Epinephrine for out of hospital cardiac arrest: A systematic review and meta-analysis of randomized controlled trials

Review

Epinephrine

for

out

of

hospital

cardiac

arrest:

A

systematic

review

and

meta-analysis

of

randomized

controlled

trials

Maria

Vargas

*

,

Pasquale

Buonanno

,

Carmine

Iacovazzo

,

Giuseppe

Servillo

a_{DepartmentofNeurosciences,ReproductiveandOdontostomatologicalSciences,UniversityofNaples“FedericoII” ,}

Naples,Italy

b

DepartmentofAnesthesiaandIntensiveCare,IstitutodiRicoveroeCuraaCarattereScientificoIstitutoNeurologico Mediterraneo,Neuromed,Pozzilli,Italy

Abstract

Objective:Toevaluatetheeffectivenessofepinephrine,comparedwithcontroltreatments,onsurvivalatadmission,ROSC,survivalatdischarge,and afavorableneurologicoutcomeinadultpatientsduringOHCA.

Datasource:MEDLINEandPubMedfrominceptiontoAugust2018.

Studyselection:Randomizedcontrolledtrials(RCTs)onadultpatientsafterOHCAtreatedwithepinephrineversuscontrols. Dataextraction:Independent,double-dataextraction;riskofbiasassessmentwithCochraneCollaboration’scriteria.

Datasynthesis:15RCTsrepresenting20716OHCAadultpatients.Epinephrine,comparedwithallpooledtreatments,wasassociatedwithabetter survivalratetohospitaldischarge(RR:1.16,95%CI:1.00–1.35)andafavorableneurologicoutcome(RR:1.24,95%CI:1.04–1.48).Nodifferencewas foundinsurvivaltohospitaladmission(RR:1.02,95%CI:0.75–1.38)andROSCwhencomparingepinephrinewithallpooledtreatments(RR:1.13,95% CI:0.84–1.53).Whenepinephrinewascomparedwithaplacebo/nodrugs,survivaltohospitaldischarge(RR:1.34,95%CI:1.08–1.67),ROSC(RR: 2.03,95%CI:1.18–3.51)andsurvivaltohospitaladmission(RR:2.04,95%CI:1.22–3.43)wereincreased,buttherewasnotafavorableneurologic outcome(RR:1.22,95%CI:0.99–1.51).

Conclusions:InOHCA,standardorhighdosesofepinephrineshouldbeusedbecausetheyimprovedsurvivaltohospitaldischargeandresultedina meaningfulclinicaloutcome.Therewasalsoaclearadvantageofusingepinephrineoveraplaceboornodrugsintheconsideredoutcomes. Keywords:Out-of-hospitalcardiacarrest,Epinephrine,Hospitalsurvival,Fragilityindex

Introduction

Overallsurvivaltohospitaldischargeafter out-of-hospitalcardiac arrest(OHCA)rangedfrom8%to10%.1_{Severalfactorsaffectedthe}

hospital survival,suchas cardiopulmonaryresuscitation ofgood quality andadequate post-resuscitation care.Standardized algo-rithmsforadvancedlifesupport(ALS)andpost-resuscitationcare havebeenimplementedintheEuropeanguidelinesfor resuscita-tion.2 _{The last guideline regarding ALS recommends using}

* Correspondingauthorat:DepartmentofNeurosciences,ReproductiveandOdontostomatologicalSciences,UniversityofNaples“FedericoII”,Via Pansini,80100,Naples,Italy.

E-mailaddress:vargas.maria82@gmail.com(V.Maria).

https://doi.org/10.1016/j.resuscitation.2019.01.016

Received22September2018;Receivedinrevisedform13January2019;Accepted14January2019 Availableonlinexxx

0300-9572/2019ElsevierB.V.Allrightsreserved.

Available

online

at

www.sciencedirect.com

Resuscitation

epinephrine (1mg) every 3–5min until return of spontaneous circulation (ROSC) is achieved. Actually, the optimal dose of epinephrineisnotknown,andtherearenohumandatasupporting the use of repeated doses. Increasing cumulative doses of epinephrine during resuscitation of patients with asystole and PEA is an independent riskfactor for an unfavorable functional outcomeandin-hospitalmortality.2_{Accordingtoprevious}

system-aticreviews,theuseofadrenalineforOHCAincreasedtheratesof ROSC but did not improve long-term survival and a positive neurologicoutcome.2Recently,thePARAMEDIC2trialshowedthat epinephrineinOHCAwasassociatedwithaslightimprovementin the30-day survival, butseveral survivors experienced asevere neurologicimpairment.3

Theaim of thissystematic reviewand meta-analysis was to evaluatetheeffectivenessofepinephrine,comparedwithcontrol treatments,onsurvivalatadmission,ROSC,survivalatdischarge, andabeneficialneurologicoutcomeinadultpatientsduringOHCA. Because we found in the literature different control treatments during OHCA, we planned (a priori comparisons) to do a sub-analysis,dividingthestudiesaccordingto(1)thetreatmentsusedin thecontrolgroups,whichwereaplacebo/nodrugs,ahighdoseof epinephrineandepinephrineplusvasopressin,and(2)thefragility index,whichwascalculatedon theprimaryoutcomedeclared by eachstudy.

Methods

This meta-analysis is registered with PROSPERO (number CRD42018114339).

Datasourcesandsearchstrategy

Weaimedtoidentifyall randomizedcontrolledtrials (RCTs)on adult patients after OHCA who were treated with epinephrine versusthe controls.Theelectronicsearchstrategy wasapplied withstandardfiltersforidentificationoftheRCTs.Thedatabases searchedwereMEDLINEandPubMed(frominceptiontoAugust 2018).We applied anEnglishlanguagerestriction. Thesearch strategy included the following keywords: cardiac arrest, out-of-hospitalcardiacarrest,circulatoryarrest,cardiopulmonary resuscita-tion, adrenaline, epinephrine, vasopressin, mortality to hospital admission,survivaltohospitaladmission,humansandrandomized clinicaltrial.

Studyselection

Weincludedonlypublishedfullpapers.WhenmorethanoneRCTwas notavailable for eachtopic, we considered observational clinical studies.Datawereindependentlyextractedfromeachstudybytwo authors(MVandPB)usingadatarecordingformdevelopedforthis purpose.

Interventions

Theinterventionsofinterestwerethecomparisonsbetweenthe standard dose of epinephrine (SDE) versus all the pooled treatments,SDEversus aplaceboornodrugs, SDEversus a highdoseofepinephrine(HDE)>1mgperdose,andSDEversus epinephrine+vasopressin. We compared the SDE with all

pooled treatments because data on the effectiveness of this drug during OHCA come from conflicting studies. All pooled treatments consisted of a placebo or no drugs, HDE, and epinephrine+vasopressin.

Outcome

Theprimaryoutcomewas thesurvivalto hospitaldischargeafter OHCA.Thesecondaryoutcomeswerethereturnof spontaneous circulation (ROSC), survival to hospital admission and a good neurologicoutcome.Agoodneurologicoutcomewasdefinedasa cerebral performance category (CPC) of 1 and 2, an overall performancecategory(OPC)of1and2,amodifiedRankinScale scoreof1and2,andanormalormoderatedisabilityatthehospital discharge.

Dataextractionandqualityassessment

Theinitialdataselectionwasperformedbyscreeningtitlesand abstractsbytwopairsofindependentreviewers(MVandPB;GS and CI). The full-text copy of potentially relevant studies was obtained for detailed evaluation. Data from each study were independently extracted by two pairsof independent reviewers (MVandPB;GSandCI)usingapre-standardizeddataabstraction form.Dataextractedfromthestudieswereindependentlychecked foraccuracybytworeviewers(MVandGS).Aqualityassessment was conductedby tworeviewers (CI and PB)with the GRADE approach. The quality evaluation included (1) the use of randomizationsequencegeneration,(2)thereportingofallocation concealment,(3)blinding,(4)reportingincompleteoutcomedata, and(5)comparabilityofthegroupsatthebaseline.Wesolvedany possibledisagreementbyconsensusthroughconsultationwithan external reviewer, ifneeded. We further calculatedthe fragility index(FI) for eachstudyto assess its robustness.The FIwas calculatedonthevariable/sthateachstudydeclaredasaprimary outcome/s.

Quantitativeanalysis

This meta-analysiswasconductedaccordingto PRISMA guide-lines.4 _{A mixed random effect with the DerSimonian and Laird}

method was used in this meta-analysis. The results were graphicallyrepresentedwithforestplotgraphs.TheRelativeRisk (RR)and95%CIforeachoutcomewereseparatelycalculatedfor eachtrialwithgroupeddatausingtheintention-to-treatprinciple. ThechoicetouseRRwasdrivenbythedesignofthemeta-analysis based on the RCTs. Tau2 defined the variance between the studies.Thedifferenceinestimatesofthetreatmenteffectbetween thetreatmentandcontrolgroupsforeachhypothesiswastested usingatwo-sidedztestwithstatisticalsignificanceconsideredata p value of less than 0.05. The homogeneity assumption was checkedbyaQtestwithadegree offreedom(df)equaltothe number of analyzed studies minus 1. The heterogeneity was measuredbyI,whichdescribesthepercentage oftotalvariation acrossstudiesthatisduetoheterogeneityratherthanchance.I2

wascalculated from basicresultsobtained fromatypical meta-analysis as I2_{=100% A}

 (Q _df)/Q,where Q is Cochran_’s heterogeneitystatisticanddfisthedegreeoffreedom.Avalueof 0% indicates no observed heterogeneity, and larger values demonstrateincreasingheterogeneity.

Weplanned(aprioricomparisons)todoasub-analysistoanalyze alltheoutcomesaccordingtothefollowingcategorieswhenpossible: SDEversusallpooledtreatments(allpooledtreatmentsincludeda placebo/nodrugs,HDEandepinephrine+vasopressin),SDEversus aplacebo/nodrugs,SDEversusHDE,andSDEversusepinephrine+

vasopressin(Epi+Vaso).

We evaluated the FI of the RCTs included in this meta-analysis using a two-by-two contingency table anda p value produced by the Fisher exact test.5 According to the FI, we definedrobustRCTswithFI>0,andnotrobustRCTswithFI=0. Wefurtheranalyzedalloutcomesaccordingto(1)robustRCTs with FI>0, and(2) not robust RCTs with FI=0. The analyses were conducted with OpenMetaAnalyst (version 6) and SPSS version20(IBMSPSS).

Toevaluatepotentialpublicationbias,aweightedlinearregression wasused,withthenaturallogoftheORasthedependentvariable, andtheinverseofthetotalsamplesizeastheindependentvariable. ThisisamodifiedMacaskill’stestthatgivesmorebalancedtype-Ierror ratesinthetailprobabilityareasincomparisontootherpublication biastests.6

Result

Studyselection

Atotalof1986studieswere identified,and783were duplicated; 108 full-text articles were assessed for eligibility; and 15 RCTs,

involving20716patients,werefinallyincludedintheanalysis.7–21 Fig.1illustratestheflowdiagramofincludedstudies.

Characteristicsoftheincludedstudy

Patients with OHCAwere randomized to receive SDE versusa placebo/no drugs in three studies.7–9 _{Six studies randomly}

compared patients receivingSDE versusHDE,10–15 whereassix studies compared SDE versus Epi+Vaso.16–21 Five studies consideredmorethanoneprimaryoutcome.11,13,15,16,18.According to the FI, only three studies had an FI>0.8,11,15 The Table 1Supplementary materials reported thecharacteristics of theincludedstudies.

Qualityassessment

All the included RCTs had a low risk of bias. The

Table 2Supplementarymaterials showed thequalityassessment foreachincludedstudy.

Primaryoutcome

TheSDEimprovedsurvivaltohospitaldischargewhencompared withallthepooledtreatments(SDEversusall,RR:1.16,95%CI: 1.00_–1.35,p=0.04)and withaplacebo/nodrugs (SDEversusa placebo/no drugs,RR:1.34,95% CI:1.08–1.67,p=0.00). These resultswereconfirmedalsobytheanalysisincludingrobustand not-robusttrials(SDEFI>0versuscontrol,RR:1.32,95%CI:1.06–

1.65,p=0.01;SDEFI=0versuscontrol,RR:1.05,95%CI:0.86_– 1.28,p=0.63).

Therewasnodifferenceinsurvivaltohospitaldischargewhen comparingtheSDEwiththeHDE(SDEversusHDE,RR:1.03,95% CI:0.75–1.41,p=0.80)andtheSDEwithEpi+Vaso(SDEversus Epi+Vaso,RR:0.99,95%CI:0.69–1.43,p=0.99).Fig.2showsthe forestplotforthesurvivaltohospitaladmission.

Secondaryoutcomes

NodifferencewasfoundintheROSCwhencomparingtheSDEwith allthepooledtreatments(SDEversusall,RR:1.14,95%CI:0.84_– 1.53,p=0.40).TheSDEimprovedtheROSCwhencomparedwitha

placebo/nodrugs(SDEversusaplacebo/nodrugs,RR:2.03,95%CI: 1.18_–3.51,p=0.01).PatientstreatedwiththeHDEhadahigherrateof ROSCcomparedwiththosetreatedwiththeSDE(SDEversusHDE, RR:0.85,95%CI:0.74–0.97,p=0.01).Therewasnodifferenceinthe rateofROSCbetweentheSDEandtheEpi+Vaso(SDEversus Epi+Vaso,RR:1.02,95%CI:0.91–1.14,p=0.71).Fig.3showsthe forestplotfortheROSC.TherewasnodifferenceintheROSC,evenin the analysis including therobust and thenot-robust RCTs(SDE FI>0versuscontrol,RR:1.21,95%CI:0.46–3.20,p=0.69;SDE FI=0versuscontrol,RR:1.10,95%CI:0.93–1.29,p=0.23).

Survivaltohospitaladmissionwasnotreducedinpatientstreated withtheSDEwhencomparedwithallthepooledtreatments(SDE versusall;RR:1.02,95%CI:0.75_–1.38,p=0.88),eveniftheanalysis Fig.2–ForestplotforsurvivaltothehospitaldischargecomparingSDEversusallthepooledtreatment.

WeightsJacobs:2.02%,Olasveengen:13.30%,Perkins:31.27%,Brown:8.46%Callaham:1.08%,Gueugniaud:12.21%, Sherman:0.14%,Stiell:0.57%,Ducros:0.25%,Gueugniaud:8.10%,Lindner:1.59%,Ong:2.72%,Wenzel:18.22%. Valueswerepresentedasrelativeriskand95%CI.SDE:standarddoseofepinephrine.

Fig.3–ForestplotforROSCcomparingSDEversusallthepooledtreatment.

Weights: Jacobs: 6.33%, Olasveengen: 7.17%, Perkins: 7.36%, Brown: 7.26%, Callaham: 7.06%, Choux: 7.08%, Gueugniaud:7.37%,Sherman:4.71%,Stiell:5.06%,Callaway:6.81%,Ducros:5.72%,Gueugniaud:7.33%,Lindner: 6.34%,Ong: 7.13%, Wenzel: 7.20%.Values werepresented as relativeriskand 95% CI. SDE:standarddose of epinephrine.

ofrobustRCTsfoundanimprovementinthisresult(SDEFI_>0versus control,RR:2.04,95%CI:1.22_–3.43,p=0.06).TheSDEimprovedthe survivalto hospitaladmission whencompared with aplacebo/no drugs(SDEversusaplacebo/nodrugs,RR:2.04,95%CI:1.22–3.43, p=0.00).Patients treated with the HDE had a bettersurvival to hospitaladmissionthanthosetreatedwiththeSDE(SDEversusHDE, RR:0.86,95%CI:0.75_–0.99,p=0.04).Nodifferenceinsurvivalto hospital admissionwasfound whencomparingtheSDEwith the Epi+Vaso(SDEversusEpi+Vaso,RR:0.87,95%CI:0.74–1.01, p=0.07).Fig.4demonstratestheforestplotforthesurvivaltohospital admission.

TheSDEincreasedtherateofpatientsdischargedwithagood neurologicoutcomewhencomparedwithallthepooledtreatments (SDEversusall,RR:1.66,95%CI:1.00–1.35,p=0.04)andinthe not-robusttrials(SDEFI=0versuscontrol,RR:1.29,95%CI:1.02–1.64, p=0.02).There was no difference in patients discharged with a

favorable neurologic outcome when comparing the SDE with a placebo/nodrugs(SDEversusaplacebo/nodrugs,RR:1.22,95%CI: 0.99–1.52,p=0.06),whencomparingtheSDEwiththeHDE(SDE versusHDE,RR:1.20,95%CI:0.73–1.95,p=0.45),whencomparing theSDEwiththeEpi+Vaso(SDEversusEpi+Vaso,RR:1.35,95% CI:0.91–2.00,p=0.13),andinrobusttrials(SDEFI>0versuscontrol, RR:1.18,95%CI:0.91_–1.53,p=0.21).Fig.5illustratestheforestplot forapositiveneurologicoutcome.

Discussion

Inthissystematicreviewandmeta-analysisincluding20 716 pa-tients treated with epinephrine during OHCA, we found that epinephrine,whencomparedtoallthepooledtreatments,improved thesurvivaltohospitaldischargeandagoodneurologicoutcome, Fig.4–ForestplotforsurvivaltothehospitaladmissioncomparingSDEwithallthepooledtreatment.

Weights: Jacobs: 7.55%, Olasveengen: 8.26%, Perkins: 8.41%, Brown: 8.20%, Callaham: 7.25%, Choux: 7.80%, Gueugniaud:8.42%,Callaway:7.48%,Ducros:5.93%,Gueugniaud:8.37%,Lindner:6.07%,Ong:7.87%,Wenzel:8.33%. Valueswerepresentedasrelativeriskand95%CI.SDE:standarddoseofepinephrine.

Fig.5_–ForestplotforthegoodneurologicoutcomecomparingSDEversusallthepooledtreatment.

Weights:Jacobs:2.58%,Olasveengen:33.34%,Perkins:31.97%,Callaham:0.32%,Gueugniaud:12.32%,Callaway: 0.79%,Ducros:0.34%,Gueugniaud:6.24%,Ong:1.98%,Wenzel:10.07%.Valueswerepresentedasrelativeriskand 95%CI.SDE:standarddoseofepinephrine.

but did not increase the ROSC and the survival to hospital admission.Inthesubgroupanalyseswefoundthat(1)thesurvival tohospital discharge,ROSC, and survival tohospital admission increased when epinephrine was compared with a placebo/no drugs;(2)therewereno differencesin theconsideredoutcomes betweenepinephrineandepinephrineplusvasopressin;(3)ahigh dose of epinephrine ameliorated the ROSC and the survival to hospitaladmissionbuthadthesameeffectasastandarddoseof epinephrine on the survival to hospital discharge and a good neurologic outcome; and (4) considering the robust RCTs, epinephrine increased theshort-and long-termsurvivals butdid notimprovethesecondaryoutcomes.

Toourknowledge,thisisthefirstmeta-analysisthat(1)compares epinephrinewithaplacebo/nodrugs;(2)stratifiestheRCTsaccording totheFI;and(3)includesthestudybyPerkinsetal.9Recently,Zhang etal.performedasystematicreviewandmeta-analysiscomparing epinephrinewithepinephrineplusvasopressinandincludingAsian, American,andEuropeanstudies.22_{Interestingly,Zhangetal.found}

thatepinephrineimprovedtheROSConlyinAsianpatients,andnotin EuropeanandAmericanpatients.22However,AsianRCTshadvery poorquality,andthemeta-analysispossessedahugeheterogeneity.22 Linetal.performedasubgroup analysisoftheRRcomparing epinephrinewithahighdoseofepinephrine,andepinephrinewith epinephrineplusvasopressin.23_{Thismeta-analysisdidnotincludethe}

studybyPerkinsetal.9_{;actually,theydidnotperformanyforestplot}

betweenepinephrineandaplacebo/nodrugs.23Twoprevious meta-analyses, including RCTs and observational studies, found that epinephrinewas noteffectiveatincreasingthesurvivalathospital discharge.24,25_{Ourpooledresultswereverydifferentfromtheprevious}

reviews,23–25_{becausewefoundabettersurvivaltohospitaldischarge}

andameaningfulneurologicoutcomebyusingepinephrineinOHCA. Previousmedicalliteraturereportedthatepinephrineversusa placebo was associated with a significant improvement in the ROSCand thesurvivalto hospitaladmission,butnodifference wasfoundinthesurvivalto hospitaldischargeandafavorable neurologicoutcome.26_{EventherecentstudybyPerkinsetal.,the}

PARAMEDICtrial,9showedthatepinephrine,whencomparedwith aplacebo,anddespitehavingapowerfuleffectontheROSCafter OHCA,producedonlyasmallabsoluteincreaseinsurvivalandno improvementinafavorablefunctionalrecovery.Inthisview,the results of the present meta-analysis may be groundbreaking, becauseepinephrine,when comparedwith aplacebo/nodrugs, improved short-term and long-term outcomes. These results should encourage further studies evaluating the realbenefit of epinephrine on short-term and long-term outcomes, even if successful long-term outcomesare also due to the in-hospital managementand therapiesthatareincreasedfor severaldays afterOHCA.27 Optimizingthe respiratoryand cardiac functions aftercardiacarrestmayimprovelong-termoutcomesbeyondthe useofepinephrineduringOHCA.28

Currentguidelinesoncardiacarreststatethatitisreasonableto consideradministering1mgofepinephrineevery3–5minduringadult cardiac arrest.2 In the present meta-analysis, a high dose of epinephrine,whencomparedwithastandarddoseofepinephrine, hadabetterrateofROSCandsurvivaltohospitaladmission,buta similareffectonsurvivaltohospitaldischargeandagoodneurologic outcome.Ahighdoseofepinephrinemayincreasecoronaryperfusion pressureandperipheralvasoconstriction.2However,ahighdoseof epinephrinemayalsohavedetrimentaleffects,suchasanincreasein myocardial oxygen consumption, ectopic ventricular arrhythmias,

transient hypoxemia from pulmonary arteriovenous shunting, im-paired microcirculation, and worsepost-cardiac arrestmyocardial dysfunction.26

Asystematicreview,ameta-analysis,andtheRCTstopthelist of the level of evidence.29 The RCTs included in the meta-analysesarecurrentlyscreenedformethodologicalqualityinorder tosearchfortheriskofbias.29_{Recently,theFIwasintroducedin}

critical care literature, with the purpose of measuring the robustnessofRCTsfromastatisticalpointofview.30According tothecurrentdefinition,RCTswithalargerFIhavemorerobust findingswhencomparedwiththestudieswithapoorFI.30AnFI was neverappliedto themeta-analysis.31 _{Becausealow FIin}

critical care trials reinforced the finding that the robustness of evidence available toclinical decisionmakers in this settingis limited,wemeasuredtheFIforalltheincludedRCTs.Althoughall theincludedRCTshadalowriskofbias,onlythreeRCTsshowed anFIofmorethan0.Intermsofhospitaldischargeandsurvivalto hospitaladmission,wefoundsimilarresultsbetweenthesubgroup analysis of the RCTs with an FI_>0 and the pooled group. Probably, anFImay addmorerobustnessto theresultsof the present meta-analysis. Nevertheless,the useofanFI in meta-analysisshould beimplementedbyfutureliterature. Inlinewith previoussystematicreviews,23 _{wealsoemployedthe Cochrane}

CollaborationtoolandtheGRADEcriteriatoassesstheriskofbias forourincludedstudies.

Limitations

Thissystematicreviewandmeta-analysishaslimitationsthatneedto be addressed. First,we evaluated treatments with guidelines for cardiacarrestthatarereportedtobestillindebate.2Second,wefound heterogeneity>25%in13outof24comparisonsfortheconsidered outcomes. Third,the results ofthe FIshould be interpreted with caution.Fourth,weincludedRCTsandexcludedprospectiveand retrospectivestudies.

Conclusions

In OHCA, standard or high doses of epinephrine should be used,becausetheyimprovedsurvivaltohospitaldischargeand meaningfulclinicaloutcomes.Therewasalsoaclearadvantage ofusingepinephrineoveraplaceboornodrugsintheconsidered outcomes.Furthertrialsareneededtoassessthebestdoseof epinephrineforOHCA,becausetheoptimaldoseofepinephrine is still unknown. According to our data reporting that a high dose of epinephrine was associated with better ROSC and survivaltohospitaladmission,futureresearchshouldinvestigate thispoint.

Conflict

of

interest

None.

Funding

Appendix

A.

Supplementary

data

Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j. resuscitation.2019.01.016.

REFERENCES

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

2.ArrichJ,SterzF,HerknerH,Testori,BehringerW.Totalepinephrine doseduringasystoleandpulselesselectricalactivitycardiacarrestsis associatedwithunfavorablefunctionaloutcomeandincreased in-hospitalmortality.Resuscitation2012;83:333–7.

3.Perkins,DeakinTQ,NolanJO,etal.Arandomizedcontrolledtrialof epinephrineinoutofhospitalcardiacarrest.NEnglJMed 2018;379:711–21.

4.MoherD,LiberatiA,AltmanDG,etal.Prefereedreportingitemsfor systematicreviewandmeta-analyses:thePRISMAstatement.PLoS Med2009;6:e1000097.

5.WalshM,SrinathanSK,McAuleyDF,etal.Thestatisticalsignificance ofrandomizedcontrolledtrialresultsisfrequentlyfragile:acasefora FragilityIndex.JClinEpidemiol2014;67:622–8.

6.PetersJL,SuttonAJ,JonesDR,AbramsKR,RushtonL.Comparison oftwomethodstodetectpublicationbiasinmeta-analysis.JAMA 2006;295:676–80.

7.JacobsIG,FinnJC,JelinekGA,etal.Effectofadrenalineonsurvivalin out-of-hospitalcardiacarrest:arandomiseddouble-blind placebo-controlledtrial.Resuscitation2011;82:1138–43.

8.OlasveengenTM,SundeK,BrunborgC,etal.Intravenousdrug administrationduringout-of-hospitalcardiacarrest:arandomizedtrial. JAMA2009;302:2222–9.

9.PerkinsGD,JiC,DeakinCD,etal.Arandomizedtrialofepinephrinein out-of-hospitalcardiacarrest.NEnglJMed2018;23:711–21.

10.BrownCG,MartinDR,PepePE,etal.Acomparisonofstandard-dose andhigh-doseepinephrineincardiacarrestoutsidethehospital,The MulticenterHigh-DoseEpinephrineStudyGroup.NEnglJMed 1992;327:1051–5.

11.CallahamM,MadsenCD,BartonCW,etal.Randomizedclinicaltrialof high-doseepinephrineandnorepinephrinevsstandard-dose epinephrineinprehospitalcardiacarrest.JAMA1992;268:2667–72.

12.StiellIG,HebertPC,WeitzmanBN,etal.High-doseepinephrinein adultcardiacarrest.NEnglJMed1992;327:1045–50.

13.ChouxC,GueugniaudPY,BarbieuxA,etal.Standarddosesversus repeatedhighdosesofepinephrineincardiacarrestoutsidethe hospital.Resuscitation1995;29:3–9.

14.ShermanBW,MungerMA,FoulkeGE,etal.High-doseversus standard-doseepinephrinetreatmentofcardiacarrestafterfailureof standardtherapy.Pharmacotherapy1997;17:242–7.

15.GueugniaudPY,MolsP,GoldsteinP,etal.Acomparisonofrepeated highdosesandrepeatedstandarddosesofepinephrineforcardiac arrestoutsidethehospital.NEnglJMed1998;339:1595–601.

16.LindnerKH,DirksB,StrohmengerHU,etal.Randomisedcomparison ofepinephrineandvasopressininpatientswithout-of-hospital ventricularfibrillation.Lancet1997;349:535–7.

17.WenzelV,KrismerAC,ArntzHR,etal.Acomparisonofvasopressin andepinephrineforout-of-hospitalcardiopulmonaryresuscitation.N EnglJMed2004;350:105–13.

18.CallawayCW,HostlerD,DoshiAA,etal.Usefulnessofvasopressin administeredwithepinephrineduringout-of-hospitalcardiacarrest. AmJCardiol2006;98:1316–21.

19.GueugniaudPY,DavidJS,ChanzyE,etal.Vasopressinand epinephrinevs.epinephrinealoneincardiopulmonaryresuscitation.N EnglJMed2008;359:21–30.

20.DucrosL,VicautE,SoleilC,etal.Effectoftheadditionofvasopressin orvasopressinplusnitroglycerintoepinephrineonarterialblood pressureduringcardiopulmonaryresuscitationinhumans.JEmerg Med2011;41:453–9.

21.OngME,TiahL,LeongBS,etal.Arandomised,double-blind, multicentretrialcomparingvasopressinandadrenalineinpatientswith cardiacarrestpresentingtoorintheEmergencyDepartment. Resuscitation2012;83:953–60.

22.ZhangQ,LiuB,ZhaoL,etal.Efficacyofvasopressin-epinephrine comparedtoepinephrinealoneforoutofhospitalcardiacarrest patients:asystematicreviewandmeta-analysis.AmJEmergMed 2017;35:1555–60.

23.LinS,CallawayCW,ShahPS,etal.Adrenalineforout-of-hospital cardiacarrestresuscitation:asystematicreviewandmeta-analysisof randomizedcontrolledtrials.Resuscitation2014;85:732–40.

24.PatanwalaAE,SlackMK,MartinJR,etal.Effectofepinephrineon survivalaftercardiacarrest:asystematicreviewandmeta-analysis. MinervaAnestesiol2014;80:831–43.

25.AtiksawedparitP,RattanasiriS,McEvoyM,etal.Effectsofprehospital adrenalineadministrationonout-of-hospitalcardiacarrestoutcomes: asystematicreviewandmeta-analysis.CritCare2014;18:463.

26.SoarJ,NolanJP,BottingerBW,etal.Europeanresuscitation guidelinesforresuscitation2015section3.Adultadvancedlife support.Resuscitation2015;95:100–47.

27.SutherasanY,PenuelasO,MurielA,etal.Managementandoutcome ofmechanicallyventilatedpatientsaftercardiacarrest.CritCare 2015;19:215–22.

28.SutherasanY,VargasM,BrunettiI,etal.Ventilatorytargetsafter cardiacarrest.MinervaAnestesiol2015;81:39–51.

29.MuradMH,AsiN,AlsawasM,etal.Newevidencepyramid.BMJEvid BasedMed2016;21:125–7.

30.RidgeonEE,YoungPJ,BellomoR,etal.TheFragilityIndexin multicenterrandomizedcontrolledcriticalcaretrials.CritCareMed 2016;44:1278–84.

31.VargasM,ServilloG.Theendofcorticosteroidinsepsis:fragileresults fromfragiletrials.CritCareMed2018;12:e1228.