ProACS Risk Score: an Early and Simple Score for Risk Stratification of Patients with Acute Coronary Syndromes

(1)

www.revportcardiol.org

Revista

Portuguesa

de

Cardiologia

Portuguese

Journal

of

Cardiology

ORIGINAL

ARTICLE

ProACS

risk

score:

An

early

and

simple

score

for

risk

stratification

of

patients

with

acute

coronary

syndromes

Ana

Teresa

Timóteo

a,∗

,

Sílvia

Aguiar

Rosa

a

,

Marta

Afonso

Nogueira

a

,

Adriana

Belo

b

,

Rui

Cruz

Ferreira

a

_,

_on

_behalf

_of

_the

_ProACS

_{Investigators}

a_Cardiology_Department,_Santa_Marta_Hospital,_Centro_Hospitalar_de_Lisboa_Central,_Lisbon,_Portugal b_National_Center_for_Data_Collection_in_Cardiology,_Portuguese_Society_of_Cardiology,_Coimbra,_Portugal

Received29March2016;accepted18May2016 Availableonline31January2017

KEYWORDS Riskstratification score; Acutecoronary syndromes; Prognosis Abstract

Introduction:There are barriers to proper implementation of risk stratification scores in patientswithacutecoronarysyndromes(ACS),includingtheircomplexity.Ourobjectivewasto developasimplescoreforriskstratificationofall-causein-hospitalmortalityinapopulation ofpatientswithACS.

Methods:ThescorewasdevelopedfromanationwideACSregistry.Thedevelopmentand inter-nalvalidationcohortswereobtainedfromthefirst31829patients,randomlyseparated(60%and 40%,respectively).Theexternalvalidationcohortconsistedofthelast8586patientsincluded intheregistry.Thiscohortissignificantlydifferentfromtheothercohortsintermsof base-linecharacteristics,treatmentandmortality.Multivariatelogisticregressionanalysiswasused toselectfourvariableswiththehighestpredictivepotential.Ascorewasallocatedtoeach parameterbasedontheregressioncoefficientofeachvariableinthelogisticregressionmodel: 1pointforsystolicbloodpressure≤116mmHg,Killipclass2or3,andST-segmentelevation; 2pointsforage≥72years;and3pointsforKillipclass4.

Results:Thenewscorehadgooddiscriminativeabilityinthedevelopmentcohort(areaunder thecurve[AUC]0.796),anditwassimilarintheinternalvalidationcohort(AUC0.785,p=0.333). Intheexternalvalidationcohort,therewasalsoexcellentdiscriminativeability(AUC0.815), withanadequatefit.

∗_{Corresponding}_author.

E-mailaddress:anatimoteo@yahoo.com(A.T.Timóteo).

http://dx.doi.org/10.1016/j.repc.2016.05.010

(2)

Conclusions:TheProACSriskscoreenableseasyandsimpleriskstratificationofpatientswith ACS for in-hospital mortality thatcan beused atthe first medical contact,with excellent predictiveabilityinacontemporarypopulation.

Score_de_risco_ProACS:_emscore_simples_e_precoce_para_estratificac_¸ão_de_risco emdoentescomsíndromescoronáriasagudas

Resumo

Introdução:Existemalgumasbarreirasàimplementaçãoadequadadosscoresdeestratificação de risco em doentes comsíndrome coronária aguda(SCA), tais como a suacomplexidade. Onossoobjetivofoidesenvolverumscoresimplesparaestratificaçãoderiscodemortalidade hospitalardetodasascausasnumapopulaçãodedoentescomSCA.

Métodos: O score foi desenvolvido a partir de um registo nacional de SCA. A coorte de desenvolvimentoedevalidaçãointernafoiobtidaapartirdosprimeiros31829doentes, aleato-riamenteseparados (60e40%,respetivamente). A coortedevalidaçãoexternaé composta pelosúltimos8586doentesincluídosnoregisto.Estacoorteésignificativamentediferentedas restantes(característicasbasais,tratamentoemortalidade).Foiutilizadaanálisederegressão logísticamultivariadaparaselecionarasquatrovariáveiscommaiorpotencialpreditivoefoi atribuídaumapontuaçãobaseadanocoeficientederegressãodecadavariávelnomodelode regressãologística:umpontoparaTAS≤116mmHg,classeKillip2ou3,eelevaçãosegmento ST,doispontosparaidade≥72anosetrêspontosparaclasseKillip4.

Resultados: Onovo scoretemuma boacapacidadepreditiva nacoortededesenvolvimento (areaundercurve[AUC]0,796),semelhanteàcoortedevalidaçãointerna(AUC0,785,p=0,333). Nacoortedevalidaçãoexternatambémapresentouumaexcelentecapacidadediscriminativa (AUC0,815),comcalibraçãoadequada.

Conclusões:Oscore deriscoProACS permite umaestratificac¸ãode riscoprecocee simples emdoentescomSCAparamortalidadehospitalar,quepodeserutilizadanoprimeirocontacto médico,comexcelentecapacidadepreditivanumapopulac¸ãocontemporânea.

Introduction

The management of acute coronary syndromes (ACS) has

changeddramaticallyinthepast20yearsfollowing

demon-strationofthebenefitsassociatedwithinvasivestrategies,

particularly in high-risk patients.1---5 _Selection _for _these

strategiesis an importanttaskin every patientwithACS.

Severalrisk scores have been developedfor this purpose,

fromboth clinicaltrials andregistries.Theydifferin

pre-dictive accuracy as well as in the number and type of

variablesincluded.ThefirsttobedevelopedwastheTIMI

riskscore,6,7_but_its_predictive_accuracy_is_usually_lower_than

morerecentscores.Themostrecentandmostwidelyused

istheGlobalRegistryofAcuteCoronaryEvents(GRACE)risk

score,developedfromtheGRACEregistry.8,9 _This_has_very

highpredictiveaccuracy,butincludesmanyvariableswith

significant complexity, which may explain why it is often

underused.10,11

Previously,ourgroupandothershavedemonstratedthat

theseriskscorescanbesimplified,withaslightreductionin

predictiveaccuracycomparedtotheGRACEscore,butthat

can beconsidered acceptable.12,13 _These _simplified_scores

showsimilaraccuracytotheTIMIriskscore.6,7,14

Ourobjectivewastodevelopasimplescoreforrisk

strat-ificationofin-hospitalmortalityinpatientswithACS,tobe

usedvery early in patientmanagement, includingat

pre-hospitallevel.

Methods

The Portuguese Registry on Acute Coronary Syndromes

(ProACS)isamulticenternationwideregistryofACS.Itisa

prospective,continuousobservationalregistry,with33

par-ticipatingcardiologydepartmentsfromPortugal(mainland

andislands).15_Patient_inclusion_in_the_registry_began_on

Jan-uary1,2002,andallconsecutiveadultpatients(≥18years)

registered until October 31, 2014 were included in the

present study.Criteriafor inclusioninthe registrywerea

historyofchestpainat restorother symptomssuggestive

of an ACS (with themost recent episodeoccurring within

48hoursofadmission)withorwithoutneworpresumednew

(3)

Jan 1, 2002 - Oct 30, 2014 40 415 patients Jan 1, 2002 - Jun 30, 2011 31 829 patients Jul 1, 2011 - Oct 30, 2014 8586 patients 60% for development 19 128 patients

17 380 patients 11 548 patients 8532 patients

Missing values

40% for internal validation 12 701 patients

External validation 8586 patients

Figure1 Flowchartofpatientinclusion.

branchblockorelevatedbiomarkersofmyocardialdamage

witha riseand/or fall inlevels. Acutemyocardial

infarc-tion(MI)wasdefinedaccordingtotheuniversaldefinitionof

type1myocardial infarction.16 _A_diagnosis_of _ST-elevation

myocardialinfarction(STEMI)wasmadeinthepresenceof

persistent(>30min) ST-segmentelevation.Allother cases

withelevatedbiomarkersofmyocardialdamagewere

con-siderednon-STEMI(NSTEMI).

Data were collected in a dedicated computer

database, and included demographic,clinical and patient

management-related characteristics, as well as clinical

outcome.Hypertension,diabetesandhyperlipidemiawere

definedaseitherpreviouslyknownoronspecifictherapy.If

thepatientshadsmoked duringtheprevious 30daysthey

were classified as smokers and were self-reported.

Deci-sions on patient management strategy, including referral

forcoronaryangiographyandmodeofmyocardial

revascu-larization,ifany,werelefttotheattendingphysicianand

site-specificprotocols.Theprimaryendpointwasall-cause

mortalityduringtheindexhospitalization.

From the main study population (40415 records), we

obtainedthreecohortsofpatients(Figure1).FromJanuary

1, 2002 to June 30, 2011, 31829 patients were included

in the registry and were randomly divided into a cohort

for risk score development(60% of the patients) and the

cohort for internal validation (the remaining 40%). All

patients included between July 2011 and October 2014

(8586 patients) were included in the external validation

cohort.Patientswithmissingvaluesforthemainvariables

wereexcluded(2955 patients---7.3%of theinitialpatient

sample).

The study complies with the 1975 Declaration of

Helsinki.The nationaldataprotectioncommitteefor

clin-ical research approved the study protocol and informed

consentwasobtainedfromeachpatient.Theregistryis

reg-isteredatClinicalTrials.govwiththeidentification number

NCT01642329.

This manuscript was written in accordance with the

Transparent Reporting of a Multivariable Predictor Model

forIndividualPrognosisorDiagnosis(TRIPOD)statementon

appropriatedescriptionofstudiesforscorevalidation.17

Statisticalanalysis

Categoricalvariablesarereportedaspercentagesand

dif-ferencesbetweengroups weretested withthechi-square

test or Fisher’s exact test, as appropriate. Continuous

variables are reported as means and standard deviation.

Normality was tested with the Kolmogorov-Smirnov test.

One-wayANOVAwasusedtocomparecontinuousvariables

withnormaldistribution.Continuousvariableswithout

nor-maldistributionarereportedasmediansandinterquartile

rangeandwerecomparedwiththeKruskal-Wallistest.

Multivariablestepwiseforwardlogisticregressionmodels

were used to select the four variables with the

high-est predictive potential for score development. Variables

wereremovedfromthemodelwhenthep-valueexceeded

0.10 and were kept in the final model when less than

0.05. The variables included were age, admission heart

rate, systolic blood pressure and Killip class, creatinine,

ST-segment deviation, gender, diabetes, smoking status,

previous myocardial infarction, percutaneous coronary

intervention(PCI),coronaryarterybypassgrafting,strokeor

peripheralarterialdisease.The selectedcontinuous

varia-bleswerecategorized(dichotomized)byreceiveroperating

characteristic(ROC)curveanalysis.Cut-offswereobtained

bymaximizingthesumofsensitivityandspecificity.Foreach

variable,appropriateweightsweredeterminedbasedonthe

regressioncoefficientofthefinallogisticmodel(tothe

clos-estwholenumber).Pointsweresummedtoobtainthefinal

riskscore.ROCcurveanalysisandtheareaunderthecurve

(AUC)wereusedtostudythepredictive valueofthenew

riskscoreineach cohort.Bootstraptechniques wereused

forROCcurvecomparisonandforconfidenceintervals.

IBMSPSSStatisticssoftware(version19.0.0.2)wasused

for all statistical analyses. All statistical tests were

(4)

Table1 Clinicalcharacteristicsofthepatientsineachcohort. Development cohort(n=17380) Internalvalidation cohort(n=11548) Externalvalidation cohort(n=8532) p Age,years 66(13) 66(13) 66(13) 0.155 Male(%) 69.9 70.1 72.1 <0.001 BMI(kg/m2₎ ₂₇₍₄₎ ₂₇₍₄₎ ₂₇₍₄₎ _0.785 Previoushistory(%) Hypertension 62.2 62.9 69.8 <0.001 Smoking 24.0 24.5 27.9 <0.001 Diabetes 27.0 27.8 30.4 <0.001 Hyperlipidemia 45.2 45.4 57.5 <0.001 Myocardialinfarction 18.4 18.6 20.2 0.002 PCI 8.0 8.3 14.5 <0.001 CABG 4.1 4.3 5.0 0.005 Stroke/TIA 7.0 7.3 7.6 0.132 PAD 3.1 3.5 5.8 <0.001 COPD 6.6 6.1 5.5 0.195 Admission STEMI(%) 43.6 42.3 40.8 <0.001 Heartrate(bpm) 79(20) 79(20) 77(19) <0.001 SBP(mmHg) 141(30) 141(30) 138(29) <0.001 Killipclass≥2(%) 18.9 19.3 15.5 <0.001 Killipclass4(%) 2.0 2.0 1.9 0.801 Creatinine(mg/dl) 0.9(0.8-1.2) 0.9(0.8-1.2) 0.9(0.8-1.2) 0.130 Bloodglucose(mg/dl) 130(108-170) 130(107-175) 129(106-174) 0.739 Hemoglobin(g/dl) 13.7(1.9) 13.8(1.9) 13.7(1.9) 0.618 Coronaryangiography(%) 68.9 69.1 86.8 <0.001 Multivesseldiseasea_(%) _50.6 _50.6 _50.7 _0.994 PCI(%) 43.8 43.4 65.0 <0.001 CABG(%) 1.4 1.5 0.5 <0.001 LVEF<50%(%) 39.9 39.9 36.7 <0.001 Outcome(%) Re-infarction 1.5 1.4 1.2 0.120 Stroke/TIA 0.8 0.9 0.8 0.807 Majorbleeding 1.3 1.2 1.4 0.298 Death 5.4 5.0 4.0 <0.001

BMI:bodymassindex;CABG:coronaryarterybypassgrafting;COPD:chronicobstructivepulmonarydisease;LVEF:leftventricularejection fraction;PAD:peripheralarterialdisease;PCI:percutaneouscoronaryintervention;SBP:systolicbloodpressure;STEMI:ST-elevation myocardialinfarction;TIA:transientischemicattack.

a_Percentage_of_patients_who_underwent_coronary_angiography.

Results

Patients were included in each cohort according to the

flowchart presented in Figure 1. Patients in the

develop-ment cohort had fewer risk factors and less past history

of coronary artery disease compared to the internal and

external validation cohorts (Table 1). On admission they

had higher heart rate and systolic blood pressure and

worse Killip class and STEMI was more frequent,

partic-ularly compared to the external validation cohort. They

lessoftenunderwentcoronaryangiographyand

revascular-izationprocedures.In-hospital mortalitywaslower inthe

validation cohorts, particularly in the external validation

cohort.

Inthe multivariatemodel,the variableswiththe

high-estpredictiveability wereage,systolicblood pressureon

admission,KillipclassonadmissionandST-segment

eleva-tion. The cut-off obtained for age was 72 years and for

systolicbloodpressureitwas116mmHg.TheProACSscore

was developed according to these cut-offs with a simple

scoringsystem(Table2).

The ProACSscore wasthen tested in the development

cohort as well as in the internal and external validation

cohorts.Inallcohortsthediscriminativeabilitywasalsovery

good(AUC>0.75)andsimilarbetweenallcohorts,although

slightly higherin theexternal validationcohort (Table 3).

The fit of the developed logistic model was also good in

thedevelopmentcohort(Figure2).Performancewassimilar

comparingpatientswithSTEMIandnon-ST-elevationacute

coronary syndrome (NSTACS). However, compared to the

GRACErisk score,thepredictive abilitywasslightlylower

(5)

Table2 ProACSriskscore.

Variable Cut-off Points

Age <72years ≥72years 0 2 SBP <116mmHg ≥116mmHg 1 0 Killipclass 1 2 3 4 0 1 1 3 ST-segmentelevation No Yes 0 1

SBP:systolicbloodpressure.

Table3 AreaunderthecurvefortheProACSriskscorein thedifferentcohorts.

Cohort C-statistic(95%CI)

Development 0.796(0.782-0.810)

Internalvalidation 0.785(0.767-0.803)

Externalvalidation 0.815(0.793-0.837)

STEMI 0.799(0.768-0.830)

NSTACS 0.809(0.774-0.845)

CI:confidenceinterval;NSTACS:non-ST-elevationacute coro-narysyndrome;STEMI:ST-elevationmyocardialinfarction.

400 Hosmer-Lemeshow chi-square (6 df)=8072 p=0.233 300 200 Predicted deaths 100 Observed deaths 0 0 100 200 300 400

Figure2 Fitofthefinalmultivariatelogisticregressionmodel inthedevelopmentcohort.df:degreesoffreedom.

OntheProACSriskscore,patientswithascore0havea

lowriskofin-hospitaldeath(<1%)andthosewithascore≥3

haveahighrisk(>4%).Patientswithscore1-2have

interme-diaterisk(1-4%)(Figure4).

Discussion

Inthepast20years,invasivestrategiesforthetreatmentof

ACShavehadasignificantimpactonprognosis.1---5_It_is_thus

importanttoidentifyandselectthepatientswhocanderive

themostbenefitfromthesestrategies.InSTEMIprimaryPCI

isnowgenerallyrecommended1,3 _and_so_risk_{stratification}

1.0 0.8 0.888 (0.865, 0.910) 0.815 (0.793, 0.837) 0.6 0.4 0.2 Sensitivity 0.0 1.0 0.8

External validation cohort

0.6 Specificity 0.4 0.2 ProACS score GRACE score 0.0

Bootstrap test for two correlated ROC curves (D=-7.6893, p<0.001) Figure3 ComparisonofProACSandGRACEriskscoresinthe externalvalidationcohort.ROC:receiveroperating character-istic. 12 10 8 6 4 2 0 0.4% Score 0 Obser ve d in-hospital mor tality Score 1-2 Score ≥3 2.1% 11.2% %

Figure4 Observedin-hospitalmortalityintheexternal vali-dationcohortaccordingtotheriskcut-offs.

forthesepatientsislessimportant,butsomecasesrequire

urgent pre-hospitalreferral to amore specialized center.

NSTACS patients are a heterogeneous group and invasive

strategiesarerecommendedinhigh-andintermediate-risk

patients.2,4 _For_this _reason,_several _risk _scores _have _been

developed in the past 15 years. The TIMI risk scores for

STEMI and unstable angina/NSTEMI were developed from

largeclinicaltrials.6,7_These_scores_were_simple_and_intuitive

withsimple variables,andtheir usespreadrapidly

world-wide.Other risk scores werealso developed but theTIMI

scoreremainedthemostuseduntil2003,whentheGRACE

registrygrouppublishedariskscoreforin-hospital

mortal-ityinpatientswithACSandlaterforsix-monthmortality.8,9

Thisscorewasdevelopedfromaregistrythatrepresented

real-world patients instead of the highly controlled and

selectedpopulations of clinical trials, fromwhich elderly

patientsandthosewithsevererenaldysfunctionwere

usu-allyexcluded.TheGRACEriskscoreshowedhighpredictive

(6)

themostusedscore,andwassubsequentlyupdated.18

How-ever,theGRACEscore’snomogramincludesseveralvariables

withawiderangeofcategories,someofwhichcanonlybe

obtainedafterinitialdiagnostictests.Thisaddscomplexity

toriskscoringandmeansitcannotbeusedatanearlystage.

Thismightexplainitsunderuseinclinicalpracticeinspite

ofthedevelopmentofseveraltools.10,11

Ourgroupdemonstratedinasingle-centerstudythatitis

possibletosimplifytheseriskscoreswithsimpleandfewer

variables,withsimilarpredictiveaccuracy,althoughwedid

notsetouttoproposeanewriskscoreduetotheinherent

limitationsofourstudy.12 _A_Canadian_group_has_developed

anotherscore, the Canada Acute Coronary Syndrome risk

score (C-ACS), based onlarge ACS registries in Canada.13

Theselectedvariableswereage,Killipclass,systolicblood

pressureandheartrate,withasimple dichotomizationof

continuousvariables.Thisscoreenablesrapidstratification

ofpatientswithACS,evenbyhealthcareprofessionals

with-outadvanced medicaltraining,in pre-hospitalsettings or

emergencydepartments.Itshowedagoodpredictivevalue

forboth short-andlong-term mortalityandinbothSTEMI

andNSTACS.The c-statistic ofthis newscoreis adequate

(≥0.75)butslightlylowerthanclassicalriskscoresforthe

predictionofshort-andlong-termmortality.13

SinceNorth Americanpopulations have different

char-acteristicsfromotherpopulations,includingEuropeanand

particularlyMediterranean(low-risk)countries,wethought

itwouldbevaluabletodevelopasimilarriskscoreforearly

useinACSthatwouldbesimpleandeasytomemorize.19,20

The ProACS risk score was developed from a nationwide

registry of ACS in Portugal. The sample is large and is

representativeofreal-worldpatients.Inourscore,weused

simplevariables (age,systolicblood pressure,ST-segment

elevation and Killip class on admission), similar to the

C-ACS score, and confirmed the validity of the selected

variables(thosewiththehighest predictivevalue).It also

showed good predictive accuracy, which although lower

thantheGRACEriskscore,isstillacceptable.Itisalsoeasy

tocalculate andcan beappliedinpre-hospitalsettingsor

emergencydepartmentsbynon-medicalhealthcare

profes-sionals.Itimprovesreferralofhigh- andintermediate-risk

patientstohospitalswithcatheterizationfacilities,

partic-ularly in NSTACS. External validation was performed in a

morerecentcohortofpatients,whichmayexplainsomeof

thedifferencesobserved,comparedwiththedevelopment

andinternalvalidationcohorts,particularlythelower

pro-portionof STEMIand higher rateof coronary angiography

andrevascularization.Itis alsoimportanttomentionthat

predictive accuracy was higher in the external validation

cohort,whichhighlightsitsutilityincontemporarycohorts.

Predictive accuracy was similar in patients with STEMI

and NSTACS, suggesting that it can also be used in

STEMI patients, although in this group the recommended

strategyisprimaryPCI.

Limitations

ThisscorewasdevelopedandvalidatedinaPortuguese

pop-ulation.Itrequiresfurthervalidationinother populations,

preferablyincontemporarycohorts.

Wechosetoexcludefromtheanalysisallpatientswith

missingvaluesofthescorevariables.Overall,only7.3%of

the patients were excluded, which is acceptable because

of our large sample size. Data from the external

vali-dationcohortweremorecompleteandfewerpatientswere

excludedfromthiscohort.

Wecouldnottestgoodnessoffitwiththenewriskscore

becausethisis acategoricalscoreandisthusnotsuitable

forthistypeofanalysis.

Conclusion

TheProACSriskscoreenableseasyandsimplerisk

stratifi-cationforin-hospitalmortalityatthefirstmedicalcontact

ofpatientswithACS.Italsohasexcellentpredictiveability

inacontemporarypopulationofpatientswithACS,although

slightlylowerthantheGRACEriskscore.Itssimplicitycan

improveimplementationof scores forrisk stratification in

clinicalpractice,althoughfurthervalidationinother

popu-lationsandcountriesiswarranted.

Ethical

disclosures

Protection of human and animal subjects.The authors

declarethatnoexperimentswereperformedonhumansor

animalsforthisstudy.

Confidentialityofdata.Theauthorsdeclarethattheyhave

followedtheprotocolsoftheirworkcenteronthe

publica-tionofpatientdata.

Right to privacy and informed consent.The authors

declarethatnopatientdataappearinthisarticle.

Conflicts

of

interest

Theauthorshavenoconflictsofinteresttodeclare.

References

1.Task Force on the management of ST-segment elevation acute myocardial infarction of the European Society of Cardiology(ESC)StegG, JamesSK,AtarD,etal.ESC Guide-lines for the management of acute myocardial infarction in patientspresenting with ST-segment elevation. EurHeart J. 2012;33:2569---619.

2.HammCW,BassandJP, AgewallS,etal.,ESCCommitteefor Practice Guidelines. ESC Guidelines for the management of acutecoronarysyndromesinpatientspresentingwithout per-sistentST-segmentelevation.EurHeartJ.2011;32:2999---3054.

3.O’GaraPT, Kushner FG, Ascheim DD, et al. 2013 ACCF/AHA Guideline for the management of ST-elevation myocardial infarction: a report of the American College of Cardiology Foundation/AmericanHeartAssociationTaskForceonPractice Guidelines.Circulation.2013;127:e362---425.

4.AmsterdamEA,WengerNK,BrindisRG,etal.2014AHA/ACC Guideline for the management of patients with non-ST-elevationacutecoronarysyndromes:areportoftheAmerican CollegeofCardiology/AmericanHeartAssociationTaskForceon PracticeGuidelines.Circulation.2014;130:e344---426.

5.Moreira D, Marmelo B, Delgado A, et al. A decisão de não revascularizar o enfarte agudo do miocárdio sem

(7)

supradesnivelamento de ST--- condicionantes e prognóstico. Arealidadenacional.RevPortCardiol.2015;34:315---28.

6.AntmanEM,CohenM,BerninkPJ,etal.TheTIMIriskscorefor unstableangina/non-ST-elevationMI.Amethodfor prognostica-tionandtherapeuticdecisionmaking.JAMA.2000;284:835---42.

7.MorrowDA,AntmanEM,CharlesworthA,etal.TIMIriskscore forST-elevationmyocardialinfarction:aconvenient,bedside, clinicalscoreforriskassessmentatpresentation.Circulation. 2000;102:2031---7.

8.GrangerCB,GoldbergRJ,DabbousO,etal.Predictorsof hos-pitalmortalityintheglobalregistryofacutecoronaryevents. ArchInternMed.2003;163:2345---53.

9.EagleKA, LimMJ,DabbousOH,et al.Avalidatedprediction modelforallformsofacutecoronarysyndromes:estimatingthe riskof6-monthpostdischargedeathinaninternationalregistry. JAmMedAssoc.2004;291:2727---33.

10.YanAT,YanRT,HuynhT,etal.Understandingphysicians’risk stratificationofacutecoronarysyndromes:insightsfrom the CanadianACSIIRegistry.ArchInternMed.2009;169:372---8.

11.BagnallAJ,GoodmanSG,FoxKA,etal.Optimalmedical ther-apyfornon-ST-elevationacutecoronarysyndromes:exploring whyphysiciansdonotprescribeevidence-basedtreatmentand whypatientsdiscontinuemedicationsafterdischarge.Cir Car-diovascQualOutcomes.2010;3:530---7.

12.TimóteoAT,PapoilaAL,LopesJP,etal.Isitpossibletosimplify riskstratificationscoresforpatientswithST-segmentelevation

myocardialinfarctionundergoingprimaryangioplasty?RevPort Cardiol.2013;32:967---73.

13.HuynhT,KrouzS,YanA,etal.CanadaAcuteCoronarySyndrome RiskScore:anewriskscoreforearlyprognosticationinacute coronarysyndromes.AmHeartJ.2013;166:58---63.

14.Gonc¸alves PA, Ferreira J, Aguiar C, et al. TIMI, PURSUIT, and GRACErisk scores:sustained prognosticvalueand inter-action with revascularization in NSTE-ACS. Eur Heart J. 2005;26:865---72.

15.FerreiraSantosJ,AguiarC,GavinaC,etal.RegistoNacional deSíndromesCoronáriasAgudas:seteanosdeactividadeem Portugal.RevPortCardiol.2009;28:1465---500.

16.ThygesenK,AlpetJS,JaffeAS,etal.Thirduniversaldefinition ofmyocardialinfarction.EurHeartJ.2012;33:2551---67.

17.CollinsGS, ReitsmaJB,AltmanDG,etal. Transparent repor-tingofamultivariablepredictionmodelforindividualprognosis or diagnosis (TRIPOD). The TRIPOD statement. Circulation. 2015;131:211---9.

18.Pieper KS, Gore JM, Fitzgerald G, et al. Validity of a risk-predictiontool for hospital mortality:theGlobal Registryof AcuteCoronaryEvents.AmHeartJ.2009;157:1097---105.

19.SansS,KestelootH,KromhoutD.Theburdenofcardiovascular diseasesmortalityinEurope.EurHeartJ.1997;18:1231---48.

20.D’AgostinoR, Grundy S,Sullivan LM,etal. Validationofthe Framinghamcoronaryheartdiseasepredictionscores:resultsof amultipleethnicgroupsinvestigation.JAMA.2001;286:180---7.