REPC-286; No.ofPages6
RevPortCardiol.2013;xxx(xx):xxx---xxx
Revista Portuguesade
Cardiologia
PortugueseJournalofCardiologywww.revportcardiol.org
ORIGINALARTICLE
Diagnosticyieldofcurrentreferralstrategiesforelective
coronaryangiographyinsuspectedcoronaryarterydisease----
AnanalysisoftheACROSSregistry
MiguelBorgesSantos∗,FerreiraAntónioMiguel,deAraújoGoncalvesPedro,
LuísRaposo,RuiCampanteTeles,ManuelAlmeida,MiguelMendes
DepartamentodeCardiologia,HospitaldeSantaCruz,Lisboa,Portugal
Received5September2012;accepted1November2012
KEYWORDS Coronary angiography; Chestpain/diagnosis; Stableangina; Myocardialischemia Abstract
IntroductionandObjectives:Thepurposeofthisstudywastoassessthediagnosticyieldof currentreferralstrategiesforelectiveinvasivecoronaryangiography(ICA).
Methods:Weperformedacross-sectionalobservationalstudyofconsecutivepatientswithout knowncoronaryarterydisease(CAD)undergoingelectiveICAduetochestpainsymptoms.The proportionofpatientswithobstructiveCAD(definedasthepresenceofatleastone≥50% stenosisonICA)wasdeterminedaccordingtotheuseofnoninvasivetesting.
Results:Thestudypopulationconsistedof1892individuals(60%male,meanage64±11years), ofwhom1548(82%)hadapositivenoninvasivetest:exercisestresstest(41%),stressmyocar- dialperfusionimaging(36%),stressechocardiogram(3%)orcoronarycomputedtomography angiography(3%).Referralwithouttestingoccurredin18%ofpatients.Theoverallprevalence ofobstructiveCADwas57%,higheramongthosewithprevioustesting(58%vs.51%withoutpre- vioustesting,p=0.026)andwhenanatomicratherthanfunctionaltestswereused(81.3%vs. 57.1%,p=0.001).Apositivetestandconventionalriskfactorswereallindependentpredictorsof obstructiveCAD,withadjustedoddsratios(95%confidenceinterval)of1.34(1.03---1.74)fornon- invasivetesting,1.05(1.04---1.06)forage,3.48(2.81---4.29)formalegender,1.86(1.32---2.62) forcurrentsmoking,1.74(1.38---2.20)fordiabetes,1.30(1.04---1.62)forhypercholesterolemia, and1.39(1.08---1.80)forhypertension.
Conclusions:Morethan40%ofpatientswithoutknownCADundergoingelectiveICAdidnot haveobstructivelesions,eventhoughfouroutoffivehadapositivenoninvasivetest.These examswererelativelyweakgatekeepers;functionaltestsweremoreoftenusedbutappeared tobeoutperformedbytheanatomictest.
©2012SociedadePortuguesadeCardiologia.PublishedbyElsevierEspaña,S.L.Allrights reserved.
∗Correspondingauthor.
E-mailaddress:[email protected](M.BorgesSantos).
0870-2551/$–seefrontmatter©2012SociedadePortuguesadeCardiologia.PublishedbyElsevierEspaña,S.L.Allrightsreserved.
Pleasecitethisarticleinpressas:BorgesSantosM,etal.Diagnosticyieldofcurrentreferralstrategiesforelective coronaryangiographyinsuspectedcoronaryarterydisease----AnanalysisoftheACROSSregistry.RevPortCardiol.2013.
http://dx.doi.org/10.1016/j.repc.2012.11.008
ARTICLE IN PRESS
+ModelREPC-286; No.ofPages6
2 M.BorgesSantosetal.
PALAVRAS-CHAVE Angiografiacoronária; Dor torácica/diagnóstico; Anginaestável; Isquémiamiocárdica
Rendimentodasatuaisestratégiasdereferenciac¸ãoparacoronariografiaeletiva porsuspeitadedoenc¸acoronária----análisedoregistoACROSS
Resumo
Introduc¸ãoeobjetivos: Oobjetivodoestudofoiavaliarorendimentodasatuaisestratégiasde referenciac¸ãoeletivaparacoronariografiainvasiva.
Métodos:Estudotransversaldeindivíduosconsecutivossemdoenc¸acoronáriaconhecidasub- metidosacoronariografiapordortorácica.Determinac¸ãodaprevalênciadedoenc¸acoronária obstrutiva(definidapelapresenc¸adepelomenosumaestenose≥50%)deacordocoma utilizac¸ãodetestesnão-invasivosparadespistedecardiopatiaisquémica.
Resultados:Foramavaliados1892indivíduos(60%homens,idademédia64±11anos),dos quais1548(82%)tinhamumtestenão-invasivopositivo:provadeesforc¸o(41%),cintigrafiade perfusãomiocárdica(36%),ecocardiogramadestress(3%)eangiografiacoronáriaportomografia computorizada(3%).Ocorreureferenciac¸ãosemtesteprévioem18%dosdoentes.Aprevalência globaldedoenc¸aobstrutivafoi57%,sendomaiselevadanosdoentessubmetidosatestesnão- invasivos(58%versus51%nosdoentessemtestesprévios,p=0,026)enaquelesemqueoteste eraanatómicoversusfuncional(81,3%versus57,1%,p=0,001).Umtestenão-invasivopositivo efatoresderiscoconvencionaisforampreditoresindependentesdedoenc¸aobstrutiva,com
odds-ratioajustado(intervaloconfianc¸a95%)de:testenão-invasivo1,34(1,03-1,74),idade 1,05(1,04-1,06),sexomasculino3,48(2,81-4,29),tabagismoativo1,86(1,32-2,62),diabetes 1,74(1,38-2,20),hipercolesterolemia1,30(1,04-1,62)ehipertensão1,39(1,08-1,80).
Conclusões:Maisde40%dosdoentessemdoenc¸acoronáriaconhecidaquerealizamcoronar- iografiaeletivanãotêmdoenc¸aobstrutiva,apesardequatroemcadacincoterumteste não-invasivopositivo.Estestestessãogatekeepersrelativamentefracos;osfuncionaisforam utilizadosmaisfrequentementemasoanatómicopareceutermelhordesempenho. ©2012SociedadePortuguesadeCardiologia.PublicadoporElsevierEspaña,S.L.Todosos direitosreservados.
Listofabbreviations
CAD coronaryarterydisease
CCTA coronarycomputedtomographyangiography ECG electrocardiogram
ICA invasivecoronaryangiography
SPECT single-photonemissioncomputedtomography
Introduction
Theevaluationofpatientswithsuspectedcoronaryartery disease(CAD)isbasedonclinicalassessment,oftensupple- mentedbynoninvasivetestswhichserveasgatekeepersfor invasivecoronaryangiography(ICA).1---3ICAisthediagnostic
goldstandardforCADbutiscostly,haslimitedavailability andcarriesariskofcomplicationsrelatedtoitsinvasive nature.4Theaimsofperformingnoninvasivetestinginthis
settingincludeminimizingunnecessaryrisksandcosts,and identifyingpatientswhowillbenefitfromrevascularization. However,despitethefrequentuseofnoninvasivetesting,a significantproportionofpatientsundergoingICAdonothave obstructiveCADorarenoteligibleforrevascularization.5,6
Thepurposeofthisstudywastoassesscurrentpatternsof noninvasivetestingandtoappraisetheirdiagnosticyield amongsymptomaticpatientsundergoingICAforsuspected CAD.
Methods
Population
Thiswasanobservational,cross-sectionalstudyperformed ata singlehospital center servinganurban population of 900 000 inhabitants in Lisbon, Portugal. The study population consisted of all patients referred for elec- tiveICAforevaluationofchestpainsymptomsbetween January 2006 and November 2010. Patients’ referral for ICA and the decision to performprevious noninva- sive testing, including the testing modality, were left to the discretion of attending physicians. Noninvasive testingwasperformedmostlyatprivatepracticefacili- ties.
The modalities ofnoninvasive testing were exercise electrocardiogram (ECG) stress testing, stress myocar- dialsingle-photonemissioncomputedtomography(SPECT), stress echocardiography and coronary computed tomo- graphy angiography (CCTA). ‘Ischemic changes’ on the resting ECG were not considered noninvasive testing. The following exclusion criteria were applied sequen- tially: non-elective setting (acute coronary syndrome), previously knownCAD (definedas previousacute coro- narysyndrome,revascularizationprocedureordocumented coronary stenosis ≥50% on previous ICA), preopera- tive evaluation, presenting symptom other than chest pain, negative noninvasive test result and incomplete information on patients’ clinical characteristics or ICA result.
Pleasecitethisarticleinpressas:BorgesSantosM,etal.Diagnosticyieldofcurrentreferralstrategiesforelective coronaryangiographyinsuspectedcoronaryarterydisease----AnanalysisoftheACROSSregistry.RevPortCardiol.2013.
http://dx.doi.org/10.1016/j.repc.2012.11.008
ARTICLE IN PRESS
+ModelREPC-286; No.ofPages6
Yieldofcurrentreferralstrategiesforelectivecoronaryangiography 3 11523 coronary angiography
procedures (2006-2010)
1892 patients referred for elective coronary angiography for the
evaluation of chest pain
1548 (81.8%) with a positive noninvasive test
344 (18.2%) without previous testing
ACS setting, n=4044 (35.1%) Prior CAD, n=2465 (21.4%)
Preop evaluation or other, n=2162 (18.8%)
Presenting symptom other than chest pain, n=927 (8.0%)
Negative noninvasive test or incomplete information, n=33 (0.3%)
Figure1 Selectionofthestudypopulation.ACS:acutecoronarysyndrome;CAD:coronaryarterydisease;preop:preoperative.
Patientevaluation
Dataon demographiccharacteristics,cardiovascularrisk factors,typeofnoninvasivetestingandresultsofcoronary angiographywereprospectivelycollectedintheongoing ACROSS(AngiographyandCoronaryRevascularizationOn SantacruzhoSpital)registry,approvedbythelocalethics committee.Thediagnosesofhypertension,hypercholester- olemiaanddiabetes(regardlessoftype,durationorcurrent treatment) were assigned if indicated in the patients’ referralletterorifthepatientwasbeingtreatedwithanti- hypertensiveorlipid-loweringdrugs,oralantidiabeticsor insulin.Toavoidunderdiagnosis,obstructivecoronarywas definedasa≥50%reductioninvesseldiameterascompared toanondiseasedproximalsegment.Thisbroaddefinitionof obstructiveCADwasnotusedasacriterionforrevasculari- zation.
Statisticalanalysis
Dataarepresentedascounts(%),medians(interquartile range)ormeans±standarddeviation.Categoricalvaria- bleswerecomparedusingFisher’sexacttest.Continuous variableswerecomparedbymeansofthettest.Patients withandwithoutobstructiveCADwerecomparedfordif- ferencesinage,gender,bodymassindex,prevalenceof cardiovascularriskfactorsanduseofnoninvasivetesting. Variablesthatshowedsignificantassociationwithobstruc- tiveCAD(p<0.10)inunivariateanalysiswereincludedina binarylogisticregressionmodeltoidentifyindependentpre- dictors.Temporaldifferencesduringthestudyperiodinthe prevalenceofobstructiveCADandtheuseofnoninvasive
testingwereassessedusingthechi-squaretestfortrend.A two-sidedp-valueoflessthan0.05wasconsideredtoindi- catestatisticalsignificance.Allanalyseswereperformed usingthestatisticalpackageSPSS®version17.0(SPSS,Inc.,
Chicago,IL).
Results
Duringthestudyperiod, 11523patientsunderwentICA atourhospital.Aftertheexclusioncriteriawereapplied (Figure1),1892patientswereincludedintheanalysis.
Most patients (81.8%,n=1548) werereferredafter a positivenoninvasivetest.OnICA,theoverallprevalence ofobstructive CAD was56.7% (1072/1892).One-vessel, two-vesselorthree-vessel/leftmaindiseasewereidenti- fiedin21.1%(n=398), 17.1%(n=323)and 18.6%(n=351) ofpatients, respectively.The prevalenceofobstructive CADwaslowerinpatientsreferredwithoutpreviousnon- invasivetestingthaninthosewithapositivetest(51.2% vs.57.9%,p=0.026).Myocardialrevascularization(percu- taneous coronary intervention or referral for coronary arterybypassgrafting)wasperformedin46.7%(n=883)of patients.
Increasingage,malegender,traditionalcardiovascular riskfactorsandpositivenoninvasivetestingwerepredictors ofobstructiveCADinunivariateandmultivariateanalysis (Table1).
Therewerenosignificanttemporaldifferencesinthe prevalenceofobstructive CADduring thestudy period, despiteasignificantincreaseintheproportionofpatients undergoingnoninvasivetesting(Figure2).Exercisestress testing and stress SPECT were the most used tests,
Pleasecitethisarticleinpressas:BorgesSantosM,etal.Diagnosticyieldofcurrentreferralstrategiesforelective coronaryangiographyinsuspectedcoronaryarterydisease----AnanalysisoftheACROSSregistry.RevPortCardiol.2013.
http://dx.doi.org/10.1016/j.repc.2012.11.008
ARTICLE IN PRESS
+ModelREPC-286; No.ofPages6
4 M.BorgesSantosetal.
Table1 Populationcharacteristics.
Total Univariateanalysis Multivariateanalysis
ObstructiveCAD (n=1072) Noobstructive CAD(n=820) p Adjusted oddsratio 95%CI p Age,years 64±11 65.7±10.4 61.7±11.0 <0.001 1.05 1.04---1.06 <0.001 Male,n(%) 1141(60.3%)769(71.7%) 372(45.4%) <0.001 3.48 2.81---4.29 <0.001
Bodymassindex,kg/m2
28.0±3.9 27.6±3.9 28.0±4.5 0.030 0.98 0.96---1.01 0.240
Cardiovascularriskfactors,%
Hypertension 78.4% 81.3% 74.6% <0.001 1.39 1.08---1.80 0.011
Diabetes 27.5% 32.5% 21.1% <0.001 1.74 1.38---2.20 <0.001
Smoking 11.4% 13.2% 8.9% 0.003 1.86 1.32---2.62 <0.001
Hypercholesterolemia 69.2% 71.9% 65.7% 0.004 1.30 1.04---1.62 0.019
Previouspositivenoninvasivetesting 81.8% 83.6% 79.5% 0.026 1.34 1.03---1.74 0.028
CAD:coronaryarterydisease;CI:confidenceinterval.
accountingformorethan90%ofnoninvasivetesting.CCTA useincreasedsignificantlyfrom2006to2010.
Therates ofobstructive coronaryarterydiseaseand myocardialrevascularizationaccordingtotypeofnoninva- sivetestingarepresentedinFigure3.Comparingfunctional andanatomic tests,the prevalence ofobstructive CAD (57.1%vs.81.3%,p=0.001)washigherinthelattergroup.
Discussion
Diagnosticyieldofthecurrentreferralstrategy InourEuropean,urbanclinicalsetting,lessthan57%of patientsreferredforelectiveICAforevaluationofchest painsymptomshadobstructivelesions(definedbyabroad criterionof≥50%luminalstenosis),despitethefactthat fouroutoffivepatientshadundergoneprevioustesting. Noninvasivetestingwasfrequentlyusedbutwasonlya weakindependentpredictorofobstructiveCAD(OR1.34, p=0.028).Thisapparentlylowperformanceofnoninvasive
Exercise ECG (n=768) 0 10 20 30 % 40 50 60 60.7 52.2 53.0 44.0 59.3 42.6 81.3 Obstructive CAD Revascularization 58.3 70 80 90 100 Stress SPECT (n=678) Stress echo (n=54) Coronary CTA (n=48)
Figure3 Rateofobstructivecoronaryarterydiseaseand myocardialrevascularizationaccordingtotypeofnoninvasive testing.CAD:coronaryarterydisease;CTA:computedtomo- graphyangiography;ECG:electrocardiogram;Echo:echocar- diogram;SPECT:single-photonemissioncomputedtomography.
Prevalence of obstructive CAD 0 10 20 30 40 50 60 p=0.514 p=0.008 p=0.457 p=0.159 p=0.603 p<0.001 2006 2007 2008 2009 2010 % 70 80 90 Noninvasive testing Exercise ECG Stress SPECT Stress echo Coronary CTA
Figure2 Trendsofnoninvasivetestingbetween2006and2010.CAD:coronaryarterydisease;CTA:computedtomography angiography;ECG:electrocardiogram;Echo:echocardiogram;SPECT:single-photonemissioncomputedtomography.
Pleasecitethisarticleinpressas:BorgesSantosM,etal.Diagnosticyieldofcurrentreferralstrategiesforelective coronaryangiographyinsuspectedcoronaryarterydisease----AnanalysisoftheACROSSregistry.RevPortCardiol.2013.
http://dx.doi.org/10.1016/j.repc.2012.11.008
ARTICLE IN PRESS
+ModelREPC-286; No.ofPages6
Yieldofcurrentreferralstrategiesforelectivecoronaryangiography 5 testsasgatekeepersforICAhasseveralpossibleexplana-
tions.Oneisthattheperformanceofthesetestsinthe‘‘real world’’isworsethanthatreportedintheliteraturefrom largeexperiencedcenters.Anotherexplanationwouldbea lowpretestprobabilityofobstructiveCADinthispopulation, resultinginarelativelylargeabsolutenumberofpatients withoutobstructivediseaseundergoingICA.7Athirdhypoth-
esis,supportedbyincreasingevidence,6,8isthatthepretest
likelihoodofangiographicallysignificantCADmaybeover- estimatedwhencalculatedonthebasisofage,genderand chestpaincharacteristicsinaccordancewiththeseminal workofDiamond,ForresterandPryor.9,10
Theyieldofanydiagnostictestdependsonthepretest likelihoodofthepatientsinwhomitisusedandonthe waythetestmodifiesthatprobability.Ideally,apositive noninvasivetestshouldincreasetheprobabilityofobstruc- tivediseasetoalevelthatjustifiesperformingICA,and anegativetest shouldreducethat likelihoodtoalevel atwhichobstructiveCADcanbesafelyruledout.While ICAwill alwaysbeperformedonsomepatientswithout coronarylesions,the 2011standards forcatheterization laboratoryaccreditationfromtheAccreditationforCardio- vascularExcellenceorganizationsuggestthattheincidence ofnon-obstructivediseaseinelectivepatientsshouldbe <40%.Intheinterestsofindividualpatientsandofoverall healthcarecost-effectiveness,extremeratesareundesir- able.Recently,Gendersetal.6reportedarateofobstructive
CADof58%(rangingfrom39.4%to75.5%)inamulticen- terstudyinvolving11Europeanhospitals.IntheUSA,Patel etal.5 reportedanoverallrateof41%ofpatientswith
obstructiveCADintheNationalCardiovascularDataReg- istry,althoughthisvariedsignificantlybetweendifferent centers,from23to100%.11Takentogether,thesestudies
suggestthatbettergatekeepersareneeded.Ourfindingsare inlinewithboththesestudies,underliningtherelativelylow prevalenceofobstructiveCADonICAinapopulationwitha highfrequencyofnoninvasivetesting.
Wewerealsoabletoassessdifferencesbetweennon- invasive tests.It should benoted that ourfindingsare mainlythe result ofusing functionaltests,particularly exerciseECGandstressSPECT,whichaccountedformore than90%oftesting,asgatekeepersforICA.Althoughthe proportion of obstructive CAD washigher in theCCTA groupthanforfunctionaltests,itis uncertainwhether theoverallresultswouldhavebeendifferentifanatomic testshadbeenusedmorefrequently.Thereissomeevi- dence that CCTA may be a useful and cost-effective gatekeeperforICA(particularlyinpatientswithinterme- diatetolow pretest probability),reducing the number ofpatients without obstructive CAD referred for inva- sivetesting.12---19Recent guidelinesfor themanagement
ofpatients with chest painfrom theUnited Kingdom’s NationalInstituteforHealthandClinicalExcellence(NICE) recommendchoosingtestsaccordingtothepretestprob- ability ofCAD. Functional imaging tests are preferred forpatientswith 30---60%pretestprobabilityofdisease, whereasCCTA(precededbycalciumscore)isthepreferred method for patients with 10---29% pretest probability.20
Accordingtotheseguidelines,ICAshould beofferedas thefirsttest topatients withpretest probabilitiesover 60%.Currently,thereisdisagreementoverwhichtypeof testshouldbe usedasfirstline.21Results fromtheUS
NationalInstitutesofHealth-sponsoredPROMISEstudy(a clinicalendpoint-drivenrandomizedstudycomparingfunc- tionalstudies with CCTA forthe evaluationofpatients withsuspectedCAD)willhopefullyshedmorelightonthis matter.22,23
Studylimitations
Severallimitationsofthisstudyshouldbeacknowledged. Sincethecharacteristicsofchestpainwerenotsystem- aticallyassessedandrecordedfor eachpatient,it was notpossibletocalculatethepretestprobabilityofCAD. Althoughthemedianageandprevalenceofriskfactorsare compatiblewithatypicalCADriskpopulation,itisnotpos- sibletoascertainwhethertheweakpredictivepowerof noninvasivetestingisrelatedtoitsapplicationtoapopula- tionwithlowpretestprobability.Anotherpitfallisrelatedto thedichotomizedclassificationofnoninvasivetestsaspos- itiveornegativeforobstructiveCAD.Inmostteststhereis acontinuumof‘positivity’whichisdifficulttoaddresswith thisstudydesign.Itshouldalsobeemphasizedthatthiswas notarandomizedtrialofnoninvasivetestingand,assuch, directcomparisonsoftestingvs.notestingandcomparisons betweennoninvasivemodalitiesshouldbeinterpretedwith caution.Thediagnosticperformanceofnoninvasivetesting isdependentonthepretestprobabilityofdisease,andthe decisiontoperformnoninvasivetestingandthechoiceofthe testitselfdependonthephysician’sperceptionofpretest probability,whichmayhavedifferedbetweenthedifferent diagnosticmodalitiesapplied.
Conclusions
NearlyhalfofpatientswithoutknownCADundergoingelec- tiveICAduetochestpaindidnothaveobstructivelesions, eventhoughfouroutoffivehadapositivenoninvasivetest. Functionaltestswerebyfarthemostcommonlyusedgate- keepersbutwererelativelyweakpredictorsofobstructive CADandappeartobeoutperformedbyCCTA.Thereiscon- siderableroomforimprovingthecurrentreferralstrategy forICA.
Ethicaldisclosures
Protectionofhumanandanimalsubjects.Theauthors
declarethatnoexperimentswereperformedonhumansor animalsforthisstudy.
Confidentialityofdata.Theauthorsdeclarethattheyhave followedtheprotocolsoftheirworkcenteronthepublica- tionofpatientdataandthatallthepatientsincludedinthe studyreceivedsufficientinformationandgavetheirwritten informedconsenttoparticipateinthestudy.
Right to privacy and informed consent.The authors
declarethatnopatientdataappearinthisarticle.
Conflictsofinterest
Pleasecitethisarticleinpressas:BorgesSantosM,etal.Diagnosticyieldofcurrentreferralstrategiesforelective coronaryangiographyinsuspectedcoronaryarterydisease----AnanalysisoftheACROSSregistry.RevPortCardiol.2013.
http://dx.doi.org/10.1016/j.repc.2012.11.008
ARTICLE IN PRESS
+ModelREPC-286; No.ofPages6
6 M.BorgesSantosetal.
References
1.GibbonsRJ,AbramsJ,ChatterjeeK,etal.ACC/AHA2002guide-
lineupdateforthemanagementofpatientswithchronicstable
angina-summaryarticle:areportoftheAmericanCollegeof
Cardiology/AmericanHeartAssociationTaskForceonPractice
Guidelines(CommitteeontheManagementofPatientsWith
ChronicStableAngina).Circulation.2003;107:149---58.
2.FoxK,GarciaMA,ArdissinoD,etal.Guidelinesontheman-
agementofstableanginapectoris:executivesummary:the
TaskForceontheManagementofStableAnginaPectorisofthe
EuropeanSocietyofCardiology.EurHeartJ.2006;27:1341---81.
3.WijnsW,KolhP,DanchinN,etal.Guidelinesonmyocardial
revascularization.EurHeartJ.2010;31:2501---55.
4.BashoreTM,BatesER,BergerPB,etal.AmericanCollegeof
Cardiology/SocietyforCardiacAngiographyandInterventions
ClinicalExpertConsensusDocumentoncardiaccatheterization
laboratorystandards.AreportoftheAmericanCollegeofCar-
diologyTaskForceonClinicalExpertConsensusDocuments.J
AmCollCardiol.2001;37:2170---214.
5.PatelMR,PetersonED,DaiD,etal.Lowdiagnosticyieldof
electivecoronaryangiography.NEnglJMed.2010;362:886---95.
6.GendersTS,SteyerbergEW,AlkadhiH,etal.Aclinicalpredic-
tionruleforthediagnosisofcoronaryarterydisease:validation,
updating,andextension.EurHeartJ.2011;32:1316---30.
7.DiamondGA,KaulS.Lowdiagnosticyieldofelectivecoronary
angiography.NEnglJMed.2010;363:93,authorreply4---5.
8.ChengVY,BermanDS,RozanskiA,etal.Performanceofthe
traditionalage,sex,andanginatypicality-basedapproachfor
estimatingpretestprobabilityofangiographicallysignificant