (n=84) differences ranged from -54.6 minutes (support or care of patient Number of university hospitals (%) 1 (14%) 9 (11%) or relatives for about 1 hour) to 79.2 minutes (mobilisation and Number of teaching hospitals (%) 4 (57%) 23 (27%) positioning with three nurses). For most (86%) nursing activities Number of non-teaching hospitals (%) 2 (29%) 52 (62%) the median converted NAS overestimated the observed time. Median number of ICU beds per ICU (IQR) 13.0 [9.0, 17.0] 12.0 [8.0, 16.0] For four activities (support or care of patient for about 1 hour, administrative tasks for less than 2 hours, administrative tasks NAS validation for about 2 hours and specific interventions outside the ICU) Excluded nursing activities the converted NAS underestimated the observed time (table 1). The following three NAS nursing activities occurred less than 10 times in all measurements and were therefore excluded Discussion from the analysis at activity level: care of a pulmonary or left Our analysis showed that the NAS overestimates the nursing atrial catheter, cardiopulmonary resuscitation and specific time needed for patients in the Dutch ICU setting. Times of interventions in the ICU (endotracheal intubation, insertion most NAS items were overestimated by the NAS, except for of a pacemaker, cardioversion, endoscopy, emergency surgery four activities (support or care of patient for about 1 hour, in the previous 24 hours, gastric lavage). Furthermore, we did administrative tasks for less than 2 hours, administrative tasks not specifically measure intravenous replacement of large fluid for about 2 hours, and specific interventions outside the ICU),

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 25 Netherlands Journal of Critical Care Validation of the NAS Score

Table 3. Patient characteristics we used in this study to calculate the converted time per NAS point. Using this approximation, the converted time would have Variable Included patients All Dutch ICU in measurements patients changed from 3.84 to 3.62 minutes per NAS point. This change Number of unique patients, N 287 100.145 does not affect the results and we therefore conclude that non- Age, median [IQR] 66.0 [56.0-76.0] 66.0 [55.0-75.0] nursing duties do not significantly influence the performance BMI, median [IQR] 26.0 [23.6-28.7] 25.9 [23.1-28.4] of the NAS. Admission type A strength of our study is that we validated the NAS with -Medical, N (%) 121 (42.2) 51,290 (52.7) time-and-motion measurements, which is considered to be [13] -Surgical: urgent and elective, N (%) 151 (52.6) 45,905 (47.2) the best technique for measuring nursing workload. To our In-hospital mortality, N (%)* 85 (22.3) 13,017 (13.0) knowledge, this has not been performed before in the context ICU LOS (in days), median [IQR]* 3.2 [0.9, 14.8] 1 [0.7-4.0] of NAS validation. Measurements for nursing activities by using Comorbidities time-and-motion measurements are more accurate compared Acute renal failure, N (%)* 37 (12.9) 9211 (9.2) with the work-sampling approach, as used for the development [24] Cardiovascular insufficiency, N (%) 16 (4.2) 4257 (4.3) of the NAS. Furthermore, since measurements took place Chronic renal failure1, N (%) 25 (6.7) 7976 (7.9) in all types of ICUs, we believe that results of this study are Chronic respiratory insufficiency, N (%)* 7 (2.4) 4620 (4.6) generalisable to all Dutch ICUs. Cirrhosis, N (%)* 1 (3.5) 1751 (1.7) One of the limitations of our study is the fact that we excluded COPD, N (%) 36 (12.5) 13,304 (13.3) NAS activities because they did not occur or occurred less than Diabetes, N (%) 68 (17.8) 16,273 (16.2) ten times. Two of these activities are usually scored in other Gastrointestinal bleeding, N (%) 2 (0.7) 2263 (2.3) categories: the activity ‘intravenous replacement of large fluid Haematological malignancy, N (%) 6 (2.1) 2143 (2.1) losses’ is mostly scored under NAS item 1 ‘bedside’. The activity Immunological insufficiency, N (%) 16 (5.6) 8290 (8.3) ‘treatment of complicated metabolic acidosis/alkalosis’ is mostly Neoplasm, N (%) 9 (3.1) 4506 (4.5) scored in NAS item 3 ‘medication’. Since these activities could be scored in other categories, we did not include them in our study. * Indicates a significant P value of <0.05. COPD = chronic obstructive pulmonary disease; IQR = interquartile range; LOS length of stay; 1 Chronic renal failure consists of Three NAS activities (left atrial monitoring, cardiopulmonary chronic renal insufficiency and chronic dialysis resuscitation after arrest, and specific interventions in the ICU, respectively) and six subcategories 1b, 1c, 4b, 4c, 7b, and 8c where the NAS gives an underestimation of the observed time. (the nursing activities that required more than 2, 3 or 4 hours This study showed that 35% of nursing time is explained by the of the nurses’ time) did not happen often enough (so, ten times NAS model (R2 = 0.35). The converted NAS time per patient or more) during the measurements, which makes the validation (202.6 minutes per shift) in our study was comparable with the of the NAS incomplete. Given the fact that the median time of converted NAS times per patient in other studies. Bernet et al.[17] nursing care per patient is 2.4 hours (144.3 minutes), nursing found 150 to 156 minutes per shift and Deberg et al.[18] found activities taking more than 2, 3 or 4 hours rarely occur in daily 180 to 228 minutes per shift. The different articles on the NAS ICU practice so it is not likely that our results have been affected give variable NAS times per shift. A full shift of work equals 480 by this situation. minutes of nursing time. Nurses took care of two or three patients in 60% of our The low correlation of Pearson’s R and R2 (0.59 and 0.35) measurements; we assume that nurses taking care of only one implicates that the NAS is not accurate enough to estimate the patient can perform nursing activities in a shorter amount of nursing time at patient level. However, it is currently still the time. We did not specifically study this but further research best nursing workload model for quantifying nursing workload could eventually point out what is the optimum time per in ICUs.5 There is no clear cut-off point from which the model nursing activity. can be identified as ‘good enough’ based on the R2. However, Furthermore, the observed patients seem to have been since the NAS is used for capacity planning, an R2 closer to 1 more severely ill and consequently had a longer length of would be more desirable. stay compared with all Dutch patients in the same time Since ICU nurses also spend time on non-nursing duties in period, which is likely caused by our selection mechanism. almost every shift, such as coaching a student or participating In order to measure as many nursing activities as possible we in an emergency team within the hospital, we performed a probably more often choose nurses who took care of patients sensitivity analysis to determine whether these non-nursing that were expected to stay the whole shift and these patients duties were affecting the correlation. According to several were probably more severely ill. This may have biased our studies nurses spend approximately 3 to 6% of their shift on non- results since our aim was to validate the NAS and check for nursing duties.[19,23] We therefore took the average of 4.5% and underestimations or overestimations compared with time-and- subtracted this from the 80% of productive nursing time, which motion measurements and it is possible that observed times in

26 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Validation of the NAS Score

sicker patients differ from those in less sick patients. However, References according to Armstrong et al. NAS scores in intermediate care 1. Hughes R. Patient safety and quality: An evidence-based handbook for nurses [25] patients did not differ from those in ICU patients. (Vol. 3). Rockville, MD: Agency for Healthcare Research and Quality 2008. Finally, our study does not correct for the nurses years of 2. Aiken LH, Clarke SP, Sloane DM, et al Hospital nurse staffing and patient mortality, nurse burnout, and job dissatisfaction. JAMA. 2002;288:1987-93. experience on the ICU or level of education. In the analysis we 3. Margadant C, Wortel S, Hoogendoorn M, et al. The Nursing Activities Score per Nurse Ratio Is Associated With In-Hospital Mortality, Whereas the Patients Per included student and registered ICU nurses but further research Nurse Ratio Is Not. Crit Care Med. 2019;48:3-9. in larger groups should investigate whether different groups 4. Palese A, Comisso I, Burra M, et al. Nursing Activity Score for estimating nursing care need in intensive care units: findings from a face and content validity study. need different weighting of NAS points. Based on our results we J Nurs Manag. 2006;24:549-59. believe there is room for improvement in the measurement of 5. Hoogendoorn ME, Margadant CC, Brinkman S. Workload scoring systems in the Intensive Care and their ability to quantify the need for nursing time: a systematic nursing workload. The NAS could be improved by adjusting the literature review. Int J Nurs Stud. 2020;101:103408. 6. Cullen DJ, Civetta JM, Briggs BA, Ferrara LC. Therapeutic intervention scoring NAS points given to the different items. The developers of the system: a method for quantitative comparison of patient care. Crit Care Med. NAS did not report the Pearson’s R or R2, but stated that the NAS 1974;2:57-60. 7. Miranda DR, Nap R, de Rijk A, et al. Nursing activities score. Crit Care Med. is reflecting 81% of total nursing time. About 11% of the nurses’ 2003;31:374-82. time is spent on personal activities. The remaining 8% comes from 8. Padilha KG, de Sousa RMC, Garcia PC, et al. Nursing workload and staff allocation in an intensive care unit: a pilot study according to Nursing Activities Score (NAS). nursing activities derived from medical interventions, related Intensive Crit Care Nurs. 2010;26:108-13. 9. Gonçalves LA, Padilha KG, Sousa RMC. Nursing activities score (NAS): a proposal exclusively to the severity of illness of the patient not measured for practical application in intensive care units. Intensive Crit Care Nurs. by the NAS.[7] The TISS takes these medical interventions into 2007;23:355-61. 10. Stafseth SK, Tønnessen TI, Diep LM, Fagerstrøm L. Testing the reliability and account, such as induced hypothermia, cardiac assist device, validity of the nursing activities score in critical care nursing. J Nurs Measure. pacemaker monitoring or ECG monitoring. For this reason, we 2018;26:142-62. 11. Urden LD, Roode JL. Work sampling: a decision-making tool for determining suggest additional research towards the merging of the TISS-28 resources and work redesign. J Nurs Admin. 1997;27:34-41. and the NAS. The models could be partly combined which could 12. Finkler SA, Knickman JR, Hendrickson G, et al. A comparison of work-sampling and time-and-motion techniques for studies in health services research. Health possibly improve the estimation of nursing workload. Our results Serv Res. 1993;28:577. 13. Burke TA, McKee JR, Wilson HC, et al. A comparison of time-and-motion and self- on observed time per patient and per nursing activity could reporting methods of work measurement. J Nurs Admin. 2000;30:118-25. be taken into consideration when assigning weighting to the 14. Dutch National Intensive Care Evaluation (NICE) foundation. Available at: https:// www.stichting-nice.nl/dd/#561 Accessed 1 June 2019. activities in this new model. Moreover, we think that expressing 15. Interview Reis Miranda, jaarboek NICE. Available at: https://www.stichting-nice. nursing activities in minutes or hours would be more informative nl/focusIC.jsp Accessed 1 September 2019. 16. R Core Team. R: A language and Environment for Statistical Computing [computer compared with points, since it is more straight forward for ICU program]. Vienna, Austria: R foundation. Available at: http://www.R-project.org Accessed May 1, 2019. managers to work with. 17. Bernat Adell A, Abizanda Campos R. Nursing Activity Score (NAS). Our experience with a nursing load calculation system based on times. Enferm Intensiva. Conclusion 2005;16:164-73. 18. Debergh DP, Myny D, Van Herzeele I, et al. Measuring the nursing workload per The NAS was developed more than 15 years ago and significantly shift in the ICU. Intensive Care Med. 2012;38:1438-44. 19. Abbey Mphil M, Chaboyer W, Mitchell M. Understanding the work of intensive overestimates the nursing time needed for ICU patients in the care nurses: a time and motion study. Aust Crit Care. 2012;25:13-22. 20. Bosman R, Rood E, Oudemans-van Straaten, H, et al. Intensive care information current daily ICU practice. Therefore we recommend a revision system reduces documentation time of the nurses after cardiothoracic surgery. of the time weighting assigned to each nursing activity to gain Intensive Care Med. 2003;29:83-90. 21. Harrison L, Nixon G. Nursing activity in general intensive care. J Clin Nurs. better insight into the true nursing workload and to enable a 2002;11:158-67. more effective nursing capacity planning. 22. Kaya H, Kaya N, Turan Y. et al. Nursing activities in intensive care units in Turkey. Int J Nurs Pract. 2011;17:304-14. 23. Norrie P. Nurses' time management in intensive care. Nurs Crit Care. 1997;2:121-5. Disclosures 24. Steyerberg EW. Evaluation of Performance. Clinical Prediction models: a practical approach to development, validation, and updating. Springer 2009; p. 83-99. The data that support the findings of this study are available from 25. Armstrong E, de Waard MC, de Grooth HJS, et al. Using nursing activities score to the National Intensive Care Evaluation (NICE) but restrictions assess nursing workload on a medium care unit. Anesth Analg. 2015;121:1274-80. apply to the availability of these data, which were used under Additionals license for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of the NICE registry. The department of medical informatics (with C.C. Margadant, S. Brinkman, and N.F. de Keizer as employees) receives funding for data processing of the NICE registry. The funding by the NICE foundation does not alter the authors' adherence to all Intensive and Critical Care Nursing policies on sharing data and materials. Four co-authors (M. Hoogendoorn, R.J. Bosman, J.J. Spijkstra, and N.F. de Keizer) are members of the board of NICE. https://nvic.nl/sites/nvic.nl/files/Margadant_SPREAD.pdf The authors declare that they have no conflict of interests.

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 27 Netherlands Journal of Critical Care Submitted February 2020; Accepted July 2020

CASE REPORT

Jaundice due to autoimmune haemolytic anaemia in a patient with Gram-negative septic shock

M.A. Lantinga1,2, F.N. Polderman1, D.E. Issa3, C.P.C. de Jager1 Departments of 1Intensive Care, 2Gastroenterology and Hepatology and 3Haematology, Jeroen Bosch Hospital, ’s-Hertogenbosch, the Netherlands

Correspondence M.A. Lantinga - [email protected]

Keywords - jaundice, E. coli, autoimmune haemolytic anaemia, AIHA

Abstract Toxic shock syndrome and sepsis-induced cholestasis are the urolithiasis and recurrent orchiditis with recent use of foremost common causes of jaundice in the intensive care unit. fluoroquinolones (ciprofloxacin and levofloxacin) for A rare cause of jaundice is autoimmune haemolytic anaemia epididymitis, presented to the emergency department (ED) (AIHA). We present a case in which recurrent Escherichia coli with severe abdominal pain and discoloured urine. Physical sepsis induced a Coombs-positive AIHA leading to jaundice. examination showed jaundice, elevated respiratory rate (22 Although AIHA induced by Gram-negative shock is rare, it breaths/min), hypotension (84/52 mmHg) and tachycardia needs to be considered in the absence of alternative causes to (100 beats/min) in the absence of fever or altered mentation. prevent delay of treatment. Abdominal examination revealed painful percussion, tenderness and guarding. Scrotal examination showed status Introduction after unilateral orchidectomy and no sign of orchiditis or Jaundice in the intensive care unit (ICU) has a broad differential epididymitis. Laboratory investigations revealed anaemia diagnosis, with toxic shock syndrome (i.e. hepatic hypoxia) (haemoglobin 6.7 mmol/l, N: 8.5-11.0) with sign of haemolysis and sepsis-induced cholestasis, predominantly linked to (haptoglobin <0.08 g/l, N: 0.40-2.40; reticulocytes 154 x109/l, infections with Gram-negative bacteria, being the foremost N: 25-120), severe leukocytosis (53.7 x 109/l, N: 4.0-10.0, common causes.[1, 2] Using the SOFA score, sepsis-induced differential: >95% neutrophils, no fragmentocytes), prolonged liver dysfunction is quantified by measuring serum total prothrombin time (38 sec, N: 12-15) and prolonged activated bilirubin, with levels ≥20 µmol/l indicating some degree of partial thromboplastin time (64 sec, N: 26-36) with only mildly hepatic involvement.[3] Sepsis-induced liver dysfunction is impaired antithrombin (65%, N: 80-120). Blood platelets were frequent and associated with increased mortality, although it is normal (249 x 109/l, N: 150-400) whereas lactate dehydrogenase uncertain if elevated serum bilirubin during sepsis truly reflects could not be determined due to haemolysis. Renal function a pathological condition or is just a mere reflection of altered was impaired (eGFR CKD-epi 29 ml/min/1.73 m², N: >90) with bile production and transport into the systemic circulation.[4] severe hyperbilirubinaemia (unconjugated bilirubin 408 µmol/l, A rare cause of jaundice is autoimmune haemolytic anaemia N: 0-16), low albumin (28 g/l, N: 35-50), mildly elevated liver (AIHA).[5] In Coombs-positive AIHA, antibodies directed enzymes (ALT 137 U/l, N: 0-44; gamma-glutamyltransferase against red blood cells are produced, leading to haemolysis. 104 U/l, N: 0-54), raised inflammatory markers (C-reactive Coombs-positive AIHA has a broad differential diagnosis, protein 80 mg/l, N: 0-8; procalcitonin >75 ng/ml, N: 0-0.5) with viral and mycoplasmal infections being the leading cause with normal serum glucose (4.5 mmol/l, N: 4.0-7.7) and of secondary AIHA.[6] We present a patient who was admitted ammonia levels (32 µmol/l, N: 10-35). In the ED, we performed to the ICU in whom Escherichia coli septic shock induced abdominal radiography and ultrasonography, both of which hyperbilirubinaemia as a result of AIHA. were unremarkable, excluding urolithiasis, and single doses of cefuroxime, metronidazole and tobramycin were given. Case Report He was admitted to our ICU under the diagnosis of severe septic A 71-year-old man, with a history of ischaemic cerebrovascular shock with multiorgan failure combined with disseminated disease, unexplained anaemia, choledocholithiasis, cholecystectomy, intravascular coagulation and haemolysis without evident

28 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Jaundice due to septic shock

Table 1. Differential diagnosis of jaundice at the intensive care unit haemolytic anaemia because the direct antiglobulin test (i.e. direct Coombs test) was strongly positive for complement, Prehepatic Intrahepatic Extrahepatic weakly positive for IgG and IgM and showed a borderline Haemolysis Sepsis Pancreatitis presence of cryoglobulins (typing performed when haemolysis Heart failure Ischaemia/hypoperfusion state Biliary tract stricture was reduced under prednisolone). We started continuous Advanced cirrhosis Drug-induced liver injury (e.g. Biliary tract obstruction acetaminophen) venovenous haemodialysis because of renal failure. Both urine and blood cultures returned positive for E. coli. Meropenem Total parenteral nutrition was switched to cefuroxime based on sensitivity results. To Postoperative cholestasis further investigate hyperbilirubinaemia (table 1), additional Viral hepatitis (A, B, C (+D), E) tests were performed. Serological testing was negative for Non-viral hepatitis (alcoholic, non-alcoholic) other autoimmune diseases (tested: antinuclear antibodies, End-stage liver disease anti-mitochondrial antibodies, anti-neutrophilic cytoplasmic Non-hepatitis infections (e.g. autoantibodies) and active viral infection (tested: parvovirus; HIV; CMV; EBV) viral hepatitis A, B, C and E; HIV; cytomegalovirus; Epstein- Haemophagocytic syndrome Barr virus; varicella-zoster virus and herpes simplex virus). CMV = cytomegalovirus; EBV = Epstein-Barr virus. Leptospirosis and capnocytophaga canimorsus infection were not suspected. To exclude lymphoma and alternative cause (Apache II score 31; Apache IV score 103; SOFA score infection sources, 18fluorodeoxyglucose positron-emission 15). Computed tomography of the thorax and abdomen were computed tomography (18F-FDG PET/CT) was performed, unremarkable, except for minimal unilateral perinephric which showed no relevant findings. Magnetic resonance stranding. Antibiotics were switched to meropenem because cholangiopancreatography showed an unremarkable biliary of previous culture results showing drug-resistant Escherichia system, specifically no biliary tract obstruction or stricture. coli. High-dose steroids (prednisolone 60 mg daily), combined Bone marrow biopsy after one week and one month showed a with folic acid and supplementation of calcium and vitamin D, reactive state and no signs of haemophagocytic syndrome or were initiated under the suspicion of secondary autoimmune lymphoma.

Figure 1. Course of serum total bilirubin over time

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 29 Netherlands Journal of Critical Care Jaundice due to septic shock

Eventually bilirubin levels decreased after reaching a peak Infectious causes lead to a transient form of AIHA, in contrast value of 581 µmol/l (figure 1). Corticosteroid treatment was to lymphoproliferative disorders. Onset of symptoms of AIHA tapered and antibiotic treatment was stopped after the patient is in line with the response of the body’s immune system and stabilised. Following antibiotic treatment the direct Coombs test is seen in week two and three after infection. In severe cases remained strongly positive for complement (C3d) and positive AIHA can lead to haemoglobinuria and transient renal failure, for IgM. He was discharged to the nephrology medical ward as was seen in our patient. Treatment of this transient cause for further recovery. After discharge, renal function recovered of AIHA consists of supportive care in the form of intensive and ultimately outpatient intermittent haemodialysis could be hydration, limited blood transfusion (as this can induce stopped. additional haemolysis) and prednisone treatment to maintain adequate renal blood flow.[6] Discussion In conclusion, we present a unique case of a patient presenting We present a patient with Gram-negative septic shock induced with Coombs-positive AIHA induced by an E.coli sepsis. Coombs-positive AIHA which led to severe jaundice. However rare, AIHA must be considered in patients presenting In retrospect, the combination of the patient’s medical history, with sepsis and signs of haemolysis without evident cause, so abdominal pain, perinephric stranding and E. coli in both urine that adequate treatment can be timely initiated. and blood cultures, without other evident explanation, could lead to the working diagnosis of E. coli urosepsis. However, as there Disclosures was a discrepancy between the level of hyperbilirubinaemia and All authors declare no conflict of interest. No funding or severity of sepsis, other potential causes of hyperbilirubinaemia financial support was received. Informed consent was obtained were investigated. Although E. coli sepsis is common, having from the patient for publication of this case report. AIHI due to E. coli is rare, suggesting a specific susceptibility in our patient for pathogen specific factors. Cold agglutinin References disease was considered, but this diagnosis was considered 1. Nesseler, N., et al., Clinical review: The liver in sepsis. Crit Care, 2012. 16(5): p. 235. unlikely because of a good response to prednisolone therapy 2. Trauner, M., P. Fickert, and R.E. Stauber, Inflammation-induced cholestasis. J and absence of typical symptoms in our patient. Ultimately, the Gastroenterol Hepatol, 1999. 14(10): p. 946-59. 3. Vincent, J.L., et al., The SOFA (Sepsis-related Organ Failure Assessment) score to diagnosis of E. coli sepsis induced AIHA was deemed the most describe organ dysfunction/failure. On behalf of the Working Group on Sepsis- likely cause for the high bilirubin level at presentation. Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med, 1996. 22(7): p. 707-10. A review of the literature revealed only one reference to E. 4. Jenniskens, M., et al., Cholestatic liver (dys)function during sepsis and other [6] critical illnesses. Intensive Care Med, 2016. 42(1): p. 16-27. coli induced Coombs-positive AIHA. As stated earlier, 5. Michel, M., Classification and therapeutic approaches in autoimmune hemolytic mycoplasmal and viral infections are typically seen as infectious anemia: an update. Expert Rev Hematol, 2011. 4(6): p. 607-18. 6. Gehrs, B.C. and R.C. Friedberg, Autoimmune hemolytic anemia. Am J Hematol, aetiology, with a peak incidence seen at the age of 70 years. 2002. 69(4): p. 258-71.

30 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Submitted January 2020; Accepted July 2020

CASE REPORT

Leptospira-induced acute disseminated encephalomyelitis

Y. Özbay1, P. Brussee2, J.G. van der Hoeven3 The first two authors contributed equally to this article 1Department of Intensive Care, Beatrix Hospital, Gorinchem, the Netherlands 2Department of Intensive Care, Bernhoven Hospital, Uden, the Netherlands 3Department of Intensive Care, Radboud University Medical Centre, Radboud University, Nijmegen, the Netherlands

Correspondence Y. Özbay - [email protected]

Keywords - acute disseminated encephalomyelitis, leptospirosis, leptospira infection

Abstract In the following three days, his symptoms worsened in We describe a patient with acute disseminated encephalomyelitis combination with darkened urine, oliguria and marked confusion following leptospira infection. This case highlights the with erratic behaviour and fever. The patient presented again to importance of complete investigation for diagnosis and early the same hospital and laboratory testing now showed severe treatment, leading to a better prognosis and reduction in rhabdomyolysis with creatine kinase 148,795 U/l (0-171 U/l), morbidity and mortality. urea 25 mmol/l (2.5-6.4 mmol/l), creatinine 494 µmol/l (53-119 µmol/l), aspartate aminotransferase 2912 U/l (0-35 U/l), alanine Introduction aminotransferase 564 U/l (0-45 U/l), lactate dehydrogenase 2955 Leptospirosis-induced acute disseminated encephalomyelitis U/l (0-248 U/l), and a C-reactive protein of 65 mg/l (0-10 mg/l). (ADEM) has rarely been reported in the literature with almost There was a mild leukocytosis of 14.4 x 109/l (4.0-10 x 109/l) and all cases originating from Asian countries. We describe a case of mild thrombocytosis of 579 x 109/l (150-400 x 109/l). ADEM after leptospirosis acquired in the Netherlands. CT angiography of the abdomen showed no apparent pathology besides possible infiltrate of the right lower lobe of the lung, for Case report which a short three-day course of cefuroxime was started. We describe a 69-year-old male patient with a previous medical No definite cause was found for this rhabdomyolysis and history of non-erosive gastritis, osteoporosis, depressive the confused state was thought to be caused by delirium. disorder and pneumonia requiring mechanical ventilation five Hyperhydration with balanced solutions, frusemide and sodium months prior to presentation. This pneumonia was complicated bicarbonate solution was commenced. The differential diagnosis by delirium and a cryptogenic organising pneumonitis requiring with these findings was cryptogenic organising pneumonia, long-term, high-dose steroids. On presentation, his medication hypersensitivity pneumonitis (extrinsic allergic alveolitis) due to consisted of venlafaxine 75 mg, prednisone 5 mg and calcium known activity with birds, or iatrogenic pneumonitis following carbonate/colecalciferol 500 mg/440 units, all once daily. the use of venlafaxine, which was withdrawn. During a heat wave in the Netherlands with maximal After hyperhydration, the rhabdomyolysis and acute kidney temperatures of 37.5 ⁰C, the patient complained of severe injury improved. No renal replacement therapy was necessary. abdominal pain with loss of appetite and uncontrollable Severe confusion resulted in exhaustion making sedation vomiting. After a day he visited his general practitioner who and mechanical ventilation necessary. Prior to intubation, considered gastritis and started metoclopramide 10 mg three tremors and myoclonus were witnessed. Lumbar puncture was times daily, omeprazole 40 mg twice daily and ranitidine 150 mg performed, showing a clear aspect with normal glucose and twice daily. The prednisone was discontinued. Due to persistent lactate but an increased white blood cell count (20 x 106/l (3-14 complaints he was seen in the emergency department of his x 106/l)) and a high protein (1620 mg/l (206-629 mg/l). Blood local hospital four days later. An ECG and laboratory tests were cultures and liquor cultures remained sterile. performed showing reassuring results and he was discharged A gastroscopy and colonoscopy were performed but showed no home without a classifying diagnosis and prescribed oxycodone. abnormalities.

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 31 Netherlands Journal of Critical Care Leptospira-induced acute disseminated encephalomyelitis

Figure 1. T2-flair, first MRI

An EEG showed diffuse delta and theta activity compatible with culture were negative, potentially due to the previous antibiotic severe encephalopathy. He was transferred to our university administration. The patient had not travelled abroad for the hospital for expert neurology consultation and advanced imaging. past two years. Approximately two weeks before disease onset, An MRI was performed (figure 1) which showed multifocal high however, he was exposed to brackish water next to a petting zoo T2 signals from the proximal cervical myelum towards the rear while he was fishing with his grandson. of the capsula interna on both sides, also including the thalamus, We concluded that the patient was suffering from leptospirosis periventricular white matter and the corpus callosum. Multifocal with abdominal complaints and rhabdomyolysis, complicated diffusion restriction was present in the corpus callosum and the by aseptic meningitis or possibly acute disseminated ependymal borders of both ventricles and the left parafalcine encephalomyelitis (ADEM). Treatment was started with cortical grey matter. We considered a differential diagnosis of ceftriaxone 2 g twice daily for seven days. We also treated the heatstroke, infective encephalitis, autoimmune encephalitis presumed ADEM after 8 days of intubation with a course of or a serotonin syndrome due to interaction of venlafaxine, methylprednisolone 1 g once daily for five days followed by 1 mg/ metoclopramide and oxycodone. kg prednisone per day as maintenance therapy and intravenous immunoglobulin 0.4 g/kg for five days. The patient remained Serological testing for autoimmune antibodies (ANA, ANCA, comatose with a Glasgow Coma Score of 3 and absent corneal, myositis blot, anti-TPO) and paraneoplastic antibodies swallow and cough reflexes. Ventilator triggering was present but in serum and liquor were all negative. Testing for human insufficient to maintain normocapnia. A second MRI (figure 2) herpesvirus 6, herpes simplex virus 1 and 2, varicella zoster 11 days after the first MRI and after completion of the course virus, toxoplasmosis, enterovirus, mycoplasma, Chlamydia of intravenous immunoglobulin showed progression of the psittaci, HIV, lues and Cryptococcus was negative. aforementioned abnormalities, with severe diffusion restriction Unexpectedly, rapid testing for leptospirosis (leptocheck) of the periventricular white matter, brainstem and high cervical was positive and hereafter we also found a positive ELISA myelum. One week later, his neurological status was unchanged, for leptospirosis. The first microscopic agglutination with a third MRI (figure 3) showing the same abnormalities with test (MAT) was negative but three days later the MAT a mildly reduced diffusion restriction and oedema. A second tested positive for leptospirosis (Leptospira interrogans five-day course of 0.4 g/kg intravenous immunoglobulin was serovar icterohaemorrhagiae Kantorowicz and serovar initiated. Five days after completing this course, there were no icterohaemorrhagiae Copenhageni Wijnberg with a titre of 1:40 changes in his clinical condition and supportive treatment was and 1:20, respectively). PCR on blood and a urine leptospirosis withdrawn. Post-mortem examination was not performed.

32 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Leptospira-induced acute disseminated encephalomyelitis

Figure 2. T2-flair, second MRI

Discussion Leptospira are subdivided into 20 species containing over 728 Leptospira spp. can cause the potentially severe zoonosis different serovars. The interrogans species (containing the leptospirosis. Leptospirosis has a worldwide distribution with icterohaemorrhagiae serogroup) accounts for over 20% of these high prevalence in developing countries due to poor sanitary serovars. The species defines for a small part the gravity of the conditions and is underreported. Globally 1500-2000 cases are disease and the clinical symptoms; the serovars do not play a being reported every year. In the Netherlands the incidence role in the clinical picture. of Leptospira spp. is around 0.5 per 100,000 population.[1] The primary reservoir of Leptospira spp. is rodents, but it can also Although initially all the symptoms seemed to resolve with be transmitted via other mammals including cattle, pigs and supportive care, a progressive neurological syndrome developed, dogs. Infection mainly occurs after human contact with urine- eventually leading to deep coma. MRI findings showed contaminated water through mucous membranes or breaks in progressive encephalitis fitting ADEM. Extensive analysis the skin. The average incubation time is about 10 days with including infectious, autoimmune and paraneoplastic tests presentation of the symptoms between 2 and 30 days. The clinical demonstrated no other potential diagnosis. As leptospirosis is spectrum of leptospirosis varies from mild and self-limiting a known aetiological trigger for ADEM, we concluded that our to severe forms, such as Weil's syndrome (renal dysfunction, patient had leptospirosis-induced ADEM. bleeding diathesis, icterus, liver failure, and rhabdomyolysis) ADEM is an inflammatory demyelating disorder of the and severe pulmonary haemorrhage syndrome. In about 10- central nervous system. In most cases ADEM is likely to be of 15% of cases neurological symptoms are seen. With severe autoimmune aetiology after a viral infection or vaccination,[6-11] presentation of the disease, mortality can be as high as 10%.[2] but many other cases are described without any association to The diagnosis can be confirmed by several techniques including an antecedent factor. It is thought that ADEM is caused by a PCR, serological tests and isolation of Leptospira spp. in blood, form of molecular mimicry of the infective agent and myelin urine and liquor. antigens, causing sensitisation of lymphocytes directed against Given the exposure to possible contaminated water at the petting the brain. The prevalence in developing countries is unknown, zoo (no source confirmation was attempted) and the positive but it is thought to be more frequent than reported.[12] The leptocheck in our case (compared with MAT: sensitivity 87.4%, neurology is multifocal and polysymptomatic, usually with an specificity 73.3%), ELISA (compared with MAT: sensitivity sub acute onset. Symptoms include headache, fever, behavioural 86%, specificity 84.5%), and MAT (sensitivity 55.3%, specificity disorders, ataxia, chorea, athetosis, decreased consciousness and 95.7%), the diagnosis is highly likely.[3-5] coma. In acute phases the mortality varies from 10 to 20%.[13,14]

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 33 Netherlands Journal of Critical Care Leptospira-induced acute disseminated encephalomyelitis

Figure 3. T2-flair, third MRI

ADEM is mainly a clinical diagnosis which needs to be confirmed leptospirosis, as early treatment is associated with a better by MRI with T2-weighted and fluid attenuated inversion recovery prognosis, reducing morbidity and mortality. approaches. These settings usually show diffuse asymmetrical multifocal lesions, with an increased T2 signal and decreased Disclosures T1 signal, in the myelum and/or intracerebral parts of the brain. All authors declare no conflict of interest. No funding or Changes usually occur asymmetrically in the white matter, basal financial support was received. ganglia, thalamus, midbrain, spinal cord and cortical grey matter. Also other pathology needs to be excluded.[13,15-19] References Spinal fluid may show nonspecific alteration of protein levels 1. Leptospirosis in the Netherlands [Internet]. Available from: https://www. and an increase in the cell count.[20,21] There is no specific atlasinfectieziekten.nl/leptospirose 2. Farr RW, Leptospirosis. Clin Infect Dis. 1995;21:1-8. laboratory test for ADEM. 3. Hartskeer RA, Smythe LD. The role of leptospirosis reference laboratories. Curr Top Leptospirosis-induced ADEM is rare, with only a few cases Microbiol Immunol. 2015;387:273-88. 4. Limmathurotsakul D, Turner EL, Wuthiekanun V, et al. Fool’s gold: Why imperfect reported; usually the outcome is good and a complete recovery reference tests are undermining the evaluation of novel diagnostics: A [19,22-24] reevaluation of 5 diagnostic tests for leptospirosis. Clin Infect Dis. 2012;55:322-31. is described in most cases. 5. Agampodi SB, Matthias MA, Moreno AC, Vinetz JM. Utility of quantitative According to the literature, the therapy for leptospirosis- polymerase chain reaction in leptospirosis diagnosis: Association of level of leptospiremia and clinical manifestations in Sri Lanka. Clin Infect Dis. induced ADEM includes intravenous immunoglobulin, high- 2012;54:1249-55. dose intravenous steroids and possibly plasma exchanges.[25-27] 6. Carreira J, Casella MI, Ascenção BB, et al. Acute disseminated encephalomyelitis, a rare post-malaria neurological complication: Case report and review of the In our case we initiated the therapy with a course of high-dose literature. Travel Med Infect Dis. 2019;28:81–5. methylprednisolone followed by maintenance therapy of 1 mg/ 7. Horie J, Suzuki K, Nakamura T, Okamura M, Iwasaki A, Hirata K. Human herpesvirus 6 encephalitis followed by acute disseminated encephalomyelitis in an kg prednisone a day and intravenous immunoglobulins for five immunocompetent adult. Rinsho Shinkeigaku. 2017;57:174-9. days. Follow-up with two MRIs and successive neurological 8. Marziali S, Picchi E, Di Giuliano F, et al. Acute disseminated encephalomyelitis following Campylobacter jejuni gastroenteritis: Case report and review of the examinations after a second round of five days of Intravenous literature. Neuroradiol J. 2017;30:65-70. 9. Gupta M, Nayak R, Khwaja GA, Chowdhury D. Acute disseminated immunoglobulins did not show any improvement in our patient, encephalomyelitis associated with dengue infection: A case report with literature after which we considered a meaningful recovery unlikely. review. J Neurol Sci. 2013;335:216-8. 10. Masoodi I, Farooq O, Ahmad I, et al. Acute disseminated encephalomyelitis as the first presentation of CNS tuberculosis: report of a case with brief review. Ger Med Although challenging with intubated patients on sedation, the Sci. 2010;8:Doc32. 11. Chen W-T, Huang Y-C, Peng M-C, Wang M-C, Lin K-P. Acute Disseminated importance of this case lies in the need for rapid diagnosis Encephalomyelitis After Influenza Vaccination: A Case Report. Crit Care Nurse. of ADEM with central nervous system involvement after 2016;36:e1-6.

34 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Leptospira-induced acute disseminated encephalomyelitis

12. Roos KL, Miravalle A. Postinfectious encephalomyelitis. In: Scheld WM, Whitley RJ, 20. Poser CM, Paty DW, Scheinberg L, et al. New diagnostic criteria for multiple Marra CM (eds). Infections of the central nervous system. Wolters Kluwer Health, sclerosis: Guidelines for research protocols. Ann Neurol. 1983;13:227-31. Philadelphia; 2014: pp 331-9. 21. Py MO, Andre C. [Acute disseminated encephalomyelitis and meningococcal A 13. Caldemeyer KS, Harris TM, Smith RR, Edwards MK. Gadolinium enhancement in and C vaccine: case report]. Arq Neuropsiquiatr. 1997;55:632-5. acute disseminated encephalomyelitis. J Comput Assist Tomogr. 1991;15:673-5. 22. Alonso-Valle H, Muñoz R, Hernández JL, Matorras P. Acute disseminated 14. Sonneville R, Demeret S, Klein I, et al. Acute disseminated encephalomyelitis in encephalomyelitis following Leptospira infection. Eur Neurol. 2001;46:104-5. the intensive care unit: Clinical features and outcome of 20 adults. Intensive Care 23. Coyle CD, Hurst EW. Acute Disseminated Encephalomyelitis Following Med. 2008;34:528-32. Vaccination. Lancet. 1929;214:1246-8. 15. Atlas SW, Grossman RI, Goldberg HI, Hackney DB, Bilaniuk LT, Zimmerman RA. 24. Lelis SSR, Fonseca LF, Xavier CC, Horta MB, Cordeiro SS. Acute Disseminated MR diagnosis of acute disseminated encephalomyelitis. J Comput Assist Tomogr. Encephalomyelitis After Leptospirosis. Pediatr Neurol. 2009;40:471–3. 1986;10:798-801. 25. Menge T, Hemmer B, Nessler S, et al. Acute disseminated encephalomyelitis: An 16. Caldemeyer KS, Smith RR, Harris TM, Edwards MK. MRI in acute disseminated update. Arch Neurol. 2005;62:1673-80. encephalomyelitis. Neuroradiology. 1994;36:216-20. 26. Schwarz S, Mohr A, Knauth M, Wildemann B, Storch-Hagenlocher B. Acute 17. Kesselring J, Miller DH, Robb SA, et al. Acute disseminated encephalomyelitis: Mri disseminated encephalomyelitis: A follow-up study of 40 adult patients. findings and the distinction from multiple sclerosis. Brain. 1990;113:291-302. Neurology. 2001;56:1313-8. 18. Lukes SA, Norman D. Computed tomography in acute disseminated 27. Keegan M, Pineda AA, McClelland RL, Darby CH, Rodriguez M, Weinshenker BG. encephalomyelitis. Ann Neurol. 1983;13:567-72. Plasma exchange for severe attacks of CNS demyelination: Predictors of response. 19. Singh S, Alexander M, Korah IP. Acute disseminated encephalomyelitis: MR Neurology. 2002;58:143-6. imaging features. Am J Roentgenol. 1999;173:1101-7.

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 35 Netherlands Journal of Critical Care Submitted July 2020; Accepted August 2020

CASE REPORT

Treacher Collins syndrome: a case report

R. Ibrahim, D. Hejazi Albasha, H. Daood Pediatric Intensive Care Unit, Children’s Damascus University Hospital, Syria

Correspondence R. Ibrahim - [email protected]

Keywords - Treacher Collins syndrome, mandibulofacial dysostosis, Berry's syndrome, Franceschetti-Zwahlen-Klein syndrome

Abstract in intubation, and many techniques aiming at successful airway Treacher Collins syndrome is an inherited and rare, autosomal management have been reported, such as intubation under dominant condition that presents several craniofacial fibre-optic bronchoscopy, the use of a laryngeal mask airway, deformities at different levels. The disorder is characterised fibre-optic intubation through a laryngeal mask airway, and by abnormalities of the auricular pinna, hypoplasia of facial even tracheostomy as a last resort.[4] bones, antimongoloid slanting palpebral fissures with coloboma The care of individuals affected by Treacher Collins syndrome of the lower eyelids and cleft palate. This condition affects requires a multidisciplinary approach and may involve an estimated 1 in 50,000 people. Upper airway obstruction intervention from a number of healthcare professionals both and difficult tracheal intubation are often encountered in preoperatively and postoperatively.[5,6] patients with this syndrome. Neonates and small infants with We describe the case of a 40-day-old baby girl with Treacher craniofacial abnormalities may represent great challenges Collins syndrome. By presenting this case, we intend to show regarding the management of the airway. We present a 40-day- that in small infants with this syndrome, in whom difficulties old baby girl with Treacher Collins syndrome. The infant was in ventilation and intubation are expected, thoughtful airway seen in our hospital for sudden-onset respiratory distress that management planning is very important. A multidisciplinary progressed rapidly to respiratory failure. Chest X-ray was team should be formed to coordinate the available expertise to consistent with pneumonia. The patient developed worsening manage these complex situations. An appropriate algorithm of respiratory distress and was transferred to the intensive care airway management should be clearly envisaged. unit and mechanically ventilated. She proved difficult to intubate and difficult to wean. Tracheostomy and gastrostomy Case report were necessary in the management of this patient. We discuss A 40-day-old baby girl, weight 2 kg, was seen in our emergency the importance of a multidisciplinary planned approach in the room with a three-day history of sudden-onset shortness of management of this rare syndrome. breath, poor feeding and lethargy. In the previous 12 hours, she had developed respiratory distress and cyanosis. Her mother Introduction denied having a history of fever or chills. The infant was started Treacher Collins syndrome, otherwise known as mandibulofacial on antibiotic therapy without any improvement. dysostosis, is a congenital disorder of craniofacial development,[1] mainly characterised by maxillary, zygomatic and mandibular History hypoplasia, a high arched palate and temporomandibular joint The baby was born vaginally at full term; weight at birth was abnormalities. Patients with this syndrome are particularly 2 kg and there was no delay in crying. This patient was the difficult or even impossible to mask, ventilate or intubate.[2] youngest of the three children born to parents with no family Treacher Collins syndrome can present one of the most history of any syndrome. She had two normal siblings. Her difficult airway management problems encountered by an parent’s marriage was not consanguineous; there was no anaesthesiologist, because children with this condition can history of exposure to known teratogenic agents or of maternal have many craniofacial abnormalities and they often require a diseases. The mother also had no history of alcohol, smoking variety of corrective surgical procedures, thus confronting the and drug abuse. The baby was having difficulty in feeding and anaesthesiologist with a possibly difficult intubation.[3] not gaining weight. She developed pneumonia at the age of 6 Treacher Collins syndrome is well known to present difficulties days and aspirated several times.

36 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Treacher Collins syndrome

Physical examination atresia of the auditory meatus, microcephaly, ‘beak-like’ nose, On her arrival to the emergency room, the infant looked ill, with macrostomia ‘fishlike mouth’ and thin hair (figures 1 and 2). a temperature of 38.5°C, a respiratory rate of 55 breaths/min, Intraoral examination revealed a high arched cleft palate and a pulse rate of 150 beats/min. The blood pressure was 85/60 without cleft lip (figure 3). mmHg and the oxygen saturation was 78% by pulse oximetry A detailed examination of this patient revealed clinical features while breathing room air. Examination of her chest revealed of mandibulofacial dysostosis, bilaterally symmetrical but crackles at both lungs, cardiac examination was normal. abnormal face characteristics. Computed tomography showed The infant had a characteristic appearance with antimongoloid bilateral choanal obstruction and bilateral atresia of the external slanting of the palpebral fissures, bilateral coloboma of the lower auditory canals. Echocardiograph and abdominal ultrasound eyelids with absence of cilia of the lower eyelids, hypertelorism revealed normal findings. of the eyes, proptosis, normal-sized pupils with a normal Based on the phenotypic and radiographic findings, the reaction to light, malar hypoplasia of the zygomatic arch with diagnosis of Treacher Collins syndrome was made. a ‘sunk-in’ appearance, sad facial expression, a ‘bird-like’ face with trigonocephaly, mandibular hypoplasia, micrognathia, Hospital course retrognathia, low-set ears, malformations of the auricular pinna, During the first day, the patient’s condition deteriorated, her

Figure 1. The clinical view of the 40-day-old baby girl with Figure 2. The clinical manifestations of Treacher Collins syndrome Treacher Collins syndrome. Consent of the patients family was obtained for the publication of these images.

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 37 Netherlands Journal of Critical Care Treacher Collins syndrome

stabilised enough for the ventilator device to be removed. The patient was extubated; however, this led to shortness of breath, a rapid heart rate and elevated blood pressure with no improvement after high flow oxygen inhalation and noninvasive ventilator-assisted respiration. Therefore, she was intubated again. During the period she was connected to the ventilator, she proved difficult to wean. Weaning was attempted 24 hours after using a T-piece and 4 l/min of oxygen was administered. She had a good cough reflex. Three attempts were made to remove the endotracheal tube but all of them failed due to immediate respiratory distress with documented low saturation requiring reintubation. Therefore, after consulting the paediatric otolaryngologist, the decision was made to perform Figure 3. A high arched cleft palate a tracheostomy. Because of feeding problems, the patient was commenced dyspnoea worsened, and she experienced respiratory distress. on central total parenteral nutrition. After a discussion with Two hours after admission to the hospital, she was transferred the paediatric gastroenterologist, the decision was made to to the paediatric intensive care unit (PICU). Examination of perform a gastrostomy to aid in growth and development, and the chest X-ray indicated bilateral infiltrates consistent with for prevention of aspiration. pneumonia. The patient was treated with a combination of The treatment of this baby girl required the coordinated antibiotics. Bronchoalveolar samples showed Streptococcus efforts of a team of specialists, including a paediatrician, pneumoniae. However, the blood culture which was collected paediatric otolaryngologist, paediatric nurse, plastic surgeon, for the bacteriological diagnosis was negative. audiologist, ophthalmologist, geneticist and other specialists to During the next day, she underwent intubation for worsening systematically and comprehensively plan the child’s treatment. hypoxia and respiratory distress, and a three-way central Multiple surgeries would have been needed to treat the various venous catheter (CVC) was inserted into the right internal craniofacial abnormalities associated with Treacher Collins jugular vein. As great difficulty was anticipated in securing the syndrome. airway, the otorhinolaryngologist and anaesthesiologist teams Unfortunately, her clinical condition deteriorated; 36 days after were urgently requested to be present at the procedure. These admission, she developed septic shock. Cultures of blood from teams were waiting in the adjacent room with all the necessary the CVC and peripheral vein, and culture of urine were obtained. equipment laid out and ready, including a paediatric intubation Because ventilator-associated pneumonia was suspected, set and a tracheostomy set. diagnostic bronchoalveolar lavage culture was performed The intubation of the infant was very difficult with two to evaluate for pneumonia as a sepsis source. The antibiotic unsuccessful attempts at orotracheal intubation. One treatment was empirically changed to wide-spectrum antibiotic intubation attempt was with a video laryngoscopy and one therapy (linezolid and colistin). She required high inotropic was through the laryngeal mask airway. Finally, we decided to support and high ventilator settings. Inotropes and vasopressors perform tracheal intubation via a fibreoptic bronchoscope. The were used to maintain her mean arterial pressure within the fibreoptic bronchoscope was checked and loaded with a 3.5 mm normal range. Chest X-ray revealed bilateral infiltrates. A endotracheal tube. We were ready to perform an emergency nosocomial, ventilator-associated pneumonia was diagnosed. tracheostomy if required with the otorhinolaryngologist and The results of blood cultures from the CVC and peripheral vein, anaesthesiologist on hand. The fibrescope was inserted into the and bronchoalveolar samples yielded Pseudomonas aeruginosa. mouth. The trachea was entered and the tube was passed into Urine culture was negative. the trachea. The CVC was removed and a new catheter was inserted into the left internal jugular vein. She was switched to high-frequency After seven days, the patient’s condition had improved. The oscillatory ventilation but the patient never responded chest X-ray showed no infiltrates. Computed tomography of adequately to any therapy. the rhinopharynx showed bilateral choanal obstruction. The Despite appropriate treatment, she remained critically ill with diagnosis of choanal atresia was confirmed by nasofibroscopy. unstable systemic blood pressure. The patient underwent successful correction of bilateral Forty days after admission, she went into cardiac arrest. choanal atresia. Although cardiopulmonary resuscitation was performed for 30 Three days after the operation, the pulmonary situation had minutes, the patient succumbed to her illness.

38 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Treacher Collins syndrome

Discussion the most relevant and challenging tasks for anaesthetists. Treacher Collins syndrome is a rare congenital disorder Many airway management techniques had been described of craniofacial dysmorphism characterised by numerous in paediatric patients with craniofacial malformations.[2] In developmental anomalies restricted to the head and neck.[7] our case we used different devices: intubation with video Early descriptions were attributed to Berry (1889), Treacher laryngoscopy was attempted once but failed, then we used Collins (1900) and Franceschetti and Klein (1949) and hence the laryngeal mask airway once which also failed and, finally, the names Berry's syndrome and Franceschetti-Zwahlen-Klein intubation was achieved by a fibreoptic bronchoscope. Each syndrome.[1,9,11] device was only used once in order to prevent the risk of airway The frequency of Treacher Collins syndrome is 1 in 50,000 trauma which is associated with multiple attempts. live births.[5,7,8] The most frequent clinical manifestations We formed a multidisciplinary team consisting of paediatric are antimongoloid palpebral fissures, malar and mandibular critical care doctor, anaesthesiologist and otorhinolaryngologist hypoplasia, malformation of auricular pinna, coloboma of to manage intubation difficulties. the lower eyelids, conductive deafness, cleft palate and dental Nagamine et al. reported a 13-year-old girl with Treacher anomalies.[9] Collins syndrome who had a history of difficult intubation and Additional abnormal structures that are occasionally found in was scheduled for plastic surgery. They took three-dimensional Treacher Collins syndrome include absent parotid glands, cervical computed tomography images to better evaluate the anatomical spine malformation, cryptorchidism, extremity malformation, features of the upper airway. The patient’s anaesthetic airway renal anomalies, and congenital heart disease.[8,10] management was influenced successfully by the findings of Treacher Collins syndrome is caused by mutations in the TCOF1, the images.[15] In our case, airway management was urgent and POLR1C or POLR1D genes that affect facial development there was not enough time for additional investigations. before birth.[9,11,12] Mutations in the TCOF1 gene ac¬count for Grohskopf et al. carried out a study to determine the prevalence 81% to 93% of all cases. POLR1C and POLR1D gene mutations of ICU-acquired infections, which is a major cause of morbidity are responsible for 2% of Treacher Collins syndrome cases.[9] in PICU patients. They concluded that age-adjusted risk factors The three genes have been involved with a dominant (TCOF1, for infection included central intravenous catheters, arterial POLR1D) or a recessive (POLR1D, POLR1C) autosomal mode catheters, total parenteral nutrition, or mechanical ventilation.[16] of inheritance.[13] In our case, the patient had multiple risk factors for infection Forty percent of the cases are associated with previous family such as central intravenous catheters, parenteral nutrition, and history and those affected have a 50% chance of passing it on mechanical ventilation to their next generation. The remaining 60% of the cases are Çelik et al. performed a study to determine the prevalence of thought to arise because of new mutation.[6,9,13] Our case did infections and the predominant organisms. They concluded not report a familial history of the syndrome and the cause was that the rate of nosocomial infection is high in ICU patients, likely a new mutation. especially for respiratory infections. The most frequently The diagnostic features of Treacher Collins syndrome include reported infection was ventilator-associated pneumonia. The abnormalities in eyes, ears, nose, mouth and facial bone. predominant bacteria were P. aeruginosa and S. aureus.[17] In The vast majority of these features were present in our case. our case, a nosocomial, ventilator-associated pneumonia was Based on these clinical features five clinical forms of Treacher diagnosed. The results of blood cultures and bronchoalveolar Collins syndrome were identified by Franceshetti and Klein: samples yielded P. aeruginosa. the complete form presenting with all known features, an Vincent et al. provided an up-to-date, international picture of the incomplete form presenting with less severe ear, eye, zygoma, extent and patterns of infection in ICUs. They concluded that and mandibular abnormalities, the abortive form with only infections are common in patients in contemporary ICUs, and lower lid pseudo coloboma and zygoma hypoplasia, the the risk of infection increases with the duration of ICU stay.[18] In unilateral form with anomalies limited to only one side of the our case, the baby girl stayed in the PICU for 40 days. face and the atypical form presenting with other abnormalities Treacher Collins syndrome can be detected using prenatal not usually part of this syndrome.[1,7,9] In our case, the patient screening ultrasound. Three-dimensional sonographic imaging presented with the complete form of this syndrome. has been shown to detect these subtle features including Treacher Collins syndrome has been recognised to be associated downslanting palpebral fissures, micrognathia, and lowset ears/ with upper airway obstruction and difficult tracheal intubation.[14] microtia. Polyhydramnios is seen as well.[10] In our case, the In neonates and other paediatric patients with predicted difficult syndrome was not detected before birth. For a good outcome ventilation and intubation, thoughtful airway management it is mandatory to anticipate problems in securing the airway planning is mandatory. Usually, difficult intubation in these and have a plan ready on how to proceed. Ideally appropriate patients is anticipated, which gives us some time to prepare.[2] prenatal planning is needed and prompt intervention after birth Management of difficult airway in children remains one of by a team of experienced medical and surgical personnel.

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 39 Netherlands Journal of Critical Care Treacher Collins syndrome

There is no cure for Treacher Collins syndrome. Treatment is Disclosures aimed at the specific needs of each individual. Many children All authors declare no conflict of interest. No funding or require a multidisciplinary approach involving a craniofacial financial support was received. team, including a paediatric otolaryngologist, audiologist, plastic surgeon, geneticist, psychologist, dental surgeons and Informed consent was obtained from the patient’s family for the other healthcare professionals.[1] publication of this case report (and the accompanying images). Of primary concern are breathing and feeding problems that present at birth as a consequence of micrognathia and tongue References obstruction of the hypopharynx.[5,9] 1. Renju R, Varma BR, Kumar SJ, Kumaran P. Mandibulofacial dysostosis (Treacher A tracheostomy may even be necessary in some cases to Collins syndrome): A case report and review of literature. Contemp Clin Dent. maintain an adequate airway. Furthermore, a gastrostomy 2014;5:532. 2. Marques-Pires R, Trindade H. The airway approach to a neonate with Treacher could be necessary to ensure an adequate caloric intake while Collins syndrome–Case report. Rev Esp Anestesiol Reanim. 2017;64:233-6. protecting the airway. Surgery to restore a normal structure of 3. Muraika L, Heyman JS, Shevchenko Y. Fiberoptic tracheal intubation through a laryngeal mask airway in a child with Treacher Collins syndrome. Anesth Analg. the face is generally performed at defined ages, depending on 2003;97:1298-9. [9] 4. Lin TC, Soo LY, Chen TI, et al. Perioperative airway management in a child with the developmental stage. Tracheostomy was required at some Treacher Collins syndrome. Acta Anaesthesiol Taiwan. 2009;47:44-7. stage in childhood in 41% of children with TCS, and the ratio 5. Trainor PA. Craniofacial birth defects: The role of neural crest cells in the etiology and pathogenesis of Treacher Collins syndrome and the potential for prevention. increased to 84% in TCS patients who also had choanal atresia Am J Med Genet. 2010;152A:2984-94. or stenosis.[4] 6. Trainor PA, Dixon J, Dixon MJ. Treacher Collins syndrome: etiology, pathogenesis and prevention. Eur J Hum Genet. 2009;17:275-83. In our case, tracheostomy, gastrostomy and surgical repair of 7. Thomas P, Krishnapillai R, Ramakrishnan BP. Treacher Collins Syndrome: A Case bilateral choanal atresia were essential in the management of Report and Review of Literature. Oral Maxillofac Pathol J. 2019;10:90-4. 8. Martelli H, Coletta RD, Miranda RT, de Barros LM, Swerts MS, Bonan PR. Orofacial the baby girl. We made a plan with our team of specialists to features of Treacher Collins syndrome. Med Oral Patol Oral Cir Bucal. 2009;14 systematically and comprehensively treat the patient. :E344-8. 9. Khodadadi E, Dehghan Z. Treacher Collins Syndrome: A Case Report and Review of Literature. J Pediatr Rev. 2019;7:45-54. Conclusion 10. Chang CC, Steinbacher DM. Treacher Collins syndrome. Semin Plast Surg. 2012;26:83-90. Each case of Treacher Collins syndrome is unique and needs to 11. Papageorgiou E, Papoulidis I, Zavlanos A, Papanikolaou E, Manolakos E, Fidani S. A novel familial mutation associated with be managed individually. Many manifestations of the disease can 12. Vincent M, Geneviève D, Ostertag, et al. Treacher Collins syndrome: a clinical and be improved by surgery and other supportive treatments. Early molecular study based on a large series of patients. Genet Med. 2016;18:49-56. 13. Shete, P, Tupkari, JV, Benjamin T, Singh A. Treacher Collins syndrome. J Oral detection and well-planned treatment, tailored to the individual Maxillofac Pathol. 2011;15:348. patient, can reduce disease symptoms and produce excellent 14. Bucx MJ, Grolman W, Kruisinga FH, Lindeboom JA, Van Kempen AA. The prolonged use of the laryngeal mask airway in a neonate with airway obstruction results for complete restoration of the form and function of the and Treacher Collins syndrome. Pediatr Anesth. 2003;13:530-3. patient. Airway management should also be specific for each 15. Nagamine Y, Kurahashi K. The use of three-dimensional computed tomography images for anticipated difficult intubation airway evaluation of a patient with individual patient. A multidisciplinary approach and careful Treacher Collins syndrome. Anesth Analg. 2007;105:626-8. planning is mandatory for the appropriate management of 16. Grohskopf LA, Sinkowitz-Cochran RL, Garrett DO, et al. A national point- prevalence survey of pediatric intensive care unit-acquired infections in the these patients. The members of the team are determined by the United States. J Pediatr. 2002;140:432-8. complexity of the case. Preparing for potential difficulties and 17. Çelik İ, İnci N, Denk A, Sevim E, Yaşar D, Yaşar MA. Prevalence of hospital acquired infections in anesthesiology intensive care unit. Fırat Tıp Dergisi. 2005;10:132-5. communicating effectively can lead to successful results and 18. Vincent JL, Rello J, Marshall J, et al. International study of the prevalence and avoid catastrophic outcomes. outcomes of infection in intensive care units. JAMA. 2009;302:2323-9.

40 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Submitted June 2020; Accepted August 2020

CASE REPORT

A case of cerebral and coronary air embolism after flushing of a pleural drain

J.W. Smithuis1, M.C.A. Müller1, F.F. Smithuis2, R.P. Weenink3 Departments of 1Intensive Care Medicine, 2Radiology, 3Anaesthesiology Amsterdam UMC, University of Amsterdam, the Netherlands

Correspondence J.W. Smithuis – [email protected]

Keywords - hyperbaric oxygen therapy, arterial gas embolism, air embolism

Abstract into the pulmonary veins. This report describes the case of a 64-year-old woman with loss of consciousness, forced deviation of the head and eyes, and signs Introduction of myocardial infarction immediately after flushing of a pleural Vascular gas embolism is predominantly an iatrogenic medical drain. A CT scan of the brain demonstrated extensive bifrontal condition.[1] It can occur in any procedure that includes the air configurations and her ECG showed ST-segment elevation. risk of air or another gas entering the bloodstream. Procedures She was diagnosed with cerebral and coronary air embolism. that carry a high risk of gas embolism include neurosurgical She responded favourably to two sessions of hyperbaric oxygen procedures in the sitting position, central venous line therapy. We conclude that arterial gas embolism can develop, placements, cardiac surgery[1] and other procedures where the even in the absence of a right-to-left shunt, if air enters directly venous circulation is exposed at a level above the heart.[2] The

Figure 1. Air configurations in the subarachnoid space, surrounding the gyri, most probably in vascular structures

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 41 Netherlands Journal of Critical Care Cerebral and coronary air embolism after flushing a pleural drain

Figure 2. ST-segment elevations predominantly in leads I, II, III, aVF and V3-5 severity of symptoms is dependent on the size and location of Case report the gas embolism. A 64-year-old woman with a previous medical history of Vascular gas embolism can occur in the venous or arterial bronchial asthma was admitted to the hospital with a diagnosis system. Arterial gas embolism can result in neurological and of pneumonia with pleural effusion. After two weeks of in- myocardial damage, usually due to obstruction of the supplying hospital treatment that involved intravenous antibiotics and vessels of these tissues. Venous gas emboli, however, are mostly placement of a pleural drain, she was slowly recovering. On filtered out by the lungs, except if they pass on to the arterial day 16 of her admission, immediately after flushing the pleural system through a right-to-left shunt, such as a patent foramen drain with saline, she reported general discomfort, followed by ovale which is present in 20-34% of the population.[3] This type acute loss of consciousness with a Glasgow Coma Scale (GCS) of embolism is referred to as paradoxical embolism.[4] Here, we of E1M1V1, forced deviation of her head and eyes to the right present a case of arterial gas embolism resulting in myocardial and a bilateral Babinski reflex. Notable was also a tachycardia and cerebral symptoms treated with hyperbaric oxygen therapy. (130 beats/min) with hypotension (80/40 mmHg).

Figure 3. Right frontal lobe swelling

42 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Cerebral and coronary air embolism after flushing a pleural drain

A CT scan of the brain showed extensive air configurations up, four months after discharge from our hospital, the paralysis between the frontal sulci on both sides, highly suspect for air of her left arm had disappeared. Paresis of her left leg had also emboli in vascular structures (figure 1). CT scan of the thorax improved except for a slight persisting foot drop. showed consolidation in the right lung and a correctly placed pleural drain. Her ECG showed signs of acute ischaemia in the Discussion anterolateral and inferior regions (figure 2), confirmed by an We present a case of coronary and cerebral arterial gas embolism elevated troponin level. The patient was diagnosed with arterial after flushing of a pleural drain. In our patient, it was quickly gas embolism in the coronary as well as cerebral arteries due recognised that the neurological abnormalities could very well to air entry in the pulmonary veins after flushing of the pleural be the result of arterial gas embolism. This resulted in timely drain. Treatment with 100% oxygen through a non-rebreathing transfer to our hospital and treatment with hyperbaric oxygen mask was started. After consultation with our hyperbaric therapy. Her neurological situation greatly recovered after two department, she was urgently transported to our hospital for courses of hyperbaric oxygen therapy. The tonic-clonic seizures treatment. that developed at the end of the second treatment session were After arrival in our hospital, vital parameters showed a normal in retrospect probably not a sign of cerebral oxygen toxicity, blood pressure and 100% oxygen saturation on 15 litres per since they reoccurred hours after hyperbaric oxygen therapy. minute of supplemental oxygen through a non-rebreather They were most likely a result of cerebral swelling due to mask. ST-elevations where still apparent on her ECG, cardiac capillary leakage caused by the arterial gas embolism. ultrasound revealed normal cardiac function with the Cerebral and coronary arterial gas embolism can develop in exception of apical hypokinesia. The neurological symptoms four ways. It can evolve from direct air entry into the pulmonary had improved; her GCS was now E3M5V2, her right arm was veins, as occurred in our patient and has been described paretic. Babinski reflexes were still present. previously.[6] This is also the mechanism in divers holding their Six and a half hours after development of her symptoms, breath during ascent, or when arterial gas embolism occurs treatment with hyperbaric oxygen therapy was started. The as a complication of mechanical ventilation.[7] Another cause treatment was performed according to U.S. Navy Treatment of arterial gas embolism is direct air entry in the left side of Table 6, which lasts approximately five hours. The treatment the heart, the aorta or carotid arteries, as can occur during table was completed without complications and it was noticed cardiac, aortic or carotid surgery.[8] A third cause is paradoxical she moved both arms afterwards. She was transferred to the embolism resulting from air entry into the systemic veins, for coronary care unit for haemodynamic monitoring. example during placement of a central venous line, followed by The next morning, neurological examination showed slight shunting into the arterial system through a cardiac right-to-left improvement. GCS was E3M4V3 with flexion of both arms to shunt or due to overflow of air from the pulmonary arteries to a painful stimulus. However, she did stick out her tongue upon the veins.[9] The last possibility is direct introduction of air into request. Because of this minor improvement, it was decided a systemic artery, for instance due to flushing of a radial artery that the patient should undergo another session of hyperbaric catheter with retrograde ascent of bubbles into the aortic arch oxygen therapy for two hours. At the end of this second session, and then into the carotid arteries.[10] Regarding our case, it is she developed a self-limiting tonic-clonic seizure lasting one remarkable that arterial gas embolism developed after flushing minute, which was thought to be the result of cerebral oxygen the drain, especially because the drain had been flushed toxicity, a rare but well-known side effect of hyperbaric oxygen regularly in the days preceding the event without complications. therapy.[5] However, back on the ward she developed another Possibly, movement of the drain had occurred, damaging the two tonic-clonic seizures accompanied with hypotension pleural viscera, making the pulmonary veins vulnerable for air and bradypnoea upon which she was intubated, antiepileptic entry. Another way of damage to the visceral pleura could have medication was started and the patient was transferred to the been through excessive force during flushing. ICU. CT scan of the brain (figure 3) showed swelling of the In our patient, neurological deterioration immediately after right frontal lobe, and no bleeding or signs of ischaemia. No flushing of the pleural drain supported the diagnosis of further seizures developed and two days after ICU admission, she arterial gas embolism. Many cases, however, are less clear was successfully extubated. Neurological examination showed cut, and the possibility of arterial gas embolism may not be maximal GCS, but paralysis of her left arm and a severely paretic considered. This might be the result of the common belief left leg, consistent with the abnormalities in the right frontal lobe that arterial gas embolism is a rare condition. Indeed, a large seen on CT. Besides the development of atrial fibrillation, the clinical study estimated the incidence of cerebral arterial gas haemodynamic parameters remained stable. On the third day embolism as being 5.7 in every 1,000,000 admitted patients.[11] of her stay in the ICU she was transferred back to the referring However, this number is likely an underestimation of the actual hospital. Remaining neurological abnormalities at that time were incidence. Contributing factors to this underestimation are the paralysis of her left arm and mild paresis of her left leg. At follow- broad differential diagnosis of acute in-hospital neurological or

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 43 Netherlands Journal of Critical Care Cerebral and coronary air embolism after flushing a pleural drain

cardiac deterioration, the difficulty in diagnosing arterial gas Conclusion embolism with imaging techniques[4] and the lack of knowledge In this case report we present a case of arterial gas embolism of the disorder. The presumptive clinical diagnosis of arterial after flushing of a pleural drain, resulting in myocardial as gas embolism should be made in all cases of acute cerebral or well as cerebral hypoxia. The patient responded favourably to coronary ischaemia after witnessed or possible vascular air hyperbaric oxygen therapy. Arterial gas embolism is rare but the introduction. CT scan of the cerebrum may be used to rule out incidence is presumably underestimated. Prompt recognition of other diagnoses, but should not be used to exclude cerebral this condition is vital, since urgent administration of hyperbaric arterial gas embolism, since air bubbles may not be visible. oxygen therapy is advantageous. Acknowledging the four [12] Cardiac ischaemia may be diagnosed with ECG and cardiac possible ways in which arterial gas embolism can develop helps enzymes, bearing in mind that coronary angiography will not in making the right diagnosis. reveal a thrombus in cases of arterial gas embolism. In all cases of proven or suspected arterial gas embolism, a hyperbaric facility Disclosures should be consulted immediately. Large case series show that a All authors declare no conflict of interest. No funding or delay of more than six to seven hours until starting hyperbaric financial support was received. oxygen therapy is associated with a reduced chance of successful outcome and therefore all efforts should be made to transport the References [11,13] patient to a hyperbaric facility as soon as possible. 1. McCarthy CJ, Behravesh S, Naidu SG, Oklu R. Air Embolism: Diagnosis, Clinical The mechanisms of action of hyperbaric oxygen therapy in Management and Outcomes. Diagn Basel Switz. 2017;7(1). 2. Duke DA, Lynch JJ, Harner SG, Faust RJ, Ebersold MJ. Venous air embolism in arterial gas embolism are fourfold. Firstly, direct pressure sitting and supine patients undergoing vestibular schwannoma resection. on the bubble causes a decrease in size following Boyle’s law Neurosurgery. 1998;42:1282-6; discussion 1286-7. 3. Calvert PA, Rana BS, Kydd AC, Shapiro LM. Patent foramen ovale: anatomy, (at a constant temperature, the volume of a fixed amount outcomes, and closure. Nat Rev Cardiol. 2011;8:148-60. [14] 4. Muth CM, Shank ES. Gas embolism. N Engl J Med. 2000;342:476-82. of a gas varies inversely with its pressure). Secondly, it 5. Yıldız Ş, Aktaş Ş, Cimşit M, Ay H, Toğrol E. Seizure Incidence in 80,000 patient promotes gas resorption by applying high concentrations treatments with hyperbaric oxygen. Aviat Space Environ Med. 2004;75:992-4. 6. Janisch T, Siekmann U, Kopp R. Cerebral air embolism after pleural streptokinase of oxygen, reducing the partial pressure of nitrogen in the instillation. Diving Hyperb Med. 2013;43:237–8. blood and tissues, thus providing a large gradient for nitrogen 7. Marini JJ, Culver BH. Systemic gas embolism complicating mechanical ventilation [4,15,16] in the adult respiratory distress syndrome. Ann Intern Med. 1989;110:699-703. absorption from the air bubble. In the third place, the high 8. Kurusz M, Butler BD. Bubbles and bypass: an update. Perfusion. 2004;19 Suppl 1:S49-55. partial oxygen tensions improve tissue oxygenation of marginally 9. Christin F, Bouffard Y, Rossi R, Delafosse B. Paradoxical symptomatic air embolism after saline contrast transesophageal echocardiography. Echocardiogr Mt Kisco perfused tissue (the penumbra), promoting cellular survival. Lastly, N. 2007;24:867-9. hyperbaric oxygen therapy has immunomodulatory properties, 10. Hatling D, Høgset A, Guttormsen AB, Müller B. Iatrogenic cerebral gas embolism-A systematic review of case reports. Acta Anaesthesiol Scand. 2019;63:154–60. for instance by decreasing inflammation by reducing leucocyte 11. Bessereau J, Genotelle N, Chabbaut C, et al. Long-term outcome of iatrogenic gas adherence.[4] Multiple hyperbaric oxygen therapy regimens embolism. Intensive Care Med. 2010;36:1180-7. 12. van Hulst RA, Klein J, Lachmann B. Gas embolism: pathophysiology and exist, but most experts advise primary treatment of arterial gas treatment. Clin Physiol Funct Imaging. 2003;23:237-46. 13. Blanc P, Boussuges A, Henriette K, Sainty JM, Deleflie M. Iatrogenic cerebral air embolism with US Navy Treatment Table 6, which consists of initial embolism: importance of an early hyperbaric oxygenation. Intensive Care Med. compression to 2.8 atmospheres absolute (ATA), the equivalent of 2002;28:559-63. 14. Cross EPM. Physics, Pharmacology and Physiology for Anaesthetists ( 2nd ed). a dive to a depth of 18 meters, followed by decompression to 1.9 Cambridge ; New York: Cambridge University Press; 2014. p. 321. ATA (9 meters water depth). Total duration is almost five hours. 15. Ziser A, Adir Y, Lavon H, Shupak A. Hyperbaric oxygen therapy for massive arterial air embolism during cardiac operations. J Thorac Cardiovasc Surg. 1999;117:818-21. Hyperbaric oxygen therapy can be repeated daily as long as there is 16. Wherrett C, Mehran RJ, Beaulieu M-A. Cerebral arterial gas embolism following stepwise improvement after each session, and is usually performed diagnostic bronchoscopy: delayed treatment with hyperbaric oxygen. Can J [17] Anaesth J. 2002;49:96-9. for two hours at 2.4-2.5 ATA (14-15 meters water depth). Our 17. Moon RE. Hyperbaric treatment of air or gas embolism: current recommendations. patient responded favourably to two courses of hyperbaric oxygen Undersea Hyperb Med J. 2019;46:673-83. therapy and was then transferred back to the referring hospital.

44 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

Abstracts & Case Reports Intensivistendagen 2021

O01 O02

The impact of nUrsiNg DEliRium Preventive Intensive care unit-specific virtual reality is INterventions in the Intensive Care Unit safe, feasible and improves mental health in (UNDERPIN-ICU): A multicenter, stepped sepsis survivors; a randomized wedge cluster randomized controlled trial controlled trial. Paul J.T. Rood1, Marieke Zegers1, Peter Pickkers1, Johan Vlake1, Jasper Van Bommel2, Evert-Jan Wils1, Tim Korevaar2, Mark van den Boogaard1 Diederik Gommers2, Michel Van Genderen2 1Radboud university medical center, Nijmegen, The Netherlands 1Franciscus Gasthuis & Vlietland, Rotterdam, The Netherlands, 2Erasmus University Medical Center, Rotterdam, The Netherlands

Background: Delirium is common in critically ill patients and associated Background: One-third of critical illness survivors develops psychological with impaired short- and long-term outcomes. Non-pharmacologic impairments after ICU treatment, such as post-traumatic stress disorder interventions are recommended in current delirium guidelines, (PTSD) and depression, resulting in a decreased quality of life. Evidence- however their effects have not been unequivocally established. Aim of based treatment modalities are currently unavailable. Virtual reality the present study was to determine the effects of a multicomponent has been demonstrated to be effective for treating non-ICU-related nursing intervention program on delirium in the Intensive Care Unit in psychological symptoms, but its effect in ICU survivors remains unknown. patients with a high-delirium risk. We therefore developed an ICU-specific virtual reality (ICU-VR) intervention, Methods: A stepped wedge cluster randomized controlled trial was and first assessed its feasibility and safety. Additionally, we explored the conducted in the Intensive care units (ICU) of 10 hospitals (clusters) in effect of ICU-VR on psychological sequelae and quality of life. the Netherlands. Adult critically ill surgical, medical or trauma patients Methods: This multicenter, randomized pilot trial included adult, at high risk to develop delirium (predicted risk >35%) that were mechanically ventilated patients with sepsis or septic shock, treated in admitted from December 31, 2016 to May 1, 2019, were included. Every the ICUs of two hospitals in Rotterdam, the Netherlands. Patients were second month, a multicomponent nursing intervention program aimed randomly assigned (1:1) to receive ICU-VR (ICU-VR group) or exposure at optimizing vision, hearing, orientation, sleep, cognition and mobility to a nature VR environment (control VR group). After initiation with was implemented as standard of care in a randomly assigned center VR, of which the moment was determined by patients’ self-reported by adjusting the local delirium protocol, formation of a local delirium readiness to undergo VR, patients were offered the intervention for the working group, tailoring of template program materials to the local consecutive seven days to determine the number of desired sessions. standards and wishes, and training of the nursing and medical staff. The latter two, and the immersion experienced by participants, were Primary outcome was the number of delirium-free and coma-free days feasibility outcomes, and changes in vital sings and cybersickness were alive in 28 days after ICU admission. safety outcomes. The effects of ICU-VR on PTSD, depression and quality Results: A total of 1749 patients were included. Time spent per 8 hours of life were assessed using the impact of event scale – revised (IES-R), shift on the multicomponent interventions was median [IQR] 38 [14- beck depression inventory (BDI) and short-form 12 (SF-12), respectively. 116] minutes in the intervention period and median 32 [13-73] minutes Results: 50 patients (median age: 61 years; 21 [42%] female) were in the control period (p=0.44). Patients in the intervention period had included. Patients underwent the initial VR intervention eight days median [IQR] 23 [4-27] delirium-free and coma-free days, compared to (median, range 4-23) after ICU discharge, and a median of two (1- median 23 [5-27] days for patients in the control group (mean difference 7) and one (1-6) sessions were desired in the control VR and ICU-VR -1.21 days, 95%CI -2.84 to 0.42 days; p=0.15). Delirium occurrence group, respectively. Patients in the ICU-VR group experienced higher remained similar: 39% in the intervention group and 40% in the control immersion than those in the control VR group. No changes in vital signs group (odds ratio 1.10, 95%CI 0.79 to 1.53; p=0.59) Also, the number of were observed, and cybersickness scores were low. ICU-VR patients delirium days was similar: median [IQR] 2 [1-4] days (ratio of medians reported less PTSD (IES-R sum score, 6 months after initial VR; control 0.90, 95%CI 0.75 to 1.09; p=0.27). VR, 23 [2-49], vs. ICU-VR, 5 [0-35], p<0.001) and less depression (BDI sum Conclusion: In this large randomized controlled trial in adult ICU score; 13 [2-34] vs. 7 [0-24], p=0.02) until six months, and a better mental patients with a high-delirium risk, a limited increase was achieved quality of life (SF-12 mental component scale: 47 [26-63] vs. 57 [36-67], of the use of nursing interventions, and no change in the number of p<0.01) until one month after initial exposure (figure 1). delirium-free and coma-free days alive in 28 days, nor in any of the Conclusion: ICU-VR is feasible and safe. It can be implemented shortly secondary outcomes could be determined. These findings do not after critical illness and has strong potential to improve short-term, support implementation of such a program in a context of relatively and possibly long-term, psychological recovery and mental health. high baseline application and short delirium duration. Therefore, ICU-VR represents a promising modality that should be further examined.

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 45 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

ventilation in the ICU, a higher BMI is not related to a different immunological response, unfavourable respiratory mechanics, or impaired outcome.

Figure 1. Post-traumatic stress disorder, depression and mental quality of life six months after receiving initial VR.

O03

A higher BMI is not associated with a different immune response and disease course in critically ill COVID-19 patients. Emma Kooistra1, Aline de Nooijer1, Wout Claassen1, Inge Grondman1, Nico Janssen1, Mihai Netea1, Frank van de Veerdonk1, Johannes van der Hoeven1, Matthijs Kox1, Peter Pickkers1 1Radboudumc, Nijmegen, the Netherlands

Background: Obesity appears to be an independent risk factor for ICU admission and a severe disease course in COVID-19 patients. An aberrant inflammatory response and impaired respiratory function have been suggested as underlying mechanisms. We investigated whether obesity is associated with differences in inflammatory, respiratory, and clinical outcome parameters in critically ill COVID-19 patients. Methods: Sixty-seven COVID-19 ICU patients were divided into obese (BMI ≥30 kg/m2, n=18, 72% class I obesity, 28% class II obesity) and non-obese (BMI <30 kg/m2, n=49) groups. Concentrations of circulating interleukin (IL)-6, IL-8, IL-10, tumor necrosis factor alpha (TNF-α), interferon gamma (IFN-γ), interferon gamma-induced protein (IP)-10, monocyte chemoattractant protein (MCP)-1, and IL-1 receptor antagonist (RA) were determined from ICU admission until 10 days afterwards, and routine laboratory as well as clinical parameters were collected daily. Results: BMI was 32.6 [31.2-34.5] and 26.0 [24.4-27.7] kg/m2 in the obese Figure 1: Circulating cytokines over time. Data presented as geometric means + 95%CI. and non-obese group, respectively. Apart from temperature, which was * p<0.05, calculated using Sidak’s post-hoc multiple comparisons tests on significantly lower in obese patients (38.1 [36.9-38.9] vs. 38.7 [38.0 -39.5] individual timepoints. °C, p=0.02), there were no between-group differences on ICU admission. Plasma cytokine concentrations declined over time (p<0.05 for all, Fig. 1), but no differences between obese and non-obese patients were observed. Also, BMI did not correlate with the cytokine response (IL-6 r=0.09, p=0.61, TNF-α r=0.03, p=0.99, IP-10 r=0.28, p=0.11). The kinetics of clinical inflammatory parameters and respiratory mechanics were also similar in both groups. Finally, no differences in time on ventilator (22 [16-40] days vs. 27 [14-40] days, p=0.41), ICU length of stay (25 [17- 40] days vs. 29 [15-40] days, p=0.53) or 40-day mortality (17% vs. 24%, p=0.50) between obese and non-obese patients were apparent. Conclusion: In conclusion, in COVID-19 patients requiring mechanical

46 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

EP01 Conclusion: The prognostic- and predictive enrichment of biological ARDS subphenotypes may be extended to critically ill patients without Biological subphenotypes of acute ARDS. Using the concept of biological subphenotypes for splitting respiratory distress syndrome show cohorts of critically ill patients could add to improving future precision- prognostic enrichment in patients without based trial strategies and lead to identifying treatable traits for all acute respiratory distress syndrome critically ill. Nanon F.L. Heijnen1, Ronny M. Schnabel1, Iwan C.C. van der Horst1, Dennis C.J.J. Bergmans1, Laura A. Hagens2, Lonneke A. van Vught2, Marry R. Smit2, Tom van der Poll2, Brendon P. Scicluna2, Marcus J. Schultz2, Lieuwe D.J.Bos2, David S.Y. Ong3, Olaf L. Cremer4 1Maastricht University Medical Centre+, Maastricht, The Netherlands, 2University Medical Center, location Academic Medical Center Amsterdam, The Netherlands, 3Franciscus Gasthuis & Vlietland, Department of Intensive Care, Rotterdam, The Netherlands, 4University Medical Center Utrecht, Utrecht, The Netherlands

Background: Biological subphenotypes have been identified in acute respiratory distress syndrome (ARDS), which have shown to provide prognostic- and possibly predictive enrichment. Differentiation of biological subphenotypes in ARDS is mainly driven by inflammatory parameters measured in plasma(1,2). Considering that this inflammatory response can also be observed in other types of acute respiratory failure and patients admitted to the ICU for other reasons, the question arises whether these subphenotypes are limited to ARDS. We aimed to determine whether these biological subphenotypes and their prognostic- and predictive enrichment could be extended to a Figure 1: Percentage of patients extubated and alive over time by cluster subphenotype . The figure depict the univariable cox proportional hazard general ICU population, irrespective of fulfilling the definition of ARDS. modelling. ARDS = Acute Respiratory Distress Syndrome. Cluster subphenotypes Methods: This was a secondary analysis of the Molecular Diagnosis and as described by Bos et al. (1). Risk Stratification of Sepsis (MARS) study (ClinicalTrials.gov identifier NCT01905033), a prospective observational study of admitted adult ICU patients. Two previously proven parsimonious models were used to determine the biological subphenotype of each patient. First, a cluster based model distinguishing an uninflamed and reactive subphenotype based on plasma levels of interleukin-6, interferon gamma, angiopoietin 1/2, and plasminogen activator inhibitor-1(1). Second, a latent class analysis (LCA) model revealing a hypo-inflammatory and hyper- inflammatory subphenotype using plasma levels of interleukin-8, protein C, and bicarbonate(2). The prognostic enrichment was assessed with logistic regression analysis (30-day mortality) while correcting for the mediating effect of ARDS and with Cox regression analysis (ICU- mortality and ventilor-free days), for both biological phenotyping models. As part of predictive enrichment, blood leukocyte gene expression profiles in biological subphenotypes were analyzed by principal component analysis. Results: We analyzed the data of 2500 patients (674 with ARDS and 1826 without ARDS). Here, we only report the results of patients without Figure 2: Principal component (PC) analysis by cluster subphenotype. On the x-axis PC1 explaining approximately 56% of variance and on the y-axis PC2 ARDS. The cluster-derived reactive subphenotype was, independently explaining approximately 8% of variance in the most differentially expressed of ARDS, associated with (1) a higher 30-day mortality (OR 1.48, CI 1.16- genes. Colors identify groups, rectangles represent the centroid per group, and dots represent individual patients. ARDS = Acute Respiratory Distress Syndrome. 1.89), (2) a higher probability of ICU mortality (HR 2.43, CI 1.90-3.11), Cluster subphenotypes as described by Bos et al. (1). and (3) a lower probability of successful extubation while alive (HR 0.66, CI 0.59-0.74; Figure 1) compared to the uninflamed subphenotype. The blood leukocyte gene expression profiles of each individual subphenotype were similar for patients with and without ARDS (all p>0.05; Figure 2). LCA-derived subphenotypes showed similar profiles (data not shown).

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 47 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

EP02

Predicting mortality of individual COVID-19 patients; a multicenter Dutch cohort Marcel Aries1 on behalf of COVID predict consortium 1Maastricht University Medical Centre+, Maastricht, The Netherlands

Background: The first wave of the COVID-19 pandemic had a dramatic effect on our society and severely disrupted our daily lives, economies and healthcare systems. During the first wave, hospitals and ICUs throughout Europe were overwhelmed and resources were exhausted. Given the novelty of the virus, accurate information about the clinical Figure 1: Panel A: Overall performance of both prognostic models per feature set. course and prognosis of individual patients is important. We aimed All models perform well above chance level. XGB generally performs better than LR, except on the premorbid feature set, where both models performed equal. to develop and validate models that predict mortality of SARS-CoV-2 The highest performance was achieved by XGB on both all features and the 10 infected patients admitted to Dutch hospitals. selected features. Panel B: The confusion matrix of the best performing models, XGB trained on the 10 selected features. The prediction threshold was tuned to Methods: In this multi-centre retrospective cohort study across ten the shortest distance to the upper left corner of the AUC plot to create the ‘optimal’ Dutch hospitals, adult SARS-CoV-19 hospital patients from February binary prediction. to June 2020 were included. The retrospective data collection was based on the WHO rapid COVID-19 case report form. After consultation EP03 with several specialist consultants and an evaluation of the COVID-19 literature, additional clinical and laboratory features were added to the case report form. We evaluated the 21-day mortality by the area under Circulating dipeptidyl peptidase 3 levels the receiver operatory curve (AUC), sensitivity, specificity, positive predict persistent vasopressor requirement, predictive value and negative predictive value. The predictive value of development of acute kidney injury, and age was explored by comparison with age-based rules used in practice prolonged ICU length of stay in cardiac and by excluding age from analysis. surgery patients. Results: 2273 patients were included, of whom 516 had died or Dirk van Lier MD1, Karine Santos PhD2, Matthijs Kox PhD1, Peter Pickkers MD PhD1 discharged to palliative care within 21 days after admission. Five feature 1Radboudumc, Nijmegen, the Netherlands, 2 24TEEN4, Hennigsdorf, Germany sets, including premorbid, clinical presentation and laboratory & radiology values, were derived from 80 admission features. Additionally, Background: Up to 30% of cardiac surgery patients develop acute an ANOVA-based data-driven feature selection selected the ten features kidney injury (AKI), whereas approximately 25% develop post- with the highest F-values: age, number of home medications, urea cardiopulmonary bypass vasoplegic syndrome, requiring prolonged nitrogen, lactate dehydrogenase, albumin, oxygen saturation (%), vasopressor therapy. Early detection and treatment of organ dysfunction oxygen saturation measured on room air, oxygen saturation measured could prevent progression to organ failure and improve outcome in on oxygen therapy, blood gas pH and history of chronic cardiac disease. these patients. Dipeptidyl peptidase 3 (DPP3) is a protease involved in A linear logistic regression (LR) and non-linear tree-based gradient the degradation of cardiovascular mediators, most notably angiotensin- boosting (XGB) algorithm fitted the data with an AUC of 0.81 (95%-CI II, and was found to be predictive of short-term outcome in patients 0.77 - 0.85) and 0.82 (0.79 - 0.85), respectively, using the ten selected with cardiogenic shock. In the present study, we assessed DPP3 kinetics features. Both models outperformed age-based decision rules used in in cardiac surgery patients and its association with the development of practice (AUC 0.69, 0.65 - 0.74 for age > 70). Furthermore, performance subsequent organ dysfunction. remained stable when excluding age as predictor (AUC 0.78, 0.75 - 0.81). Methods: Adult cardiac surgery patients admitted postoperatively 384 patients (17%) were admitted to the ICU. The XGB model showed to the ICU were included from September 2018 until December 2019. the highest performance on a selected feature set (AUC 0.69, 0.59 - 0.79) Blood for determination of DPP3 was sampled within 2 hours after ICU Conclusion: Both models showed excellent performance and had admission (day 1) and on the next morning (day 2). Demographic and better test characteristics than age-based decision rules, using ten medical data were extracted from electronic health records. Area under admission features readily available in Dutch hospitals. The models hold the receiver-operator curves (AUCs) with 95% confidence-intervals promise to aid decision making during a hospital bed shortage. were calculated to assess predictive capacity. Optimal Youden-cutoff values were established to define ‘high’ and ‘low’ DPP3 levels at different timepoints. Results: Two-hundred-and-three patients were included. DPP3 levels on days 1 and 2 predicted prolonged vasopressor dependency (>72 hours), with respective AUCs of 0.64 (0.57-0.71) and 0.73 (0.66-0.79), see

48 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

Figure 1. DPP3 was also predictive of AKI, with AUCs of 0.61 (0.53-0.70) Prolonged vasopressor dependency and 0.70 (0.62-0.78) for levels at day 1 and 2, respectively (Figure 1), 60%

and had similar predictive value for prolonged ICU admission (>4 days, d e p

o p = 0.012 p < 0.001 p < 0.001 l

e 40%

Figure 1). Among patients with high DPP3 levels at day 1, those who v e d e

reverted to low concentrations at day 2 had a markedly reduced risk of m o c t u prolonged vasopressor dependency (4.2% vs. 27.8%, p=0.02), AKI (9.5% o 20% % vs 37.3%, p=0.02) as well as prolonged ICU length of stay (8.3% vs 30.4%, p=0.045) compared to patients with persistently high DPP3 levels L H L H L - L H - L L - H H - H (Figure 2). In contrast, patients with low day 1 DPP3 levels and high (10/118) (18/82) (2/95) (25/100) (1/71) (1/24) (9/45) (15/54) concentrations on day 2 had increased risk of prolonged vasopressor Acute kidney injury 60% dependency (20.0% vs 1.4%, p=0.001), AKI (34.1% vs 8.8%, p=0.001) as d e

well as prolonged ICU length of stay (26.7% vs 1.4%, p<0.001) compared p

o p = 0.025 p < 0.001 p < 0.001 l

e 40% v to patients with sustained low DPP3 levels (Figure 2). e d e m

Conclusions: DPP3 levels after cardiac surgery are associated with o c t u prolonged vasopressor dependency, development of acute kidney o 20% % injury, and prolonged ICU admission. Swift normalization of DPP3 following surgery reflects low chances of a complicated ICU admission L H L H L - L H - L L - H H - H due to development of secondary organ failure. (25/118) (30/84) (8/89) (35/96) (6/68) (2/21) (15/45) (19/51) Prolonged ICU length of stay 60% Prolonged vasopressor dependency Acute kidney injury Prolonged ICU length of stay 1.0 1.0 1.0 d e p

o p = 0.008 p < 0.001 p < 0.001

0.8 0.8 0.8 l

e 40% v y y y t t t e i i i v v d v i 0.6 i

t i 0.6

t 0.6 t i i i e s s s m n n n e o e S e S c 0.4 S 0.4 0.4 t u

o 20% DPP3 AUC 0.64 (0.57-0.71) DPP3 AUC 0.61 (0.53-0.70) DPP3 AUC 0.64 (0.57-0.71) 0.2 day1 0.2 day1 0.2 day1 % DPP3day2 AUC 0.73 (0.66-0.79) DPP3day2 AUC 0.70 (0.62-0.78) DPP3day2 AUC 0.73 (0.66-0.79) 0.0 0.0 0.0 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 1 - Specificity 1 - Specificity 1 - Specificity L H L H L - L H - L L - H H - H Figure 2. Area under the receiver operator curves (AUROCs) of DPP3 measurements performed on day 1 and day 2 of ICU admission. Separate graphs display different study (13/118) (22/84) (3/95) (30/102) (1/71) (2/24) (12/45) (17/56) Figureoutcomes. Vasopressor 1: Area dependency under was defined the as thereceiver need for vasopressor operator therapy to maintain curves adequate (AUROCs) blood pressure for of>72 houDPP3rs after admission. measurements Acute kidney injury was defined according to the Kidney Disease: Improving Global Outcomes (KDIGO) criteria. Prolonged ICU stay was defined as the need for ICU admission for >4 days. performedDPP3 = dipeptidyl peptidase on 3, dayAUC = Area 1 andunder the day curve, ICU 2 =of Intensive ICU Care admission. unit. Separate graphs display different FigureFigure 2: 3. IncrementalIncremental properties properties ofof DPP3DPP3 kinetics kinetics on on the the prediction of different study outcomes. Vasopressor dependency was defined as the need for vasopressor negativeprediction study of different outcomes. negative DPP3 study was categorizedoutcomes. DPP3 into high was (H) or low (L) using therapy to maintain adequate blood pressure for >72 hours after admission. Acute calculatedcategorized optimal into high Youden (H) or cutoffs. low (L) Patientsusing calculated below defined optimal cutoffs during both kidney injury was defined according to the Kidney Disease: Improving Global measurementsYouden cutoffs. were Patients categorized below definedas low-low cutoffs (L-L), during while both Outcomes (KDIGO) criteria. Prolonged ICU stay was defined as the need for ICU patientsmeasurements above cutoffs were categorizedduring both asmeasurements low-low (L-L), were while categorized as high-high admission for >4 days. (H-H).patients Patients above above cutoffs cutoff during on both day 1,measurements who subsequently were reduced DPP3 to below DPP3 = dipeptidyl peptidase 3, AUC = Area under the curve, ICU = Intensive Care cutoffcategorized on day as2 were high categorized-high (H-H). as Patients high-low above (H-L). cutoff Patients on who were below cutoff unit. onday day 1, 1,who but subsequently displayed DPP3 reduced levels above DPP3 cutoff to below on day cutoff 2 were on categorized as low- day 2 were categorized as high-low (H-L). Patients who were high (L-H). AKI = Acute kidney injury, ICU = Intensive Care unit below cutoff on day 1, but displayed DPP3 levels above cutoff on day 2 were categorized as low-high (L-H). AKI = Acute kidney injury, ICU = Intensive Care unit

EP04

Specialized pediatric intensive care retrieval teams for transport of non-ventilated patients - just a pie in the sky or today’s reality? Mirl Burger1, Anneliese Nusmeier1 1Radboud university medical center, Department of Intensive Care Medicine, Nijmegen, The Netherlands

Background: The present-day Dutch standard1 on pediatric intensive care (PICU) advises to perform transport of ventilated patients by a specialized PICU retrieval team (PRT) and transport of non-ventilated patients by the referring hospital physician (RHP). However, in everyday practice transport of (potentially life-threateningly ill) non-ventilated patients to a PICU is up for debate. The significance of a specialized transport team on patient outcome has been proven before2, but

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 49 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

not specifically for non-ventilated patients yet. Aim of present study EP05 is to evaluate patient, transport and outcome characteristics for non- ventilated transports, and to assess differences among transportation Short-term outcomes of COVID-19 patients groups. after review of a Medical Emergency Team in Methods: The cohort of children (<18 years) transported from a a large teaching hospital in the Netherlands nontertiary hospital to the PICU of Radboudumc between January 1st Willian van Dijk1, Calvin J. de Wijs1, Ralph K.L. So1 2018 and December 31st 2019 was subdivided by mode of transportation 1Albert Schweitzer Hospital, Dordrecht, the Netherlands groups (PRT vs. RHP). Non-ventilated patients were selected, defined as: patients with any form of respiratory support except invasive Background: During the COVID-19 pandemic in march-may 2020, mechanical ventilation at moment of referral acceptance, without being several patients with respiratory distress were admitted to the general given the advice at referral to be intubated before transport. Patient wards of our hospital. A proportion of these patients deteriorated and characteristics, transport details, and short-term patient outcome were were assessed by a Medical Emergency Team (MET). Patients subject to analyzed and compared between groups. MET review in general have a high in-hospital mortality e.g. 25%. Aim of Results: In total 337 acute pediatric interhospital transports were the present study is to compare the short-term outcomes of COVID-19 conducted, of which 42 non-ventilated by PRT and 109 non-ventilated patients after MET review with a non-COVID-19 cohort of MET reviewed by RHP. Patients transported by PRT seemed more severely ill (higher patients. predicted mortality rates, more often referred from ICU location), more Methods: This was a retrospective cohort study. All proven COVID-19 in need of specialized care (mechanical ventilation during transport), patients above the age of 18 years admitted in the months of march-may with worse outcome (longer length of PICU stay and more ventilator 2020 who were reviewed by a MET were included. During this period days). Despite travelling longer distances, patients transported by PRT COVID-19 patients on the ICU were treated with chloroquine and best had faster access to pediatric intensive care (table 1). supportive care, according to best practices. All MET reviewed patients Conclusion: A specific subcategory of non-ventilated patients may in march-may 2019 were included in a second cohort. Comorbidity was benefit from transportation by a specialized PICU retrieval team. scored with the Charlson Index (CCI). Degree of illness was measured More research on defining this subcategory is needed to suggest a by the National Early Warning Score (NEWS). Outcomes were 30-day modification of the current Dutch guideline on pediatric transport care. mortality and hospital length of stay. Results: A total 257 (PCR proven) COVID patients were admitted to our References: 1. Sectie Intensive Care voor Kinderen (SICK), NVK. Zorg voor Kwaliteit van Zorg hospital in the study period. After a mean 1.8 days (Standard deviation voor Kritiek. Zieke Kinderen. Utrecht; 2011. (SD) 1.6), 27 of these patients (mean age 62 years (SD 8.6), 78% male) 2. Ramnarayan P, Thiru K, Parslow RC, Harrison DA, Draper ES, Rowan KM. Effect of specialist retrieval teams on outcomes in children admitted to paediatric were reviewed by a MET and all were admitted to the ICU. Mean CCI was intensive care units in England and Wales: a retrospective cohort study. Lancet. 2.6 (SD 1.4) and NEWS was 8.0 (SD 2.1). Mortality was 11% after 30 days 2010;376(9742):698-704. and mean hospital length of stay was 34.6 (SD 18.7) days.After a mean of

Table 1: Overview of significant differences in patient characteristics, transport 11.1 (SD 17.6) days 31 non-COVID patients (mean age 67 years (SD 11.8), characteristics and patient outcome between non-ventilated transports by RHP 42% male) were reviewed by a MET and 65% were admitted to the ICU. (Referring Hospital Physician) versus by PRT (PICU Retrieval Team). Mean CCI was 4.5 (SD 2.3) and NEWS was 8.3 (2.8). Mortality was 36%

RHP (n=109) PRT (n=42) p-value after 30 days and mean hospital length of stay was 18.6 (SD 22.1) days. PIM (%; med, range) 0,75 (0,21 – 1,29) 1,3 (0,78 – 4,06) <0,0001 Location at referral (%) Intensive care 1,8 26,2 <0,0001 Conclusion: In conclusion MET reviewed COVID patients were more General ward 72,5 45,2 often male, suffered less comorbidities, deteriorated faster after Other 25,7 28,5 Transport distance (km; med, IQR1-3) 31 (23 – 52) 52 (31 – 71) 0,002 admission to the hospital, but were as seriously ill as non-COVID Time until ICU care available (min; med, IQR1-3) 115 (90 – 151,25) 80 (60 – 105) <0,0001 Respiratory support during transport𝛂𝛂 (%) None 29,4 11,9 <0,0001 patients; furthermore, mortality was lower but hospital length of stay Low flow 65,1 35,7 High flow 4,6 4,8 was twice as high in COVID patients. We suggest to consider modalities NIV 0,9 0 Invasive 0 47,6 of continuous monitoring in COVID-19 patients directly from admission Ventilator days (mean ± sd) 0,73 ± 1,94 3,2 ± 4,47 <0,0001 Length of PICU 𝛃𝛃stay (h; med, IQR1-3) 35,9 (20,7 – 66,8) 71,9 (25,0 – 153,9) 0,008 to the hospital with the goal of early recognition of deterioration, early time until ICU care available𝛃𝛃 = mobilisation time of transport team plus one-way time on the road. 𝛂𝛂 calculated for survivors of PICU admission (n=107 for RHP, n=40 for PRT). 𝛃𝛃 treatment and possibly shorten length of stay. Table 1 - Overview of significant differences in patient characteristics, transport characteristics and patient outcome between non-ventilated transports by RHP (Referring Hospital Physician) versus by PRT (PICU Retrieval Team).

50 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021

Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

EP06 )

2 Cardiometabolic risk /m

Computed Tomography reference values 2 250 for Visceral Obesity and increased cardiometabolic risk in a Caucasian cohort 200 Michelle Baggerman1, Steven Olde Damink1, Marcel van de Poll1, Bjorn Winkens2, Inge Dekker3, Peter Weijs3 150 1Maastricht UMC+, Maastricht, The Netherlands, 2Maastricht University, Maastricht, The Netherlands, 3Amsterdam UMC locatie AMC, Amsterdam, The Netherlands. 100 Background: Visceral obesity is a presumed risk factor for the metabolic syndrome. It’s relation to the outcome of critically ill patients following 50 Intensive Care Unit (ICU) admission is unknown. As recently showed that the relation between sarcopenia and ICU outcome is determined by a 0 definition based in external reference values from an otherwise healthy Visceral Adipose Tissue (cm le le e a al cohort1. To investigate the relation between obesity and outcome in sk male sk ma ri i critically ill patients, it is important to use external reference values to point isk fem sk fem h w r o ri out risk factors for dying in the hospital1. This is also important to enhance ig L H igh r ow H comparability in research regarding body composition and outcome in L Figure 1: Mean Visceral Adipose Tissue Index and cardiometabolic risk per gender critically ill. Therefore, the aim of this study was to define gender specific High cardiometabolic risk is defined as: the use of antihypertensive drugs and/or reference values for visceral obesity in a representative Caucasian cohort statines. Low cardiometabolic risk is defined as: no use of antihypertensive drugs or statines. Data are presented with mean (95% confidence interval). Visceral based upon the association between the amount of visceral adipose Adipose Tissue Index (VATI) measured at the level of L3 on an abdominal CT-scan tissue (VAT) and markers of increased cardiometabolic risk. and indexed for height and is given in cm2/m2. Methods: Visceral Adipose Tissue Area Index (VATI cm2/m2) at the level High cardiometabolic risk (male) of vertebra L3 was analyzed using CT scans of 416 healthy kidney donor 100 VATI candidates. The use of antihypertensive drugs and/or statins was used 2 2 VATI 38.7 cm /m BMI as an indicator for increased cardiometabolic risk. Gender specific cut- 80 off values for VATI to identify people with an increased cardiometabolic VATI 60 AUC = 0.702 risk with a sensitivity ≥80% was calculated using receiver operating P = 0.002 characteristic (ROC) curve. 40

Sensitivity % BMI Results: In both men and women who used antihypertensive drugs, 20 AUC = 0.556 P = 0.385 statins or both, VATI was higher than in those who did not use these drugs 0 (P≤0.013) (Figure 1). In males and females respectively, a value of VATI 0 20 40 60 80 100 of ≥38.7 cm2/m2 and ≥24.9 cm2/m2 was associated with an increased 100% - Specitivity % cardiometabolic risk profile with a sensitivity of 80%. ROC analysis showed that VATI was a better predictor of increased cardiometabolic risk than BMI (area under ROC curve (AUC)=0.702 vs AUC=0.556 in males and AUC=0.757 vs AUC=0.630 in females) (Figure 2). High cardiometabolic risk (female) Conclusion: Quantification of visceral adipose tissue may be more 100 VATI reliable in determining cardiometabolic risk than BMI. Although the VATI 24.9 cm2/m2 80 BMI definition requires further validation, gender and ethnicity specific reference values are important for comparability and reproducibility of 60 VATI AUC = 0.757 body composition research, especially for body composition research in P < 0.001 40

critically ill patients. Sensitivity % BMI 20 AUC = 0.630 P = 0.012 References: 0 1. Baggerman MR, van Dijk DPJ, Winkens B, van Gassel RJJ, Bol ME, Schnabel 0 20 40 60 80 100 RM, et al. Muscle wasting associated co-morbidities, rather than sarcopenia 100% - Specitivity % are risk factors for hospital mortality in critical illness. Journal of critical care. 2019;56:31-6. Figure 2: Cut-off value for VATI for a higher cardiometabolic risk per gender High cardiometabolic risk is defined as: the use of antihypertensive drugs and/or statines. The risk is given per gender using ROC analysis. Visceral Adipose Tissue Index (VATI) measured at the level of L3 on an abdominal CT-scan and given in cm2/ m2. Body Mass Index (BMI) is given in kg/m2. The values with a sensitivity of at least 80% is used as a cut-off value for a higher cardiometabolic risk. VATI showed to be a predictor with a higher sensitivity and specificity for cardiometabolic risk than BMI.

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 51 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

EP07

4000 Hyperinflammation affects midazolam

clearance in critically ill adults with 3000

COVID-19 IL6

10 pg/ml 1 1 1 1 T.J.L. Smeets , A.J. Valkenburg , S.D.T. Sassen , B.C.P. Koch , 2000 116 pg/ml M. van der Jagt1, H. Endeman1, D. Gommers1 300 pg/ml 1Erasmus MC University Medical Center Rotterdam, Rotterdam, The Netherlands

Midazolam concentration (ug/L) 1000 Background: Patients treated for coronavirus disease 2019 (COVID-19) pneumonia in the intensive care unit (ICU) tend to be heavily sedated 0 with midazolam. Midazolam is associated with more days spent in 0 50 100 150 200 coma, higher incidence of delirium and increased length of ICU stay Time (h) (LOS). Midazolam undergoes extensive metabolism by CYP3A enzymes Figure 1: Simulated plasma profiles of midazolam for patients with an estimated IL-6 of 10 pg/mL, 116 pg/mL and 300 pg/mL After a 5 mg midazolam loading dose to form two pharmacologically active metabolites (1-OH-midazolam followed by 25 mg/h continuous infusion. and 1-OH-midazolam-glucuronide) and is a known indicator of CYP3A activity. A previous clinical study in critically ill children showed an inhibitory effect of inflammation level on midazolam metabolism[1]. Since an exaggerated proinflammatory response seems to play an EP08 important role in COVID-19, an effect on midazolam pharmacokinetics is likely. In other words, inflammation may affect the sedative effect Anakinra treatment in critically ill COVID-19 of midazolam and associated complications (LOS and delirium). The patients: a prospective cohort study aim of our study was to determine the effect of inflammation on the Emma Kooistra1, Emma Kooistra1, Inge Grondman1, Nico Janssen1, 1 1 1 pharmacokinetics of midazolam. Aline de Nooijer , Mihai Netea , Frank van de Veerdonk , Johannes van der Hoeven1, Matthijs Kox1, Peter Pickkers1, Methods: A retrospective analysis of 189 blood samples from 31 Esther Ewalds2 patients with COVID-19 was performed. Samples were collected as part 1Radboudumc, Nijmegen, The Netherlands, 2Bernhoven Hospital, Uden, The Netherlands of COVID-19 research (approved by ethics committee) in the ICU once a week during midazolam administration. Patients with concomitant Background: A subset of critically ill COVID-19 patients develop a use of the strong CYP3A4 inhibitor erythromycin (up to a week before hyperinflammatory state. Anakinra, a recombinant interleukin-1 receptor admission) and continuous renal replacement therapy were excluded. antagonist, is known to be effective in several hyperinflammatory Midazolam and metabolites were analyzed by an ultra-performance diseases. We investigated the effects of anakinra on inflammatory liquid chromatography–tandem mass spectrometry (UPLC-MS/MS) FDA parameters and clinical outcomes in critically ill, mechanically ventilated validated method. A population pharmacokinetic model was developed, COVID-19 patients with clinical features of hyperinflammation. using nonlinear mixed effects modelling (NONMEM). Interleukin 6 (IL-6), Methods: In this prospective cohort study, twenty-one critically ill a marker of inflammation, was analyzed as covariate. COVID-19 patients treated with anakinra were compared to a group Results: The data were described by a two-compartment model for of standard care. Serial data of clinical inflammatory parameters and midazolam and the metabolites 1-OH-midazolam and 1-OH-midazolam- concentrations of multiple circulating cytokines were aligned on glucuronide. The population mean estimate for midazolam clearance start day of anakinra in the treatment group, and median start day of was 7.3 L/h (4.2-10.5 L/h). The midazolam clearance is reduced by 25% anakinra in the control group. Analysis were performed for day -10 to (7.3 L/h to 5.5 L/h) when IL-6 increased from population median 116 pg/ +10 relative to alignment day. Clinical outcomes were analyzed during mL to 300 pg/mL. We simulated the midazolam plasma concentrations 28 days. Additionally, three sensitivity analyses were performed: (1) with various IL-6 levels (Figure 1). using propensity score-matched groups, (2) selecting patients who Conclusion: Inflammation reflected by IL-6 reduces midazolam did not receive corticosteroids and (3) using a subset of the control clearance in critically ill patients with COVID-19. This is probably caused group aimed to match the criteria (fever, elevated ferritin) for anakinra by the inhibitory effect of IL-6 on CYP3A activity. Be aware of prolonged treatment. sedation caused by increased plasma concentrations of midazolam in Results: Baseline characteristics and clinical parameters on ICU critically ill COVID-19 patients. However, in this study we did not relate admission were similar between groups. Due to bias by indication, the pharmacokinetics to the pharmacodynamics of midazolam. plasma ferritin (p=0.009) and temperature (p=0.001) were significantly higher in the anakinra group on alignment day. Following treatment, References: 1. Vet et al. Inflammation and Organ Failure Severely Affect Midazolam Clearance no relevant differences in circulating cytokine levels were observed in Critically Ill Children. Am J Respir Crit Care Med. 2016 Jul 1;194(1):58-66. between both groups. Decreases in clinical laboratory parameters, including temperature (p=0.03), white blood cell counts (p=0.02), and plasma levels of ferritin (p=0.003), procalcitonin (p=0.001), creatinine

52 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

(p=0.01), and bilirubin (p=0.007) were more pronounced in the samples were serially obtained to construct time-concentration curves anakinra group (Fig. 1). There were no significant differences between of cytokines tumour necrosis factor (TNF)-α, interleukin (IL)-6, IL-8, IL- the anakinra and control group in duration of mechanical ventilation 10, IL-1 receptor antagonist (IL-1RA), macrophage inflammatory protein (23 [10-29] days vs. 17 [7-29] days, p=0.79) ICU length of stay (24 [10- (MIP)-1α, monocyte chemoattractant protein (MCP)-1, granulocyte 29] days vs. 17 [6-29] days, p=0.59), or 28-day mortality (19% vs. 18%, colony-stimulating factor (G-CSF) and interferon gamma-induced p=0.87). Sensitivity analyses confirmed these results. protein (IP)-10. Area under the time-concentration curves (AUCs) were Conclusion: In conclusion, anakinra is effective in reducing clinical signs calculated to provide an integral measure of the cytokine response. of hyperinflammation in critically ill COVID-19 patients. A randomized Every 30 minutes, flu-like symptoms were scored on a 25-point scale controlled trial is warranted to draw conclusions about the efficacy of and tympanic temperature was measured. anakinra on clinical outcomes. Results: Median [interquartile range] age was 23 [21-25] years for males and 23 [21-24] years for females (p=0.18), whereas BMI was 23.0 [20.8- 25.1] and 23.6 [21.9-25.7] kg/m2, respectively (p=0.12). For both sexes, LPS administration resulted in a profound increase in all measured cytokines (Figure 1). However, compared to male subjects, females displayed markedly higher levels of TNF-α (41% higher median AUC, p<0.0001) IL-6 (+50%, p<0.0001), IL-8 (+28%, p<0.0001), IL-1RA (+112%, p<0.0001), G-CSF (+45%, p=0.02) and IP-10 (+47%, p<0.0001). In line, peak symptom scores were 8 [6-11] points for females vs. 6 [4-9] points for males (p=0.014, Figure 2A). Nevertheless, the LPS-induced increase in body temperature was less pronounced and less prolonged in females than in males (p=0.0009, Figure 2B). Conclusion: In a large cohort of healthy volunteers, we demonstrate Figure 1: Clinical inflammatory parameters over time. Data are presented as that females mount a more pronounced cytokine response and report geometric mean + 95% CI and were analyzed using mixed-models analysis (time*group interaction factor). P-values under graph titles reflect overall more flu-like symptoms after administration of bacterial LPS compared between-group differences (day -10 until day 10). Between-group p-values for day with males. However, the fever response was less outspoken in females. -10 until day 0 and day 0 until day 10 are shown on the left and right of each panel, respectively. These findings indicate sex-specific regulation of the immune response, which may be of relevance for the future development of a more personalized approach for the treatment of systemic inflammatory conditions. EP09

The systemic inflammatory response induced by LPS administration is more pronounced in women than in men Aron Jansen1, Niklas Bruse1, Nicole Waalders1, Peter Pickkers1, Matthijs Kox1 1Radboud University Medical Center, Nijmegen, The Netherlands

Background: Over the last couple of decades, a multitude of failed clinical trials has made it increasingly clear that systemic inflammatory conditions such as sepsis are very heterogeneous. Individual characteristics such as sex may influence the manner in which these conditions manifest and thereby warrant different treatment approaches. This is again emphasized by the current COVID-19 pandemic, a systemic inflammatory disease which affects males more than females. In the present work, we aimed to investigate whether sex affects the immune response in a large cohort of healthy volunteers undergoing experimental human endotoxemia, a standardized, Figure 1: Plasma concentrations of several measured cytokines after LPS administration (at t=0). TNF = tumour necrosis factor, IL = interleukin, RA = receptor controlled, and reproducible model of systemic inflammation. antagonist, MCP = monocyte chemoatrractant protein, MIP = macrophage Methods: Fifty-five male and 54 female healthy volunteers were inflammatory protein, G-CSF = granulocyte colony stimulating factor, IP = interferon gamma induced protein, pg/mL = picograms per milliliter, m = minutes. included in this prospective experimental cohort study. A transient Data are represented as geometric mean and 95%-confidence intervals. p-values systemic inflammatory response was induced by intravenous bolus were calculated using two-way analysis of variance (time*sex interaction term) on administration of 1 ng/kg bacterial lipopolysaccharide (LPS). Blood log-transformed data.

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 53 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

showed ground glass opacities, commonly progressing to nondependent consolidations. We observed four distinct histopathological patterns: organizing pneumonia in four out of seven patients (fig. 1.A); fibrosis (figure 1.B), diffuse alveolar damage (fig 1.C) and acute fibrinous and organizing pneumonia (AFOP) (fig 1.D). None of the biopsy specimens showed any signs of invasive aspergillosis. Conclusion: In this case series common late histopathology in critically ill COVID patients is not classic diffuse alveolar damage but heterogeneous with predominant pattern of organizing pneumonia. Figure 2: A. Change in symptom score after LPS administration (at t=0). B. Change Postmortem biopsy investigations in critically COVID-19 patients with in body temperature after LPS administration (at t=0). oC = degrees Celcius, m = probable COVID-19 associated pulmonary aspergillosis obtained no minutes. Data are represented as median and interquartile range in panel A and mean ± standard error of the mean in panel B. p-values were calculated using a evidence for invasive aspergillosis. Mann-Whitney test on peak symptom score for panel A and by two-way analysis Based on these late postmortem histopathologic findings, we suggest of variance (time*sex interaction term) for panel B. that the empiric use of corticosteroid therapy in COVID-19 related ARDS should be reconsidered.

EP10 References: 1. Verweij PE, Rijnders BJA, Brüggemann RJM, Azoulay E, Bassetti M, Blot S, et Late histopathologic characteristics of al. Review of influenza-associated pulmonary aspergillosis in ICU patients and proposal for a case definition: an expert opinion. Intensive Care Med. critical ill COVID-19 patients: different 2020;46(8):1524–35. Available from: https://doi.org/10.1007/s00134-020- phenotypes without evidence of invasive 06091-6

aspergillosis, a case series Table 1: Patient characteristics, clinical course and histopathological diagnosis Antine W. Flikweert1, Marco J.J.H. Grootenboers1, David C.Y. Yick1, Arthur W.F. du Mée1, Nardo J.M. van der Meer1, Thijs C.D. Rettig1, Merijn K.M. Kant1 1Amphia, Breda, The Netherlands

Introduction: Pathological data of critical ill coronavirus disease 2019 (COVID-19) patients is essential in the search for optimal treatment options. The aim of this study is to relate the histopathological findings of post-mortem lungbiopsy with the clinical course of COVID-19 in search for new insight of the pathogenesis of COVID-19. Methods: Patients with laboratory confirmed severe acute respiratory syndrome coronavirus 2 (SARS-COV-2) admitted to the ICU due to acute hypoxemic respiratory failure between 22 March 2020 and 30 April 2020 and with available postmortem needle core biopsy of the lung were eligible for inclusion in this case series. Patients received routine ICU management, which included, among other things, high dose anticoagulation with low-molecular-weight heparin (LMWH) (nadroparin 87 IE/kg twice daily). We used a case definition of influenza associated aspergillosis to classify patients with COVID-19 associated pulmonary aspergillosis (CAPA).(1) Diagnostic criteria include proven Figure 1: Microscopic lung findings. A, intra-alveolar fibroblastic bodies, consistent with organizing pneumonia, (H&E, 10x) B: loss of alveolar architecture SARS-CoV-2 infection with clinical symptoms and a galactomannan with extensive fibrosis (H&E; 10x) C: inflammatory infiltrate with neutrophils and index of ≥1 on bronchoalveolar lavage (BAL) or of≥0 5 on serum; or lymphocytes, note the hyaline membranes (arrows), suggestive for diffuse alveolar damage. (H&E, 20x) D: prominent fibrinous exudate in alveoli as seen in AFOP Aspergillus spp. cultured from BAL. (H&E; 10x) E: thrombus in larger vessel (H&E; 5x) F: cytopathogenic changes in With permission from the patients families, postmortem core needle pneumocytes due to viral infection seen in all cases: hyperplasia, multinucleaned biopsies were performed on both lungs within an hour after death. giant cells (asterisks), intranuclear inclusion bodies (H&E, 40x). Results: Patients age ranged from 58 to 83 years, five males and two females were included. Time from hospital admission to death ranged from 12 to 36 days, with a mean of 20 ventilated days. ICU stay was complicated by pulmonary embolism in five patients and positive galactomannan on BAL fluid in six patients, suggesting COVID-19 associated pulmonary aspergillosis (table 1). Chest CT in all patients

54 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

EP11 EP12

Large variability of unbound active fraction Effect of Fully–automated versus Non– of ceftriaxone in contrast to ciprofloxacin in automated Ventilation on Mechanical plasma of critically ill patients Power of Ventilation in COVID–19 – Tim Ewoldt1, Alan Abdulla1, Diederik Gommers1, Henrik Endeman1, an observational crossover study Birgit Koch1, Anouk Muller2 Laura Buiteman1, Hassan Mkadmi1, Pim van der Heiden1, 1Erasmus University Medical Centre, Rotterdam, The Netherlands, Frederique Paulus2, Marcus Schultz2, Ary Serpa Neto3 2Haaglanden Medical Centre, The Hague, The Netherlands 1Reinier de Graaf, Delft, The Netherlands, 2Amsterdam UMC, location AMC, Amsterdam, The Netherlands 3Monash University, Melbourne, Australia Background: Traditional antibiotic dosing is not designed for critically ill patients. Severe illness, frequent hypoalbuminemia, and renal failure Background: Patients with COVID-19 often need admission to an results in aberrant pharmacokinetics. Protein binding of drugs may intensive care unit for invasive ventilation. This new disease often vary significantly in critically ill patients, which can lead to high drug challenged the caregiver to initiate ventilation in a lung protective clearance and therefore low plasma concentrations and therapeutic manner. The ‘mechanical power of ventilation’, a summary variable that failure. Additionally, bound fractions of antibiotics have no therapeutic includes all components that play a role in worsening of lung injury, is effect. For ceftriaxone and ciprofloxacin the unbound fractions have associated with outcome in patients with ARDS from another cause. been described as 5-15% and 60-70%, respectively. However, data on With fully automated ventilation the components are under control of protein binding of these antibiotics are scarce in critically ill patients. the ventilator algorithms and has been shown to reduce MP in patients Our objectives were to determine unbound fractions of ceftriaxone not having COVID-19[1]. We hypothesized MP is lower with INTELLiVENT- and ciprofloxacin in critically ill patients, and to determine predictors ASV compared to conventional ventilation in patients with COVID-19. affecting the unbound fractions. Methods: This was an observational crossover study performed on 2 Materials/methods: Samples were obtained from an ongoing ICU’s of the PRoVENT-COVID collaborative group, during the first months multicentre randomized controlled trail (DOLPHIN) at the intensive care of the outbreak from March till May 2020. Ventilation variables and units. Peak and trough samples were collected at 1, 3, and 5 days after parameters, necessary for calculating MP with the simplified equation initiation of antibiotic therapy. Total and unbound concentrations were of Gattinoni[2], were collected in 67 patients who had switched from determined with a validated mass spectrometry (LC-MS/MS) method. conventional ventilation to INTELLiVENT–ASV. Data was collected when For unbound fractions, we used linear regression to find predictors the ventilation mode was changed to INTELLiVENT–ASV: 120 and 60 (p-value <0.10). These predictors were used in a multivariate linear minutes before the change, and 60 and 120 minutes after the change. regression (stepdown) for each predictor (p-value <0.05). A mixed-effect generalized linear model with Gaussian distribution, the Results: A total of 38 patients (137 samples) receiving ceftriaxone or ventilation mode as fixed effect, and considering patients as random ciprofloxacin were included. Unbound fractions that exceeded more effect to account for repeated measurements was used to compare than 3 standard deviations (outliers) were removed from analysis MP between the two ventilation modes. Paired T-test was used for the (N=20). The median (IQ25-IQ75) unbound fractions were 20.2 (15.4- comparison of ventilation parameters. 29.4)% and 71.1 (69.4-76.5)% in peak concentrations for ceftriaxone Results: Mean MP under conventional ventilation was 27 J/min (SD and ciprofloxacin, respectively. For trough concentrations the fractions 12.3) and 19.9 J/min (SD 7.9) under fully–automated ventilation with were 12.3 (8.5-20)% and 69.1 (66.8-73.5)%. Using multivariate analysis, a significant mean difference of 7.1 J/min (P = <0.0001, 95% CI -8.4 to decreased serum albumin, increased serum creatinine, and septic shock -5.9). In addition, VT increases (P = 0.03), but ΔP, RR and Pmax decreases were found to be positive predictors for the percentage of unbound (respectively P = 0.001, P = <0.0001 and P = <0.0001) fraction of ceftriaxone trough concentration. Septic shock was the Conclusion: INTELLiVENT-ASV results in a lower MP in COVID-19 major predictor. No patient characteristics were identified as predictors patients, due to a lower Pmax, respiratory rate and driving pressure, but for unbound fractions of ciprofloxacin. not due to a lower tidal volume. Conclusions: In contrast to the moderately and fairly constant bounded ciprofloxacin, the fraction of unbound concentration was extremely References: 1. Arnal et.al. A. Airway and transpulmonary driving pressures and mechanical variable in ceftriaxone and especially for trough concentrations, powers selected by INTELLiVENT-ASV in passive, mechanically ventilated ICU higher than previously reported, resulting in fluctuations in effective patients. Heart Lung, 2019. 2. Gattinoni et.al. Ventilator-related causes of lung injury: the mechanical power. exposure. At the moment unbound fraction is not considered when ICM, 2016. dosing ceftriaxone, but therapeutic drug monitoring unbound trough concentrations might increase the likelihood of therapeutic success.

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 55 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

destruction or cavitation, not present at baseline scan, or obvious deterioration of pre-existent emphysema (Figure 1). Residual lesions were predominantly located in the areas that showed crazy paving (ground glass with reticulation) at presentation, while areas with consolidations observed during admission appeared to be spared at three months. Diffusion capacity was significantly correlated with both TLC (rho 0.56, p <0.001) and 6-MWT (rho 0.53, p < 0.001), but not with CTSS.

Table 1: Baseline characteristics, pulmonary function tests (PFT) and high- resolution chest tomography (HRCT) results at three months follow-up. Data are presented as median [IQR] or numbers (percentage), unless indicated otherwise. For laboratory results and ventilator settings, the worst value for the first 24 hours Figure 1: Shows the difference between the ventilation parameters and MP in of admission was recorded. IMV = Invasive mechanical ventilation, LLN = lower percentages, under conventional and fully–automated ventilation; black line = limit of normal, FEV1 = forced expiratory volume in 1 second, VC = vital capacity, conventional ventilation, grey dotted = INTELLiVENT-ASV; green = target zone, red FVC = forced vital capacity, RV = residual volume, TLC = total lung capacity, DLCOc = unsafe zone = diffusing capacity for carbon monoxide adjusted for hemoglobin, 6-MWT = 6-minute walk test. a Defined as receiving steroid treatment for at least 2 days or more. b 3 patients were still admitted to a rehabilitation centre at the moment EP13 of follow-up. c DLCOc failed in 2 patients. d Two patients were on supplemental oxygen while performing the 6-MWT. High prevalence of pulmonary sequelae Baseline characteristics on admission N = 48 at 3 months in mechanically ventilated Age (years) 63.00 [55.00 - 68.00] Sex (Male) 33 (68.8%) COVID-19 survivors BMI (kg/m2) 27.68 [25.18 - 30.47] APACHE II score 15.0 [13.0 - 17.3] 1 1 1 1 Rob J.J. van Gassel , Julia L.M. Bels , Anne Raafs , Bas C.T. van Bussel , Pre-existing lung disease 1 (14%) 1 1 1 Marcel C.G. van de Poll , Sami O. Simons , Lieke W.L. van der Meer , Asthma 1 (14%) 1 1 1 Hester A. Gietema , Rein Posthuma , Susanne van Santen COPD 3 (6.2%) 1Maastricht University Medical Centre, Maastricht, The Netherlands 3 (6.2%) 1 (14%) 0 (0.0%) 1 (14%) Smoking status Background: Severe COVID-19 is characterized by acute hypoxemic Current smoker 0 (0%) respiratory failure and areas with ground glass on chest computed Former smoker 23 (48%) PaO2/FiO2 ratio (mmHg) 116 [92, 156] tomography (CT). Aim of this study was to assess respiratory sequelae in Proned during ICU admission 22 (45.8%) Inspiratory pressure (cm H O) 26 [24 - 28] severe COVID-19 survivors at three months following hospital discharge. 2 PEEP (cm H2O) 14 [12 - 14] Methods: We included all COVID-19 patients treated at our ICU during VT/kg bodyweight (ml/kg) 5.46 [4.98 - 6.02] Dynamic compliance (ml/cm H O) 37.25 [29.80 - 48.85] the first pandemic wave (March-June) into the Maastricht Intensive Care 2 Received steroids during admissiona 12 (25.0%) Covid (MaastrICCht) cohort (NTR NL8613). At three months after hospital IMV duration (days) 18.5 [9.0 - 28.5] discharge, respiratory outcomes with pulmonary function testing (PFT) Days in ICU 20.5 [10.8 - 33.3] Days in Hospital 32.0 [21.0 - 40.0] including diffusing capacity (DLCOc), high resolution chest computed Hospital discharge location tomography (HRCT), and 6-minute walk test (6-MWT) were assessed Home 7 (14.9%) Nursing home 1 (2.1%) in survivors. Two experienced radiologists systematically scored chest Rehabilitation centre 39 (83.0%) HRCTs for the presence of pulmonary abnormalities, presence of new Days in rehabilitation centreb 14.0 [7.0 - 27.3] ECMO during admission? 3 (6.2%) emphysematous and cystic changes, extent of lobe involvement, Pulmonary function test (PFT) N = 43 and total lung involvement. The CT Severity Score (CTSS) was used to Absolute value % of predicted Below LLN, N (%) reflect the extent of lobe involvement. HRCTs were compared to scans FEV1 (L) 2.9 [2.6 - 3.5] 95.0 [77.0 - 104.5] 11 (25.6%) performed at presentation (n=33) at the emergency department or FEV1/VC (%) 79.9 [76.1 - 86.6] - 0 (0.0%) FVC (L) 3.6 [3.1 - 4.2] 87.0 [70.0 - 106.0] 16 (37.2%) during admission (n=5) depending on availability. All data are presented RV (L) 2.0 [1.6 - 2.2] 88.0 [70.0 - 103.0] 9 (20.9%) as median [interquartile range (IQR)]. Correlations between CTSS, PFT TLC (L) 5.6 [4.6 - 6.7] 84.0 [71.5 - 102.5] 23 (53.5%) DLCOc (L)c 5.4 [4.6 – 6.3] 61.0 [50.0 – 69.0] 36 (87.8%) results, and 6-MWT were assessed using Spearman’s rank correlation. 6-MWT (meters)d 480.0 [386.0 - 536.0] 81.5 [69.5 - 99.5] - Results: Fifty-two patients were alive three months after hospital discharge, HRCT results N = 46 of which 48 (92%) participated in the follow-up. Baseline characteristics and Fibrosis 42 (91.3%) results are detailed in Table 1. We found diminished total lung capacity Ground glass 41 (89.1%) Atelectasis 15 (32.6%) (TLC) and diffusion capacity in 54% and 89% of participants respectively, Dominant pattern but no airway obstruction on PFT. HRCTs showed ground glass opacities Reticular 31 (67.4%) Ground glass 13 (28.3%) in 89% of cases. Signs of reticulation including course fibrous bands No abnormalities 2 (4.3%) either with or without obvious parenchymal distortion, bronchiectasis, Decreased attenuation 25 (54.3%) Due to small-airways disease 21 (45.7%) and bronchiolectasis, were seen in 67% (n=31) of cases and assumed to Due to new emphysema 12 (25.0%) represent fibrosis. A quarter of the survivors showed new emphysematous CT Severity Score (CTSS) 11.0 [5.0 - 15.0)

56 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

Conclusion: The majority of invasive mechanically ventilated COVID-19 Risk factors for HFNC failure were assessed using univariate logistic survivors still had abnormal pulmonary function tests and residual regression. The accuracy of respiratory parameters to predict HFNC changes on HRCT at three months after hospital discharge. Diminished failure at successive time points was analysed with ROC curves. diffusion capacity, total lung capacity, and fibrosis on HRCT were the Results: 29 HFNC-treated patients were propensity-matched to 29 non- dominant features. Our findings warrant intensive respiratory follow-up HFNC patients. Treatment with HFNC was associated with lower ICU of mechanically ventilated COVID-19 patients. admission (69.0 vs. 96.6%; p=0.012) and lower intubation rate (58.6 vs. 86.2%; p=0.011). Thirty-day mortality (3.4 vs. 10.3%, p=0.133) was not statistically different (figure 1). Both hospital (median 16 vs. 28 days; p=0.023) and ICU LOS (median 2 vs. 16 days; p=0.001) were shorter for the HFNC group. The ICU (15 vs. 17 days; p=0.296) and hospital LOS (28 vs. 37 days; p=0.675) were similar for intubated patients, with or without HFNC pre-treatment. In 24 out of 39 patients (61.5%) HFNC failed. Respiratory rate was significantly higher (AUC 0.72 (95%CI [0.54-0.89], p=0.031)) and ROX index lower (AUC 0.80 (95% CI [0.65-0.95], p=0.003)) two hours after HFNC initiation in patients with HFNC failure. Figure 1: Representative HRCT images of two of the survivors. Panel A: HRCT Conclusion: HFNC for COVID-19-related hypoxemic respiratory failure performed at admission (upper row) and at 3-month follow-up (lower row). Chest CT at admission shows typical bilateral subpleural ground glass opacities. No signs in the non-ICU setting may lower ICU admission and intubation rates of previous emphysema were detected. However, follow-up HRCT shows obvious and shorten hospital and ICU LOS. Respiratory rate and ROX-index two emphysematous destruction. Panel B: CT image at presentation at emergency department with evident ground areas with reticulation (crazy paving). Follow- hours after starting HFNC therapy might be reliable predictors of HFNC up reveals diffuse areas of persistent ground glass without reticulation, as well as failure. areas with low density in previously normal areas, possibly due to hypoperfusion. References: Figure 1: 1. Demoule, Cumulative A., etincidence al., High of iMV, Flow ICU admissionNasal Canula and mor intality Critically during admission Ill Severe of patients COVID-19 who received treatment with HFNC (HFNC: black) and who did not received treatment with HFNC (No HFNC:Patients. red). Am J Respir Crit Care Med, 2020. EP14

The effectiveness of High Flow Nasal Cannula in the non-ICU setting for severe hypoxemic COVID-19 patients Yasemin Türk1, Yaar Aga1, Wessel Hanselaar1, Evert-Jan Wils1, Julien van Oosten2, Marlies van Bochove-Waardenburg2, Hendrik Endeman2, Louise Urlings-Strop3 1Franciscus Gasthuis & Vlietland, Rotterdam, The Netherlands, 2Erasmus MC, Rotterdam, The Netherlands, 3Reinier de Graaf, Delft, The Netherlands

Background: High-Flow Nasal Cannula (HFNC) is increasingly used in Number at risk Days 0 0.5 1.0 2.0 4.0 6.0 8.0 10 15 20 30 35 severe hypoxemic COVID-19 patients in ICU and may prevent invasive No HFNC 29 25 18 13 5 3 3 2 1 1 1 0 [1] HFNC 29 27 22 19 15 12 12 8 4 1 0 0 mechanical ventilation . However, data on the role of HFNC in the non-

ICU setting is limited. The aim of this study is to assess the contribution Figure 1: Cumulative incidence of iMV, ICU admission and mortality during admission of patients who received treatment with HFNC (HFNC: black) and who of HFNC in the non-ICU setting on the outcome of these COVID-19 did not received treatment with HFNC (No HFNC: red). patients and evaluate possible predictors of HFNC failure. Methods: A retrospective, observational study was conducted in 2 hospitals in the Netherlands. Severe hypoxic COVID-19 patients defined as SpO2 < 92% despite 6L O2/min via nasal cannula were included. At the pulmonary ward, HFNC was initiated and combined with contact- free monitoring of heart rate, respiratory rate and O2 saturation. ICU was notified on initiation of HFNC and consulted on clinical deterioration. Severe hypoxic COVID-19 patients treated with HFNC were compared to a matched control group without HFNC treatment. Matching was performed by calculating propensity scores based on baseline characteristics, comorbidity, and severity of illness. Study endpoints were ICU admission, HFNC failure (defined as initiation of mechanical ventilation), hospital and ICU length of stay (LOS) and 30-days mortality (Kaplan-Meier; log-rank test).

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 57 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

EP15 Table 1: Baseline characteristics of the study population. Data presented as median [IQR] and between group differences tested with Mann-Whitney U or Fisher’s exact test as appropriate. Enteral provision of free amino acids as opposed to intact protein results in Nutrition group increased systemic availability of exogenous Overall (n = 14) Protein (n=7) Amino acid (n=7) p Age (y) 60 [47, 70] 55 [44, 60] 71 [54, 72] 0.096 amino acids in mechanically ventilated ICU Gender (Male) 9 ( 64.3) 4 ( 57.1) 5 ( 71.4) 1.000 patients Weight (kg) 86 [75, 90] 85 [78, 90] 87 [75, 92] 1.000 BMI (kg/m2) 26.6 [24.8, 28.9] 26.3 [25.7, 27.7] 26.9 [24.4, 29.3] 0.949 Rob van Gassel1, Steven Olde Damink1, Marcel van de Poll1, Origin of admission 0.427 Michelle Weijzen2, Imre Kouw2, Luc van Loon2 ER 8 ( 57.1) 5 ( 71.4) 3 ( 42.9) Other hospital 5 ( 35.7) 2 ( 28.6) 3 ( 42.9) 1Maastricht UMC+, Maastricht, The Netherlands, 2Maastricht University Medical Ward 1 ( 7.1) 0 ( 0.0) 1 ( 14.3) Center, Maastricht, The Netherlands, APACHE II score 16 [11, 19] 18 [12, 20] 14 [12, 18] 0.654 SAPS II score 27 [21, 33] 28 [18, 30] 26 [23, 37] 0.405 Background: Enteral nutrition is part of routine care for mechanically Day of ICU admission during study 15 [10, 22] 15 [11, 26] 15 [8, 22] 0.564 Fecal volume at baseline (ml/day) 475 [362, 600] 450 [375, 600] 500 [375, 600] 0.948 ventilated, critically ill patients. Gastrointestinal dysfunction and Lenght of mechanical ventilation (days) 32 [18, 47] 20 [14, 42] 40 [35, 52] 0.749 intestinal malabsorption are common among ICU patients, further ICU Lenght of stay (days) 39 [26, 46] 41 [24, 52] 38 [28, 42] 0.136 Hospital length of stay (days) 66 [60, 87] 61 [54, 66] 80 [67, 98] 1.000 exacerbating protein-energy deficits during ICU admission. It has been suggested that free amino acids are more effectively absorbed and released in the circulation when compared to protein bound amino acids. The present study applied state-of-the-art tracer methodology to assess the systemic availability of enteral nutrition containing intact protein or an equivalent amount of free amino acids in critically ill patients. Methods: In a parallel study design, 16 mechanically ventilated patients at risk for malabsorption (diarrhea >350 g/day) were randomly assigned to receive enterally 20 g intrinsically L-[1-13C]-phenylalanine labeled milk protein (n=8) or an equivalent amount of free amino acids fully matched to the milk protein labeled with L-[1-13C]-phenylalanine (n=8) over a 2-hour period. Primed continuous L-[ring-2H5]-phenylalanine and L-[ring-3,5-2H2]-tyrosine infusions were applied, with frequent collection of arterial blood samples. The applied methodology allowed us to assess protein digestion and/or amino acid absorption over a 6-hour Figure 1: Systemic availability of enteral nutrition derived phenylalanine, period (2 hours of nutrient administration and a 4-hour postprandial expressed as % of the provided phenylalanine. Vertical dashed lines represent start period). Exogenous phenylalanine release into the circulation over time and end of enteral nutrient infusion. Data presented as mean ± SE. Between group and between groups were assessed with linear mixed effects analysis differences are derived from the linear mixed effect model and marked with * for p < 0.05, ** p < 0.01 or *** p < 0.001. with subjects as random effects with a random intercept. Data were analyzed and visualized using R (v3.6.2). Results: Two patients, one from the protein and one from the amino EP16 acid group, were excluded from the analysis due to loss of the arterial sampling line and an incomplete intravenous tracer prime, respectively. Baseline characteristics of the analyzed cohort are detailed in table 1. Frequent CRRT circuit failure in COVID-19: Following the start of enteral nutrient infusion, exogenous amino acids a novel anticoagulation approach were rapidly released into the systemic circulation, with a significant Hinke de Lau1, Susan Kilian1, Melissa Vink1, Guus Middelaar1, Adriaan Dooms1, Eline Vlot1, Arend Jan Meinders1 increase in exogenous phenylalanine release from 60 min onwards 1St Antonius Hospital, Utrecht, The Netherlands (95% CI 1,1 – 14,1 % of phenylalanine provided released, p = 0.035, Figure 1). Between group differences became evident at the end of Background: Continuous renal replacement therapy in the treatment of nutrient infusion (T = 120 min) and continued to expand postprandially. patients with active COVID-19 is frequently complicated by thrombosis After 4 hours of postprandial sampling, the phenylalanine release as in the extracorporeal circuit, necessitating a revisited anticoagulation a percentage of the amount provided was nearly 20% higher when strategy. compared to the protein group (95% CI 8,4 – 26,8 % of phenylalanine Methods: The standard regional citrate anticoagulation dose of 2.2 provided released, p<0.001). mmol/l was increased to a starting dose of 3.0 mmol/l and routine Conclusion: Enteral provision of free amino acids in mechanically ventilated, thrombosis prophylaxis with low molecular weight heparin was critically ill patients leads to more rapid and greater post-prandial systemic adjusted from a single dose of Nadroparin 2850 IU to 5700 IU twice daily. amino acid release when compared to the provision of intact protein. In a non-randomized cohort study, the safety and efficacy before and

58 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

after the introduction this high dose anticoagulation were reviewed. EP17 Results: Twelve COVID-19 patients requiring continuous renal replacement therapy were included, 40 filter sets using standard dose COVID-19 immediately increases moral and 44 filter sets in the high dose anticoagulation group (table 1). The distress and burnout symptoms in ICU median filter lifespan increased from 19.3 to 39.5 hours (p=0.001) using professionals: a longitudinal cohort study high dose anticoagulation Niek Kok1, Jelle Van Gurp1, Hans Van der Hoeven1, Cornelia Hoedemaekers1, Marieke Zegers1, Malaika Fuchs2 Conclusion: High dose RCA and increased thrombosis prophylaxis, 1Radboudumc, Nijmegen, The Netherlands, 2 Canisius Wilhelmina Hospital, appears to substantially prolong filter lifespan in patients with COVID-19 Nijmegen, The Netherlands while preserving systemic acid-base balance and metabolic homeostasis. Background: Intensive care units (ICUs) have been overwhelmed by the catastrophic outbreak of the coronavirus disease 2019 (COVID-19). Bear the additional psychological and moral burdens of the pandemic, ICU professionals are at risk of developing burnout. This study assessed prevalence and incidence rates of burnout symptoms and moral distress in ICU professionals before and during the COVID-19 crisis. Methods: ICU professionals in six ICUs of two hospitals received a baseline survey (October 2019; before the first peak of the COVID-19 pandemic) with 252 respondents (response rate 53%) and a follow- up survey (May 2020; immediately after the first peak) with 233 respondents (response rate 50%). Burnout symptoms and moral distress were measured with the Maslach Burnout Inventory and the Moral Figure 1: Boxplot of the CRRT filter lifespan with the low dose anticoagulation Distress Scale, respectively. Generalized Estimating Equations were used group on the left and high dose group on the right. The box shows the IQR (p25 - p75) with the line marking the median lifespan. The dashed line indicates the to analyze repeated measurements. historic mean CRRT filter lifespan of non-COVID-19 patients. Results: The prevalence of burnout symptoms was 23.0% before

Table 1: Overview of the anticoagulation used and main results. 1 indicates COVID-19 and 36.1% at post-peak time, with higher rates in nurses statisticallyTable1 significant difference (38.0%) than in physicians (28.6%). Reversely, the incidence rate of  NVCNRDFQNTO HFGCNRDFQNTO new burnout cases among physicians was higher (26.7%) compared to nurses (21.9%). Higher prevalence of burnout symptoms were observed

DO@QHMOQHLHMF NMD °««« Ž in the post-peak-COVID-19 period (OR 1.83, 95% CI 1.32 to 2.53), for nurses (OR 1.77 95% CI 1.03 to 3.04), for professionals working overtime

DFHNM@KBHSQ@SDOQD§EHKSDQ ­­§®LLNKŽK ®–®°LLNKŽK (OR 2.11, 95% CI 1.48 to 3.02) and for professionals directly engaged with care for COVID-19 patients (OR 1.87, 95% CI 1.35 to 2.60). Nurses

  ¬CC­³°«  ­CC°²«« §­CC were less likely than physicians to develop burnout symptoms at post- peak time (OR 0.28 95% CI 0.08 to 0.94) (Table 1). Moral distress from ´°««  scarcity and having to work with colleagues perceived to be less skilled or act unsafe was higher in the post-peak period and was associated NTQRNE ´°³° ¬³¯¬° with burnout (Figure 1). Conclusion: This study showed that working long hours under HKSDQRDSR ¯« ¯¯ conditions of scarcity of staff, time and resources has an immense impact on the mental health of ICU professionals. Stakeholders, ¬ DCH@MEHKSDQKHEDRO@M~GNTQR ¬´®‡­¬~³ˆ ®´°‡¯°~¯ˆ  including ICU management and policymakers, should be aware that

‡ ˆ increasing working hours due to a crisis goes hand in hand with an increasing risk for dropout of personnel due to burnout. At a time when

DQTLB@KBHTL~LLNKŽK‡Q@MFDˆ ¬¬¯‡«®°–¬­³ˆ ¬¬¬‡«®­–¬­°ˆ¬ ICU professionals are enduringly called upon, it is of utmost importance to preserve their mental and moral integrity. ICUs are advised to set

O­~J@‡Q@MFDˆ ²¯¬‡¯–¬®ˆ ²¯«‡¯–¬®ˆ up peer support systems and stimulate grassroots dialogues on moral requirements in pandemic times.

HB@QANM@SD~LLNKŽK‡Q@MFDˆ ­±¬‡¬³­–®­²ˆ ­³¯‡­«±–®³²ˆ¬

O ‡Q@MFDˆ ²®«‡²«³–²°¯ˆ ²®¯‡²«³–²±«ˆ¬



NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 59 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

TableTable 1: 1:General Generalizedized estimating estimating equation of associationsequation between of associations time, respondent between characteristics time, Objectives: To identify views, experiences, and needs for SDM in the respondentand burnout symptoms characteristics and burnout symptoms ICU according to ICU physicians, ICU nurses, and former ICU patients Variable Univariate analysis Multivariable analysis N = 436 and their close family members. OR [95% CI] P value OR [95% CI] P value Time (reference = pre-COVID-19) 1.83 [1.32-2.53] .00 3.01 [.29-2.41] .10 Methods: This is a qualitative study. A total of 19 face-to-face interviews Profession (reference = physician) 1.77 [1.03-3.04] .04 2.63 [1.00-6.89] .05 Interaction: time*profession .63 [.22-1.76] .38 .28 [.08-.94] .04 were conducted with 29 participants when the heterogeneity of the Working more hours than in contract 2.11 [1.48-3.02] .00 1.75 [.91-3.35] .09 population was felt to be adequately covered. Of the 19 interviews, Engaged in direct diagnosis, treatment and/or 1.87 [1.35-2.60] .00 .93 [.36-2.38] .88 care of COVID-19 patients (reference = no) seven were with ICU physicians from two tertiary centers, five with ICU Moral distress Scarcity – resources, time, staff 1.29 [1.19-1.40] .00 1.25 [1.14-1.37] .00 nurses from one tertiary center, and nine with former ICU patients, of Colleague believed not skilled enough 1.21 [1.11-1.32] .00 .98 [.83-1.16] .82 Colleague believed to act unsafely 1.23 [1.12-1.34] .00 1.16 [1.00-1.35] .05 whom seven brought one or two of their close family members who had Intercept .24 [.07-.85] .03 been involved in the ICU stay. A semi-structured topic guide was used. Figure 1: Mean values of morally distressing situations pre- and post-peak COVID-19 Interviews were analyzed using Atlas.ti. 0 0,5 1 1,5 2 2,5 3 3,5 4 Results: A total of 18 categories were identified pertaining to the Providing suboptimal care due to insufficient funding, time or staff* struggles of ICU physicians, needs of former ICU patients and their Initiating life-saving actions that I think will only postpone death* family members and the preferred role of ICU nurses (see: table 1). ICU Witnessing that a patient is suffering due to a lack of continuity by caregivers physicians mainly spoke of the struggles they encountered with SDM, Having to work with nurses or doctors who are not skilled enough to care for the patient*** such as the uncertainty about long-term outcomes, time constraints, Witnessing inadequate patient care due to poor communication within the team a prioritization of medical facts and a fear of losing control. Former Continuing treatment for a seriously ill patient with no prospects of recovery patients and family members mainly expressed aspects they missed, Working together with a colleague who, in my opinion, does not provide adequate care such as not feeling included in the ICU process and decisions, as well as a Agreeing to continue life-prolonging treatment event though this is not in the interest of the patient** lack of information about long-term outcomes and recovery. ICU nurses, Having to work with colleagues who I believe are acting unsafely*** who viewed themselves as physician-patient liaisons who provided Conducting examinations and treatments that I think are unnecessary physicians with non-medical information about the patient, among Witnessing that colleagues give ‘false hope’ to patients or their loved ones*** Not discussing the patient’s prognosis with the patient and other things, reported communication struggles with ICU physicians his/her loved ones Providing care to patients who I think should not receive regarding ICU decision making, primarily around end-of-life. care Providing care that does not alleviate the patient’s Conclusions: Interviewed ICU physicians, former patients and family suffering Not reporting anything if a doctor or nurse has made a members and ICU nurses reported struggles, needs and an elucidation mistake that has not been reported** Accepting the request from relatives not to talk about of their current and preferred role in the SDM process in the ICU. This death with a dying patient*** Witnessing medical students using painful procedures on study has corroborated and expanded knowledge to further future patients just to improve their skills Due to fear of a lawsuit, accepting a request from relatives implementation of SDM in the ICU. to provide care** Increasing the dose of morphine given to unconscious patients, which will only accelerate the dying process Table 1: Themes and categories derived from the interviews Ignoring situations in which information for obtaining informed consent is not provided** Not taking action after witnessing an ethically dubious Themes Categories situation Struggles of ICU physicians Uncertainty, interpretability and unpredictability of long-term Pre-COVID-19 Post-peak COVID-19 outcomes Responsibility *p<.05, **p<.01, ***p<.001 Struggles in communication between physicians and nurses Unwillingness of patients to participate in decision making Figure 1: Mean values of morally distressing situations pre- and post-peak Physicians prioritize medical facets COVID-19 What constitutes a good quality of life Trust in patients and families Physicians fear a loss of control Time Needs of former ICU patients and family members Treatment (non-medical sense) EP18 Feeling included in the process and knowing what’s coming Communication about long-term outcomes Coordination and cooperation between medical staff and Shared Decision Making in the ICU: a View patients and families from Three Perspectives Post-ICU and quality of aftercare Role of the ICU nurse Liaison between physician, and patient and family Nina Wubben1, Mark van den Boogaard1, Johannes van der Hoeven1, Translator between physician, and patient and family Marieke Zegers1 Questioning physician’s decision to continue treatment Incorporating non-medical information in ICU decision- 1 Radboudumc, Nijmegen, The Netherlands making Introduction: Shared decision making (SDM) is a cooperative process between clinicians and patients and their family members that enables a way of healthcare decision making that combines both the clinician’s expertise as well as the patient’s values and health goals. Though international critical care societies agree about the importance of SDM, there remains work to be done to achieve its successful implementation in the ICU.

60 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

EP19 Table 1: Perceived barriers and opportunities for involvement in ICU aftercare by GPs. Summary of the barriers and opportunities for increased involvement in ICU Post-Intensive Care Syndrome in General aftercare, based on the outcomes of the questionnaire. Practice; Barriers and Opportunities. Abbreviations: GPs, general practitioners; ICU, intensive care unit; PICS(-F), post- intensive care syndrome(-family). Johan Vlake1, Evert-Jan Wils1, Paul Vlake2, Jasper Van Bommel3, Matilde Hobers3, Diederik Gommers3, Michel Van Genderen3 1Franciscus Gasthuis & Vlietland, Rotterdam, The Netherlands, 2Pax et Bonum, Vught, The Netherlands, 3Erasmus MC, Rotterdam, The Netherlands

Background: A substantial proportion of ICU survivors suffers from physical, cognitive and psychological impairment following ICU treatment, collectively referred to as the post-intensive care syndrome (PICS). Additionally, relatives of these patients are also at risk for psychological distress, referred to as PICS Family (PICS-F). These ICU survivors, and their relatives, have more consultations with their general practitioner (GP) in the period after ICU discharge. We hypothesized that GPs could play a more pivotal role in ICU aftercare, and aimed to identify opportunities and barriers for GPs’ involvement in post-ICU care. Methods: GPs from the Rotterdam-Rijnmond region and randomly selected GPs from other Dutch regions were approached to participate EP20 in this questionnaire-based survey study. The questionnaire used was based on the Knowledge, Attitudes and Practice questionnaire, was validated in a sample of 5 GPs, and consisted of three main subjects; Burnout, resilience and work engagement demographics/general information, the current state of information among Dutch intensivists in the aftermath sharing between the ICU and GP and knowledge about and experience of the COVID-19 crisis: a nationwide survey with the consequences of ICU treatment. Iwan Meynaar1, Thomas Ottens1, Marieke Zegers2, Margo van Mol3, Iwan van der Horst4 Results: A total of 277 out of 769 approached GPs responded. Median 1HagaZiekenhuis, The Hague, The Netherlands, 2Radboud university medical center, age of GPs was 46 years (range: 31-67), 146 (53%) were female and Nijmegen, The Netherlands, 3Erasmus Medical Center, Rotterdam, The Netherlands, 4 GPs had a median of 14 years of experience (1-40). Only 111 GPs (43%) Maastricht University Medical Center+, Maastricht, The Netherlands were familiar with the term PICS, and 143 GPs (56%) with the long- Background: The COVID-19 crisis put a strain on intensive care resources term consequences for relatives of ICU patients, prior to receiving the everywhere in the world increasing the risk of burnout. Previously, the questionnaire. However, 219 GPs (86%) expressed a wish to gain more prevalence of burnout among Dutch intensivists was found to be low. [1] knowledge regarding PICS and PICS-F. Annually, 10 patients (95% Engagement and resilience among intensivists have not previously range, 2-50) per GP were estimated to be admitted to the ICU, but an been studied quantitatively, however, both are related to burnout ICU admission was documented by only 39 GPs (16%). A median of 2 and provide a possible way to mitigate burnout. Our objective was to patients (95% range, 1-10) and 2 relatives (95% range, 0-20) consulted study burnout and its association with work engagement and resilience their GP annually with PICS-related problems. GPs report that they among Dutch intensivists in the aftermath of the COVID-19 crisis. feel responsible for both recognition (62%) as treatment (63%) of Methods: A survey study among the members of the Dutch Society for PICS. Information sharing between the hospital/ICU and the GP in its Intensive Care. The questionnaire consisted of questions on personal present state is judged suboptimal, and may for instance be improved and work-related characteristics; burnout was assessed using the Dutch by informing GPs earlier, provide information targeted for the GP and version of the Maslach Burnout Inventory, the Utrecht Burnout Scale including information regarding PICS in the ICU discharge letter. (UBOS); work engagement using the Utrecht Engagement Scale (UBES); Conclusion: A majority of GPs is not familiar with PICS(-F), feels and resilience using the Resilience Evaluation Scale (RES). responsible for recognition and treatment of PICS, and wants to become Results: In total, 177 responded, resulting in a response rate of 177/649 more involved in ICU aftercare. However, barriers, such as inadequate (27.2%), with 162 complete cases. Of all respondents, 60.5% were male, knowledge, suboptimal information sharing and lack of experience limit 57.4% were 46 years or older, 63.0% had 10 years of experience or more this involvement. Opportunities include educational programs, and and 75.3% worked in a large teaching hospital or an academic hospital. improvement of information provision. According to the UBOS classification rules, burnout was diagnosed in 13 intensivists (8.0%). Out of 162 respondents, 63 (38.9%), 82 (50.6%) and 17 (10.5%) were classified as having a high, intermediate and low work engagement. No significant risk factors for burnout or work engagement were found in the personal or work-related characteristics. We found significantly lower scores for engagement and resilience in

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 61 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

intensivists with burnout than intensivists without burnout. Burnout Conclusion: The findings during cardiac arrest POCUS can be used to and engagement were negatively correlated (Pearson’s R = - 0.706, R2 = decide on starting specific therapeutic modalities, and can provide 49.8%, p < 0.001). Burnout and resilience were also negatively correlated valuable data for prognosis estimation. POCUS-studies have shown (Pearson’s R = - 0.569, R2 = 32.3%, p < 0.001). Resilience and engagement promising results for improving the decision-making process. were positively correlated (Pearson’s R = 0.533, R2 = 28.4%, p < 0.001). Conclusion: In the aftermath of the 2020 COVID-19 crisis, we found a References: 1. Long B, Alerhand S, Maliel K, Koyfman A. Echocardiography in cardiac arrest: raised prevalence of burnout among intensivists, however this is still low An emergency medicine review. American Journal of Emergency Medicine. in international comparisons. Intensivists with burnout scored low on 2018;36:488–493. 2. Lalande E, Burwash-Brennan T, Burns K, Atkinson P, et al. Is point-of-care resilience and low on work engagement. ultrasound a reliable predictor of outcome during atraumatic, non-shockable cardiac arrest? A systematic review and meta-analysis from the SHoC References: investigators. Resuscitation 2019; 139:159–166. 1. Meynaar IA, Saase JLCM Van, Feberwee T, Aerts TM, Bakker J, Thijsse W. Burnout among Dutch intensivists – a nationwide survey. Neth J Crit Care 2016;24:12–7.

EP21

POCUS-skills for the intensive care: quick assessment ultrasound during cardiopulmonary resuscitation Tijmen Gijtenbeek1, David van Westerloo1 1Leiden University Medical Centre, Leiden, The Netherlands

Background: Point of care ultrasound (POCUS) can provide bedside confirmation of several life threatening diagnoses. Figure 1: Subxiphoid view, showing both the atria, ventricles and the pericardium. Figure 1: Subxiphoid view, showing both the atria, ventricles and the During advanced life support, starting timely specific therapies can pericardium. improve outcome. In this practical review we describe how POCUS can be applied during cardiopulmonary resuscitation. Methods: The most suitable moment for obtaining the preferred subxiphoid view is during rhythm checks (Fig 1,2). This timeframe can be improved by lodging the phased array probe into the subxiphoid space in anticipation. Other cardiac views generally interfere more with the resuscitation procedure due to their location. Results: The likelihood ratio of diagnosing pulmonary embolism can be increased by POCUS. This can reduce the time to therapy. Cardiac tamponade can be easily visualised during POCUS examination, with a sensitivity of 96% and specificity of 98%.1 Ultrasound can be used to diagnose hypovolemia irrespective of the underlying cause. Tension pneumothorax can be diagnosed with a sensitivity of 92% and specificity Figure 2: Subxiphoid view: inflow from the vena cava into the right atrium. Figure 2: Subxiphoid view: inflow from the vena cava into the right atrium. of 99%.1 When a ‘lung point’ is found, the sensitivity of pneumothorax is close to 100%.1 Abdominal ultrasound can be used to detect free fluid. EP22 The ultrasound sensitivity and specificity for ruptured aortic aneurysm are 99% and 98% respectively.1 POCUS can be used for estimating the prognosis during resuscitation Establishing the glucosuria threshold in efforts. In a systematic review, visible cardiac activity on ultrasound was glucose-controlled ICU patients associated with on odds ratio for survival to hospital discharge of 8.03 Maxim H. de Rooij MD1, Maarten W. Nijsten MD PhD1 (CI-95%:3.01–21.39) compared to absence of cardiac activity.2 Cardiac 1University Medical Center Groningen, Groningen, The Netherlands combined with carotid ultrasound was able to better distinguish sustained circulation from pseudo-PEA and pseudo-asystole than Department of Critical Care, University Medical Center Groningen, standard ECG and pulse palpation in several small studies. Groningen, The Netherlands The risk of delays in rhythm checks during POCUS, can be reduced with Background: Glucosuria is frequently observed in critically ill patients. training. In healthy persons glucosuria occurs only when the blood glucose rises

62 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

above a threshold of 10 mmol/L. It has been claimed that glucosuria is EP23 not quantitatively related to renal function in critically ill patients[1]. Since glucosuria may also constitute an marker of proximal tubular Severe phosphine poisoning and the added dysfunction we studied the relation between blood glucose (BG), value of high-dose insulin euglycemia therapy. urinary glucose excretion (UGE) and measured creatinine clearance Johanna de Groot1, Anne Kan1, Maaike Sikma1, Martine Otten2, Suzette Brouwers2 (mCC) in a large cohort of patients. 1University Medical Center Utrecht, Utrecht, The Netherlands, Methods: Patients who were at least a whole day admitted to our 2Diakonessen Hospital, Utrecht, The Netherlands ICU between 2016 and 2020 were examined on the first day after ICU admission. The 24-hour urine was collected and was assessed for UGE and Phosphine, the active gaseous ingredient of aluminium phosphide, is an urinary creatinine excretion to determine the mCC. Only patients with extremely toxic compound used as a fumigant pesticide.1 Mortality may an mCC of >30 ml/min were included. Glucose regulation was performed occur within one hour of exposure to a phosphine concentration of 3.6 according a computer-assisted glucose control protocol [2]. Mean BG, ppm. Toxicity is mediated through 3 pathways; (1) acetylcholinesterase maximal BG and mean insulin dose were determined for each patient. inhibition, (2) disruption of mitochondrial adenosine triphosphate (ATP) Results: 3565 patients (66% men) with a mean±SD age of 60±16 years production, and (3) generation of reactive oxygen species, causing DNA were included. damage and cell death (Figure 1).1 The mean±SD BG was 7.7±1.3 mmol/L with a mean±SD insulin dose of Supportive care is key, focusing on limiting mitochondrial dysfunction 1.2±1.7 U/h. and enhancing acetylcholine neurotransmission.1 We present a case The mean±SD mCC was 94±48 ml/min and the mean±SD UGE was of severe phosphine poisoning and discuss the added value of HIET. 5.4±28.6 mmol/day. The median (IQR) UGE was 0.9 (0.53 – 1.99) mmol/d. A healthy 39-year-old woman experienced gastrointestinal symptoms Multivariate linear regression analysis with mCC, mean BG, max BG and five hours after loading sunflower pellets from a bulk carrier into her insulin dose as independent variables and UGE as dependent variable, cargo vessel. Eighteen hours later, she was admitted to the hospital with only showed mean BG and mCC as significantly related with UGE cardiogenic shock; blood pressure 79/55 mmHg, heart rate 144/min, (P<0.0001). When we assessed the relation between mean BG and UGE lactic acidosis (lactate 8.6 mmol/L), elevated kidney and liver enzymes, in an univariate analysis (figure), a threshold between 8 and 10 mmol/L and a left ventricular ejection fraction <15%. was observed, above which UGE sharply increased. Univariate analysis In the ICU fluid resuscitation, inotropes, magnesium sulphate, of the relation between mCC and UGE (figure not shown) showed a N-acetylcysteine and L-carnitine were started, based on measured significant more linear positive correlation between mCC and UGE. phosphine concentrations of 20 ppm on the cargo vessel by the Conclusion: The blood glucose threshold for renal glucose excretion Dutch Environmental Service. Mechanical ventilation, veno-arterial that we observed in ICU-patients is similar to that seen in healthy extracorporeal membrane oxygenation (VA-ECMO) and intra-aortic persons. In contrast to reported earlier, our much larger study shows a balloon pump (IABP) were initiated. HIET was started with an insulin clear relation between urine glucose excretion and creatinine clearance. infusion rate of 4 IU/kg/h. Within five days, cardiac function improved significantly. All supportive treatments were discontinued. Two months References: 1. Brunner, R., Kurz, A., Adelsmayr, G., et al. Basal glucosuria in critical care. later, echocardiography showed full recovery. Nephrology, 2015;20:293-296. Normally, myocytes use free fatty acids to generate ATP. During cardiac 2. Vogelzang M., Loef B.G., Regtien J.G., et al. Computer-assisted glucose control in critically ill patients. Intensive Care Med. 2008;34(8):1421-1427. stress, myocytes switch to glucose. When tissue perfusion diminishes, the glucose delivery to myocytes decreases, leading to hemodynamic deterioration and cell death.3 Large doses of insulin, enabling glucose to enter cells, improves cell metabolism and restores myocardial contractility.3 Low insulin doses (0.2-0.5 IU/kg/h) improve the outcome in phosphine poisoning.2 However, doses up to 10 IU/kg/h are used in beta-blocker and calcium channel-blocker intoxications.3 Based on comparable mechanisms of toxicity, higher insulin doses could be more effective in severe phosphine poisoning. Although it is impossible to assess the clinical efficiency of HIET from a single case report, our patient fully recovered and we showed that insulin doses of 4 IU/kg/h can be safely administered.

References: 1. Nath NS, et al. Mechanisms of Phosphine Toxicity. J Toxicol. 2011;494168.1- 494168.9 2. Pannu AK, et al. Glucose-insulin-potassium infusion for the treatment of acute aluminum phosphide poisoning: an open-label pilot study. Clin Toxicol. Figure 1: Relation between mean blood glucose and urinary glucose excretion 2020;58(10):1004-1009. (UGE; ± standard error), demonstrating that above a threshold between 8 and 10 3. Engebretsen KM, et al. High-dose insulin therapy in beta-blocker and calcium mmol/L UGE rapidly increases (P<0.0001). channel-blocker poisoning. Clin Toxicol. 2011;49(4):277-283.

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 63 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

A. recovered but shortly afterwards she developed a SE. Treatment with Cholinergic synapse a combination of antiepileptic drugs and sedation (figure 1.) was started locally, but as soon sedation was tapered, SE returned. She was transferred to our hospital at day 12. After failure of many attempts, including treatment with barbiturates, continuous infusion of ketamine B. combined with midazolam was started. This infusion scheme was Cytosol continued for 72 hours with simultaneous administration of pregabalin, levetiracetam and lacosamide. After discontinuation of ketamine and midazolam, the patient regained consciousness and made a complete neurological recovery. Repeated electroencephalography (EEG) showed no epileptic activity. Discussion: In approximately 22% of patients with postanoxic encephalopathy after successful cardiopulmonary resuscitation, an electroencephalographic SE is described.1 There is no evidence for the optimal pharmacological treatment strategy. Furthermore, it is uncertain whether successful treatment of the SE will also lead to a good neurological outcome. Treatment of a SRSE, defined as continuous or recurrent seizures lasting 24 hours following anesthetic medications, is even more doubtful. However, recent studies indicate that some patients achieve a good neurologic outcome with a timely diagnosis and appropriate treatment consisting of a combination of antiepileptic Figure 1: Mechanisms of phosphine toxicity drugs and aneasthetics.1 The EEG patterns and other prognostic factors A. Phosphine inhibits acetylcholinesterase, which is responsible for clearing acetylcholine from the cholinergic synapse by converting it into inactive can probably help to identify the patients with a chance of a good metabolites: choline and acetate. This results in accumulation of acetylcholine outcome. The main reason to treat our patient extensively was the short causing excessive acetylcholine signalling. B. Phosphine disrupts the ability of the mitochondria to produce adenosine period of recovery of consciousness immediately after TTM treatment. triphosphate (ATP) through inhibition of oxidative phosphorylation by non- As all combinations failed, we decided to give her a last chance with competitive blockage of cytochrome c oxidase (CytC). Subsequently, phosphine ketamine. Our experience with ketamine for this indication is very stimulates the generation of reactive oxygen species (ROS). It does so by directly increasing hydrogen peroxide due to induction of superoxide dismutase limited, but it has been suggested that ketamine and benzodiazepines (SOD) and inhibition of peroxidase and catalase, and indirectly by inhibition of have a strong synergistic effect in SE.2 glycerophosphate dehydrogenase and blockage of cytochrome c oxidase. This causes cellular oxidative stress, leading to cell dysfunction and death. Conclusion: SRSE after cardiac arrest is rare and often leads to a poor ADP/ATP: adenosine di/triphosphate, Complex IV: cytochrome c oxidase, CoQ: outcome. Our patient shows that with aggressive treatment, a good coenzyme Q, CytC: cytochrome C, FAD/FADH2: Flavin adenine dinucleotide outcome can be achieved. Identification of the patients with a chance oxidized/reduced, NAD+/NADH: nicotinamide adenine dinucleotide oxidized/ reduced, ROS: reactive oxygen species, SOD: superoxide dismutase, TCA: of a good recovery needs to be improved. tricarboxylic acid cycle. References: 1. Beretta S, et al. Epilepsy & Behavior 101 (2019) 106374 2. Holtkamp M. Drugs (2018)78:307-326 EP24

Super-refractory status epilepticus after cardiac arrest: a case report of successful treatment with ketamine Mariëlle Blacha1, Daniël van Poppelen1, Daan Velseboer1, Janneke Horn1 1Amsterdam UMC, location AMC, Amsterdam, The Netherlands

Status epilepticus (SE) after cardiac arrest is strongly associated with a poor outcome. Whether treatment can improve outcome remains Figure 1: time schedule of all different antiepileptic and anaesthetic frugs used. uncertain. Here we present a patient who was treated successfully using Day 1 = day of admission after cardiac arrest. Day 2 = day of refferal to our tertiary centre an aggressive approach. v = variable doseges * = loading dose Case: A previously healthy 48-year old woman was admitted to our tertiary clinic with super-refractory status epilepticus (SRSE) after a cardiac arrest due to ventricular fibrillation (VF). After targeted temperature management (TTM) and sedation her consciousness

64 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

EP25

Acute respiratoiry failure after caesarean section Jasmijn Selten1, Walter van den Tempel1, Susanne Stads1 1Ikazia Hospital, Rotterdam, The Netherlands

Introduction: Acute respiratory failure after childbirth can be provoked by many life-threatening conditions. It is a diagnostic challenge and a multidisciplinary approach is essential. We present an intriguing case with failure to wean. Presentation: A healthy 29 year old primiparous woman presented at our hospital at 36 weeks gestation. She had headaches, hypertension, low platelets and proteinuria. HELLP (Hemolysis Elevated Liver Enzymes and Low Platelets) was diagnosed and treated. Maternal deterioration necessitated an emergency ceaserian section under general anesthesia. A healthy boy was born during an uncomplicated procedure with no respiratory or hemodynamic problems. However, minutes after extubation, her breathing was strenuous, saturation declined and she was immediately reintubated. Chest X-ray showed extensive atelectasis bilaterally, a CT-scan excluded a pulmonary embolism and bronchoscopy was normal. Creatine kinase levels were elevated up to 15.000 U/L but troponin levels remained normal. The ECG showed no signs of ischemia Figure 1: Hydrocephalus with enlarged central ventricles without signs of or right ventricular overload and ultrasound of the heart showed a bleeding or ischemia normal left ventricular function. A CT-scan of the brain showed extensive ventriculomegalia without herniation; intracranial pressure and spinal tap were normal. Neurological consultation concluded no clear signs EP26 of muscle weakness. Ventilator settings improved after administration of diuretics. However, two consecutive attempts at extubation failed within minutes due to shallow and strenuous breathing. Acute tamponade from intracardiac cement Elaborate anamnesis from the patients parents revealed she experienced embolus after vertebroplasty a declining physical fitness since approximately 3 years. Her parents P.H. Schouwenberg1, M. ter Horst1, F.R. van Schaagen1, F.B. Oei1, C.A. den Uil1, W.B. Meijboom declared that our patient had an excessive daytime sleepiness, swallowing 1Erasmus MC, Rotterdam, The Netherlands, 2IJsselland Hospital, difficulties and had a mouthpiece for an obstructive sleep apnea. Capelle aan den Ijssel, The Netherlands Diagnostic procedure: Electromyography (EMG) showed characteristics of Dystrophia Myotonica (Morbus Steinert). Genetic testing confirmed A 70-year-old male with a recent history of thoraco-lumbar the diagnosis. vertebroplasty was admitted with dyspnea. Echocardiography revealed Follow up: A third attempt at extubation directly followed by non- pericardial effusion and right-sided intracardiac high density elongated invasive ventilation also failed. A tracheotomy was performed and structures (Fig. 1a), confirmed by CT scan (Fig. 1b). The patient developed she started weaning of the ventilator and sedatives. After 4 weeks of a classic cardiac tamponade. Pericardiocentesis was performed but intensive physiotherapy, she was decannulated and transferred home; characterized by persistent drainage of blood. During open heart she remains NIV dependent during the night. surgery (Fig. 1c), several intracardiac emboli (ICE), apparently venous Conclusion: Failure to wean after caesarean section due to luxation of a cement casts (Fig. 1d) were evacuated from the right atrium (causing pre-existing undiagnosed Dystrophia Myotonica after general anesthesia. perforation) and the right ventricle. Discussion: always be concerned about the possibility of undiagnosed Vertebroplasty is a widely used treatment of symptomatic vertebral muscular disorders when administering neuromuscular blocking compression fractures. During this procedure, low-viscosity bone agents.1,2 cement is injected in the damaged vertebra.[1] Migration of bone cement to the right heart may occur after accidental cement leakage References: in the paravertebral veins, migrating into the azygos venous system 1. Owen et al. Emergency caesarean section in a patient with myotonic dystrophy: a case of failed postoperative extubation in a patient with mild disease, Anaesth or directly into the vena cava. Being a rare complication, however, Intensive Care, 2011; 293-298 cardiorespiratory symptoms following vertebroplasty should prompt 2. Johnson et al. The Impact of Pregnancy on Myotonic Dystrophy: A Registry- [2] Based Study, Journal of Neuromuscular Diseases, 2015; 447–452 urgent echocardiography to detect ICE.

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 65 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

References: initiated. Ventilatory settings were decreased to minimal settings (3 1. Buchbinder R. et al. Percutaneous vertebroplasty for osteoporotic vertebral compression fracture. Cochrane Database of Systematic Reviews 2018, Issue ml/kg) and prone position was continued in combination with daily 11. Art. No.: CD006349. bronchoscopy for sputum evacuation. Antibiotic therapy was expanded 2. Hassani S.F. et al. Intracardiac cement embolism during percutaneous to include linezolid 600 mg BID in addition to intravenous vancomycin. vertebroplasty: incidence, risk factors and clinical management. European Radiology 2019 29: 663–673. After 16 days, clinical response was absent, with persistent MRSA positive respiratory and cultures and extension of bilateral consolidations with new cavitary lesions. In a multidisciplinary meeting it was decided to add vancomycin nebulization and to increase antibiotics dosage during ECMO treatment. Cultures became negative, with declining consolidations and cavitation’s followed by pulmonary improvement. 56 days later, he was free of ventilator support, he was soon discharged home, currently doing well. Discussion: A community acquired NP- MRSA is a rare condition in the Netherlands, aside from COPD, no other risk factors were identified in our patient. Available antibacterial agents for the treatment of NP-MRSA include glycopeptides (vancomycin and teicoplanin) and linezolid. Although SWAB guidelines advise Vancomycin as first choice (intravenous) therapy, there are concerns about its limited intrapulmonary penetration in ICU patients and Linezolid demonstrated superiority to Vancomycin in clinical efficacy1. For this reason it was

decided to add Linezolid but evidence for combination therapy is FigFigure 1 1: lacking. Toxic effects of Linezolid were not observed in our patient. In a Echocardiography subxyphoid view, bone cement cast in atrial wall (arrow) a Echocardiography subxyphoid view, bone cement cast in atrial wall (arrow) b Axial heart CT with right atrial addition, awareness of altered pharmacokinetics during ECMO support contrastb Axial and heart intra cardiac CT with cement right embolus, atrial ICEcontrast (arrow) and c ste rnotomyintra cardiac with right cement atrial perfo embolus,ration (arrow) ICE d two retrieved(arrow) polymethylmethacrylate (PMMA) bone cement casts and therapeutic drug monitoring (TDM) to guide antibiotic dosing in *c = pericardialsternotomy effusion with right atrial perforation (arrow) d two retrieved polymethylmethacrylate (PMMA) bone cement casts this patient population appears useful2. As well accurate source control * = pericardial effusion by continuous drainage of purulent collections is important. Conclusion: Beyond our expectations, this patient fully recovered from a NP- MRSA due to prolonged broad antibiotic treatment with vancomycin and linezolid combined with intensive drainage of purulent EP27 collections. A case of rapidly progressive community acquired necrotizing pneumonia caused by Methicillin-Resistant Staphylococcus aureus (NP-MRSA)

1 1 Jeanine van Dam-Kastelein1 , Jeannine Hermens 1UMC Utrecht, Utrecht, The Netherlands

Case: A 52-yo man, with a history of GOLD III chronic obstructive pulmonary disease (COPD) was presented at a peripheral hospital with dyspnea and fever for 4 days. Chest X ray showed a large consolidation in the right lung, for he was treated with ceftriaxone. Surprisingly, sputum cultures revealed MRSA strains on day 3 for which antibiotic treatment was escalated to vancomycin. Atypical pathogens and viral panels including COVID-19 were all negative. Despite this antibiotic regime, peaking fever persisted and his respiratory status rapidly declined necessitating mechanical ventilation on day 10. Pulmonary CT Figure 1: Computed Tomography: severe cavitary lung destruction, after 17 days of therapy. scan showed large cavitating lesions in the context of necrotizing MRSA pneumonia. After 10 days of conventional ventilatory management including prone position, progressive severe hypercapnia and hypoxia developed and he was transferred to our center where ECMO was

66 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care Abstracts & Case Reports Intensivistendagen 2021

EP28 EP29

An uphill battle in a case: efficacy of digital A case of Streptococcal Toxic Shock subtraction angiography for stopping Syndrome rebleeds after acute non-traumatic A.W.M. Stolwijk1, D.C.J.J. Bergmans1 subdural haemorrhage 1Department of Intensive Care, University Medical Centre+ Maastricht, The Netherlands Tijmen Gijtenbeek1, Jort van Gent1, Radboud Koot1, Jacinta Maas1 1Leiden University Medical Hospital, Leiden, The Netherlands Introduction: Streptococcal Toxic Shock Syndrome (STSS) is a rare but life-threatening disease caused by an infection with Group A Background: Acute non-traumatic subdural hemorrhage (ANSDH) Streptococcus (GAS). The Streptococcal pyrogenic exotoxins can caused by a dural arterial-venous fistula (DAF) is a rare event, with high activate the immune system by bypassing the usual antigen-mediated mortality rates. Digital subtraction angiography (DSA) is an angiographic immune response sequence, resulting in a release of large quantities of procedure that can succesfully embolise DAF in 92.6% of elective inflammatory cytokines. This causes capillary leak and tissue damage patients.1 This case-report illustrates a patient for whom DSA was leading to refractory distributive shock and multi organ failure. It is undertaken in an attempt to stop recurrent intracranial hemorrhages associated with high mortality, despite aggressive therapy. To suspect after ANSDH. and diagnose STSS at an early stage is crucial in deciding the most Methods: A 75-year-old male patient presented with a first episode of effective and early treatment because of the beneficial effect on ANSDH. Decompressive craniotomy was performed urgently. In total outcome1. five craniotomies were for recurrent intracranial hemorrhages. The first Case report: A 63-year-old man with a medical history of hypertension DSA was at day 4. Repeat DSA at day 5 revealed and embolised the was admitted to the emergency department with complaints of fever, culprit DAF. Within 12 hours of this procedure, the patient died from coughing, dyspnea, swollen throat, diarrhea and oliguria. His first new cerebellar hemorrhages. In total, three case reports were found symptoms started about two weeks earlier after a vacation to Istanbul. with similar presentations and good post-DSA prognosis. The cause for His blood pressure was 100/60 mmHg, heart rate 150 beats/min and the fatal cerebellar hemorrhage in our case could be related to the DSA. temperature 39.7°C. The operator dependency of the angiographic technique may have He was admitted to the ICU with the diagnosis of sepsis of unknown played a role in missing the fistula at first. The outcome may have been origin and treated with antibiotics (Ceftriaxone and Ciprofloxacin), better in our case if DSA was performed earlier. Cerebellar hemorrhages volume resuscitation, low dose Norepinephrine and oxygen therapy. have not been reported before as a potential side effect, possibly due to Extended microbiological testing was performed. Laboratory results publication bias. revealed a dramatic increase in leukocyte count and C-reactive protein, Conclusions: Although evidence from literature is limited, DAFs can be increased sedimentation rate, elevated blood urea nitrogen, creatinine, succesfully embolised during DSA, preventing rebleeds after ANSDH. It creatinine kinase, liver enzymes, bilirubin and lactate levels. A few hours is therefore recommended to consider performing DSA in an early stage later, after a sudden cardiac arrest (PEA) and ROSC in four minutes, his after ANSDH. The risk of adverse events from DSA need to be weighted condition deteriorated fast with refractory distributive shock, despite against the potential benefits. aggressive medical supportive therapy. During a second cardiac arrest and under the suspicion of abdominal compartment syndrome the References: surgeon performed an emergency decompressive laparotomy; only 1. Chen CJ et al. Neurosurg Focus 2018; 45: 1-7. a large amount of intra-abdominal fluid was visible without signs of ischemia of the intestines. Despite mechanical ventilation, high doses of inotropic, antibiotics, hydrocortisone and renal replacement therapy, the patient died 16 hours after admission to the hospital after a third cardiac arrest with asystole. The next day, all microbiology results were negative, except one blood culture which showed the presence of Streptococcus pyogenes. Conclusion: This case was presented to draw attention to STSS and illustrates the importance of early recognition and treatment of this rare clinical entity because of rapid progression to mortality despite aggressive therapy1.

References: 1. Schmitz, M., Roux, X., Huttner, B. et al. Streptococcal toxic shock syndrome in the intensive care unit. Ann. Intensive Care 8, 88 (2018). Figure 1: Close-up angiographic view of the fistula connecting the middle meningealFigure 1: arteryClose-up to a duralangiographic vein (arrow). view of the fistula connecting the middle meningeal artery to a dural vein (arrow).

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 67 Netherlands Journal of Critical Care TitelKlein NVIC CONFERENCE AND COURSE AGENDA

COURSES AND CONFERENCES | NVIC

NVIC Basiscursus Echografie Wednesday 17 March - Thursday 18 March 2021 Monday 4 October - Tuesday 5 October 2021 Tuesday 15 June - Wednesday 16 June 2021 Tuesday 12 October - Wednesday 13 October 2021 Wednesday 8 September - Thursday 9 September 2021 Wednesday 3 November - Thursday 4 November 2021 Tuesday 9 November - Wednesday 10 November 2021 Wednesday 24 November - Thursday 25 November 2021 Tuesday 14 December - Wednesday 15 December 2021 NVIC Cursus Luchtwegmanagement op de IC Monday 19 April - Tuesday 20 April 2021 NVIC Intensivistendagen Tuesday 16 November - Wednesday 17 November 2021 Thursday 4 February – Friday 5 February 2021

NVIC FCCS cursus NVIC Najaarscongres Tuesday 9 February - Wednesday 10 february 2021 Thursday 16 September 2021 Thursday 4 March - Friday 5 March 2021 Tuesday 30 March - Wednesday 31 March 2021 Webinar AI op de IC Wednesday 14 April - Thursday 15 April 2021 Friday 19 March 2021 Wednesday 19 May - Thursday 20 May 2021 Monday 31 May - Tuesday 1 June 2021 Wednesday 9 June - Thursday 10 June 2021 Tuesday 29 June - Wednesday 30 June 2021 COURSES AND CONFERENCES | OTHER Tuesday 6 July - Wednesday 7 July 2021 Tuesday 31 August - Wednesday 1 September 2021 EuroNeuro 2021 - Parijs Wednesday 22 September - Thursday 23 September 2021 Wednesday 14 April 2021 - Friday 16 April 2021

A THANK YOU TO REVIEWERS

The following colleagues have devoted their time to review for the Netherlands Journal of Critical Care in 2020

H. Aardema D. Dos Reis Miranda D. Kleinveld M. Platenkamp S. Achterberg M. van Eijk H. Knoester A. de Pont S. Akin E. van Essen M. Kok P. Roekaerts A. Akkermans P. de Feiter E. Koomen P. Roeleveld M. Aries T. Gartner J. Kuijlen M. Sikma I. Bartelink R. Gerritsen M. Kuiper M. Smit R. Bem T. Gijtenbeek M. Kuizenga P. Spronk W. van den Bergh B. Grady N. Kusadasi M. Strachinaru D. Bergmans J. de Haan D. de Lange U. Strauch Y. Bilgin J. van der Heijden G. de Lange D. Tjwa W. Boer S. Heines M. Lont E. Torn C. Bollen J. Hermens I. van Loo H. Touw M. Boogaard J. Hjortnaes J. Maas J. Tulleken F. Bosch J. Horn R. Mahomedradja D. Velseboer M. Boulaksil I. van der Horst L. van Manen B. Vermin E. Bredewold A. van der Horst-Schrivers J. Lopez Matta A. Vlaar J. van den Brule M. Houwert R. Mauritz P. van Vliet H. Buter N. Hunfeld C. Meuwese M. Voskuil R. Crane M. van der Jagt H. Moeniralam S. Vreeswijk M. van Dam V. de Jong F. Nijhoff S. Wilmer J. van Dam-Kastelein E. de Jonge T. Olgers J. van Woensel D. Dekker B. Kapitein M. Onrust T. Delnoij J. Karemaker A. Oude Lansink-Hartgring L. Derde S. Karlsson Stafseth K. Padilha

We wholeheartedly thank the reviewers for their hard work, thereby keeping up the standards and quality of the Netherlands Journal of Critical Care. Dirk Donker, Editor in Chief

68 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care

Editorial board of the Netherlands Journal of Critical Care Dirk Donker, Editor in Chief Maarten van Eijk, Associate Jozef Kesecioglu, Section Editor Ilse van Stijn, Managing Editor Dept. of Intensive Care Medicine, Section Editor Anesthesiology Ethics Dept. of Intensive Care Medicine Div. of Anesthesiology, Intensive Dept. of Intensive Care Medicine Dept. of Intensive Care Medicine OLVG Care and Emergency Medicine University Medical Center Utrecht University Medical Center Utrecht PO Box 95500 University Medical Center Utrecht PO Box 85500 PO Box 85500 1090 HM Amsterdam PO Box 85500 3508 GA Utrecht 3508 GA Utrecht

3508 GA Utrecht Eleonora Swart, Section Editor Janneke Horn, Section Editor Michael Kuiper, Section Editor Pharmacology Walter van den Bergh, Section General Neurology Dept. of Pharmacy Editor General Dept. of Intensive Care Dept. of Intensive Care Medicine Amsterdam UMC location AMC Dept. of Critical Care Amsterdam UMC location AMC Medical Center Leeuwarden University of Amsterdam University of Groningen University of Amsterdam PO Box 888 Meibergdreef 9 Hanzeplein 1 Meibergdreef 9 8901 BR Leeuwarden 1105 AZ Amsterdam 9700 RB Groningen 1105 AZ Amsterdam Dept. of Clinical Pharmacology and Nuray Kusadasi, Associate Section Dennis Bergmans, Section Editor Can Ince, Section Editor Physiology Pharmacy Editor Hemato-Oncology Infection and Inflammation Dept. of Physiology Amsterdam UMC location Vumc Dept. of Intensive Care Medicine Dept. of Intensive Care Amsterdam UMC location AMC De Boelelaan 1117 University Medical Center Utrecht Maastricht University Medical University of Amsterdam 1081 HV Amsterdam PO Box 85500 Center+ Meibergdreef 9 3508 GA Utrecht Pieter Roel Tuinman, Section P. Debyelaan 25 1105 AZ Amsterdam Editor General 6229 HX Maastricht Evert de Jonge, Section Editor Christiaan Meuwese, Associate Dept. of Intensive Care Medicine Frank Bosch, Section Editor Scoring and quality assessment Section Editor Cardiology Amsterdam UMC location Vumc Imaging Dept. of Intensive Care Medicine Dept. of Intensive Care Medicine PO Box 7057 Dept. of Internal Medicine Leiden University Medical Center University Medical Center Utrecht 1007 MB Amsterdam Rijnstate Hospital PO Box 9600 PO Box 85500 PO Box 9555 2300 RC Leiden 3508 GA Utrecht David van Westerloo, Section 6800 TA Arnhem Editor General Nicole Juffermans, Section Editor Marike van der Schaaf, Section Dept. of Intensive Care Medicine Diederik van Dijk, Section Editor Hemostasis and Thrombosis Editor Rehabilitation Leiden University Medical Center Cardioanesthesia Dept. of Intensive Care Dept. of Rehabilitation PO Box 9600 Dept. of Intensive Care Medicine, Amsterdam UMC location AMC Amsterdam UMC location AMC 2300 RC Leiden Div. of Anesthesiology, Intensive University of Amsterdam University of Amsterdam Care and Emergency Medicine Meibergdreef 9 Meibergdreef 9 Job van Woensel, Section Editor University Medical Center Utrecht 1105 AZ Amsterdam 1105 AZ Amsterdam Pediatrics PO Box 85500 Pediatrics Intensive Care Unit 3508 GA Utrecht Emma children’s hospital Amsterdam UMC location AMC University of Amsterdam Meibergdreef 9 1105 AZ Amsterdam

International advisory board Jan Bakker Charles Hinds Paul Marik Xavier Monnet Professor of Intensive Care Professor of Intensive Care Associate Professor Service de réanimation médicale Columbia University Medical Medicine Dept. of Medicine and Medical Centre Hospitalier Universitaire Center, New York University St. Bartholomew’s Hospital Intensive Care Unit de Bicêtre Medical Center, New York, USA West Smithfield, University of Massachusetts Le Kremlin-Bicêtre, France Pontificia Universidad Católica de London, UK St. Vincent’s Hospital, Jean-Charles Preiser Chile, Santiago, Chile Worcester, USA Patrick Honoré Dept. Intensive Care Erasmus MC University Medical Professor of ICU Medicine Greg Martin Erasme University Hospital Center, Rotterdam, Netherlands Director of Critical Care Dept. of Medicine Brussels, Belgium Charles Gomersall Nephrology Platform Division of Pulmonary, Allergy and Yasser Sakr Dept. of Anaesthesia and Intensive ICU department Critical Care Dept. of Anaesthesiology and Care Universitair Ziekenhuis Brussel, Emory University School of Intensive Care The Chinese University of Hong VUB University Medicine Friedrich-Schiller University Kong, Prince of Wales Hospital Brussels, Belgium Atlanta, USA Hospital Hong Kong, China Alun Hughes Ravindra Mehta Jena, Germany Frank van Haren Professor of Clinical Pharmacology Professor of Clinical Medicine Hannah Wunsch A/ Professor, Australian National University College London Associate Chair for Clinical Dept. of Anaesthesia University Medical School London, UK Research New York Presbyterian Columbia Department of Intensive Care Department of Medicine Manu Malbrain New York, USA Medicine UCSD Medical Centre Dept. of Intensive Care Unit The Canberra Hospital San Diego, USA Hospital Netwerk Antwerp PO Box 11, Woden, ACT 2606 Campus Stuivenberg Canberra, Australia Antwerp, Belgium

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 69 Netherlands Journal of Critical Care

Information for authors

The Netherlands Journal of Critical Care (NJCC) is the official journal of Editorials the Dutch Society of Intensive Care (Nederlandse Vereniging voor Intensive Editorials are always commissioned by the Editors and comment on one or Care-NVIC). The journal has a circulation of about 1750 copies bimonthly more articles in the same issue of the Journal or to a subject with high news in the Netherlands and Belgium. value. Editorials should not exceed 1500 words and may include up to 15 High-quality reports of research related to any aspect of intensive care references. Editorials have a maximum of 3 authors and no abstract. Please medicine, whether laboratory, clinical, or epidemiological, will be considered provide 2-3 key words. for publication in the NJCC. This includes original articles, reviews, and meta-analyses. Case reports, clinical images, book reviews, editorials, letters Case reports to the editor, clinical problem solving, research news and correspondence are The text of a case report should include an abstract, introduction, case also welcome. All manuscripts pass through an independent review process report/case history, discussion, tables and figures (2 in total), and references. managed by the editorial board. The journal does not have any publication The main text may be up to 2000 words; the abstract should not exceed 150 fees, and colour figures are reproduced free of charge. words and may be unstructured. Please provide a minimum of 3 keywords The journal is indexed by Embase, Emcare and Scopus. A Medline annotation and a list of not more than 30 references. Please include an informed consent is in preparation. statement from the patient described in the case.

Ethical standards Clinical problem-solving Manuscripts reporting original research must contain a statement that all These manuscripts consider the step-by-step process of clinical decision- human and animal studies have been approved by the appropriate ethics making. Information about a patient is presented to an expert clinician committee and have therefore been performed in accordance with the or clinicians in stages (indicated by boldface type in the manuscript) to ethical standards laid down in the 1964 Declaration of Helsinki and its later simulate the way such information emerges in clinical practice. The clinician amendments. Where appropriate, it should also be stated in the text that all responds (in regular type) as new information is presented, sharing his or her persons gave their informed consent prior to their inclusion in the study. If reasoning with the reader. The text should not exceed 2500 words, and there possible, the Journal aims to also include informed consent of individuals should be no more than 15 references. Please include an informed consent described in case reports. Details that might disclose the identity of the statement from the patient described in the case. subjects under study should be omitted. The editors reserve the right to reject manuscripts that do not comply with Research news the above-mentioned requirements. The author will be held responsible for Research news should be a review of a manuscript which has appeared in the false statements or failure to fulfil the above-mentioned requirements. The past two months. It contains sections on why this study was done, the research editors adhere to the guidelines laid down by the International Committee question, how this was investigated, conclusions and the impact of the study of Medical Journalist Editors (www.icmje.org), concerning authorship and on clinical practice. The text should not exceed 800 words with a maximum scientific conduct. of 5 references. Contributions for this section will be commissioned; however, inquiries about contributions can be sent to [email protected]. Types of papers The following manuscript types are considered for publication: original Clinical images articles, review articles, systematic reviews, meta-analyses, editorials, case A clinical image should contain one or two pictures with a legend and a short reports, clinical images, book reviews, letters to the editor, clinical problem case history, and should preferably not be referenced. The manuscript should solving, research news and correspondence. succinctly present relevant clinical information, including a short description of the patient’s history, relevant physical and laboratory findings, clinical course, Original articles response to treatment (if any), and condition at last follow-up. Please provide a Original articles should not exceed 3000 words (excluding abstract, minimum of 3 keywords. The text should not exceed 500 words. Please include references, tables and legends). The manuscript should be clear in outline an informed consent statement from the patient described in the case. (with subheadings) for maximum clarity. The text should follow the IMRAD format and contain an abstract, introduction, materials (patients) and Photo quiz methods, results, discussion section and references. This should be followed In this section relevant images for critical care medicine (e.g. flow and pressure by tables and figures (maximum of 6 in total) with titles above and legends curves of mechanical ventilation or haemodynamic indices, radiological images below these elements. The abstract should not exceed 250 words and should or laboratory results) will be accompanied by a short introduction of the context. be structured: background, methods, results, conclusions. Do not include The introduction will be followed by ‘what is your diagnosis?’. The answer will references in the abstract. Authors should provide a minimum of 3 keywords, include a brief discussion of the literature. A photo quiz should not exceed 500 a running title, and a list of not more than 30 references. words and contain no more than two figures, and five references conform the Original articles must meet the following criteria: Vancouver style. Abbreviations of measurements should be quoted in SI units. - the manuscript presents the results of primary scientific research; - the results have not been published in full elsewhere; Book reviews - analyses are described in full in the manuscript; A book review should not exceed 300 words. Please mention in the header: - conclusions are presented in a clear and concise manner and are supported title, author, edition and year. Scan the cover in high resolution (300 dpi/1 by the data; mb) and send with the text. With an online review, the cover can usually - the research meets all applicable ethical standards; be downloaded. Details with the cover: title, author, edition, year, publisher, - the article adheres to appropriate reporting guidelines and community number of pages, price and ISBN number. Conclude with the name and standards for full data disclosure; affiliation(s) of the reviewer. - when reporting the results of a randomised controlled trial, author(s) should use the CONSORT statement as a guide to preparing the Letters to the editor manuscript (http://www.consort-statement.org/); Letters to the editor provide an opportunity to present results of scientific - conflicts of interest should be clearly stated in the manuscript (see below). value where a short format is most appropriate. They should not exceed 1000 words, 5 references and 1 figure or table. The authors are encouraged to refer to national and international registries of trials in their papers (such as clinicaltrials.gov), where applicable. Correspondence Correspondence provides an opportunity to debate published articles. Review articles This should not exceed 500 words, 5 references and 1 figure or table. Review articles, systematic reviews and meta-analyses should not exceed Correspondence is sent to the authors for rebuttal, and a final decision on 3000 words (excluding abstract, references, tables and legends). The publication is made at the end of this process, by the editor. manuscript should contain subheadings. A maximum of 6 tables and figures (in total) are allowed. Please provide titles above and legends below General information these elements. The abstract should not exceed 250 words and should be Each manuscript should be accompanied by a cover letter stating the structured: background, methods, results, conclusions, with the exception following: the complete postal address, email address and telephone number of a non-systematic review, which may contain a non-structured abstract. of the corresponding author and, if it is a re­submission, the previous No references should be included in the abstract. Authors should provide Neth J Crit Care number and year. The language of the journal is British a minimum of 3 keywords, a running title, and a list of not more than 70 English. Authors who are unsure of proper English usage should have their references. The authors are encouraged to refer to national and international manuscript checked by someone proficient in the English language. All text registries of trials in their papers (such as clinicaltrials.gov), where applicable. should be double spaced. The manuscript pages, including references and legends, should be sequentially numbered throughout

70 NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 Netherlands Journal of Critical Care

General guidelines on house style ownership of the manuscript (title of article) to the Netherlands Journal of - The title of the manuscript should be in typeface Times New Roman, size 20. Critical Care’. Any relevant papers that may be considered as duplicating in With the exception of the first word and proper nouns, initial capitals are not part the current submission should be reported. used in the title. - The names of departments should be in typeface Times New Roman, size 12. References - The names of hospitals should be written in English. Neth J Crit Care uses the Vancouver style of referencing. Only articles cited - Write ‘the Netherlands’, without capitalising the t. in the text are to be listed. They should be arranged in order of appearance - Generally, abbreviations should not be used in the title (see Table of standard in the text and numbered consecutively. Only the reference number should abbreviations for exceptions). appear in the text between brackets. […] Include all author names (unless - The corresponding author only provides his/her email address on the title page. there are more than six, in which case abbreviate to three and add ‘et al.’), and - Please provide a minimum of three keywords and a running title. page numbers. Use the Medline abbreviation for names of journals. - The abstract of original and review articles should be written in a structured format. Article in journals: Calandra T, Cometta A. Antibiotic therapy for Gram­ - Unstructured abstracts should take the form of a single paragraph. negative bacteremia. Infect Dis Clin North Am. 1991;5:817-34. - Headings must be in bold. Use no more than two levels of headings. Books (sections): Thijs LG. Fluid therapy in septic shock. In: Sibbald WJ, - Paragraphs starting immediately under headings and subheadings should begin Vincent J­L (eds). Clinical trials for the treatment of sepsis. (Update in at the left margin. Subsequent paragraphs should be indented. intensive care and emergency medicine, volume 19). Berlin Heidelberg New - Non-standard abbreviations (see table of standard abbreviations) should always York: Springer; 1995, pp 167-90. be explained and their use kept to a minimum. Conference meetings: Rijneveld AW, Lauw FN, te Velde AA, et al. The role - Use British English spelling – except in titles of institutions that have chosen to of interferongamma­ in murine pneumococcal pneumonia. 38th Interscience use US spelling, e.g. Academic Medical Center, Amsterdam. Examples: anaemia Conference on Antimicrobial Agents and Chemotherapy (ICAAC). San (instead of anemia), oesophagus (instead of esophagus), litre (instead of liter), Diego, Ca.; 1998, pp 290. colour (instead of color), labelling (instead of labeling), practice (noun), and Article/ information on website: Samenwerken aan duurzame zorg, landelijke practise (verb). This should be used consistently. Use the s-form spelling, e.g. monitor proeftuinen: Rijksinstituut voor Volksgezondheid en Milieu, 2015. minimise, randomisation. (Accessed 11 August 2015, at http://www.rijksoverheid.nl/documenten-en- - Do not use full stops in initials, abbreviations and academic titles. publicaties/rapporten/2015/07/10/samen-werken-aan-duurzame-zorg.html.) - References are numbered sequentially in the text and placed in square brackets after the punctuation. [..] How to submit - Genus names should be written in italics, e.g. Staphylococcus aureus, S. aureus. Please submit manuscripts directly to the Editorial Office through our online - Numbers under 10 are spelled out except for measurements with a unit (10 submission system at www.njcc.nl. mmol/l) or age (4 weeks old), or when in a list with other numbers (5 mice, 6 rats, 12 gerbils). Peer review - When referring to tables or figures in the text, use italics; do not use a capital All papers are subject to a peer-review system handled by the editors. Authors are letter, e.g. see table 2. encouraged to resubmit, when invited, the revised paper within two weeks after the editorial decision. The changes made in the revised paper should be highlighted Tables and the manuscript accompanied by a letter with a point-to-point rebuttal. Tables are to be numbered independently of the figures with Arabic numbers and are uploaded as separate documents. Proofs - Tables should be laid out in Word, using the table function. Other tables (e.g. The corresponding author will receive proofs of accepted papers by email. in pdf format or PowerPoint) will not be accepted; Corrected proofs should be returned within 48 hours of receipt. - Do not use internal horizontal or vertical lines; - Do not use spaces, tabs or hard returns in tables; Production process - Each piece of data must be contained in its own cell; Decisions of the editors are final. All material accepted for publication is - Numbers and percentages are presented in the same cell; subject to copyediting. The Neth J Crit Care reserves the right to edit for - Tables should always be cited in the text in consecutive numerical order; house style, clarity, precision of expression, and grammar. Authors review - For each table, please supply a title explaining the components of the table; these changes at the proof stage but must limit their alterations in the proof - Any abbreviations used in the table must be defined in a legend; to correcting errors and to clarifying misleading statements. - Tables should not exceed the printed area of the page (174 x 234 mm). Table of commonly used abbreviations Figures AIDS acquired immunodeficiency syndrome Figures should also be numbered with Arabic numbers and are uploaded in ALI acute lung injury separate documents. Legends should be given in the document that contains the ARDS adult respiratory distress syndrome text, references, and tables. Authors wishing to include figures or tables that have APACHE acute physiology and chronic health evaluation already been published elsewhere are required to obtain permission from the BIPAP biphasic positive airways pressure copyright owner and provide evidence that such permission has been granted CCU coronary care unit when submitting their paper. Colour figures can be published. Short, clear COPD chronic obstructive pulmonary disease legends make additional description in the text unnecessary. Figures should be CPAP continuous positive airway pressure provided in electronic format (TIFF or JPEG). CT computed tomography ECG electrocardiogram Conflict of interest ECMO extracorporeal membrane oxygenation Authors must indicate any conflict of interest. This includes a financial EEG Electroencephalogram relationship with an organisation that sponsored the research (funding, ELISA enzyme-linked immunosorbent assay speakers fee, consultancy fee), management relations with the organisation that ETCO2 end-tidal carbon dioxide sponsored the research (consultant, member of board). All sources of funding HIV human immunodeficiency virus obtained for the research should also be stated. A conflict of interest statement IC intensive care can be downloaded from the website. The completed and signed form should ICU intensive care unit be uploaded as a separate document when submitting the manuscript. If no IM Intramuscular conflict exists, authors should state: All authors declare no conflicts of interest. INR international normalised ratio IPPV intermittent positive pressure ventilation No funding or financial support was received. IV Intravenous Author agreement MAP mean arterial pressure MODS multiorgan dysfunction syndrome All authors must certify they have seen and approved the manuscript being MRI magnetic resonance imaging submitted. All authors warrant that the article is the authors’ original work, has PACU post anaesthesia care unit not been published previously and is not under consideration for publication PEEP positive end expiratory pressure elsewhere. The author agreement form can be downloaded from the website. An PET positron emission tomography article will only be published when this form is completed, signed and returned. SARS severe adult respiratory syndrome SIRS systemic inflammatory response syndrome Copyright SOFA sequential organ failure assessment Copyright ownership is to be transferred in a written statement, which must SPECT single-photon emission computed tomography accompany all manuscript submissions and must be signed by all authors. TIA transient ischaemic attack The agreement should state: ‘The undersigned authors transfer all copyright TRALI transfusion-related acute lung injury

NETH J CRIT CARE - VOLUME 29 - NO 1 - JANUARY 2021 71