Delirium after cardiac arrest: Phenotype, prediction, and outcome

Clinical

paper

Delirium

after

cardiac

arrest:

Phenotype,

prediction,

and

outcome

$

Hanneke

M.

Keijzer

a,b,

*

,

Marjolein

Klop

a,d,1

,

Michel

J.A.M.

van

Putten

c,e,2

,

Jeannette

Hofmeijer

a,c,1

a

DepartmentofNeurology,RijnstateHospital,Arnhem,TheNetherlands

b_{DepartmentofNeurology,DondersInstituteforBrain,Cognition,andBehaviour,RadboudUniversityMedicalCentre,Nijmegen,TheNetherlands} c_{DepartmentofClinicalNeurophysiology,TechnicalMedicalCenter,UniversityofTwente,Enschede,TheNetherlands}

d_{TechnicalMedicine,UniversityofTwente,Enschede,TheNetherlands} e

DepartmentsofNeurologyandClinicalNeurophysiology,MedicalSpectrumTwente,Enschede,TheNetherlands

Abstract

Aim:Toestablishincidence,phenotype,long-termfunctionaloutcome,andearlyEEGpredictorsofdeliriumaftercardiacarrest.

Methods:Thisisanadhocanalysisofaprospectivecohortstudyonoutcomepredictionofcomatosepatientsaftercardiacarrest.Patientswith recoveryof consciousness,whosurviveduntilhospitaldischarge,weresubdividedingroupswith andwithoutdelirium basedon psychiatric consultation.Deliriumphenotypeandmedicaltreatmentwereretrievedfrompatientfiles.Allotherdatawereprospectivelycollected.Weused univariateanalysesofbaselineandearlyEEGcharacteristicsforidentificationofpossibledeliriumpredictors.Associationofdeliriumwithneurological recoveryatsixmonthswasanalyzedwithmultinomiallogisticregressionanalysis.

Results:Of233patients,141surviveduntilhospitaldischarge,ofwhom47(33%)werediagnosedwithdelirium.Therewerenodifferencesinbaseline characteristicsbetweenpatientswithandwithoutdelirium.Alldeliriouspatientsweretreatedwithrelativelyhighdosagesofpsychopharmaceuticals, mostlyhaloperidolandbenzodiazepineagonists.Prevalentcharacteristicsweredisturbedcognition,perceptionandpsychomotorfunctioning(98%). Halfofthepatientshadlanguagedisordersorshouting.DeliriumwasassociatedwithlongerICUandhospitaladmission,andmorefrequentdischarge torehabilitationcentreornursinghome.Therewasatrendtowardspoorerneurologicalrecovery.EEGmeasurementswithin12haftercardiacarrest couldpredictdeliriumwith91%specificityand40%sensitivity.

Discussion:Deliriumiscommonaftercardiacarrest,andprobablyleadstolongerhospitalizationandpooreroutcome.Optimaltreatmentisunclear. EarlyEEGholdspotentialtoidentifypatientsatrisk.

Keywords:Delirium,cardiacarrest,Postanoxicencephalopathy,Postanoxiccoma,Neurologicalrecovery,Electroencephalogram

Introduction

Patientswhosurviveacomatosestateaftercardiacarrestareatriskof delirium.Deliriumisaclinicalsyndromedefinedbydisturbancesin

cognitionandbehaviour inmedicallyillpatients.Thediagnosisof deliriumisbasedonfluctuatingdisturbancesinattention,awareness, and cognition, without a direct cause, such as specific medical conditionsorneurocognitivedisorders.Thedisturbancesshouldnot occurduringaseverelyreducedlevelofarousal,suchascoma.1
3 $

ThisstudywasconductedattheRijnstatehospital,Arnhem,TheNetherlands,incollaborationwiththedepartmentofClinicalNeurophysiologyofthe UniversityofTwente,Enschede,TheNetherlands.

* Correspondingauthorat:DepartmentofNeurology,RijnstateHospital,P.O.Box9555,6800TAArnhem,TheNetherlands. E-mailaddress:hmkeijzer@rijnstate.nl(H.M. Keijzer).

1

P.O.Box9555,6800TA,Arnhem,TheNetherlands. 2

P.O.Box217,7500AEEnschede,TheNetherlands.

https://doi.org/10.1016/j.resuscitation.2020.03.020

Received28January2020;Receivedinrevisedform26February2020;Accepted28March2020 Availableonlinexxx

Available

online

at

www.sciencedirect.com

Resuscitation

Thereportedincidenceofdeliriumis30
80%inpatientsadmitted toanintensivecareunit(ICU)4,5_{andupto100%incardiacarrest} survivorstreatedwithmildtherapeutichypothermia.6_Theoriginof deliriumisoftenmultifactorial.Inpatientsaftercardiacarrest,both postanoxicencephalopathyandtreatmentslikeinducedhypothermia orsedativemedicationmayplayarole.Incriticallyillpatients,delirium isassociatedwithhighermortality,prolongedhospitalization,andan increasedriskofcognitiveimpairmentafterdischarge.5

Mostpatientsaftercardiacarresthaveneurologicaldisturbances asaresultofpostanoxicencephalopathyandmanyhavedelirious symptomsduringthefirstweeksafterresuscitation.However,patients aftercardiacarrestareclassicallyexcludedfromdeliriumstudies.This resultsinalackofinformationaboutthephenotype,riskfactorsand optimal treatment of delirium after cardiac arrest. Confusion, disorientation, attention deficits, psychomotor agitation, hallucina-tions,sleep-wakecycledisturbance,andaffectivesymptomsareoften observedaftermyocardialinfarction.7_{Itisunknownwhetherthese} representtruedelirium,wheretreatmentwithantipsychotic medica-tionmayimproveoutcome,orareadirectexpressionofpostanoxic encephalopathy,wheretreatmentmaybefutileorinterferewiththe recoveryofthepostanoxicbrain.

Identification of patients at risk may help early diagnosis and treatmentofthe delirium.Clinicalpredictorsofdelirium includean advancedage,intensivesmoking,dailyuseofalcohol,pre-existent cognitiveimpairment,andaprecedingperiodofsedation,coma,or mechanicalventilation.4,7,8VariousEEGcharacteristicshavebeen describedinrelationtodelirium,eitherduringorprecedingdelirium. Theseincludeslowing ofthedominantfrequency,9,10_{time spent inburst} suppressionpatterns,11
13andmeasuresoffunctionalconnectivity.14 Withthisstudy,weestimateincidenceofdeliriuminpatientsthat recoverfromcoma aftercardiacarrest.Wedescribe thedelirium phenotype, treatment, and association with long-term functional outcome.Inaddition,weinvestigateifparticularEEGfeaturesinthe comatose phase are predictive of delirium after recovery of consciousness.

Methods

Thisisanadhocanalysisofamulticentreprospectivecohortstudyon EEGforoutcomepredictionofcomatosepatientsaftercardiacarrest. Methodsfordatacollectionhavebeendescribedpreviously,15and partofthedataofpatientsincludeduptoNovember2017havebeen usedin earlierpublications.15
25Inshort,we appliedprospective patientinclusionandcollectionof demographic,baseline,clinical, EEG,andoutcomedata.Forthecurrentanalysisweuseddatafrom patientsintheRijnstatehospital(Arnhem,TheNetherlands)collected betweenJanuary2015andDecember2018.

Patients

Weincluded consecutivepatients whowere admittedto theICU departmentinacomatosestateaftercardiacarrestandsurvivedtill discharge from the primary hospital. Exclusion criteria were concomitant acute stroke, traumaticbrain injury,and progressive neurodegenerative disease. All patients had continuous EEG registrationsstarting assoon aspossibleafterarrivalattheICU, uptoawakeningoraminimumof72haftercardiacarrest.Duringthis comatose phase,patients were mostlytreated with propofol and morphine,sometimescomplementedwithmidazolam.

Delirium

Patientswithrecoveryofconsciousness,whosurviveduntilhospital discharge,weresubdividedingroupswithandwithoutdeliriumbased onpsychiatricconsultation.InRijnstatehospital,accordingtolocal hospitalprotocols,diagnosis andtreatment ofdeliriumare always donebyaconsultinghospitalpsychiatrist.Thepsychiatristconfirmsor rejectsthediagnosisofdeliriumbasedontheDSM-Vcriteria,andisin the lead with regard to medical treatment of delirium. Patients wereincludedinthedeliriumgroupifapsychiatristwasconsulted during hospital stay, and delirium was diagnosed, and at least one of the following criteria was reported in the patient's file: disturbances of consciousness, disturbances of cognition, or perceptualdisturbances.

Ofpatientsthatwereincludedinthedeliriumgroup,wecollected dataondeliriumsymptomsandmedicationfrompatienthospitalfiles. Presence or absence of characteristics of delirium was scored accordingtotheDSM-V,1,2_{andtheICD-10Version:2016,section} F05,26,27_{complementedwithadditionaldetailsofthesecharacteristic} symptoms,basedonourownexpertisewithpatientsaftercardiac. Characteristics are summarized and explained in Table 1. We collected which pharmaceuticals wereused to treatdelirium, the averagemaximaldoseofthesepharmaceuticals,themeanduration oftreatment,andwhethertreatmentwascontinuedafterdischarge fromtheprimaryhospital.

Outcome

Ourprimaryoutcomemeasurewasneurologicalrecovery,definedas the Cerebral Performance Category (CPC-score) measured six months after cardiac arrest. CPC scores were obtained usinga telephoneinterviewbasedonaDutchtranslationoftheEuroQol-6D questionnaire.Otheroutcomemeasuresweredurationofstayatthe ICU,durationofhospitaladmission,anddischargedestination.

EEGregistrationandanalysis

ContinuousEEGregistrationswereperformedusingaNihonKohden device(VCMMedical,Leusden,TheNetherlands),using21silver/ silver chloride electrodes according to the international 10
₂₀ system, and a sampling frequency of 500Hz. Recordings were startedassoonaspossibleafterarrivalattheICUandcontinuedfor threetofivedays,oruntilthepatientregainedconsciousnessordied. Weextractedfive-minuteEEGepochsat12,24,48and72hafter resuscitationfromtheEEGusinganautomatedalgorithm.23_Ifthe EEG epoch atone ofthese hours was unavailable,we used an available EEG epochat the closest hour within a range of 2h. Allepochswerereferencedtothelongitudinalbipolarmontageand pre-processedtoremoveEEGchannelscontainingartefactsdueto muscleactivity,flatsignals,andunrealisticallyhighamplitudepeaks, usinganautomatedartefactrejectionalgorithm.21_{EEGepochswere} includedforfurtheranalysiswhenatleast12artefact-freechannels wereavailable.TheEEGsignalsofthesechannelswerebandpass (0.5
30Hz)filteredwithazero-phasesixthorderButterworthfilter. We extracted 90-s epochs from these filtered five-minute EEG segments,toremovefilteringedgeeffectsatthebeginningandthe endofthesegment.

With data from these EEG epochs, we calculated two EEG features:thescaledalpha-to-deltaratio(sADR)andthebackground continuityindex(BCI).ThesADRwasdefinedastheratiobetweenthe

EEGpowerinthealphaband(8
13Hz)anddeltaband(1
4Hz), scaledbetween
1and1,givenby:

sADR¼

a

d

a

þ

d

(1)

where

a

representstheEEGpowerinthealphabandand

d

theEEG powerinthedeltaband.Thepowerspectraldensitywasestimated usingWelch'smethodwith 50%overlapandaHamming window lengthof2s.23_{TheBCIwasdefinedasthefractionofEEGsignalthat} was not spent in suppression.21 _{Suppression was defined as} segmentsofatleast500mswithamplitudes<10

m

V.

AlldataanalyseswereperformedwithMATLAB(2019,MathWorks Inc.,Natick,USA).

Statisticalanalysis

Baselineandclinicalcharacteristics,durationofstayattheICU,and durationofhospitaladmissionarepresentedinadescriptivewayfor patientswithandwithoutdelirium.Foranalysisofdifferencesbetween groupsofpatientswithandwithoutdeliriumweusedchi-squaredtests for ordinal variables and Mann
Whitney U tests for continuous variablesusingSPSS22(IBMCorp.,Armonk,NY).Differencesin outcomebetweenbothgroupswereanalyzedbymultinomiallogistic regressionanalysiswithMATLAB.

We compared sADR and BCI at 12, 24, 48 and 72h after resuscitationbetweenpatientswithandwithoutdeliriumusingMann
WhitneyUtestsinMATLAB.Apreliminarylogisticregressionmodel forpredictionofdelirium,basedonsADRandBCI,wascreatedatthe timepointwithlargestgroupdifferences.Discriminativevaluesofthis

model are expressed as the area under the receiver-operator characteristics (ROC) curve (AUC). Sensitivity and specificity to predictdeliriumwereestimatedattheoptimalcut-off,byminimizinga cost-functiontotheROC.Forallcomparisons,ap-value<0.05was consideredstatisticallysignificant.

Table1_–CharacteristicsofdeliriumaccordingtotheICD-10andadditionaldetails.

Characteristic Definition

ICD-10criteria Impairmentofconsciousnessandattention Adecreasedlevelofconsciousnesswithreducedabilitytodirect,focus, sustain,andshiftattention.

Globaldisturbanceincognitionandperception Thisincludesproblemswithformingofnewlong-termmemories, impairmentsofabstractthinkingandcomprehension,possibleillusions anddisorientationinplace,timeandperson.

Psychomotordisturbances Thiscanbesubdividedinpsychomotoractivation(restlessness,

repetitivemovements,frequentchangesofposition)andpsychomotor retardation(staring,slowandlittlemovements).

Disturbancesofthesleep-wakecycle Varyingfrominsomniatoreversionofthesleep-wakecycle.

Emotionaldisturbances Includessymptomslikedepression,anxiety,fear,apathyoranger.

Additionaldetails Extremerestlessness Motorrestlessnessnecessitatingbodilyfixationtothebedorchair,

becauseofdangerforthepatientormedicalstaff.Thisincludesfalling andwalkingawaywithgettinglost.

Disinhibition Lackofrestraint,resultinga.o.inpoorriskassessment,impulsive

behaviouranddisregardofsocialconventions.

Languagedisorder Problemsinproducingand/orunderstandingspokenlanguage.

Bilingualpatientsmaytemporallyforgetoneoftheirlanguages.

Wandering Restlessnessresultingininabilitytostayintheroomandwandering

fromthedepartmentandanabnormalurgetomove.

Shouting Intentionalshoutingaimingtogetsomething.

Aggression Verbalorphysicalaggressiontomedicalstaff,familymemberorother

patients.

Incontinence Urinaryoffaecalcontinenceproblems

Paranoia Anxiousorfearfulfeelingsandthoughts,oftenrelatedtothreator

conspiracy.

Hallucinations Visualorauditoryhallucinations.

Headshaking Constantshakingwiththehead.

Excessivedrinking Excessivefluidintake,obsessionwithdrinking.

Table2–Baselinecharacteristicsofpatientsafter cardiacarrestwhorecoveredfromcomaandsurvived untilhospitaldischarge.

Delirium (n=47) Nodelirium (n=94) p-value Male 40(85%) 78(83%) 0.48 Age(year) 61(10) 62(12) 0.33 Cardiaccause 43(91%) 90(96%) 0.25 Shockablerhythm 45(96%) 84(89%) 0.17 Pooroutcomea _{9(19%) 13(14%) 0.27} Propofolinfirst24h 45(96%) 87(93%) 0.72

Max.Propofoldose(mg/kg/h) 2.93(0.98) 2.78(1.13) 0.53

Midazolaminfirst24h 13(28%) 47(50%) 0.07

Max.Midazolamdose(mg/kg/h) 138(99.2) 126(84.9) 0.79

Morphineinfirst24h 47(100%) 91(97%) 0.55

Max.Morphinedose(mg/kg/h) 24.6(4.45) 28.1(15.2) 0.58 p-values were obtained using two-sided chi-square tests for nominal variablesandMann
WhitneyUtestsforcontinuousvariables.Nominal variablesarepresentedasn(%).Continuousvariablesarepresentedas mean(standarddeviation).

a_{PooroutcomedefinedasaCerebralPerformanceCategory(CPC)score3}

Results

Duringtheinclusionperiod,233comatosepatientsaftercardiacarrest wereadmitted.All patientsweretreated with24
72hoftargeted temperature management (33Cuntil January2014, 36C since February2014)afterarrivalattheICUdepartment.Ofthesepatients, 141(61%)surviveduptodischargefromtheprimaryhospitalandwere includedinthecurrentanalysis.Deliriumwasdiagnosedin47(33%)of thesurvivors.Therewerenodifferencesinbaselinecharacteristics betweensurvivingpatientswithandwithoutdelirium(Table2).

Deliriumphenotypeandtreatment

Theincidenceofdeliriumsymptomsinthedeliriumgroupispresented in Fig. 1. Treatment is presented in Table 3. Pharmacological treatmentwasappliedtoalldeliriouspatients.Mostpatientswere treatedwithmultipleantipsychoticdrugs(median(IQR)numberof drugs:3(2)).Ofnote,thisindicatestreatmentafterthecomatosestate

ondeICUandcardiaccareunit(CCU),in‘awake’patients,prescribed by a consulting psychiatrist. In 21 patients (45%), anti-delirium treatmentwascontinuedafterdischargefromtheprimaryhospital. Mann
Whitney Utests revealed thatdosages of the mostoften prescribeddrugs,haloperidol,lorazepam,oxazepam,temazepam, and zopiclone,showed no statisticallysignificant associationwith outcomeatsixmonthsaftercardiacarrest.

Dischargeandoutcome

Patientswithdeliriumstayedsignificantlylongeronintensivecare units (delirium (median (IQR)) 6(9) days,non-delirium 3(4) days, p<0.01)andinthehospital(Delirium(median(IQR))24(21)days, non-delirium 15(15) days, p_<0.01). Due to transfer to another hospital,totalstayatICUwasunknownfor5patients,totalstayin hospitalwasunknownfor7patients.Patientswithdeliriumweremore oftendischargedtoarehabilitationcentre(delirium19%,non-delirium 3%,p<0.01)orchronicnursinghome(delirium15%,non-delirium 4%,p=0.03),SupplementaryTable1).Patientswithdeliriumhada

Fig.1–Incidenceofclinicalcharacteristicspresentinthedeliriumgroup.

Table3–Medicationtosuppressdelirioussymptomsinthedeliriumgroup. Numberofpatients

n(%)

Highestdose(mg/day) Mean(SD)

Daysofuseinhospital Mean(SD)

Continuedafterdischarge n(%) Haloperidol 45(96%) 7.8(5.2) 9.6(6.9) 12(26%) Lorazepam 24((51%) 2.5(3.0) 7.4(6.0) 1(2%) Oxazepam 24(51%) 30.2(31.1) 6.2(7.36) 5(11%) Zopiclone 24(51%) 7.5(0.0) 6.3(5.89) 4(9%) Temazepam 11(23%) 11.8(4.1) 9.5(8.54) 4(9%) Valproicacid 5(11%) 1420(239) 16.0(2.92) 0(0%) Diazepam 3(6%) 6.7(2.9) 1.3(0.577) 0(0%) Clonazepam 1(2%) 0.50(
) 3.0(
) 0(0%) Zolpidem 1(2%) 10.0(
) 5.0(
) 0(0%) Quetiapine 1(2%) 200.0(
) 21.0(
) 1(2%) Risperdal 1(2%) 2.00(
) 6.0(
) 0(0%)

largerchanceofapooroutcomethanpatientswithoutdelirium,butthis differencewasnotstatisticallysignificant(p=0.15forordinalanalysis ofanyshiftinthedirectionofapooreroutcomeontheCPC,p=0.19for proportionofpatientswithCPC1,p=0.32fortheproportionofpatients withCPC1or2,Fig.2).

EEGanalysis

At12haftercardiacarrest,themediansADRwaslowerinpatients withdelirium
0.82(IQR:
0.90to
0.68)thaninpatientswithout delirium(
0.59(IQR:
0.77to
0.14),p<0.01).ThemedianBCIat 12haftercardiacarrestwasalsosignificantlylowerforpatientswith delirium(0.59 (IQR:0.39
0.78))thanforpatientswithoutdelirium (0.89(IQR:0.53
_0.98),p=0.02).Latermeasurementsshowedonlya significantlylowermediansADR forpatientswith delirium(
0.78 (IQR:
0.88 to
0.56)) than for patients withoutdelirium (
0.69 (IQR:
0.81to
0.35),p=0.03,supplementaryTable2).

TheROCcurveofapredictivemodelbasedonsADRandBCIat12h aftercardiacarrest discriminatedbetweenpatients withand without deliriumwithanareaunderthecurveof0.76(0.62
0.8895%CI)and predictedpooroutcomewithasensitivityof40%(19
62%95%CI)ata specificityof91%(75
9795%CI,Fig.3).

Discussion

Weconfirmthattheincidenceofdeliriumamongstsurvivorsafter cardiacarrestishigh.Inourstudy,onethirdofpatientsrecovering fromacomatosestatehadclinicalsymptomsclassifyingasdelirium. Disturbancesofcognitionorperceptionandpsychomotor disturban-ces were present in almost all delirious patients. Characteristic featuresincludedprominentrestlessness,withaggression, wander-ing,orgettinglost,withdangerforthepatienthimselforothers.Halfof thepatientshadlanguagedisordersorshouting.Headshakingor excessive drinking were present in 10
_20% of the patients. All patientsweretreatedwithanti-psychoticdrugs,mostlywithmorethan onedrug,inrelativelyhighdoses.Deliriumwasassociatedwithlonger stayinICUandhospital.Therewasatrendtowardspoorerfunctional recovery.

Ingeneral, deliriumisassociated with highermortality, longer intensivecareandhospitalstay,andalargerchanceofcognitive impairmentsafterdischarge.5,28Thiscorrespondswithourresults, showingsignificantlylongerhospitalizationofpatientswithdelirium. Ourlackofstatisticalsignificancewithregardtofunctionalrecoveryat sixmonthsisprobablyrelatedtothelimitedsensitivityoftheCPCto detectlongtermcognitivedisturbances.

Ourresultsareinlinewiththeonlypreviousstudyondeliriumafter cardiac arrest, where the reported incidence was even higher.6 Apparently, psychomotor, cognitive, and mental disturbancesare commonaftercardiacarrest,especiallyinthefirstdaystoweeksafter recovery ofconsciousness in patients thatsurvive the comatose phase.6 This combination of clinical symptoms may classify as deliriumaccordingtoDSM-VandICD-10andgiverisetotreatment withanti-psychoticdrugs.However,inpatientsaftercardiacarrest, thesesymptomsareprobablyoftenadirectexpressionofpostanoxic encephalopathy and/or reperfusion damage. With postanoxic encephalopathy,effectsof(highdosesof)anti-psychoticandsedative drugsareuncertain.

A more extended differential diagnosis should probably be

consideredwithdelirioussymptomsinpatientsaftercardiacarrest. Fig.

2 – Outcome a t 6 months after cardiac arrest for patients with and without delirium, classified according to the Cerebral Performance Category (CPC). p -value of difference between both groups assessed by multinomial logistic regression analysis is 0.15.

EEGstudieshaveshownthatanon-convulsivestatusepilepticusis presentin10
20%ofthepatientsduringthecomatosephase.22,29 Impairedconsciousness withpsychomotor disturbancesand rest-lessnessmaybecausedbyepilepticseizures,aswell.30_However, systematic EEG measurementsof deliriouspatients aftercardiac arrestarescarce.Thetypicalfeatureofexcessivedrinkingcouldbea signofSyndromeofInappropriateAntidiureticHormoneSecretion (SIADH).Thissyndromemayarisefromvariousfactorspresentin patientsaftercardiacarrest,suchasencephalopathy,pneumonia, vasopressin, and narcotic drugs.31 _{Non-convulsive seizures and} SIADHrequireadifferenttreatmentthanplaindelirium.

TheEEGwithin24hafterresuscitationallowsreliablepredictionof poororgoodoutcomeofapproximatelyhalfofallcomatosepatients aftercardiacarrest.15,23HereweshowthatearlyEEGpatternsmay be predictive of delirium in patients that recover from coma. In accordance with prediction of long term outcome, apparently, predictivevaluesareespeciallyhighwithinthefirst24haftercardiac arrest.AlowersADRandlowerBCIwereseenmoreofteninthe patientsdevelopingadelirium.SmallervaluesofthesADRandBCI reflectEEGpatternsthatarerelativelyslowanddiscontinuous.21_This impliesamoreseverepostanoxicencephalopathy,thatisapparently associatedwithalargerlikelihoodofdevelopingadelirium.Further research in a larger cohort with strict screening for delirium is necessarytoconfirmpredictivevaluesofmodelsbasedontheEEG. Ourstudyhascertainlimitations.First,weonlydiagnoseddelirium whenahospitalpsychiatristwasconsultedbecauseofasuspected delirium.Thismayhaveledtoinclusionofdeliriouspatientsinthe non-deliriumgroup,whenapsychiatristwasneverconsulted,especiallyin patientswithhypoactivedelirium.4,6Also,developmentofdelirium after hospital discharge from the primary hospital is missed. Otherwise, diagnosis by a consulting psychiatrist has probably contributedtoahighspecificityofthediagnosisofdelirium.

Second,thisprospectivecohortstudyfocussedonpredictionof neurological outcome after cardiac arrest.15,23 Thestudy did not includeprospectivecollectionofclinicaldataondelirioussymptoms. Delirioussymptomswerecollectedretrospectively,frompatientfiles.

Therefore,thelistofdelirioussymptomsmaybeincomplete.Third,we didnotincludeextensivecognitivefollowup.

Conclusion

Inconclusion,deliriumiscommonaftercardiacarrestandcharacterized byprominentpsychomotordisturbances.Itisunknowntowhichextent delirioussymptomsrepresentadirectexpressionofpostanoxicbrain damage.Specificcauses,suchasnon-convulsiveseizuresandSIADH, needtobeexcluded.Toestablishoptimaldiagnosesandtreatment, prospectivestudieswithfollowupofcognitivefunctioningareneeded.

Conflict

of

interest

MvPisco-founderofclinicalsciencesystems,providingsoftwarefor EEGrecordingandanalysis.

HMKisfundedbytheRijnstate-Radboudpromotionfund.

Authors

’

contribution

Hanneke M. Keijzer: conceptualization; methodology; software; formalanalyses;writingoriginaldraft.MarjoleinKlop:methodology; software;formalanalyses;writing
review&editing.MichelJ.A.M. van Putten:methodology;supervision;writing
review& editing. JeannetteHofmeijer: conceptualization;methodology;supervision; writing
review&editing.

Acknowledgments

TheauthorsthankKimberleyKamphuis,studentBiomedicalSciences at the Radboud University (Nijmegen, The Netherlands), for assistancewithscreeningpatientfilesandpreliminaryanalyses. Fig.3_–ROCcurveoflogisticregressionmodel,containingsADRandBCIat12haftercardiacarrest,topredictdelirium. Theareaunderthecurveis0.76(0.62
0.8895%CI).At91%specificity(75
97%95%CI)deliriumcouldbepredicted with40%sensitivity(19
62%95%CI).

Appendix

A.

Supplementary

data

Supplementarydataassociatedwiththisarticlecanbefound,inthe onlineversion,athttps://doi.org/10.1016/j.resuscitation.2020.03.020.

REFERENCES

1.EuropeanDeliriumAssociation,AmericanDeliriumSociety.The DSM-5criteria,levelofarousalanddeliriumdiagnosis:inclusivenessis safer.BMCMed201412:.

2.AmericanPsychiatricAssociation.NeurocognitiveDisorders. Diagnosticandstatisticalmanualofmentaldisorders:DSM-5: AmericanPsychiatricAssociationPublishing..p.947.

3.BrownTM,BoyleMF.Delerium.BMJ2002;325:644
7.

4.BoncykCS,RengelKF,PandharipandePP,HughesCG.IntheICU
deliriumpostcardiacarrest.CurrOpinCritCare2019;25:218
25.

5.SalluhJI,WangH,SchneiderEB,etal.Outcomeofdeliriumincritically illpatients:systematicreviewandmeta-analysis.BMJ2015;350: h2538.

6.PollockJS,HollenbeckRD,WangL,etal.Deliriuminsurvivorsof cardiacarresttreatedwithmildtherapeutichypothermia.AmJCrit CareOffpublAmAssocCritCareNurs.2016;25:e81
9.

7.UguzF,KayrakM,CicekE,KayhanF,AriH,AltunbasG.Delirium followingacutemyocardialinfarction:incidence,clinicalprofiles,and predictors.PerspectPsychiatrCare2010;46:135
42.

8.VanRompaeyB,ElseviersMM,SchuurmansMJ,Shortridge-Baggett LM,TruijenS,BossaertL.Riskfactorsfordeliriuminintensivecare patients:aprospectivecohortstudy.CritCare2009;13:R77.

9.PalancaBJA,WildesTS,JuYS,ChingS,AvidanMS.

Electroencephalographyanddeliriuminthepostoperativeperiod.Brit JAnaesth2017;119:294
307.

10.EvansJL,NadlerJW,Preud’hommeXA,etal.Pilotprospectivestudy ofpost-surgerysleepandEEGpredictorsofpost-operativedelirium. ClinNeurophysiolOffJIntFedClinNeurophysiol2017;128:1421
5.

11.SoehleM,DittmannA,EllerkmannRK,BaumgartenG,PutensenC, GuentherU.Intraoperativeburstsuppressionisassociatedwith postoperativedeliriumfollowingcardiacsurgery:aprospective, observationalstudy.BMCAnesthesiol2015;15:61.

12.FritzBA,KalarickalPL,MaybrierHR,etal.Intraoperative electroencephalogramsuppressionpredictspostoperativedelirium. AnesthAnalg2016;122:234
42.

13.AndresenJM,GirardTD,PandharipandePP,DavidsonMA,ElyEW, WatsonPL.Burstsuppressiononprocessedelectroencephalography asapredictorofpostcomadeliriuminmechanicallyventilatedICU patients.CritCareMed2014;42:2244
51.

14.NumanT,SlooterAJC,vanderKooiAW,etal.Functionalconnectivity andnetworkanalysisduringhypoactivedeliriumandrecoveryfrom anesthesia.ClinNeurophysiolOffJIntFedClinNeurophysiol 2017;128:914
24.

HofmeijerJ,BeerninkTMJ,BoschFH,BeishuizenA, Tjepkema-CloostermansMC,vanPuttenMJAM.EarlyEEGcontributesto

15.multimodaloutcomepredictionofpostanoxiccoma.Neurology. 2015;85:137
43.

16.CloostermansMC,vanMeulenFB,EertmanCJ,HomHW,vanPutten MJAM.Continuouselectroencephalographymonitoringforearly predictionofneurologicaloutcomeinpostanoxicpatientsaftercardiac arrest:aprospectivecohortstudy.CritCareMed2012;40:2867
75.

17.SondagL,RuijterBJ,Tjepkema-CloostermansMC,etal.EarlyEEG foroutcomepredictionofpostanoxiccoma:prospectivecohortstudy withcost-minimizationanalysis.CritCare2017;21:111.

18.Tjepkema-CloostermansMC,HofmeijerJ,HomHW,BoschFH,van PuttenM.Predictingoutcomeinpostanoxiccoma:aretenEEG electrodesenough?JClinNeurophysiolOffPublAm

ElectroencephalogrSoc.2017;34:207
12.

19.Tjepkema-CloostermansMC,HindriksR,HofmeijerJ,vanPuttenMJ. Generalizedperiodicdischargesafteracutecerebralischemia: reflectionofselectivesynapticfailure?ClinNeurophysiolOffJIntFed ClinNeurophysiol2014;125:255
62.

20.Tjepkema-CloostermansMC,HofmeijerJ,BeishuizenA,etal. Cerebralrecoveryindex:reliablehelpforpredictionofneurologic outcomeaftercardiacarrest.CritCareMed2017;45:e789
97.

21.RuijterBJ,HofmeijerJ,Tjepkema-CloostermansMC,vanPuttenM. TheprognosticvalueofdiscontinuousEEGpatternsinpostanoxic coma.ClinNeurophysiolOffJIntFedClinNeurophysiol

2018;129:1534
43.

22.RuijterBJ,vanPuttenMJ,HofmeijerJ.Generalizedepileptiform dischargesinpostanoxicencephalopathy:quantitative

characterizationinrelationtooutcome.Epilepsia2015;56:1845
54.

23.RuijterBJ,Tjepkema-CloostermansMC,TrompSC,etal.Early electroencephalographyforoutcomepredictionofpostanoxiccoma:a prospectivecohortstudy.AnnNeurol2019;86:203
14.

24.Tjepkema-CloostermansMC,daSilvaLourencoC,RuijterBJ,etal. Outcomepredictioninpostanoxiccomawithdeeplearning.CritCare Med2019.

25.GlimmerveenAB,RuijterBJ,KeijzerHM,Tjepkema-Cloostermans MC,vanPuttenM,HofmeijerJ.Associationbetweensomatosensory evokedpotentialsandEEGincomatosepatientsaftercardiacarrest. ClinNeurophysiolOffJIntFedClinNeurophysiol2019;130:2026
31.

26.WorldHealthOrganization.ICD-10:internationalstatistical classificationofdiseasesandrelatedhealthproblems.Geneva:World HealthOrganization;2016(Availablefrom:https://icd.who.int/ browse10/2016/en).

27.WorldHealthOrganization.TheICD-10classificationofmentaland behavriouraldisorders.

28.PandharipandePP,ElyEW,AroraRC,etal.Theintensivecare deliriumresearchagenda:amultinational,interprofessional perspective.IntensiveCareMed2017;43:1329
39.

29.RossettiAO,OddoM,LiaudetL,KaplanPW.Predictorsofawakening frompostanoxicstatusepilepticusaftertherapeutichypothermia. Neurology2009;72:744
9.

30.BeyenburgS,ElgerCE,ReuberM.Acuteconfusionoralteredmental state:considernonconvulsivestatusepilepticus.Gerontology 2007;53:388
96.

31.EllisonDH,BerlT.Thesyndromeofinappropriateantidiuresis.NEngl JMed2007;356:2064
72.