A nationwide overview of 1-year mortality in cardiac arrest patients admitted to intensive care units in the Netherlands between 2010 and 2016

Clinical

paper

A

nationwide

overview

of

1-year

mortality

in

cardiac

arrest

patients

admitted

to

intensive

care

units

in

the

Netherlands

between

2010

and

2016

Loes

Mandigers

*

,

Fabian

Termorshuizen

,

Nicolette

F.

de

Keizer

,

Diederik

Gommers

,

Dinis

dos

Reis

Miranda

,

Wim

J.R.

Rietdijk

,

Corstiaan

A.

den

Uil

a

DepartmentofIntensiveCare,ErasmusMCUniversityMedicalCenter,Rotterdam,TheNetherlands

b

NationalIntensiveCareEvaluation(NICE)Foundation,Amsterdam,TheNetherlands

c_{DepartmentofMedicalInformatics,AmsterdamUMC,AmsterdamPublicHealthResearchInstitute,UniversityofAmsterdam,Amsterdam,The}

Netherlands

d_{DepartmentofCardiology,ErasmusMCUniversityMedicalCenter,Rotterdam,TheNetherlands}

Abstract

Aim:Worldwide,cardiacarrest(CA)remainsamajorcauseofdeath.Mostpost-CApatientsareadmittedtotheintensivecareunit(ICU).Theaimofthis studyistodescribemortalityratesandpossiblechangesinmortalityratesinpatientswithCAadmittedtotheICUintheNetherlandsbetween2010and 2016.

Methods:Inthisstudy,weincludedalladultCApatientsregisteredintheNationalIntensiveCareEvaluation(NICE)registrywhowereadmittedtoICUs intheNetherlandsbetween2010and2016.Theprimaryoutcomewas1-yearmortalitywhichwasanalysedbyCoxregression.Thesecondary outcomeswereICUmortalityandhospitalmortality.Hospitalmortalitywasanalysedbybinarylogisticregressionanalysis.Patientswerestratifiedby whethertheyexperiencedin-hospitalcardiacarrest(IHCA)orout-of-hospitalcardiacarrest(OHCA).Finally,theoutcomeovercalendartimewas assessedforbothgroups.

Results:Weincluded26,056CApatients:10,618(40.8%)IHCApatientsand14,482(55.6%)OHCApatients.The1-yearmortalityratewas57.5%: 59%forIHCAand56.4%forOHCA,p<0.01.Thismortalityrateremainedstablebetween2010and2016forIHCA(p=0.31)anddeclinedforOHCA patients(p=0.01).Thehospitalmortalityratewas50.3%:50.5%forIHCAand50.2%forOHCA,p=0.66.Thismortalityrateremainedstablebetween 2010 2016forIHCA(p=0.21)anddecreasedforOHCApatients(p<0.01).Anadditionalanalysiswithcalendaryearasacontinuousvariableshowed amortalitydeclineof1.56%percalendaryearfor1-yearmortality.

Conclusion:Thisnationwideregistrycohortstudyreporteda57.5%1-yearmortalityrateforCApatientsadmittedtotheICUbetween2010and2016. Wereportedadeclinein1-yearmortalityforOHCApatientsintheseyears.

Keywords:Cardiacarrest,Heartarrest,Intensivecareunits,ICU,Mortality,1-yearmortality

* Correspondingauthorat:ErasmusMCUniversityMedicalCenter,DepartmentofAdultIntensiveCare,DoctorMolewaterplein40,Rotterdam,3015GD, TheNetherlands.

E-mailaddress:l.mandigers@erasmusmc.nl(L.Mandigers).

1

Sharedlastauthor.

https://doi.org/10.1016/j.resuscitation.2019.12.029

;Accepted27December2019

0300-9572/©2020TheAuthor(s).PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/ licenses/by-nc-nd/4.0/).ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).

Available

online

at

www.sciencedirect.com

Resuscitation

Introduction

Suddencardiacarrest(CA)remainsamajorcauseofdeathworldwide (World Health Organization, WHO). In Europe, approximately 375,000adultssufferannuallyfromCA.1,2Mortalityratesinpatients admittedtothehospitalduetoCAhavebeenreportedtobebetween 58 61%.3 5_{IftheseCApatientsareadmittedtotheintensivecare}

unit(ICU),mortalityratesvarybetween53 66%.6 8_{Ofallsurviving}

CApatients25 75%havepoorneurologicaloutcomesandalarge portionsuffersfromlong-termsideeffects,9 11suchaspost-traumatic stress disorder (PTSD),12,13 impaired quality of life,14 and lower physicalandmentalfunctioning.15

Recentstudieshaveexaminedcharacteristicsthatareassociated withthemortality ofCApatients. Morespecifically, these studies focusedondifferencesinpatientcharacteristics,16,17thelocationof theCA(in-orout-of-hospital),18 20andhospitalcharacteristics.4,6 Generally,thesestudiesprovidedataonallCApatients,including patientswhodiedbeforehospitaladmission.However,thereisno recentstudyshowinganationwideoverviewofCApatientsadmitted totheICU.

IntheNetherlands,4.6%ofallpatientsadmittedtotheICUhave CAasprimarydiagnosis.21_{Theoutcomesofthesepatientsarehighly}

interestingbecausetheysurvivedthefirstepisodeofCA(namely, cardiopulmonaryresuscitation).However,theyareproneto haemo-dynamicdeterioration/instability, ischaemia/reperfusion injury, and neurologicaldamage.Inaddition,large-scaledataontheoutcomesin thispatientgrouparelacking.Thisinformationwouldbeparticularly relevantgivenrecentchangesinguidelinesandtreatments,suchas targetedtemperaturemanagement(TTM).22_{Therefore,weanalysed}

alargenationaldatabaseto investigatethemortalityratesin CA patients(bothIHCAandOHCA)admittedtoICUsintheNetherlands.

Methods

Patientdata

Inthisstudy,weusedpatientdataincludedintheNationalIntensive CareEvaluation(NICE)registry.21_{Thisisanationalqualityregistryin}

theNetherlandsforICUcare,inwhichdemographics,physiological anddiagnosticdata,patientoutcomes,andICUcharacteristicsare registered.Thedataareprospectivelycollectedwithaprimaryfocus on monitoring the qualityof care in theICU. We retrospectively analysedthedatafromapproximately85%ofICUdepartmentsin 2010to100%ofICUdepartmentsin2016intheNetherlands.21,23

WeincludedalladultCApatients(_18years)whowereadmittedto theparticipatingICUsfrom 2010to2016.Thisperiodwasselected becausestartingin2010,wewereabletodetermine1-yearmortality.21,23 TheScientificBoardoftheNICEFoundation(number2018-01)andthe Medical Ethics Committee of the Erasmus MC, Rotterdam, the Netherlands(numberMEC-2018-1228)approvedthisstudyandthe needforinformedconsentwaswaived.

Characteristicsandclinicaloutcomes

Weincludedpatientcharacteristics(i.e.,sex,age,bodymassindex (BMI),andhistory(e.g.,renalinsufficiency/dialysis,chronicobstructive pulmonarydisease(COPD)/chronicrespiratoryinsufficiency, cardio-vascularinsufficiency,livercirrhosis,(haematologic)malignancy,and immunologic insufficiency)), admission characteristics (i.e., Acute

Physiology and Chronic Health Evaluation (APACHE IV) score, estimatedmortalityrateanddiagnoseswithin24hofICUadmission (e.g., acutekidney injury(AKI), the useof mechanicalventilation, infection, the administration of thrombolytic therapy, vasoactive medicationuse,andacademic/non-academichospital),andclinical outcomes(lengthofstay,ICUmortality,hospitalmortality,and1-year mortality).SupplementaryMaterialTableAshowsthedefinitionsof thesevariables.

First, we included all patients (henceforth referred to as CA patients)registeredwithanadmissiondiagnosisofCAor cardiopul-monaryresuscitation(CPR).Next,westratifiedthecharacteristics andclinicaloutcomesforIHCAandOHCApatients.However,IHCA andOHCAwerenotencodedin theNICEregistry.Therefore,we definedIHCAasanadmissiondiagnosisofCAorCPR,withadmission originwithinthehospital,excludingtheemergencydepartment(ED). OHCAwasdefinedasanadmissiondiagnosisofCAorCPR,withan admissionoriginintheEDorhome.Hospitaland1-yearmortalityrates werecalculated,andhazardratios(HRs)peryearwereassessed relativetotheyear2010.Wedeterminedthe1-yearmortalitybyusing anadministrativeclaimsdatabasethatislinkedtotheNICEregistry (i.e.,Vektisdata).23,24

Statisticalanalysis

Allcharacteristicsandclinicaloutcomesweredescribedascounts(%) andmedians(interquartilerange,IQR),asappropriate.Thedataare shownforthetotalsampleandwerestratifiedbyIHCAandOHCA. Patients with unknown locations of CA were excluded from the stratifiedanalyses.TotestfordifferencesbetweenIHCAandOHCA patients,weusedChi-squareandWilcoxontestsforcategoricaland continuousvariables,respectively.

Theprimary outcomewas 1-year mortality. Weexaminedthe mortalitytrendovertimeusingaCoxproportionalhazardmodelwith calendaryearsfrom2010to2016asanindependentvariable.For eachcalendaryear,adummyvariablewasincluded.Thevariableswe included in each of the regression models are stated in the Supplementary Material Appendix 1. We built these regression modelsintwoways:(1)univariableanalysis and(2)multivariable analysisadjustingfordemographicandclinicalcharacteristicsand includingarandominterceptforhospital.Thiswasdonetotakethe correlationbetweenpatientsfromthesamehospitalintoaccount.To checkwhethertheselectionofdifferenthospitalsacrossthecalendar yearsmighthaveinfluencedtheresults,weperformedasensitivity analysisrestrictedtohospitalsthatregisteredtheirpatientsineach calendar year of the studyperiod (2010 2016). We presentthe hazardratios(HR)and95%ConfidenceIntervals(95%CI)ofmodel (2),asthiswasthebestpossiblerepresentationofalltheavailable data.Theadjustedeffectofcalendaryearasacategoricalvariableon thehazardofdeathwastestedbymeansofapost-estimationWald test with numberof degrees of freedomequal to the numberof calendaryearsminus1(forthereferencecategory=2010).Inthistest, calendaryearwasincludedascategoricalvariablewithoutassuming a linear trend. As an additional analysis, we performed Cox proportional hazards regression, including calendar year as a continuous variable for the different outcomes. These analyses enabledustoestimatetheaveragechangeinmortalityratesperyear overtime.AllmodelswereanalysedforIHCAandOHCAseparately. IHCAandOHCAwerenotcomparedintheseanalysesbecauseofthe non-registeredcharacteristicsrelevanttoCA.

Our secondary outcomes were ICU mortality and hospital mortality.ForICUmortality,weonlypresentapercentagetomake aroughcomparisonwithhospitalmortality.Forhospitalmortality,we analysed the data using a binary logistic regression model and presentedtheOddRatios(ORs)and95%CIs.All analyseswere performedusingR-studio.Ap-value<0.05wasconsidered statisti-callysignificant.

Results

Between2010and2016,atotalof567,856patientswereincludedin the NICE registry, and 26,056 (4.6%) of those patients were admitted due to CA: 10,618 (40.8%) IHCA patientsand 14,482 (55.6%) OHCA patients. Data from 956 (3.6%) patients were excluded fromthecomparisonof OHCAversusIHCAdue toan unknownlocationoftheCA.

Descriptivestatistics

Table1presentsthepatientcharacteristicsofallCApatientsandthen thoseofthepatientsstratifiedintotheIHCAandOHCAgroups.The majorityoftheCApatientsweremale,withanaverageageof67years atthetimeofthearrest.IHCApatientswereolderandhadoverallmore comorbiditiesthanOHCApatients.Table2presentstheadmission characteristics.ThemedianAPACHEIVestimatedmortality proba-bilitywas0.75.ThemajorityofpatientshadaGlasgowcomascale (GCS) at admission of 5 (61.7%). At admission, 87.1% of the patientsweremechanicallyventilated.Table3presentstheclinical outcomes.ThemedianlengthofICUstaywas64h(IQR21-134)and themedianhospitallengthofstaywas6days(IQR2-15).

Primaryoutcome:1-yearmortality

WithinoneyearafterICUadmission,14,974(57.5%)CApatients died.This1-yearmortalitywassignificantlyhigherinIHCApatients (59.0%)versusOHCApatients(56.4%,p<0.01).InSupplementary Material Fig. A, we present the Kaplan Meier curve for 1-year mortalityforIHCAandOHCApatients.Fig.1presentstheanalysis ofmodel(2),asdescribedinthemethodssection,of1-yearmortality for IHCA and OHCA patients separately. In IHCA patients no significantdifferencesovertimewereobserved.InOHCApatientsa

significantdecreasein1-yearmortalitybetween2010and2016was observed (all p<0.02). The sensitivity analyses restricting to hospitalsthatrecruited patientsthefullstudyperiodgavesimilar results.

Secondaryoutcomes:ICUmortality,hospitalmortality,and additionalanalyses

Ofthetotalsample,11,681(44.8%)CApatientsdiedintheICU.The ICU mortality rates were slightly lowerfor IHCA than for OHCA patients(44.2%versus45.4%,respectively,p=0.05).Duringtheir hospitaladmission,13,072(50.3%)CApatientsdied.Thishospital mortalityratewascomparableinIHCAandOHCApatients(50.5% versus50.2%,respectively,p=0.66).

For hospital mortality, there was no significant trend in IHCA patientsovertime(Waldtest(df)8.39(6),p=0.21).However, the analysisforhospitalmortalityinOHCApatientswithcalendaryearasa categorised variable showed significant differences between the calendaryearsWaldtest(df)22.78(6),p_<0.01).InspectingtheOdds Ratioswefoundadecreasingtrend,asshowninFig.2.

Next,theresultsoftheadditionalanalysiswithcalendaryearasa continuousvariableshowednosignificanttrendin1-yearmortality overtimeforIHCApatients(HR0.99,95%CI0.98-1.00,p=0.13). Furthermore,itconfirmed theobserveddeclinein1-yearmortality overtimeforOHCApatients(HR0.98,95%CI0.97-1.00,p_<0.01). ThisHRshowsareductionin1-yearmortalityof1.56%peryearover thestudyperiod.

As wefoundadecreasing trendin 1-yearmortalityfor OHCA patients,wedecidedtoperformanadditionalCoxregressionusingleft truncationatthetimeofhospitaldischarge(i.e.,onlyselectingthe hospitalsurvivors),seeSupplementaryMaterialFig.B.Thisshows thatasignificanttrendovertimeremainspresent(p<0.01).

Discussion

Thisisthefirstlargenationwidestudyon1-yearmortalityratesofCA patientsadmittedtotheICU,anditwasperformedintheNetherlands between2010and2016.Thesecondaryoutcomeswedescribedwere ICUmortalityandhospitalmortality.Overall,wefoundreasonable mortalityratesandasignificantdecreasein1-yearmortalityofOHCA patientsbetween2010and2016.

Table1–Patientcharacteristics.

Totalsamplea IHCA OHCA p-value

Patients2010 2016 26,056 10,618 14,482 Age(IQR) 67(57 76) 69(59 77) 66(55 75) <0.01 Gender,male(%) 17,320(66.5) 6731(63.4) 9967(68.8) <0.01 BMI(IQR) 25.7(23.4 28.7) 25.7(23.4 28.9) 25.7(23.4 28.4) 0.99 BMImissing(%) 711(6.6) 985(6.8) History(N=26,056) Cardiovascularinsufficiency(%) 2291(8.8) 1160(10.9) 1030(7.1) <0.01

COPD/respiratoryinsufficiency(%) 4070(15.6) 1894(17.8) 2042(14.1) <0.01

Renalinsufficiency(%) 1802(6.9) 976(9.2) 765(5.3) <0.01

Livercirrhosis(%) 273(1) 138(1.3) 125(0.9) 0.01

(Hematologic)malignancy(%) 1092(4.2) 692(6.5) 374(2.6) <0.01

Immunodeficiency(%) 1316(5.1) 779(7.3) 500(3.5) <0.01

Table2_–AdmissioncharacteristicsinCApatientsadmittedtotheICUbetween2010 2016.

Totalsamplea _{IHCA OHCA p-value}

APACHEIVestimatedmortalityrate(IQR) 0.75(0.45 0.89) 0.71(0.32 0.90) 0.77(0.54 0.88) <0.01

Admissiontype <0.01 Medical 23,156(88.9) 7968(75) 14,291(98.7) Urgentsurgical 1682(6.5) 1517(14.3) 122(0.8) Electivesurgical 1103(4.2) 1088(10.2) 6(<1) Diagnosisonadmission GCS5(%) 16,066(61.7) 5333(50.2) 10,167(70.2) <0.01 GCS6 14(%) 4057(15.6) 1736(16.3) 2148(14.8) GCS15(%) 5391(20.7) 3327(31.3) 1851(12.8) GCSmissing(%) 222(2.1) 316(2.2) Dysrhythmia(%) 13,343(51.2) 5138(48.4) 7676(53) <0.01 Mechanicalventilation(%) 22,701(87.1) 8696(81.9) 13,189(91.1) <0.01 CVA(%) 1053(4.1) 455(4.3) 558(3.9) 0.10 Intracranialmass(%) 708(2.7) 264(2.5) 421(2.9) 0.04

Gastrointestinalbleeding(%) 480(1.8) 277(2.6) 189(1.3) <0.01

Diabetes(%) 4438(17) 2045(19.3) 2221(15.3) <0.01

Diagnosisat24hofICUadmission

GCS5(%) 13,541(52) 4615(43.5) 8420(58.1) <0.01 GCS6 14(%) 4198(16.1) 1621(15.3) 2407(16.6) GCS15(%) 7664(29.4) 4095(38.6) 3312(22.9) AKI(%) 4617(17.7) 2263(21.3) 2188(15.1) <0.01 Mechanicalventilation(%) 23,666(90.8) 9229(86.9) 13,592(93.9) <0.01 Infection(%) 2768(10.6) 1641(15.5) 1062(7.3) <0.01 Vasoactivemedication(%) 18,962(72.8) 7640(72) 10,652(73.6) <0.01 Thrombolytictherapy(%) 1354(5.2) 560(5.3) 756(5.2) 0.87 Academichospital(%) 7956(23.9) 3102(22.2) 4656(26.1) <0.01

a_{In956patients,itwasunknownifthecardiacarrestoccurredin-oroutsidethehospital;thesepatientswereexcludedfromtheanalyses.}

Table3_–ClinicaloutcomesinCApatientsadmittedtotheICUbetween2010 2016.

Totalsamplea _{IHCA OHCA p-value}

LengthofICUstayinhours(IQR) 64.2(21.4 133.8) 52.1(17.2 135.2) 69(26.4 133.2) <0.01

Hospitallengthofstayindays(IQR) 5.9(1.8 14.8) 6.1(1.5 15.1) 5.9(2 14.6) 0.27

ICUmortality(%) 11,681(44.8) 4690(44.2) 6579(45.4) <0.05

Hospitalmortality(%) 13,072(50.3) 5346(50.5) 7256(50.2) 0.66

1-yearmortality(%) 14,974(57.5) 6263(59.0) 8169(56.4) <0.01

a_{In956patients,itwasunknownifthecardiacarrestoccurredinoroutsidethehospital;thesepatientswereexcludedfromtheanalyses.}

Fig.1–MultivariableCoxregressionanalysisof1-yearmortalityforIHCA(p=0.31)andOHCA(p=0.01)overcalendar time.

Asaprimaryoutcome,wereporteda1-yearmortalityrateof57.5% inthetotalcohort.This1-yearmortalityratewasslightlyhigherinIHCA patientsthaninOHCApatients.Thoughinasmallersample,Engsig etal.25_{investigatedlong-term outcomesandshowedcomparable}

1-yearmortalityrates,whichweresimilarforIHCAandOHCApatients (47%and51%,respectively).Wefoundastable1-yearmortalityrate between2010and2016forIHCApatients.Remarkably,the1-year mortalityofOHCApatientsdecreasedinthistimeperiod,andthis decreasepersistedafterlimitingtheanalysistothehospitalsurvivors (seeSupplementaryMaterialFig.C).Thiscouldpointtoahealthier dischargedCApatientpopulation.

Asasecondaryoutcome,westudiedhospitalmortality,whichwas approximately50%inthetotalcohort,aswellasforbothIHCAand OHCApatientsseparately.Inaddition,wefoundthathospitalmortality wasstableforIHCApatientsbetween2010and2016.ForOHCA,we foundadecreasingtrendinhospitalmortalityinthistimeperiod.These mortalityratesarein linewiththose inpreviousstudiesreporting hospitalmortalityratesinCApatients.7,26 28

Inthisstudy,wefoundamedianAPACHEIVestimatedmortality rateof0.75,whichismuchhigherthantheobservedhospitalmortality rate.Zimmermanet al.29 validated thisAPACHE IVscore forCA patientsintheUnitedStates.However,Brinkmanetal.30showedpoor APACHEIVscoreperformanceinCApatientsintheNetherlands.Our studywasperformedwiththesamedataregistryaswasthestudyby Brinkmanetal.30_{BecausetheAPACHEIVscoreisnotvalidatedforCA}

patientsintheNetherlands,thisdifferenceinexpectedandobserved mortalitycouldbeexplained.

Giventhelimitationsofanationwideobservationaldataset,we couldnotfullystudytheunderlyingcauseforthereductioninmortality among the OHCA patients. However, we would like to discuss possibleexplanationsforthedifferencesinmortalityratesbetween IHCAandOHCApatients andthedecline in1-year andhospital mortalityofOHCApatients.

First,althoughwehavecorrectedouranalysesforrelevantpatient characteristics, CAcharacteristics may have changed during the studyperiod.Forexample,intheNetherlandspublicawarenessofCA isincreasingovertimeduetonationwideeducation.31Thiscouldhave resulted in a higher percentage of bystander life support and automated externaldefibrillator(AED)use.In turnthismayhave

resultedinashortertimetothereturnofspontaneouscirculation (ROSC),whichwewereunabletoexamineinthepresentstudy.

Second,AEDusemayhavecontributedtothe decreaseinthe mortality rate in OHCA patients. Despite the promising results, worldwide AED use is still quite limited compared to the Netherlands.18,20,32,33Blometal.32showedthatintheNetherlands, theuseofAEDsandhospitalsurvivalbothincreasedovertimeinthe period2006 2012.Inthissameperiod,Ringhetal.34_reportedan

increase inthenumberanduseofpublic AEDsanditseffectonmortality ratesinSweden.Takentogether,inouropinion,thewidespreaduseof AEDsintheNetherlandsprobablycontributedtothedecreasein1-year mortalityinOHCApatientsfoundinourstudy.

Anotherpossibleexplanationforthedecreaseinthemortalityrate in OHCApatientsisthepredominantuseoftargetedtemperature management(TTM),comparedtoitslimiteduseinIHCApatients. Mountingresearchhasbeenperformedontemperaturemanagement in CA patients,8,22,27,28,35 37 mostly these studies have been performedinOHCApatients.Asstatedinthe2015guidelines,since 2010,atemperatureof32 34Cisrecommended,whichchangedin 2015 to36C.22 _{Severalstudies showeddifferent results ingoal}

temperaturesfortheTTM,buttheyallrecommendusinghypothermia or normothermiaand preventinghyperthermia.8,22,35 37 However, Engsigetal.25_{showednodifferenceinIHCAandOHCApatients.}

Wangetal.38_{showedabenefitforTTMinIHCApatients,butTTMwas}

performedinonly3.2%ofthepatients.Chanetal.39_{alsoshowedalow}

TTMrateinIHCApatientsandtheyfoundnoassociationbetween TTMandsurvivalorneurologicaloutcomes.IntheNetherlands,itis difficulttopinpointtheexactdateoftheimplementationoftheTTM guidelines,asshownbyPickkersetal,8_{whileatthesametimemany}

hospitals started usingTTM forOHCA patientsduring our study period.Thus,thismayhavecontributedtothedifferencesinmortality trendsbetween2010and2016inIHCAandOHCApatientsfoundin ourstudy.Takentogether,TTMisusedmoreofteninOHCApatients andhasshownsomepromisingresults.However,TTMisnotusedas ofteninIHCApatients,althoughthismayactuallybe apromising treatmentstrategyforthesepatientstoo.

Fourth,intheNetherlands,nostudieshavebeenconductedon theeffectofcardiacrehabilitationonthemortalityratespecificallyin CA patients. However, some studies on the effect of cardiac Fig.2–BinarylogisticregressionanalysisofhospitalmortalityforIHCA(p=0.21)andOHCA(p<0.01)overcalendar time.

rehabilitationon mortality in acute coronary syndrome and cardiac surgerypatientsshowedalowermortalityrateforthosewhoreceived cardiacrehabilitationcomparedtothosewhodidnot.40,41_OHCApatients

aremorelikelytoreceivecardiacrehabilitation.10,42Forthesepatients,a cardiac cause such as coronary disease, was most likely the reason for the CA.Giventhedecreasein1-year mortalityratesinourstudy,which persisted after selecting only hospital survivors, it is likely that rehabilitation therapycontributedtothislower1-yearmortalityinOHCApatients.Itmay clarifythedifferencebetweenIHCAandOHCAoutcomes.

Finally,duringthelastdecadethetreatmentofcoronarydiseases improved.Advancesincoronaryrevascularizationandadherenceto secondarypreventionguidelines(includinginternalcardiacdefibrillatory therapy)mayhavecontributedtoanimportantimprovementinmortality ratesinOHCApatients,whilethisisnotthecaseinIHCApatients.

Future research may study the possible effects of pre-ICU characteristicsandin- andpost-hospitaltreatmentson short-and long-termmortalityratesofCApatientsadmittedtotheICU.Ourlinear trendshould beregardedasacrudeaverageofvariousup- and downwardmovementsofthegraph.Intherestrictedtimeframeofour analysis,thesemovementsarehardtointerpretascoincidenceordue tospecificcauses.Forthisreason,nononlineartrendswereincluded intheanalysis.Incaseastudycanbeperformedwithmoredata relevanttoCAfromalongertimeperiodamorespecifictrendanalysis (e.g.non-lineartrend)willbeinformative.

Limitations

Asinevery(retrospective)study,thisresearchhadseverallimitations. First,asmentionedbefore,theNICEregistryisaimedatqualityofcare attheICUandnotallcharacteristicsrelevanttoCAareregistered.In particular,wedidnot haveaccesstothe followingcharacteristics: witnessed/unwitnessed CA,CPR delay, time until ROSC,primary cardiacrhythm,cardiacinterventions,AEDuse,mechanical compres-sion deviceuseand cause of arrest. Thesecharacteristicsare important whenstudyingthedeterminantsoftheoutcomeofCA.Therefore,in futureresearch,thesecharacteristicsmustbetakenintoaccount.

Second,theNICEregistrydoesnotrecordwhethertheCAtook placein-orout-of-hospital.Wehadtodeterminethiswiththebest possibleapproximation.Withthismethod,weexpectthatsomeofthe IHCApatientswere misclassifiedasOHCA. Mostofthe patients admittedattheICUwithadmission-originEDwereOHCApatients; however,someofthemexperiencedCAwhileintheED.Weassume thatthisissuehaslimitedconsequencesfortheresults,butwecannot excludethepossibilityofsomebias.

Third, in this study, we could only report mortality rates. Unfortunately,therewerenodataavailableonneurologicaloutcomes; therefore,wewereunabletoreportsurvivalwithgoodneurological outcome.

Finally,insomehospitalsintheNetherlands,post-CApatientsare admittedtothecardiaccareunitinsteadoftheICU.Wewereunableto estimatehowmanypatientswereinthisgroupasthisisnotrecorded orthesedatawerenotavailabletotheauthors.Asaconsequence,this couldresultindifferentnumbersofpatientsincomparisontotheCA numbersreportedinotherstudies.

Conclusion

This nationwide registry cohort study reported a 57.5% 1-year mortalityrateforCApatientsadmittedtotheICUbetween2010and

2016.Wereportedadeclinein1-yearmortalityforOHCApatientsin theseyears.

Conflicts

of

interest

None.

Acknowledgments

None.

Appendix

A.

Supplementary

data

Supplementarymaterialrelatedtothisarticlecanbefound,inthe onlineversion,atdoi:10.1016/j.resuscitation.2019.12.029.

REFERENCES

1.AtwoodC,EisenbergMS,HerlitzJ,ReaTD.Incidenceof EMS-treatedout-of-hospitalcardiacarrestinEurope.Resuscitation 2005;67:75 80.

2.NolanJP,SoarJ,SmithGB,etal.Incidenceandoutcomeofin-hospital cardiacarrestintheUnitedKingdomNationalCardiacArrestAudit. Resuscitation2014;85:987 92.

3.GutermanEL,KimAS,JosephsonSA.Neurologicconsultationand useoftherapeutichypothermiaforcardiacarrest.Resuscitation 2017;118:43 8.

4.KurzMC,DonnellyJP,WangHE.Variationsinsurvivalaftercardiac arrestamongacademicmedicalcenter-affiliatedhospitals.PLoSOne 2017;12:e0178793.

5.RushB,AshkananiM,RomanoK,HertzP.Utilizationof electroencephalogrampostcardiacarrestintheUnitedStates:a nationwideretrospectivecohortanalysis.Resuscitation

2017;110:141 5.

6.SchoberA,HolzerM,HochrieserH,PoschM,SchmutzR,MetnitzP. Effectofintensivecareaftercardiacarrestonpatientoutcome:a databaseanalysis.CritCare2014;18:R84.

7.StraneyLD,BrayJE,FinnJ,BernardS,PilcherD.Trendsinintensive careunitcardiacarrestadmissionsandmortalityinAustraliaandNew Zealand.CritCareResusc2014;16:104 11.

8.vanderWalG,BrinkmanS,BisschopsLL,etal.Influenceofmild therapeutichypothermiaaftercardiacarrestonhospitalmortality.Crit CareMed2011;39:84 8.

9.PeruckiWH,HiendlmayrB,O’SullivanDM,GunaseelanAC,FayasF, FernandezAB.Magnesiumlevelsandneurologicoutcomesin patientsundergoingtherapeutichypothermiaaftercardiacarrest.Ther HypothermiaTempManag2018;8:14 7.

10.RadeschiG,MinaA,BertaG,etal.Incidenceandoutcomeof in-hospitalcardiacarrestinItaly:amulticentreobservationalstudyinthe PiedmontRegion.Resuscitation2017;119:48 55.

11.WangCH,HuangCH,ChangWT,etal.Outcomesofadultswith in-hospitalcardiacarrestafterimplementationofthe2010resuscitation guidelines.IntJCardiol2017;249:214 9.

12.GamperG,WilleitM,SterzF,etal.Lifeafterdeath:posttraumatic stressdisorderinsurvivorsofcardiacarrest—prevalence,associated factors,andtheinfluenceofsedationandanalgesia.CritCareMed 2004;32:378 83.

13.MongardonN,DumasF,RicomeS,etal.Postcardiacarrest syndrome:fromimmediateresuscitationtolong-termoutcome.Ann IntensiveCare2011;1:45.

14.WachelderEM,MoulaertVR,vanHeugtenC.Lifeaftersurvival: long-termdailyfunctioningandqualityoflifeafteranout-of-hospitalcardiac arrest.Resuscitation2009;80:517 22.

15.BunchTJ,WhiteRD,GershBJ,etal.Long-termoutcomesof out-of-hospitalcardiacarrestaftersuccessfulearlydefibrillation.NEnglJ Med2003;348:2626 33.

16.Al-DuryN,RawshaniA,IsraelssonJ,etal.Characteristicsand outcomeamong14,933adultcasesofin-hospitalcardiacarrest:a nationwidestudywiththeemphasisongenderandage.AmJEmerg Med2017;35:1839 44.

17.LibunganB,LindqvistJ,StromsoeA,etal.Out-of-hospitalcardiac arrestintheelderly:alarge-scalepopulation-basedstudy. Resuscitation2015;94:28 32.

18.HawkesC,BoothS,JiC,etal.Epidemiologyandoutcomesfrom out-of-hospitalcardiacarrestsinEngland.Resuscitation2017;110:133 40.

19.KazaureHS,RomanSA,SosaJA.Epidemiologyandoutcomesof in-hospitalcardiopulmonaryresuscitationintheUnitedStates,2000

2009.Resuscitation2013;84:1255 60.

20.RoYS,ShinSD,SongKJ,etal.Atrendinepidemiologyandoutcomes ofout-of-hospitalcardiacarrestbyurbanizationlevel:anationwide observationalstudyfrom2006to2010inSouthKorea.Resuscitation 2013;84:547 57.

21.NICE.Datainbeeld.2016.

22.NolanJP,SoarJ,CariouA,etal.EuropeanResuscitationCounciland EuropeanSocietyofIntensiveCareMedicine2015guidelinesfor post-resuscitationcare.IntensiveCareMed2015;41:2039 56.

23.vandeKlundertN,HolmanR,DongelmansDA,deKeizerNF.Data resourceprofile:theDutchNationalIntensiveCareEvaluation(NICE) RegistryofAdmissionstoAdultIntensiveCareUnits.IntJEpidemiol 2015;44:1850-h.

24.KarakusA,HaasLEM,BrinkmanS,deLangeDW,deKeizerNF. Trendsinshort-termand1-yearmortalityinveryelderlyintensivecare patientsintheNetherlands:aretrospectivestudyfrom2008to2014. IntensiveCareMed2017;43:1476 84.

25.EngsigM,SoholmH,FolkeF,etal.Similarlong-termsurvivalof consecutivein-hospitalandout-of-hospitalcardiacarrestpatients treatedwithtargetedtemperaturemanagement.ClinEpidemiol 2016;8:761 8.

26.CarrBG,GoyalM,BandRA,etal.Anationalanalysisofthe relationshipbetweenhospitalfactorsandpost-cardiacarrestmortality. IntensiveCareMed2009;35:505 11.

27.JonesD,BhasaleA,BaileyM,PilcherD,AnsteyMH.Effectofa nationalstandardfordeterioratingpatientsonintensivecare admissionsduetocardiacarrestinAustralia.CritCareMed 2018;46:586 93.

28.SalterR,BaileyM,BellomoR,etal.Changesintemperature managementof cardiac arrestpatients followingpublicationofthetarget temperaturemanagementtrial.CritCareMed2018;46:1722 30.

29.ZimmermanJE,KramerAA,McNairDS,MalilaFM,ShafferVL. Intensivecareunitlengthofstay:benchmarkingbasedonAcute

PhysiologyandChronicHealthEvaluation(APACHE)IV.CritCare Med2006;34:2517 29.

30.BrinkmanS,Bakhshi-RaiezF,Abu-HannaA,etal.Externalvalidation ofAcutePhysiologyandChronicHealthEvaluationIVinDutch intensivecareunitsandcomparisonwithAcutePhysiologyand ChronicHealthEvaluationIIandSimplifiedAcutePhysiologyScoreII. JCritCare201126:105e11-8.

31.Hartstichting.Forsetoenamesnellehulpbijeenhartstilstand. Hartstichting;2019.

32.BlomMT,BeesemsSG,HommaPC,etal.Improvedsurvivalafter out-of-hospitalcardiacarrestanduseofautomatedexternaldefibrillators. Circulation2014;130:1868 75.

33.GirotraS,vanDiepenS,NallamothuBK,etal.Regionalvariationin out-of-hospitalcardiacarrestsurvivalintheUnitedStates.Circulation 2016;133:2159 68.

34.RinghM,JonssonM,NordbergP,etal.Survivalafterpublicaccess defibrillationinStockholm,Sweden—astrikingsuccess.

Resuscitation2015;91:1 7.

35.DonninoMW,AndersenLW,BergKM,etal.Temperature managementaftercardiacarrest:anadvisorystatementbythe advancedlifesupporttaskforceoftheInternationalLiaisonCommittee onResuscitationandtheAmericanHeartAssociationEmergency CardiovascularCareCommitteeandtheCouncilonCardiopulmonary, CriticalCare,PerioperativeandResuscitation.Resuscitation 2016;98:97 104.

36.KalraR,AroraG,PatelN,etal.Targetedtemperaturemanagement aftercardiacarrest:systematicreviewandmeta-analyses.Anesth Analg2018;126:867 75.

37.SchenoneAL,CohenA,PatarroyoG,etal.Therapeutichypothermia aftercardiacarrest:Asystematicreview/meta-analysisexploringthe impactofexpandedcriteriaandtargetedtemperature.Resuscitation 2016;108:102 10.

38.WangCH,HuangCH,ChangWT,etal.Outcomesofadultin-hospital cardiacarresttreatedwithtargetedtemperaturemanagement:a retrospectivecohortstudy.PLoSOne2016;11:e0166148.

39.ChanPS,BergRA,TangY,CurtisLH,SpertusJA.AmericanHeart Association’sGetWiththeGuidelines-ResuscitationI.Association BetweenTherapeuticHypothermiaandSurvivalAfterIn-Hospital CardiacArrest.JAMA2016;316:1375 82.

40.SunamuraM,TerHoeveN,vandenBerg-EmonsRJG,BoersmaE, vanDomburgRT,GeleijnseML.Cardiacrehabilitationinpatientswith acutecoronarysyndromewithprimarypercutaneouscoronary interventionisassociatedwithimproved10-yearsurvival.EurHeartJ QualCareClinOutcomes2018;4:168 72.

41.deVriesH,KempsHM,vanEngen-VerheulMM,KraaijenhagenRA, PeekN.Cardiacrehabilitationandsurvivalinalargerepresentative communitycohortofDutchpatients.EurHeartJ2015;36:1519 28.

42.DumasF,ReaTD.Long-termprognosisfollowingresuscitationfrom out-of-hospitalcardiacarrest:roleofaetiologyandpresentingarrest rhythm.Resuscitation2012;83:1001 5.