Validation of the GRACE Risk Score for Predicting Death ...

The GRACE score allows us to estimate both mortality and the combined events of death or reinfarction during the hospital stay and at 6 months after discharge. RevistaEspañoladeCardiología(EnglishEdition) ISSN:1885-5857 RevistaEspañoladeCardiologíaisaninternationalscientificjournaldevotedtothepublicationofresearcharticlesoncardiovascularmedicine.Thejournal,publishedsince1947,istheofficialpublicationoftheSpanishSocietyofCardiologyandfounderoftheRECPublicationsjournalfamily.ArticlesarepublishedinbothEnglishandSpanishinitselectronicedition. Seemore Indexedin: JournalCitationReportsandScienceCitationIndexExpanded/CurrentContents/MEDLINE/IndexMedicus/Embase/ExcerptaMedica/ScienceDirect/Scopus Seemore Followus: Subscribe: Impactfactor TheImpactFactormeasurestheaveragenumberofcitationsreceivedinaparticularyearbypaperspublishedinthejournalduringthetwoprecedingyears. ©ClarivateAnalytics,JournalCitationReports2021 Seemore Impactfactor2020 4.753 Citescore CiteScoremeasuresaveragecitationsreceivedperdocumentpublished. Seemore Citescore2020 6.5 SJR SRJisaprestigemetricbasedontheideathatnotallcitationsarethesame.SJRusesasimilaralgorithmastheGooglepagerank;itprovidesaquantitativeandqualitativemeasureofthejournal'simpact. Seemore SJR2020 0.955 SNIP SNIPmeasurescontextualcitationimpactbywightingcitationsbasedonthetotalnumberofcitationsinasubjectfield. Seemore SNIP2020 0.703 Viewmoremetrics Hide JournalInformation Previousarticle | Nextarticle Vol.63.Issue6.Pages640-648(June2010) Léaloenespañol Share Share Print DownloadPDF Morearticleoptions ePub Statistics Vol.63.Issue6.Pages640-648(June2010) DOI:10.1016/S1885-5857(10)70138-9 Fulltextaccess ValidationoftheGRACERiskScoreforPredictingDeathWithin6MonthsofFollow-UpinaContemporaryCohortofPatientsWithAcuteCoronarySyndrome ValidaciónenunacohortecontemporáneadepacientesconsíndromecoronarioagudodelscoreGRACEpredictordemortalidadalos6mesesdeseguimiento Visits ... DownloadPDF EmadAbu-Assia,JoséM.García-Acuñaa,CarlosPeña-Gila,JoséR.González-JuanateyaaServiciodeCardiologíayUnidadCoronaria,HospitalClínicoUniversitariodeSantiagodeCompostela,InstitutodeInvestigacionesSanitarias(IDIS),SantiagodeCompostela,ACoruña,Spain Thisitemhasreceived ... Visits (Dailydataupdate) Articleinformation Abstract FullText Bibliography DownloadPDF Statistics Figures(4)ShowmoreShowlessTables(4)TABLE1.RiskCategoriesBasedontheGRACEScorefortheTotalPopulation,theSTEMIGroupandtheNSTACSGroupTABLE2.BaselineCharacteristicsofourPopulationComparedWiththeDerivationCohortfortheGRACEScoreTABLE3.Evolution,HospitalStayManagementandTreatmentUponDischargeinourPopulationComparedWiththeGRACEScoreDerivationCohortTABLE5.MortalityObservedandPredictedbytheGRACEModelforMortalityat6Months,BrokenDownbyRiskCategoriesintheTotalSampleandbyACSTypeShowmoreShowless Introductionandobjectives.TheGlobalRegistryofAcuteCoronaryEvents(GRACE)riskscoreprovidesanestimateoftheprobabilityofdeathwithin6monthsofhospitaldischargeinpatientswithacutecoronarysyndrome(ACS).OuraimwastoassessthevalidityofthisriskscoreinacontemporarycohortofpatientsadmittedtoaSpanishhospital.Methods.Thestudyinvolved1,183consecutivepatientswithACSevaluatedbetweenFebruary2004andFebruary2009.Theirvitalstatuswasdetermined6monthsafterhospitaldischargeandthevalidityoftheGRACEriskscorewasevaluatedbyassessingitscalibration(Hosmer-Lemeshowtest)anditsdiscriminatorycapacity(areaunderthereceiveroperatingcharacteristic[ROC]curve).Results.Intotal,459(38.8%)patientswereadmittedforST-elevationmyocardialinfarction(STEMI)and724(61.2%)fornon-ST-elevationmyocardialinfarction(NSTEMI).Percutaneousrevascularizationwasperformedin846(71.5%).ThemedianGRACEriskscorewas121[interquartilerange,96-144].Mortality6monthsafterdischargewas4.4%.ThecalibrationoftheGRACEriskscorewasacceptable(Hosmer-Lemeshow,P>.2)anditsdiscriminatorycapacitywasexcellent:theareaundertheROCcurvewas0.86(95%confidenceinterval[CI],0.807-0.916)forallpatients,0.9(95%CI,0.829-0.975)forthosewithSTEMIand0.86(95%CI,0.783-0.927)forthosewithNSTEMI.Conclusions.TheGRACEriskscoreforpredictingdeathwithin6monthsofhospitaldischargewasvalidatedandcanbeusedinpatientswithACS.ItwouldbewisetoincludetheGRACEriskscoreinthemedicalrecordsofthesepatients.Keywords:GRACEriskscoreAcutecoronarysyndromePrognosisIntroducciónyobjetivos.ElscoreGRACEpermiteestimarlaprobabilidaddemuertealos6mesesdelaltahospitalariaenlospacientesconsíndromecoronarioagudo(SCA).NuestroobjetivofuevalidarestescoreenunacohortecontemporáneadepacientesingresadosenunhospitaldeEspaña.Métodos.Desdefebrerode2004afebrerode2009,seevaluóa1.183pacientesconsecutivosconSCA.Sedeterminóelestadovitalalos6mesesdelaltayserealizóunestudiodevalidacióndelmencionadoscoreapartirdesucapacidaddecalibración(pruebadeHosmer-Lemeshow)ydediscriminación(áreabajolacurvareceiveroperatingcharacteristic[ROC]).Resultados.Ingresaron459(38,8%)pacientesporinfartoagudodemiocardioconelevacióndelST(IAMCEST)y724(61,2%)porSCAsinelevacióndelST(SCASEST);846(71,5%)fueronrevascularizadosporvíapercutánea.LamedianadelscoreGRACEfuede121[96-144]puntos.Lamortalidada6mesesfuedel4,4%.ElscoreGRACEmostróadecuadacalibraciónyexcelentediscriminaciónenelconjuntodelapoblación,enelIAMCESTyenelSCASEST(Hosmer-Lemeshow,p>0,2;áreabajolacurvaROC,0,86[0,807-0,916],0,9[0,829-0,975]y0,86[0,783-0,927],respectivamente).Conclusiones.ElscoreGRACEdeprediccióndemortalidadalos6mesesdelaltatrassufrirunSCAhasidovalidadoadecuadamenteypuedeutilizarseparaestimarelriesgodemuertea6mesesenestospacientes.EnnuestrapoblacióntienesentidoincluirenlosinformesdeatenciónporSCAelscoreGRACE.Palabrasclave:ScoreGRACESíndromecoronarioagudoPronóstico FullText INTRODUCTIONApredictedprognosisforpatientswithacutecoronarysyndrome(ACS)isconsideredtobeanecessaryinstrumentforstratifyingriskforthesepatientsandevaluatingthepotentialimpactofdifferenttherapeuticinterventions.Recently,researcherswiththeGRACEstudy(GlobalRegistryofAcuteCoronaryEvents)developedascorewhichallowsustopredictmortality6monthsafterbeingdischargedfollowinganACSepisode.1Thisscoreisapredictivelogisticalmodelwhichuses9prognosticvariables(Figure1)todeterminethepatient'sprobabilityofdeathduetoanycauseduringthefirst6monthsfollowingdischarge.Thismodelwascreatedandvalidatedformorethan20000patientsrecruitedbetween1999and2003.1UseoftheGRACEscoreforstratifyingriskinACSwasrecommendedsoonafterbyclinicalpracticeguides.2,3 Figure1.RiskscoreandnomogramfromtheGRACEstudy,forpredictingmortalityduetoanycauseinthe6monthsafterdischarge.AdaptedfromEagleetal.1AMIindicatesacutemyocardialinfarction;HR,heartrate;SAP,systolicarterialpressure.However,mindfulofthegeographicvariabilitythatgivesrisetodifferencesinpatientcharacteristicsandmedicalattentionmodelsfordifferenthealthsystems,4,5validationsinourareawillbenecessarybeforewecanuseandapplytheprobabilisticmodeloutsideoftheenvironmentinwhichitwascreated.Thiswillensurethatwedonotreceiveerroneousprobabilities.6,7ThepurposeofourstudywastovalidatetheGRACEmortalitypredictionscoreat6monthsfollowinganACSepisode,basedonitscapacitiesforcalibrationanddiscrimination.Inthisway,wewillbeabletoverifythatthemodelisagoodpredictorofmortalityinacontextsuchasacontemporarycohortofACSpatientsadmittedtoaSpanishhospital,inwhichmostpatientsweretreatedwithapercutaneousrevascularisationstrategy.METHODPatientsInthisvalidationstudy,alldemographic,clinicalandangiographicvariables,aswellasthoserelatingtomanagementandcomplicationsduringthehospitalstay,weregatheredprospectivelyinourdepartment'shospitalisationunitdatabasebetweenFebruary2004andFebruary2009.Theinformationwasgatheredbythedepartment'scardiologists.Patientsweretreatedaccordingtotheirmaindoctor'scriteriaandthestudywasapprovedbyourcentre'sclinicalresearchethicscommittee.ThestudypopulationconsistsofallconsecutivepatientsadmittedduetoACSbetweenFebruary2004andFebruary2009.Theinitialcohortcontained1361patients.Eighty-sixpatientsdiedwhileadmitted.DataisavailableforallprognosticvariablesincludedintheGRACEscorefor1262(99%)ofthepatientswhosurvivedtheinitialepisode.Wewereabletodeterminevitalstageinallpatientsexceptfor79(6.3%).Forthevalidationstudy,weexcludedpatientswhodidnothavevaliddatafortheGRACEscoreandforvitalstate6monthsafterdischarge.Therefore,thefinalcohortforthestudyconsistedof1183patients.VariablesandDefinitionsTheGRACEscoreallowsustoestimatebothmortalityandthecombinedeventsofdeathorreinfarctionduringthehospitalstayandat6monthsafterdischarge.TheGRACEscoremodalitywhichwearevalidatinginthisstudyistheonethatenablesustoestimateriskofdeath6monthsafterdischargefollowinganACSepisode.ACSepisodesareclassifiedasST-elevationmyocardialinfarction(STEMI)andACSwithoutSTelevation(NSTACS)(unstableanginaandnon-STelevatedMI[NSTEMI]).ST-elevationmyocardialinfarctionisdefinedwheretherearesuggestivesymptoms,persistent(>20minutes)STsegmentelevation≥1mmin≥2contiguousderivationorinthepresenceofleftbundle-branchblockpresumedtobeanewoccurrence,andelevatedcardiactroponinI(cTnI≥0.2ng/dL(cut-offpointfordiagnosingAMI≥0.2ng/dL,Flex®cartridges,Dimension®system,SiemensInc.,UnitedStates).WedefineNSTEMIasthepresenceofsuggestivesymptoms,cTnI≥0.2ng/dLand/ordynamicST-segmentchanges(STdecrease≥1mmornon-persistentelevationin≥2contiguousderivations).Unstableanginaisdefinedasthepresenceofsuggestivethoracicpainwithorwithoutrepolarisationabnormalitiesinthebaselineelectrocardiogram.TheserumlevelsofcTnImustbe<0.2ng/dL24hoursafterthefirstsymptomsappear.StatisticalAnalysisQuantitativevariablesarepresentedasthemean(standarddeviation),orasthemedianandinterquartilerange.Qualitativevariablesareexpressedasfrequenciesandpercentages.ThetestorFisher'sexacttestwereused,asapplicable,tocomparequalitativevariables,andStudentttesttocomparequantitativevariables.CalculatingtheGRACEScoreTheGRACEscorewascalculatedforeachpatientbyassigningtheappropriatenumberofpointsforeachofthe9prognosticvariablesthatentersintothecalculation:age,historyofheartfailure,historyofAMI,heartrateandsystolicbloodpressureatadmission,STsegmentdepression,serumcreatinineatadmission,elevatedmyocardialnecrosismarkersorenzymesandlackofpercutaneouscoronaryrevascularisationduringadmission(Figure1).ThreeriskcategorieswereestablishedusingthecutoffpointssetoutintheGRACEstudy.8Therefore,inthelow-riskcategory,theGRACEscorewas27-99pointsforSTEMIand1-88forNSTACS;intheintermediateriskcategory,thescoreforSTEMIwas100-127,and89-118forNSTACS;andinthehigh-riskcategory,thescoreforSTEMIwas128-263and119-263forNSTACS.CalibrationThemodel'scalibrationwasevaluatedusingtheHosmer-Lemeshowgoodness-of-fittest.9,10Thistestismainlyusedforvalidatingnewlycreatedmodels,butitisequallyusefulforvalidatinganexistinglogisticalmodelwithanexternaldatabase,asisthecasewiththisstudy.TheHosmer-LemeshowtestestimatesaCstatisticbasedonthedifferencebetweenthemortalityvaluesobservedinrealityandthosepredictedbythemodelfordifferentriskgroups.Thesmallerthestatisticalvalue,thebettercalibratedthemodel.APvalue>.05indicatesthatthemodeliswelladjustedtothedataandthereforeisagoodpredictorofpatients'probabilityofdeath.TheGRACEscore'scalibrationwasevaluatedfortheentirepopulationandpersubgroupsrepresentinganACStype(STEMIandNSTACS).Next,thescorecalibrationwastestedforlow,intermediateandhigh-riskcategoriesinthethreepreviouslylistedgroups(totalsample,STEMIandNSTACS).DiscriminationGRACEscorediscriminationwasbasedonitsabilitytodistinguishbetweenpatientswhowilldieinthe6monthsafterdischargeandthosewhowillsurvive.ThecapacityfordiscriminationwasanalysedbycalculatingtheareaundertheROC(receiveroperatingcharacteristic)curve(AUCROC).AmodelwithanAUC-ROCbetween0.8and0.9isconsideredtobeamodelwithagoodcapacityfordiscrimination.11TheAUC-ROCvaluewascalculatedforthetotalpopulation,forpatientswithSTEMIandforthosewithNSTACS.SincetheproportionofpercutaneousrevascularisationprocedurescarriedoutduringthehospitalstayinourstudywasmuchhigherthanthatintheoriginalGRACEcohort,wealsotestedthevalidityofthemodelingroupswithandwithoutpercutaneouscoronaryrevascularisationduringthehospitalstay.Follow-upTheclinicalfollow-uptodeterminevitalstateat6monthsafterdischargewascarriedoutretrospectivelybetweenJuneandAugust2009byreviewingourcentre'sclinicalhistoriesanddatabases.Wehadaccesstoinformationaboutthevitalstageinallpatientsexcept79(6.3%).StatisticalanalysiswasperformedusingSPSS®softwarev.15.0andMedCalc®v.9.2.0.RESULTSBaselineCharacteristicsAtotalof1183patientswasevaluated.STEMIwaspresentin459(38.8%)ofpatients,and724(61.2%)hadNSTACS(166[22.9%]unstableangina).Theproportionofpatientsineachofthe3riskcategoriesisshowninTable1.Thehigh-riskcategorycontained580(49%)ofthepatients(Table1).ThefrequencycountsfortheGRACEscorevariablesinourcohortandinthederivationcohortissummarisedinTables2and3.Thepercentageofmales,diabetesmellitusandelevatedmyocardialnecrosismarkerswassignificantlyhigherinourseries;thepercentageofsmokers/formersmokers,previoushistoryofAMIandheartfailurewassignificantlyhigherintheGRACEscorederivationcohort.Themeansystolicbloodpressureandserumcreatininelevelatadmissionwasslightlybutsignificantlylowerinourcohort. Management,HospitalStayComplications,andTreatmentUponDischargeTherewasahigherpercentageofpercutaneouscoronaryrevascularisationproceduresduringhospitalisationinourseries(71.5%versus26.6%,PPPP=.05). Upondischarge,patientsinourseriesweremorelikelytoreceivetreatmentwiththienopyridines,ACEinhibitors,orAIIRAandstatins(PMortalityBy6monthsaftertheirdischarge,52(4.4%)ofpatientshaddied(17[3.7%]intheSTEMIgroupand35[4.8%]intheNSTACSgroup[P=.4]).Figure2showsthedistributionofmortalitybyanycause6monthsafterdischarge,brokendownintolow,intermediateandhigh-riskcategories.Weobservethatthemortalityrateincreasedsignificantlyastheriskcategoryincreased.Thedifferencesweobservedprovedtobesignificantuponcomparingthe3categoriesineachgroup(totalpopulation,STEMI,andNSTACS)andcomparingtheintermediaterisktothelowriskgroupforthetotalsampleandforSTEMI.However,therewerenosignificantdifferencesinthemortalityratebetweentheintermediateandlow-riskcategoriesforNSTACS(Figure2). Figure2.Mortalityat6monthsforthe3riskcategoriesinthetotalpopulation,forSTEMIandNSTACS.*ShowsP-valuesfromthecomparisonbetweenthe3categoriesineachgroup.Thecomparisonoftheintermediateandlowriskcategorieswas.009inthesampletotal,.02intheSTEMIgroup,and.6intheNSTACSgroup.CalibrationandDiscriminationModelcalibrationwasexcellentforthepopulationasawhole,forSTEMIandforNSTACS(Table4).Thespecificanalysisforgroupsthatdidanddidnotundergopercutaneouscoronaryrevascularisationduringthehospitalstaydidnotaffectthemodel'spredictiveabilityorcalibration(Table4).Themortalitypredictedbythemodelapproachesrealvaluesforthe3riskcategories;theapproximationismuchcloserforthehigh-riskcategories(Table5).Thevalidatedmodelshowsanadequatecapacityfordiscrimination,withanAUC-ROCof0.86(95%CI,0.897-0.916;P Figure3.ROCcurveforthemodelfortheentirestudypopulation. DISCUSSIONClinicalpracticeguidesemphasisetheimportanceofstratifyingriskforACSpatients.2,3,12Thisfactisparticularlyimportantforhigh-riskpatientswhobenefitmorefromaggressivetreatment,andwho,paradoxicallyenough,seldomreceiveit.13-16ApplicationoftheGRACE1scorewasrecentlyrecommendedbyclinicalpracticeguides.2,3However,properuseofacertainscorerequireswell-contextualisedvalidationsinorderforustobesurethatitisnotprovidingfaultyprobabilities.ThemainfindingofthepresentstudyisthefirstvalidationoftheGRACEscorebyaSpanishhospital.ThevalueobtainedfromtheCstatisticintheHosmer-Lemeshowtestwas>0.1inalloftheanalysesweperformed.Thispermitsustostatethatinourcohort,themodeloffersagoodcalibrationoftheprobabilityofdeathwithin6monthsofdischargefollowinganACSepisode.ThediscriminatorycapacityofthemodelwastestedbyusingtheAUCROC,whichwas>.80inallofthetestsweran.Withthesedata,itispossibletoconcludethattheGRACEscorehasbeenvalidated,andthatitisauseful,reliableclinicaltool.Theresultingpredictionsoftheprobabilityofdeathwithin6monthsofdischargearevalidandconformtotherealriskpresentedbypatientsinourseries,forthetotalpopulation,theSTEMIgroup,andtheNSTACSgroup.Althoughweobservedthatthemodelisbetteradjustedtotheriskofdeathofpatientsinhighriskcategorieswhenwegroupedthepopulationintoriskgroups,thetendencytowardover-orunderestimatingriskinothercategorieswasveryslight(<0.7%)(Table5). Inourstudy,totalmortalityby6months(4.4%)isconsistentwiththatobservedintheGRACEregister(4.8%).1ThemortalityobservedforpatientswhopresentedSTEMI(3.7%)andNSTACS(4.8%;P=.04)wasalsosimilartothatfoundintheGRACEregister(4.8%inbothgroups).17ObservingincreasedmortalitywheretheGRACEscoreincreasedwasconsistentforthetotalpopulation,andtheSTEMIandNSTACSgroups.Thatis,whenthecalculatedscoresweregroupedintolow,intermediateandhighriskcategories,the3categorieswereassociatedwithsignificantdifferencesinmortalityforeachofthe3studygroups.ThesefindingsareconsistentwiththosefromotherstudiesinwhichmortalitywasseentoincreaseastheGRACEscoreincreased.18-21Althoughcomparingmortalityamongthe3riskcategoriesturnedupsignificantdifferencesinthetotalsample,theSTEMIgroupandtheNSTACSgroup(Figure2),whenwecomparemortalitybetweentheintermediateandlow-riskcategoriesforNSTACS,wefindnosignificantdifferences.InananalysisofresultsfromasubstudyintheMASCARAregister,22nodifferenceswerefoundbetweentheintermediateandlowriskcategoriesforNSTACS.Therefore,bothweregroupedinasinglecategoryandsubsequentlycomparedwiththehigh-riskcategory.Thisalsosupportsourstudy,inwhichtherewerenosignificantdifferencesinmortalityat6monthsbetweenthelowandintermediate-riskcategoriesintheNSTACSgroup(1.08%and1.55%,respectively;P=.6).ThemedianGRACEscoreforthetotalsampleandfortheNSTACSgroupwassimilartothatcalculatedbyGonzález-Ferreiraetal22andElbarounietal.23Inourstudy,however,themedianscorefortheSTEMIgroupwaslowerthanthatfoundbyElbarounietal.23Thismaybepartiallyduetoourstudy'sdifferentdistributionamongsomeoftheprognosticvariablesthatconstitutethemodel.Themodel'scapacityfordiscriminationandcalibrationwerebothexcellentinourseries,andtheyagreewithanotherstudycarriedoutinaneighbouringcountry.21Furthermore,ourseriesoffersamore"contemporary"evaluationoftheGRACEmodelforpredictingmortalitywithina6-monthfollow-upperiod.AnotherpeculiarityofourworkisthatthevalidationstudywascarriedoutinacohortindependentfromtheGRACEregister.Thismeansthatitisanexternalvalidationstudyinthetruesenseoftheterm,unliketheonecarriedoutbyElbarounietal23inwhichpartofthecohortusedinthevalidationstudyhadalsocontributedtodevelopingtheGRACEscore.Thehighproportionofpercutaneouscoronaryrevascularisationfoundinourcohort(71.5%)mostlikelyreflectshowaggressivetreatmentofACShasbecomewidespreadinrecentyears.Inaddition,ourstudywascarriedoutinacentrewithaninterventionalcardiologistinresidence,unlikethemulti-centreGRACEstudy,inwhichtherewasnouniformavailabilityofsucharesource.4GiventhenotabledifferenceinthepercentageofpercutaneousrevascularisationsintheGRACEstudy,weperformedaspecificanalysisofthegroupswhodidanddidnotundergopercutaneouscoronaryrevascularisationwhilehospitalised.Theresultsofthepreliminaryanalysisdidnotchangethemodel'spredictivecapacityorcalibration(Table4).OtherGRACEscorevalidationshavebeencarriedoutinPortugal,21Canada,18-20,23,24andmostrecentlyintheUK.25InPortugal,AraújoGonçalvesetal21didasingle-centrestudyof460AMIpatientsrecruitedbetween1999and2001tovalidatetheGRACEscoreforpredictingriskofdeathorAMIwithinfollow-upperiodsof30daysand1year.Validationindicesfoundinthisstudywereadequate(Hosmer-LemeshowPvalue>.1andAUC-ROCat0.67and0.72forestimatingriskofdeathorAMIwithinfollow-upperiodsof30daysand1yearrespectively).IntheUK,Galeetal25observedintheirmulti-centrestudythattheGRACEscoreoffersanexcellentcapacityfordiscrimination(AUC-ROC,0.80)forpredictingmortality6monthsafterdischarge.CalibrationforthisGRACEscoremodalitywasnottestedinthisstudy.Lastly,thevalidityoftheGRACEscorewasevaluatedin5multi-centrestudiesinCanada.18-20,23,24TwoofthesestudiesassessedthevalidityoftheGRACEscoreforpredictingmortalityat6months.Themodelshowedexcellentdiscriminationinbothtests(AUC-ROC,0.80).However,whilecalibrationwasgoodinoneofthesestudies(Hosmer-Lemeshow,P>.3),19itwasnotoptimalintheotherstudy18whichobservedthemodel'stendencytowardslightlyoverestimatingtheriskofdeathofNSTEMIpatients(Hosmer-Lemeshow,P=.06).LimitationsTheGRACEscoreallowsustoestimatebothmortalityandthecombinedeventsofdeathorreinfarctionduringthehospitalstayandat6monthsafterdischarge.26Oneofourstudy'slimitationsisthatitdoesnotvalidatetheGRACEprognosticscoreacrossallofitsmodalities.Anotherlimitationofourstudyisthatitwascarriedoutinasinglecentre.Forthisreason,ourresultsshouldonlybeappliedwithcautiontoothercentreswithdifferentpopulationandmedicalserviceprofiles.Onelimitationofanyriskscoreisthatwhileitmaydiscriminatewellbetweendifferentriskgroups,itdoesnotnecessarilypredictindividualriskcorrectly.CONCLUSIONSInourpopulation,theGRACEscoreforpredictingmortalityat6monthsfromdischargewasproperlyvalidated,andmaybeusedforestimatingriskofdeathat6monthsforACSpatients.Inourpopulation,itmakessensetoincludetheGRACEriskscoreonACScarereports.Morestudies,preferablymulti-centreones,willbenecessarytoconfirmourfindingsandvalidatealloftheGRACEscoremodalities.ABBREVIATIONSACS:acutecoronarysyndromeAUC:areaunderthecurveGRACE:GlobalRegistryofAcuteCoronaryEventsNSTEMI:non-ST-elevationmyocardialinfarctionROC:receiveroperatingcharacteristicSTEMI:ST-elevationmyocardialinfarctionSEEARTICLEONPAGES629-32Correspondence:Dr.E.Abu-Assi.HospitalClínicoUniversitario.ServiciodeCardiología.Avda.Choupana,s/n.15706SantiagodeCompostela.ACoruña.España.E-mail:[email protected],2009.AcceptedforpublicationDecember23,2009. Bibliography[1]EagleKA,LimMJ,DabbousOH,PieperKS,GoldbergRJ,VandeWerfF,etal..Avalidatedpredictionmodelforallformsofacutecoronarysyndrome:estimatingtheriskof6-monthpostdischargedeathinaninternationalregistry.GRACEInvestigators..JAMA,291(2004),pp.2727-33http://dx.doi.org/10.1001/jama.291.22.2727|Medline[2]ArdissinoD,BoersmaE,BudajA,Fernández-AvilésF,FoxKA,HasdaiD,etal..GuíadePrácticaClínicaparaeldiagnósticoytratamientodelsíndromecoronarioagudosinelevacióndelsegmentoST..RevEspCardiol,60:1070.e1-e80(2007),[3]AndersonJL,AdamsCD,AntmanEM,BridgesCR,CaliffRM,CaseyDEJr,etal..ACC/AHA2007guidelinesforthemanagementofpatientswithunstableangina/non-ST-Elevationmyocardialinfarction..AmCollCardiol,50(2007),pp.e1-e157[4]FoxKA,GoodmanSG,KleinW,BriegerD,StegPG,DabbousO,etal..Managementofacutecoronarysyndromes.Variationsinpracticeandoutcome.FindingsfromtheGlobalRegistryofacutecoronaryevents(GRACE)..EurHeartJ,23(2002),pp.1177-89Medline[5]FoxKA,GoodmanSG,AndersonFAJr,GrangerCB,MoscucciM,FlatherMD,etal..Fromguidelinestoclinicalpractice:theimpactofhospitalandgeographicalcharacteristicsontemporaltrendsinthemanagementofacutecoronarysyndromes.TheGlobalRegistryofAcuteCoronaryEvents(GRACE)..EurHeartJ,24(2003),pp.1414-24Medline[6]CooneyMT,DudinaAL,GrahamIM..Valueandlimitationsofexistingscoresfortheassessmentofcardiovascularrisk:areviewforclinicians..JAmCollCardiol,54(2009),pp.1209-27http://dx.doi.org/10.1016/j.jacc.2009.07.020|Medline[7]BrindleP,BeswickA,FaheyT,EbrahimS..Accuracyandimpactofriskassessmentintheprimarypreventionofcardiovasculardisease:asystematicreview..Heart,92(2006),pp.1752-9http://dx.doi.org/10.1136/hrt.2006.087932|Medline[8]GRACEACSRiskScore..Disponibleen:http://www.outcomes-umassmed..org/grace/grace_risk_table,(cfm),[9]LemeshowS,HosmerD..Areviewofgoodnessoffitstatisticforuseinthedevelopmentoflogisticregresiónmodels..AmJEpidemiol,115(1982),pp.92-106Medline[10]LemeshowS,KlarJ,TeresD..Outcomepredictionforindividualintensivecarepatients:useful,misused,orabused?IntensiveCareMed..,21(1995),pp.770-6[11]LópezdeUllibarriGalparsoroI,PítaFernández,S..CurvasROC..AtenPrimaria,5(1998),pp.229-35[12]VandeWerfF,BaxJ,BetriuA,LundqvistC,CreaF,FalkV,etal..GuíasdePrácticaClínicadelaSociedadEuropeadeCardiología.ManejodelinfartoagudodemiocardioenpacientesconelevaciónpersistentedelsegmentoST..RevEspCardiol,62(2009),pp.e1-e47http://dx.doi.org/10.1016/S0300-8932(09)70373-2|Medline[13]BhattDL,RoeMT,PetersonED,LiY,ChenAY,HarringtonRAetal..Utilizationofearlyinvasivemanagementstrategiesforhigh-riskpatientswithnon-ST-segmentelevationacutecoronarysyndromes:resultsfromtheCRUSADEQualityImprovementInitiative..JAMA,292(2004),pp.2096-104http://dx.doi.org/10.1001/jama.292.17.2096|Medline[14]YanRT,YanAT,TanM,McGuireDK,LeiterL,FitchettDH,etal..CanadianAcuteCoronarySyndromeRegistryInvestigators.Underuseofevidence-basedtreatmentpartlyexplainstheworseclinicaloutcomeindiabeticpatientswithacutecoronarysyndromes..AmHeartJ,152(2006),pp.676-83http://dx.doi.org/10.1016/j.ahj.2006.04.002|Medline[15]TricociP,PetersonE..D,MulgundJ,NewbyLK,SaucedoJF,KleimanNS,etal.Temporaltrendsintheuseofearlycardiaccatheterizationinpatientswithnon-ST-segmentelevationacutecoronarysyndromes(resultsfromCRUSADE)..AmJCardiol,98(2006),pp.1172-6http://dx.doi.org/10.1016/j.amjcard.2006.05.047|Medline[16]YanAT,YanRT,HuynhT,CasanovaA,RaimondoFE,FitchettDH,etal..CanadianAcuteCoronarySyndromeRegistry2Investigators.Understandingphysicians''riskstratificationofacutecoronarysyndromes:insightsfromtheCanadianACS2Registry..ArchInternMed,169(2007),pp.372-8http://dx.doi.org/10.1001/archinternmed.2008.563|Medline[17]GoldbergRJ,CurrieK,WhiteK,BriegerD,StegPG,GoodmanSG,etal..Six-monthoutcomesinamultinationalregistryofpatientshospitalizedwithanacutecoronarysyndrome(theGlobalRegistryofAcuteCoronaryEvents[GRACE])..AmJCardiol,933(2004),pp.288-93[18]BradshawPJ,KoDT,NewmanAM,DonovanLR,TuJV..ValidityoftheGRACE(GlobalRegistryofAcuteCoronaryEvents)acutecoronarysyndromepredictionmodelforsixmonthpost-dischargedeathinanindependentdataset..Heart,92(2006),pp.905-9http://dx.doi.org/10.1136/hrt.2005.073122|Medline[19]AlterDA,VenkateshV,ChongA..EvaluatingtheperformanceoftheGlobalRegistryofAcuteCoronaryEventsrisk-adjustmentindexacrosssocioeconomicstrataamongpatientsdischargedfromthehospitalafteracutemyocardialinfarction..AmHeartJ,151(2006),pp.323-31http://dx.doi.org/10.1016/j.ahj.2005.07.013|Medline[20]YanAT,YanRT,TanM,CasanovaA,LabinazM,SridharK,etal..Riskscoresforriskstratificationinacutecoronarysyndromes:usefulbutsimplerisnotnecessarilybetter..EurHeartJ,28(2007),pp.1072-8http://dx.doi.org/10.1093/eurheartj/ehm004|Medline[21]DeAraújoGonçalvesP,FerreiraJ,AguiarC,Seabra-GomesR..TIMI,PURSUIT,andGRACEriskscores:sustainedprognosticvalueandinteractionwithrevascularizationinNSTE-ACS..EurHeartJ,26(2005),pp.865-72http://dx.doi.org/10.1093/eurheartj/ehi187|Medline[22]Ferreira-GonzálezI,Permanyer-MiraldaG,HerasM,CunátJ,CiveiraE,ArósF,etal..Patternsofuseandeffectivenessofearlyinvasivestrategyinnon-ST-segmentelevationacutecoronarysyndromes:AnassessmentbypropensityscorefortheMASCARAstudygroup..AmHeartJ,156(2008),pp.946-53http://dx.doi.org/10.1016/j.ahj.2008.06.032|Medline[23]ElbarouniB,GoodmanSG,YanRT,WelshRC,KornderJM,DeyoungJP,etal..ValidationoftheGlobalRegistryofAcuteCoronaryEvent(GRACE)riskscoreforin-hospitalmortalityinpatientswithacutecoronarysyndromeinCanada.CanadianGlobalRegistryofAcuteCoronaryEvents(GRACE/GRACE(2))Investigators..AmHeartJ,158(2009),pp.392-9http://dx.doi.org/10.1016/j.ahj.2009.06.010|Medline[24]YanAT,JongP,YanRT,TanM,FitchettD,ChowCM,etal..CanadianAcuteCoronarySyndromesregistryinvestigators.Clinicaltrial??derivedriskmodelmaynotgeneralizetoreal-worldpatientswithacutecoronarysyndrome..AmHeartJ,148(2004),pp.1020-7http://dx.doi.org/10.1016/j.ahj.2004.02.014|Medline[25]GaleCP,MandaSO,WestonCF,BirkheadJS,BatinPD,HallAS..EvaluationofriskscoresforriskstratificationofacutecoronarysyndromesintheMyocardialInfarctionNationalAuditProject(MINAP)database..Heart,95(2009),pp.221-7http://dx.doi.org/10.1136/hrt.2008.144022|Medline[26]GRACEACSRiskScore..Disponibleen:http://www..outcomes-umassmed,(org/grace), Subscribetoournewsletter Tools Print Sendtoafriend Exportreference Mendeley Statistics Recommendedarticles DeterminantsandPrognosticImpactofHeartFailureand... RevEspCardiol2018;71:820-8 AssociationBetweenClinicalPathwaysLeadingtoMedical... RevEspCardiol2017;70:817-24 New-onsetorPre-existingAtrialFibrillationinAcute... RevEspCardiol2019;72:383-91 Publishin RevistaEspañoladeCardiología Instructionsforauthors Submitanarticle EthicsinpublishingInformationforreviewers Frequentlyaskedquestions Lastarticles Filosofíapararesidentes.Pasadoy... 10.1016/j.rccl.2022.01.003 Larespuestadelafilosofía... 10.1016/j.rccl.2021.12.004 Riskstratificationinheartfailure.Anew... 10.1016/j.rccl.2022.01.010 Seemore Lastarticles Coronaryembolismduetocaseousmitralannular... 10.24875/RECICE.M21000219 TheReCrossdual-lumenmicrocatheterversatility... 10.24875/RECICE.M21000211 Long-termresultsofaprimaryangioplastyprogram... 10.24875/RECICE.M21000214 Seemore Editor'sPick(Spanishonly) AlineamientocomisuralduranteelTAVIconACURATEneo.AlfredoRedondoeIgnacioJ.Amat-Santos ImageBank Unaformainusualdeestimulaciónendocárdicaizquierda ImagessubjecttoCopyright,toapplyforpermissiontoreprint,[email protected]. ThisjournalsubscribestotheprinciplesandguidelinesoftheCommitteeonPublicationEthics(COPE)www.publicationethics.org IberoamericanCardiovascularJournalsEditors'Network(RCVIB) Readmore ESCNationalSocietiesCardiovascularJournalsEditors’Network Readmore Idiomas Español English RevistaEspañoladeCardiología(EnglishEdition) Home Articlesinpress CurrentIssue Archive Previousissues ClinicalPracticeGuidelines RegistriesoftheScientificSections Editor'sPick(Spanishonly) MostOftenRead Supplements RECpublications EditorialBoard Publishinthisjournal Instructionsforauthors Submitanarticle EthicsinpublishingInformationforreviewers Frequentlyaskedquestions Editorialnetworks Subscribetoournewsletter Articleoptions Léaloenespañol DownloadPDF Bibliography Tools Print Sendtoafriend Exportreference Mendeley Statistics es en ¿Esustedprofesionalsanitarioaptoparaprescribirodispensarmedicamentos? Areyouahealthprofessionalabletoprescribeordispensedrugs?