xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID ERHUW3831_DROS9F INGEST_TIME 2017-06-28T21:15:14Z PACKAGE IR00009846_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
PAGE 1
EarlyInvasiveStrategyandIn-HospitalSurvivalAmongDiabeticsWith Non-ST-ElevationAcuteCoronarySyndromes:AContemporary NationalInsightAhmedN.Mahmoud,MD;*IslamY.Elgendy,MD;*HendMansoor,PharmD,MS;XuerongWen,PhD,MPH,MS;MohammadK.Mojadidi, MD;AnthonyA.Bavry,MD,MPH;R.DavidAnderson,MD,MSBackground- Therearelimiteddataonthemeritsofanearlyinvasivestrategyindiabeticswithnon-ST-elevationacutecoronary syndrome,withunclearin uenceofthisstrategyonsurvival.Theaimofthisstudywastoevaluatethein-hospitalsurvivalof diabeticswithnon-ST-elevationacutecoronarysyndrometreatedwithanearlyinvasivestrategycomparedwithaninitial conservativestrategy. MethodsandResults- TheNationalInpatientSampledatabase,years2012 2013,wasqueriedfordiabeticswithaprimary diagnosisofnon-ST-elevationacutecoronarysyndromede nedaseithernon-ST-elevationmyocardialinfarctionorunstableangina (unstableangina).Anearlyinvasivestrategywasde nedascoronaryangiography revascularizationwithin48hoursofadmission. Propensityscoreswereusedtoassembleacohortmanagedwitheitheranearlyinvasiveorinitialconservativestrategybalanced on > 50baselinecharacteristicsandhospitalpresentations.Incidenceofin-hospitalmortalitywascomparedinbothgroups.Ina cohortof363500diabeticswithnon-ST-elevationacutecoronarysyndrome,164740(45.3%)weretreatedwithanearlyinvasive strategy.Propensityscoringmatched21681diabeticsinbotharms.Incidenceofin-hospitalmortalitywaslowerwithanearly invasivestrategyinboththeunadjusted(2.0%vs4.8%;oddsratio[OR],0.41;95%CI,0.39 0.42; P < 0.0001)andpropensitymatchedmodels(2.2%vs3.8%;OR,0.57;95%CI,0.50 0.63; P < 0.0001).Thebene twasobservedacrossvarioussubgroups, exceptforpatientswithunstableangina( Pinteraction= 0.02). Conclusions- Anearlyinvasivestrategymaybeassociatedwithalowerincidenceofin-hospitalmortalityinpatientswith diabetes.Thebene tofthisstrategyappearstobesuperiorinpatientspresentingwithnon-ST-elevationmyocardialinfarction comparedwithunstableangina. ( JAmHeartAssoc .2017;6:e005369.DOI:10.1161/JAHA.116.005369) KeyWords: acutecoronarysyndrome earlyinvasivestrategy mortality propensityscoreanalysis Diabetesmellitus(DM)isarapidlygrowingglobalhealth burden.In2014,theprevalenceofDMwasestimatedto be422millionworldwide,doublinginfrequencysince1980.1IntheUnitedStates,theprevalenceofDMincreased dramaticallyfrom3.5%in1990to8.3%in2012.2DMisa predominantriskfactorforatheroscleroticcoronaryartery disease(CAD)andacutemyocardialinfarction(MI).3,4DiabeticshaveahigherincidenceofmultivesselCAD, acceleratedatherosclerosis,atheroscleroticplaquerupture, andincreasedplateletactivity,allofwhichincreasethe incidenceofacuteMIcomparedtonondiabetics.4Additionally,DMisindependentlyassociatedwithahigherincidence ofearlyandlatemortalityfollowinganon-ST-elevationacute coronarysyndrome(NSTE-ACS).5,6Thereiscurrentlyapaucityofdataonthebene tofan earlyinvasivestrategyindiabeticpatientswithNSTE-ACS, FromtheDivisionofCardiovascularMedicine,DepartmentsofMedicine(A.N.M.,I.Y.E.,M.K.M.,A.A.B.,R.D.A.)andDepartmentofPharmaceuticalOutcomesandPolicy (H.M.),UniversityofFlorida,Gainesville,FL;CollegeofPharmacy,UniversityofRhodeIsland,Kingston,RI(X.W.);CardiologySection(111D),NorthFlorida/South GeorgiaVeteransHealthSystemMalcomRandallVeteransAdministrationMedicalCenterMedicalService,Gainesville,FL(A.A.B.). AnaccompanyingTableS1isavailableathttp://jaha.ahajournals.org/content/6/3/e005369/DC1/embed/inline-supplementary-material-1.pdf *DrMahmoudandDrElgendycontributedequallytothisarticle. AnabstractofthecurrentstudywasacceptedforpresentationattheAmericanCollegeofCardiology66thAnnualScienti cSession&Expo,March17 19,2017,in Washington,DC. Correspondenceto: R.DavidAnderson,MD,MS,UniversityofFloridaHealthScienceCenter,1600SWArcherRd,POBox100277,Gainesville,FL32610. E-mail:david.anderson@medicine.u .edu ReceivedDecember18,2016;acceptedFebruary1,2017. 2017TheAuthors.PublishedonbehalfoftheAmericanHeartAssociation,Inc.,byWileyBlackwell.ThisisanopenaccessarticleunderthetermsoftheCreative CommonsAttributionLicense,whichpermitsuse,distributionandreproductioninanymedium,providedtheoriginalworkisproperlycited. DOI:10.1161/JAHA.116.005369 JournaloftheAmericanHeartAssociation1 ORIGINALRESEARCH by guest on May 12, 2017 http://jaha.ahajournals.org/ Downloaded from
PAGE 2
giventhelimitednumberofdiabeticpatientsenrolledinmost randomized,clinicaltrials.7Multipleobservationalstudiesand meta-analysesofrandomizedtrialsdemonstratethebene tof anearlyinvasivestrategyinmanagementofdiabeticNSTEACSpatients,mainlythroughreductionofcompositeclinical eventratesandMI.8,9Asadirectconsequence,theAmerican CollegeofCardiologyFoundation/AmericanHeartAssociation(ACCF/AHA)TaskForcerecommendsaroutineinvasive strategy(within72hoursofhospitalization)fordiabetics presentingwithNSTE-ACS.10However,evidencefromreal-life registrydataindicatesthatdiabeticsarelessfrequently offeredacutereperfusiontherapyoracuterevascularization comparedtonondiabetics.6,11,12Todate,thereisinsuf cientevidencetosupportasurvival bene tofanearlyinvasivestrategyindiabeticpatientswith NSTE-ACS.Theaimofthisstudyistoevaluatetheeffectofan earlyinvasivestrategyinNSTE-ACSdiabeticpatients,with emphasisonsurvival,lengthofhospitalstay,andcost.MethodsStudyDataSourcesTheNationalInpatientSample(NIS)database(years2012and 2013)wasusedtocollectdataforthecurrentstudy.Years 2012and2013werechosenbecausethosewerethemost recentNISdatabasereleasesatthetimethisstudywas conductedandthuswouldmostcloselyre ectcontemporary managementofNSTE-ACS.In2012,theNISwasredesignedto includearandomsampleofpatientdischargesfromall hospitals,ratherthanarandomsampleofhospitalsretaining theirdischarges,whichresultedinmore-precisenational estimates.TheNISisthelargestall-payerdatabaseinthe UnitedStatesavailableforpublicuseandsponsoredbythe AgencyforHealthcareResearchandQuality(AHRQ)asapartof theHealthcareCostandUtilizationProject.13Itrepresentsa 20%strati edsampleofalldischargesfromtheUScommunity hospitals.Eachindividualpatienthospitalizationisde-identi ed andmaintainedasaspeci centry.DataavailableintheNIS include1primaryand24secondarydiagnoses,25procedure diagnoses(inInternationalClassi cationofDiseases,Ninth Edition,ClinicalModi cation[ICD-9-CM]codingformat), patientdemographiccharacteristics(eg,sex,age,race,and medianhouseholdincome),hospitalcharacteristics(eg,ownership),expectedpaymentsource,totalhospitalcharges, dischargestatus,lengthofhospitalstay,aswellas,severity andcomorbiditymeasures.Dischargeweightsarealsoavailableforeachpatient srecord,whichcouldbeusedinnational weightedestimatesgeneration.Thisstudywasdeemedtobe exemptfromtheInstitutionalReviewBoard(IRB)becauseof thepublicnatureoftheNISdatabase,withtheabsenceofany personalidentifyinginformation.ValidationandDataControlTheNISperformsannualdataqualityassessments,toensure theinternalvalidlyofitsdata.Ithadbeenpreviously comparedwithalternativedatabases,suchasAmerican HospitalAssociationAnnualSurveyDatabase,theNational HospitalDischargeSurveyfromtheNationalCenterforHealth Statistics,andtheMed-PARinpatientdatabasefromCenters forMedicareandMedicaidServices,withcomparable estimatestoallofthepreviouslystateddatabases.14PatientPopulationAllpatientswithaprimarydiagnosiscodeofnon-ST-elevation myocardialinfarction(NSTEMI)(ICD-9-CMcodeof410.7x)or unstableangina(UA;ICD-CM9codeof411.1)andsecondary diagnosisofdiabetesmellitus(accordingtotheElixhauser comorbiditysoftwarede nedbytheAHRQ)wereincluded.15Theanalysiswaslimitedtopatientswiththeprimarydiagnosis ofNSTE-ACSasitisusuallyconsideredthemainreasonfor admissionintheNISdatabase.Thiswouldallowtheexclusion ofpatientswithtype2NSTEMIsecondarytootherconcomitant diseases.Anearlyinvasivestrategygroupwasde nedas coronaryangiography(ICD-9-CMcodesof88.55,37.22,or 37.23)withorwithoutrevascularization,thatis,percutaneous coronaryintervention(PCI;ICD-9-CMcodesof00.66,36.01, 36.02,36.05,36.06,and36.07)orcoronaryarterybypass grafting(CABG;ICD-9-CMcode36.1x)withproceduretime beingwithin48hoursofadmission(ie,day0or1).The remainingpatientswerede nedasaninitialconservative strategygroup.Thisde nitionofanearlyinvasivestrategyin NSTE-ACShaspreviouslybeenusedforotherstudiesutilizing datafromtheNISdatabase.16,17StudyOutcomesThemainoutcomeofthisstudywasin-hospitalmortality, referredtointheNISdatabaseasthe DIED variable.The mainoutcomewascomparedinboththeearlyinvasiveand initialconservativestrategygroupsafteradjustingformultiple patientandhospitalcharacteristics.Otheroutcomesof interestwerelengthofhospitalstay,referredintheNIS databaseas LOS variableandtotalhospitalcharges, referencedasthe TOTCHG variable.Thetotalhospital chargesrepresentthetotalamountbilledbythehospitalfor serviceratherthantheactualpaymentreceived.PatientsandHospitalCharacteristicsDatavariablescollectedwerepatients demographics,includingage,sex,race,medianhouseholdincomebyZIPcode, weekendversusweekdayadmission,andprimaryexpected payer(Medicare,Medicaid,Privateinsurance,Uninsured,or DOI:10.1161/JAHA.116.005369 JournaloftheAmericanHeartAssociation2 EarlyInvasiveStrategyinDiabeticsWithNSTE-ACS MahmoudetalORIGINALRESEARCH by guest on May 12, 2017 http://jaha.ahajournals.org/ Downloaded from
PAGE 3
Other)andtheElixhauserlistofcomorbidities.Otherrelevant diagnoseswerealsoextractedbytheircorrespondingICD-9CMcodes,includingacuteischemicstroke,intracranial hemorrhage,gastrointestinalbleeding,cardiogenicshock, familyhistoryofcoronaryarterydisease(CAD),historyof pastMI,pastPCIorCABG,paststrokeortransientischemic attack(TIA),CAD,carotidarterydisease,smoking,dyslipidemia,atrial brillation,anddementia.Hospitalcharacteristics,suchasbedsize(small,medium,andlarge),urban location,andteachingstatus,werealsocollected.18Afulllist ofICD-9-CMcodesforallvariablescollectedinthecurrent studyissuppliedinTable1.StatisticalAnalysisWeightednationalestimateswerecalculatedusingthe dischargeweightssuppliedbytheNIS.Frequenciesand percentageswereusedforestimationofcategoricalpatient andhospitalcharacteristicsandmeanswithSDormedians, with25thto75thpercentileranges,forcontinuousones. CategoricalvariableswerecomparedbyPearson schi-square test,andanindependent-sampleStudent t testwasusedfor meanscomparisoninbothgroups.Medianswerecompared usingMood smediantest.Amultivariablelogisticregression modelwasconstructedusingallpreviouslystatedpatientand hospitalcharacteristicsasindependentvariables,withearly invasivestrategybeingthedependentvariable.Theresultant individuallymatchedprobabilityscorewasthenusedfor propensityscorematchingof2similargroups(anearly invasivestrategyvsinitialconservativestrategy)with1:1ratio andmatchtoleranceof0.01.Aclinicallysigni cantdifference betweenbothgroupswasconsideredpresentiftheabsolute differenceinfrequencyormeanswas > 5%postmatching.The propensity-matchedcohortofpatientswasthenusedto comparetheincidenceofdifferentoutcomesofinterest.The oddsratio(OR)ofin-hospitalmortalitywascalculatedinthe propensity-matcheddatausingabinaryconditionallogistic regressionmodel. AsubgroupanalysiswasconductedcomparingtheORof in-hospitalmortalityaccordingtotheuseofanearlyinvasive strategyacrossvariousprede nedsubsets,including NSTEMIversusUA,historyofcongestiveheartfailure (CHF),presenceofcardiogenicshock,malesversusfemales, andpasthistoryofMI,CAD,orrevascularization(PCIor CABG).Asecondarypropensityadjustedmultivariablelogisticregressionanalysiswasalsocalculatedforin-hospital mortalitywithallpreviouslystatedvariablesalongwiththe useofearlyinvasivestrategyandthepropensityscoreas independentvariables.Theregressionwasperformedbya backwardstep-wiseapproachwithacut-offlevelof0.05for entryand0.1forremoval,followedbycalculationofan adjustedORforin-hospitalmortality. Totakeintoaccountthepossibilityofanimmortality timebias,asensitivityanalysiswasconductedafter exclusionofNSTE-ACSpatientswhohadalengthofhospital staylessthan48hours(ie,day0and1)inthepropensitymatchedcohort.Immortaltimebiasreferstoaperiodof follow-upduringwhich,bydesign,thestudyoutcomecannot occur.19Limitingtheanalysistopatients,whohadalength ofhospitalstaymorethan48hours,wasanindirectmethod ofanalyzingpatientswhosurvivedforatleast48hours,as thetimeofdeathwasnotareportedvariableintheNIS database. Finally,anotherpropensityscoreanalysiswasconducted usingatightermatchtoleranceof10e 5tocon rmthe ndingsoftheprimaryanalysis,becauseastatistically signi cantdifferenceinthefrequencyofsomecategorical variableswasdetectedintheprimaryanalysis.Allstatistical analyseswereperformedusingIBMSPSSStatisticssoftware (version23.0;IBMCorp.,Armonk,NY)with2-sided P valueof < 0.05forstatisticalsigni canceassessmentforallanalyses andORwith95%CIasameasureofeffectsizereportedby logisticregression. Table1. InternationalClassi cationofDiseases,Ninth Edition,ClinicalModi cation(ICD-9-CM)Codesofthe VariablesIncludedinthePropensityScoreMatching*Variable ICD-9-CMCodeAcuteischemicstroke433.01,433.11,433.21,433.31, 433.81433.91,434.01,434.11, 434.91,435.x,436 Intracranialhemorrhage430,431,432.x Gastrointestinalbleeding153 Cardiogenicshock785.51 FamilyhistoryofCADV17.3 PastMI412 PastPCIV45.82 PastCABGV45.81 Paststroke/TIAV12.54 CAD414.00,414.01,414.02,414.03,414.04, 414.05,414.06,414.07 Carotidarterydisease433.10 SmokinghistoryV15.82,305.1 Dyslipidemia53 Atrialfibrillation427.31 Dementia290.xx,294.1x,294.2x,294.8, 331.0 331.12,331.82,797CABGindicatescoronaryarterybypassgrafting;CAD,coronaryarterydisease;MI, myocardialinfarction;PCI,percutaneouscoronaryintervention;TIA,transientischemic attack. *OthervariablesnotreportedinthetablewerecollectedusingElixhauserComorbidity Software.15 DOI:10.1161/JAHA.116.005369 JournaloftheAmericanHeartAssociation3 EarlyInvasiveStrategyinDiabeticsWithNSTE-ACS MahmoudetalORIGINALRESEARCH by guest on May 12, 2017 http://jaha.ahajournals.org/ Downloaded from
PAGE 4
ResultsBaselineCharacteristicsWeidenti ed363500diabeticpatientswhowereadmitted intheUnitedStateswithaprimarydiagnosisofNSTE-ACS (meanage,68.1 12.6years;42.9%female;64.5%being white;65.5%coveredbyMedicare)duringtheyears2012 2013.Mostofthepatientswerehypertensive(82.8%)and hadapasthistoryofCAD(77.9%).Atotalof229435 patients(63.1%)underwentaninvasivestrategy(earlyor delayed)duringtheiradmission(coronaryangiography,PCI, orCABG).Ofthetotalpatientpopulation,164740patients (45.3%)hadanearlyinvasivestrategyperformedduring theirhospitalstay.Totalincidenceofin-hospitalmortality was12925patients(3.9%).Patientandhospitalcharacteristicsoftheincludedpatientsaresummarizedin Table2. Comparedwithaninitialconservativestrategy,patients undergoinganearlyinvasivestrategywereyounger(65.2 [SD = 11.8]vs70.6[SD = 12.8]years; P < 0.0001),lessfrequentlyfemale(38.9%vs46.3%; P < 0.0001),andhadfewer comorbidities(Table2).Theuseofanearlyinvasivestrategy wasmorecommoninlarge-bed-sizehospitalsandinurban teachinghospitals(Table2).Patientscarryingadiagnosisof UAwerelesslikelytoundergoanearlyinvasivestrategy comparedtothosewithNSTEMI(21.3%vs45.3%; P < 0.0001). Thefrequencyofrevascularizationwashigherintheearly invasivestrategygroupbybothPCI(96.9%vs30.1%; P < 0.0001)andCABG(15.6%vs5.9%; P < 0.0001)(Table2). Nonetheless,allpatientsintheinitialconservativestrategy groupwhohadacoronaryangiographyperformedhada revascularizationprocedure(i.e.PCIand/orCABG)aswell. Propensityscorematchingyielded21681patientsinboth groupswithasimilardistributionofpatientandhospital characteristicsbetweenthe2groupsofinterest,exceptfora fewcategoricalvariablessuchasexpectedprimarypayer ( P < 0.0001),race( P < 0.0001),medianhouseholdincome ( P = 0.02),andhospitallocation( P < 0.0001).However,all variableshadabsolutefrequencydifferenceslessthan5% andthusdeemedclinicallyinsigni cant(Table2).Tocon rm ourresults,asecondarypropensityscoreanalysiswas constructedwithatightermatchtoleranceof10e 5that yielded12363patientsintheearlyinvasivestrategygroupand 12367intheinitialconservativeapproachgroup,matchedin alltheformerlystatedcategoricalvariables(TableS1).In-HospitalMortalityTheincidenceofin-hospitalmortalitywaslowerwithanearly invasivecomparedwithaninitialconservativestrategyinthe unadjustedcohort(2.0%vs4.8%;OR,0.41;95%CI,0.39 0.42; P < 0.0001).Thisbene twasmaintainedinthe propensity-matchedcohort(2.2%vs3.8%;OR,0.57;95%CI, 0.50 0.63; P < 0.0001)andinthepropensity-adjustedmultivariablelogisticregressionanalysis(OR,0.54;95%CI,0.52 0.57; P < 0.0001;Figure1).Theincidenceofin-hospital mortalityalsowaslowerwithanearlyinvasivestrategyin thesecondarypost-hocanalysisusingatightermatch tolerance(2.5%vs3.7%;OR,0.65;95%CI,0.56 0.75; P < 0.0001)andinthesensitivityanalysisafterexcludingthe patientswithlengthofhospitalstaylessthan48hoursinthe propensity-matchedcohort(2.1vs3.3;OR,0.63;95%CI, 0.56 0.72; P < 0.0001). Onsubgroupanalysis,thebene tofanearlyinvasive strategywasdemonstratedamongawiderangeofprespeci edsubgroupsexceptinpatientswithUA,wheretherewas noapparentevidenceofsurvivalbene twithanearlyinvasive strategy(0.5%vs0.1%;OR,7.86;95%CI,0.82 75.72; P = 0.07),withevidenceofheterogeneitywhencomparedto NSTEMIpatients( Pinteraction= 0.02;Figure2).OtherOutcomesAnearlyinvasivestrategywasassociatedwithashorter lengthofhospitalstaycomparedwithaninitialconservative strategywithamedianof3(2 7)versus4days(2 7), respectively( P < 0.0001),andhighertotalhospitalcharges withamedianof66042USdollars($40315 112895)versus 39265$(19193 77832$),respectively( P < 0.0001; Figure1).DiscussionInthecurrentpropensity-matchedanalysisofcontemporary real-lifedata,anearlyinvasivestrategywasassociatedwith anincreasedin-hospitalsurvivalinNSTE-ACSpatientswith concomitantDM.Theseresultssupportthe2014ACCF/AHA guidelinerecommendationsforanearlyinvasivestrategyin diabetics,especiallythosewithhigh-riskfeatures(eg,NSTEMI andcardiogenicshock).10Meanwhile,theuseofthisstrategy inlowerriskpatients,suchasthosewithUA,maynotbe associatedwithimprovedsurvival. Thesurvivalbene tofanearlyinvasivestrategyinthe NSTE-ACSpopulationremainsamatterofongoingdebate.20 22Whereasnoneofthelandmarktrialscomparinganearly invasivewithaninitialconservativestrategyillustrateda statisticallysigni cantreductioninmortality,thesetrialswere notstatisticallypoweredtoanswerthatquestion.21 24We calculatedtheminimalsamplesizerequiredbyarandomized trialtodetectthedifferenceinproportionsillustratedinour studyandfoundthatthesamplesizeofalmost3500patients wouldbenecessarytoobtainthesameresults,25,26whichis morethandoublethenumberofdiabeticNSTE-ACSpatients DOI:10.1161/JAHA.116.005369 JournaloftheAmericanHeartAssociation4 EarlyInvasiveStrategyinDiabeticsWithNSTE-ACS MahmoudetalORIGINALRESEARCH by guest on May 12, 2017 http://jaha.ahajournals.org/ Downloaded from
PAGE 5
Table2. PatientandHospitalCharacteristicsofBoththeUnmatchedandPropensity-Score MatchedCohortofPatientsVariable UnmatchedPatients P Value Propensity-ScoreMatched P Value* EarlyInvasive(%)InitialConservative(%)EarlyInvasive(%)InitialConservative(%)Totalnumberofpatients 164740(100)198760(100) 21681(100)21681(100) Primarydiagnosis NSTEMI161020(98)185055(93) <0.000121130(98)20247(93) <0.0001 UA3720(2)13705(7) <0.0001550(3) 1434(7) <0.0001 Patientdemographics Age,meany(SD) 65(12) 71(13) <0.000167(11) 68(13) 0.09 Femalesex 64130(39)91945(46) <0.00019044(42)9176(42) 0.20 Race <0.0001 <0.0001* White 109040(70)128515(69) 15063(70)14983(69) Black 20000(13)26500(14) 2768(13)3163(15) Hispanic 16295(10)20015(11) 2227(10)2093(10) AsianorPacificIslander4400(3)5660(3) 606(3) 577(3) Other 6110(4)5655(3) 868(4) 710(3) Primaryexpectedpayer <0.0001 0.02* Medicare 95625(58)142420(72) 14157(65)14262(66) Medicaid 12845(8)14590(7) 1477(7)1798(8) Privateinsurance 39630(24)28845(14) 4327(20)3902(18) Uninsured 10160(6)7390(3) 1037(5)1075(5) Other 5110(3)4565(2) 581(3) 554(3) Weekendadmission 33325(20)59330(30) <0.00015279(24)5379(25) 0.27 Householdincome(median) <0.0001 <0.0001* 0to25thpercentile 53420(33)66370(34) 7384(34)7280(34) 26to50thpercentile43660(27)51265(26) 5821(27)5668(26) 51to75thpercentile36995(23)43555(23) 4895(23)4921(23) 76to100thpercentile27175(17)32760(17) 3581(17)3812(18) Patientcharacteristics Smoking 54065(33)44355(22) <0.00016067(28)5901(27) 0.08 Dyslipidemia 125075(76)131195(66) <0.000115786(73)15666(72) 0.20 Obesity 45060(27)43055(21) <0.00015449(25)5369(25) 0.38 KnownhistoryofCAD 146755(89)136395(69) <0.000118399(85)18493(85) 0.21 FamilyhistoryofCAD 13100(8)7365(4) <0.00011148(5)1115(5) 0.48 PastMI 20705(13)25895(13) <0.00012922(14)3011(14) 0.21 PastPCI 137975(84)172170(87) <0.00013519(16)3548(16) 0.71 PastCABG 12640(8)24895(13) <0.00012178(10)2269(11) 0.15 PaststrokeorTIA 9150(6)13390(7) <0.00011316(6)1321(6) 0.92 Carotidarterydisease 4080(3)4750(2) 0.09573(3) 575(3) 0.95 Peripheralvasculardisease25335(15)36090(18) <0.00013819(18)3817(18) 0.98 Pulmonarycirculationdisease145( <1)275( <1) <0.000121( <1)29( <1) 0.26 Dementia 3000(2)15350(8) <0.0001562(3) 508(2) 0.10 Atrialfibrillation 22725(14)39970(20) <0.00013536(16)3578(17) 0.59 Alcoholabuse 3525(2)3830(2) <0.0001454(2) 434(2) 0.50 Deficiencyanemia 32325(20)55310(28) <0.00015062(23)5101(24) 0.66Continued DOI:10.1161/JAHA.116.005369 JournaloftheAmericanHeartAssociation5 EarlyInvasiveStrategyinDiabeticsWithNSTE-ACS MahmoudetalORIGINALRESEARCH by guest on May 12, 2017 http://jaha.ahajournals.org/ Downloaded from
PAGE 6
Table2. ContinuedVariable UnmatchedPatients P Value Propensity-ScoreMatched P Value* EarlyInvasive(%)InitialConservative(%)EarlyInvasive(%)InitialConservative(%) Collagenvasculardisease3505(2)4410(2) 0.06480(2) 471(2) 0.79 Chronicbloodlossanemia1295(1)2085(1) <0.0001197(1) 200(1) 0.88 Congestiveheartfailure690( <1)2005(1) <0.0001124( <1)145( <1) 0.20 Valvulardisease 195( <1)615( <1) <0.000136( <1)47( <1) 0.23 Chronicpulmonarydisease35540(22)52430(26) <0.00015252(24)5322(25) 0.43 Coagulopathy 9480(6)11610(6) 0.271313(6)1280(6) 0.50 Liverdisease 2895(2)4385(2) <0.0001420(2) 416(2) 0.89 Renaldisease(chronic)41815(25)78900(40) <0.00016967(32)7005(32) 0.70 Electrolytesabnormalities34070(21)56890(29) <0.00015170(24)5203(24) 0.71 AIDS 120( <1)175( <1) 0.1216( <1)14( <1) 0.72 Drugabuse 3050(2)3425(2) <0.0001392(2) 394(2) 0.94 Depression 14495(9)19320(10) <0.00011971(9)2040(9) 0.25 Hypertension 138115(84)162920(82) <0.000118088(83)18086(83) 0.98 Hypothyroidism 19215(12)29180(15) <0.00012801(13)2819(13) 0.80 Lymphoma 705( <1)1240(0.6) <0.0001110( <1)114( <1) 0.79 Metastaticcancer 625( <1)2050(1) <0.0001114( <1)106( <1) 0.59 Solidtumorwithoutmetastasis1720(1)3480(2) <0.0001289(1) 295(1) 0.80 Otherneurologicaldisorder7385(5)16065(8) <0.00011142(5)1192(6) 0.29 Paralysis 2735(2)5445(3) <0.0001431(2) 457(2) 0.38 Psychoses 4040(3)6860(4) <0.0001622(3) 597(3) 0.47 Pepticulcer(nonbleeding)35( <1)35( <1) 0.434( <1) 2( <1) 0.41 Weightloss 3330(2)6725(3) <0.0001542(3) 518(2) 0.46 Cardiogenicshock 6100(4)5220(3) < 0.0001678(3) 654(3) 0.50 Intracranial hemorrhage150( <1)255( <1) <0.000121( <1)24( <1) 0.66 Acuteischemicstroke 2315(1)3185(2) <0.0001328(2) 337(2) 0.73 Gastrointestinalbleeding2545(2)5530(3) <0.0001433(2) 455(2) 0.46 Invasiveprocedure PCI159585(97)59810(30) <0.000110415(48)3862(18) <0.0001 CABG25650(16)11760(6) <0.00013037(14)1820(8) <0.0001 Hospitalcharacteristics Hospitalbedsize <0.0001 0.05 Small 13150(8)26530(13) 2062(10)2203(10) Medium 40040(24)54740(28) 5568(26)5553(26) Large 111550(68)117490(59) 14051(65)13925(64) Hospitallocation <0.0001 0.14 Urbanteaching 92320(56)91430(46) 11133(51)11532(53) Urbannonteaching 61130(37)81580(41) 8811(41)8213(38) Rural 11290(7)25750(13) 1737(8)1936(9) In-hospitalmortality 3320(2.0)9605(4.8) <0.0001475(2.2)826(3.8) <0.0001Allpercentagesareapproximatedtothenearestinteger.AIDSindicatesacquiredimmunede ciencysyndrome;CABG,coronaryarterybypassgraftsurgery;CAD,coronaryarterydisease; NSTEMI,non-ST-elevationmyocardialinfarction;MI,myocardialinfarction;PCI,percutaneouscoronaryintervention;TIA,transientischemicattack;UA,unstableangina. *Although P valueissigni cantthedifferencein%betweenbotharmswas < 5%andthusdeemedclinicallynonsigni cant.Variableswerenotincludedinthepropensityscorematching. DOI:10.1161/JAHA.116.005369 JournaloftheAmericanHeartAssociation6 EarlyInvasiveStrategyinDiabeticsWithNSTE-ACS MahmoudetalORIGINALRESEARCH by guest on May 12, 2017 http://jaha.ahajournals.org/ Downloaded from
PAGE 7
includedinallprevioustrialscombined.9Althoughourresults suggestanassociationbetweenanearlyinvasivestrategyand improvedsurvivalindiabeticswithNSTE-ACS,thisdoesnot establishcausality,giventheretrospectivenatureofthedata; thus,largerrandomized,clinicaltrialspoweredfordifferentiationbetweenshort-andlong-termmortalityarenecessary tocon rmour ndings. Earliermeta-analysesofrandomizedtrialshavedemonstratedapossiblereductioninall-causemortalityfollowingan earlyinvasivestrategy,withameanfollow-upof1.5to 2yearsaftertheinitialevent.20,22However,subsequent meta-analysesfailedtoreplicatethesameresults.7,27Ametaanalysisof9904patientswithNSTE-ACSfrom9randomized trialsinvestigatedthebene tofanearlyinvasivestrategyuse indiabetics(17%ofthetotalpopulation)comparedto nondiabetics.9Althoughthestudydidnotshowanyadded mortalitybene twithanearlyinvasivestrategy,itshoweda reductionintherateofnonfatalMI.9Interestingly,thebene tofanearlyinvasivestrategywas leastevidentinthesubgroupofdiabeticsdiagnosedprimarily withUA.Thisisconsistentwithpreviousevidenceshowinga lackofbene tofanearlyinvasivestrategyinlow-riskNSTEACSpatientswithlowtroponinlevels28andsupports choosinganinitialconservativestrategyasavalidoptionin thissubsetofpatients.Another ndingwastheinfrequentuse ofanearlyinvasivestrategyinhigh-riskdiabeticpatients admittedwithNSTE-ACSandcardiogenicshock.Only,54%of thepatientswithcardiogenicshockunderwentanearly invasivestrategy,however,thosepatientsweretheoneswho bene tedthemostfromthisstrategywiththelowestoddsof in-hospitalmortalitycomparedwithbothconservatively managedcardiogenicshockpatientsandnoncardiogenic shockpatientsmanagedbyanearlyinvasivestrategy. Asigni canteffectmodi cationin uencedbytheadmission daywasevident,withlessbene tnotedwithanearlyinvasive strategy,ifthepatientswereadmittedduringweekends comparedwithweekdays( Pinteraction= 0.02).Althoughtheexact reasonsforsuchdifferencesarenotentirelyclear,itmaybe attributedtolackofareadilyavailablecatheterizationstaff togetherwithsomehumanfactors,suchassleepdeprivation andfatigueasobservedinpreviousstudies.29Despitethis,the incidenceofin-hospitalmortalityremainedsigni cantlylower withanearlyinvasivestrategywhencomparedwithaninitial conservativestrategyduringtheweekends(OR = 0.71;95%CI, 0.57 0.90; P < 0.0001). Eventhougharoutineinvasivestrategyisrecommended fordiabeticsbymajorcardiovascularguidelinecommittees,10,30nationalregistriesworldwidehaveillustratedthat thisstrategyisstillunderutilized.6,10Althoughtheexact explanationforthis ndingisunclear,ouranalysisindicates thatanearlyinvasivestrategywasusedmorefrequentlyin youngerpatientswithlesscomorbidityatbaseline,whichis usuallynotthecaseinmostdiabeticsatthetimeofNSTEACSpresentation. Althoughthede nitionoftimingforanearlyinvasive strategy(ie,within48hours)appearstobesomewhat Figure1.Aclustercolumngraphcomparingalloutcomesofinterestbetweenanearlyinvasiveandan initialconservativestrategy.LOS,lengthofhospitalstay;$,USdollars. Bothlengthofhospitalstayandtotalhospitalchargeswerederivedfromthepropensity-matchedcohortof patients. DOI:10.1161/JAHA.116.005369 JournaloftheAmericanHeartAssociation7 EarlyInvasiveStrategyinDiabeticsWithNSTE-ACS MahmoudetalORIGINALRESEARCH by guest on May 12, 2017 http://jaha.ahajournals.org/ Downloaded from
PAGE 8
differentinthecurrentstudycomparedwiththeACC/AHA guidelines,thesamede nitionwasadoptedbythepivotal trialscomparinganearlyinvasivestrategytoaninitial conservativestrategy,suchasTreatAnginawithAggrastat andDetermineCostofTherapywithanInvasiveorConservativeStrategy ThrombolysisinMyocardialInfarction18 (TACTICS-TIMI18;4 48hours),31InvasiveversusConservativeTreatmentinUnstableCoronarySyndromes(ICTUS; 24 48hours)21andFRagminandFastRevascularisation duringInStabilityinCoronaryarterydiseaseII(FRISCII; 24 48hours)23trials.Aproceduretimeof24hourswasused mainlytode neanearlyinvasivestrategyintrialscomparing earlywithdelayedinvasivestrategy.32Thus,wepreferredto usethe48-hourde nitionofanearlyinvasivestrategytobe consistentwithprevioustrialscomparingearlyinvasiveto initialconservativestrategiesinNSTE-ACSpatients. Toourknowledge,thisstudyrepresentsthelargest contemporaryanalysiscomparinganearlyinvasivewithan initialconservativestrategyindiabeticswithNSTE-ACS. AlthoughtheNISisanadministrativedatabase,thelarge samplesizeofpatientsthatallowedadequatepowerto assessthemainoutcomeofinterest,togetherwiththe diversedemographicsoftheincludedpatients,largenumber ofpatients andhospitalcharacteristics,andvariousinhospitalcomplicationsrecorded,madethisdatabasean excellentsourcetoexploreourquestionofinterest.Multiple similartrendandoutcomestudieshavebeenpublishedusing theNISdatabase,whichvalidatesitsreliabilityinaddressing variouspracticeissues.14,16,17,33Despitethesestrengths,the currentstudyhassomelimitations. First,althoughweadjustedforover50independent variablesinthepropensity-matchedanalysis,thecurrent studyisretrospectiveinnatureandissubjecttobiases attributedtounmeasuredconfounders.Second,theinherent limitationoftheNISdatabase,suchaserrorincodingor misdiagnosis,mayhaveoccurred;however,giventhelarge Figure2.Forestplotrepresentingsubgroupanalysisofin-hospitalmortalityaccordingtovariouspatientriskfactors.ACSindicatesacute coronarysyndrome;CABG,coronaryarterybypassgraft;CAD,coronaryarterydisease;CHF,congestiveheartfailure;MI,myocardialinfarction; NSTEMI,non-ST-elevationmyocardialinfarction;OR,oddsratio;PCI,percutaneouscoronaryintervention. DOI:10.1161/JAHA.116.005369 JournaloftheAmericanHeartAssociation8 EarlyInvasiveStrategyinDiabeticsWithNSTE-ACS MahmoudetalORIGINALRESEARCH by guest on May 12, 2017 http://jaha.ahajournals.org/ Downloaded from
PAGE 9
patientsample,webelievethatsucherrorswouldbelimited andshouldnotaffecttheintegrityofourresults.Also,theNIS isanadministrativedatabasethatreliesmainlyonICD-9 codesratherthanclinicallyadjudicatedoutcomesordiagnoses.Third,becauseoftheadministrativenatureoftheNIS database,multipleclinicaldatawerenotaccountedfor,such asmedicationsadministered(eg,antiplateletand-coagulant therapy),laboratory(eg,troponinsandbrainnatriuretic peptidelevels),andimaging(eg,echocardiography)results. Thislimitedtheabilitytoassessthemeritsofanearly invasivestrategyinvariousrisksubgroups(eg,accordingto theThrombolysisinMyocardialInfarction[TIMI]score).We attemptedtoassesstheimpactofpastCADandpast revascularizationhistoryinoursubgroupanalysisasan indirectassessmentofpatientswithhigherTIMIscore,and therewasnoevidenceofeffectmodi cationbyeither.Lack ofdataregardingvarioustypesofmedicationsadministered isanimportantlimitation,giventhattheinabilitytoadministercertainmedications(eg,anticoagulantand-platelet agents)thatareknowntoimprovesurvivalinNSTE-ACS, becauseofthepatients underlyingcomorbidities,mighthave affectedthechoiceofinitialtreatmentstrategy.Fourth,the individualsubgroupanalysiscomparisonswerenotpropensity matched,andthusthepossibilityofunmeasuredbiascannot beexcluded.Finally,giventhenatureofNISdata,wewere limitedtoin-hospitalmortalityandcouldnotcompareboth strategieswithregardtolong-termoutcomes.ConclusionsInthislargepropensityscorematchedanalysis,anearly invasivestrategywasfoundtobeassociatedwithimprovedinhospitalsurvivalindiabeticpatients,especiallythosewithhighriskfeatures,suchasNSTEMIorcardiogenicshock.Theuseof anearlyinvasivestrategyindiabeticswithUAdoesnotappear tobebene cialwithapossiblesignalofharm,andthusaninitial conservativestrategymaybeasaferapproachforthese patients.Our ndingsshouldbecon rmedwithfuturetrialsthat arepoweredfordetectionofshort-andlong-termsurvival bene tofanearlyinvasivestrategyindiabeticswithNSTE-ACS.SourcesofFundingPublicationofthisarticlewasfundedinpartbytheUniversity ofFloridaOpenAccessPublishingFund.DisclosuresDrAndersonisaconsultantforBiosenseWebster,aJohnson &JohnsonCompany.DrBavryreceivedanhonorariumfrom AmericanCollegeofCardiology.Allotherauthorshaveno potentialcon ictsofinteresttodisclose.References1.RoglicG.WHOGlobalreportondiabetes:asummary. IntJNoncommunicable Dis .2016;1:3 8. 2.GeissLS,WangJ,ChengYJ,ThompsonTJ,BarkerL,LiY,AlbrightAL,GreggEW. Prevalenceandincidencetrendsfordiagnoseddiabetesamongadultsaged20 to79years,UnitedStates,1980 2012. JAMA .2014;312:1218 1226. 3.EmergingRiskFactorsCollaboration,SarwarN,GaoP,SeshasaiSR,GobinR, KaptogeS,DiAngelantonioE,IngelssonE,LawlorDA,SelvinE,StampferM, StehouwerCD,LewingtonS,PennellsL,ThompsonA,SattarN,WhiteIR,Ray KK,DaneshJ.Diabetesmellitus,fastingbloodglucoseconcentration,andrisk ofvasculardisease:acollaborativemeta-analysisof102prospectivestudies. Lancet .2010;375:2215 2222. 4.JacobyRM,NestoRW.Acutemyocardialinfarctioninthediabeticpatient: pathophysiology,clinicalcourseandprognosis. JAmCollCardiol 1992;20:736 744. 5.DonahoeSM,StewartGC,McCabeCH,MohanaveluS,MurphySA,CannonCP, AntmanEM.Diabetesandmortalityfollowingacutecoronarysyndromes. JAMA .2007;298:765 775. 6.RousanTA,PappyRM,ChenAY,RoeMT,SaucedoJF.Impactofdiabetes mellitusonclinicalcharacteristics,management,andin-hospitaloutcomesin patientswithacutemyocardialinfarction(fromtheNCDR). AmJCardiol 2014;114:1136 1144. 7.ElgendyIY,KumbhaniDJ,MahmoudAN,WenX,BhattDL,BavryAA.Routine invasiveversusselectiveinvasivestrategiesfornon-ST-elevationacute coronarysyndromes:anupdatedmeta-analysisofrandomizedtrials. Catheter CardiovascInterv .2016;88:765 774. 8.NorhammarA,MalmbergK,DiderholmE,LagerqvistB,LindahlB,Ryd enL, WallentinL.Diabetesmellitus:themajorriskfactorinunstablecoronaryartery diseaseevenafterconsiderationoftheextentofcoronaryarterydiseaseand bene tsofrevascularization. JAmCollCardiol .2004;43:585 591. 9.O DonoghueML,VaidyaA,AfsalR,AlfredssonJ,BodenWE,BraunwaldE, CannonCP,ClaytonTC,deWinterRJ,FoxKA,LagerqvistB,McCulloughPA, MurphySA,SpacekR,SwahnE,WindhausenF,SabatineMS.Aninvasiveor conservativestrategyinpatientswithdiabetesmellitusandnon ST-segment elevationacutecoronarysyndromes:acollaborativemeta-analysisof randomizedtrials. JAmCollCardiol .2012;60:106 111. 10.AmsterdamEA,WengerNK,BrindisRG,CaseyDEJr,GaniatsTG,HolmesDRJr, JaffeAS,JneidH,KellyRF,KontosMC,LevineGN,LiebsonPR,MukherjeeD, PetersonED,SabatineMS,SmallingRW,ZiemanSJ;ACC/AHATaskForce Members.2014AHA/ACCguidelineforthemanagementofpatientswithnonST-elevationacutecoronarysyndromes:areportoftheAmericanCollegeof Cardiology/AmericanHeartAssociationTaskForceonPracticeGuidelines. Circulation .2014;130:e344 e426. 11.NorhammarA,Lindb ackJ,Ryd enL,WallentinL,StenestrandU;(RIKS-HIA) RoIaKaSHICA.Improvedbutstillhighshort-andlong-termmortalityratesafter myocardialinfarctioninpatientswithdiabetesmellitus:atime-trendreport fromtheSwedishRegisterofInformationandKnowledgeaboutSwedishHeart IntensiveCareAdmission. Heart .2007;93:1577 1583. 12.GustafssonI,HvelplundA,HansenKW,GalatiusS,MadsenM,JensenJS, Tilst edHH,TerkelsenCJ,JensenLO,JrgensenE,MadsenJK,AbildstrmSZ. Underuseofaninvasivestrategyforpatientswithdiabeteswithacute coronarysyndrome:anationwidestudy. OpenHeart.2015;2:e000165.DOI: 10.1136/openhrt-2014-000165. 13.NIS,HCUPNationwideInpatientSample.HealthcareCostandUtilization Project(HCUP).Availableat:https://www.hcup-us.ahrq.gov/nisoverview.jsp. AccessedOctober25,2016. 14.EpsteinAJ,PolskyD,YangF,YangL,GroeneveldPW.CoronaryrevascularizationtrendsintheUnitedStates,2001 2008. JAMA .2011;305:1769 1776. 15.ElixhauserA,SteinerC,KruzikasD.Comorbiditysoftwaredocumentation. 2004.HCUPMethodsSeriesReport#2004-1.ONLINEFebruary6,2004.US AgencyforHealthcareResearchandQuality.Availableat:https://www. vhcup-us.ahrq.gov/toolssoftware/comorbidity/comorbidity.jsp.Accessed June8,2014. 16.KolteD,KheraS,PalaniswamyC,MujibM,FonarowGC,AhmedA,JainD, FrishmanWH,AronowWS.Earlyinvasiveversusinitialconservativetreatment strategiesinoctogenarianswithUA/NSTEMI. AmJMed .2013;126:1076 1083. 17.KheraS,KolteD,AronowWS,PalaniswamyC,SubramanianKS,HashimT, MujibM,JainD,PaudelR,AhmedA,FrishmanWH,BhattDL,PanzaJA, FonarowGC.Non-ST-elevationmyocardialinfarctionintheUnitedStates: contemporarytrendsinincidence,utilizationoftheearlyinvasivestrategy,and in-hospitaloutcomes. JAmHeartAssoc .2014;3:e000995.DOI:10.1161/ JAHA.114.000995. 18.Descriptionofdataelements.Availableat:https://www.hcup-us.ahrq.gov/ db/vars/hosp_bedsize/nisnote.jsp.NIS.AccessedSeptember15,2016. DOI:10.1161/JAHA.116.005369 JournaloftheAmericanHeartAssociation9 EarlyInvasiveStrategyinDiabeticsWithNSTE-ACS MahmoudetalORIGINALRESEARCH by guest on May 12, 2017 http://jaha.ahajournals.org/ Downloaded from
PAGE 10
19.L evesqueLE,HanleyJA,KezouhA,SuissaS.Problemofimmortaltimebiasin cohortstudies:exampleusingstatinsforpreventingprogressionofdiabetes. BMJ .2010;340:b5087. 20.BavryAA,KumbhaniDJ,RassiAN,BhattDL,AskariAT.Bene tofearlyinvasive therapyinacutecoronarysyndromes:ameta-analysisofcontemporary randomizedclinicaltrials. JAmCollCardiol .2006;48:1319 1325. 21.deWinterRJ,WindhausenF,CornelJH,DunselmanPH,JanusCL,BendermacherPE,MichelsHR,SandersGT,TijssenJG,VerheugtFW;Invasiveversus ConservativeTreatmentinUnstableCoronarySyndromes(ICTUS)Investigators.Earlyinvasiveversusselectivelyinvasivemanagementforacutecoronary syndromes. NEnglJMed .2005;353:1095 1104. 22.MehtaSR,CannonCP,FoxKA,WallentinL,BodenWE,SpacekR,WidimskyP, McCulloughPA,HuntD,BraunwaldE,YusufS.Routinevsselectiveinvasive strategiesinpatientswithacutecoronarysyndromes:acollaborativemetaanalysisofrandomizedtrials. JAMA .2005;293:2908 2917. 23.FRagminFastRevascularisationduringInStabilityinCoronaryarterydisease (FRISCII)Investigators.Invasivecomparedwithnon-invasivetreatmentin unstablecoronary-arterydisease:FRISCIIprospectiverandomisedmulticentre study. Lancet .1999;354:708 715. 24.FoxKA,Poole-WilsonPA,HendersonRA,ClaytonTC,ChamberlainDA,Shaw TR,WheatleyDJ,PocockSJ;RandomizedInterventionTrialofunstableAngina Investigators.Interventionalversusconservativetreatmentforpatientswith unstableanginaornon-ST-elevationmyocardialinfarction:theBritishHeart FoundationRITA3randomisedtrial.RandomizedInterventionTrialofunstable Angina. Lancet .2002;360:743 751. 25.HulleySB,CummingsSR,BrownerWS,GradyD,NewmanTB. Designing ClinicalResearch:AnEpidemiologicApproach .4thed.Philadelphia,PA: LippincottWilliams&Wilkins;2013:75.Appendix6B. 26.FleissJL,TytunA,UryHK.Asimpleapproximationforcalculatingsamplesizes forcomparingindependentproportions. Biometrics .1980;36:343 346. 27.ElgendyIY,MahmoudAN,WenX,BavryAA.Meta-AnalysisofRandomized TrialsofLong-TermAll-CauseMortalityinPatientsWithNon-ST-Elevation AcuteCoronarySyndromeManagedWithRoutineInvasiveVersusSelective InvasiveStrategies. AmJCardiol. 2017;119:560 564. 28.MorrowDA,CannonCP,RifaiN,FreyMJ,VicariR,LakkisN,RobertsonDH, HilleDA,DeLuccaPT,DiBattistePM,DemopoulosLA,WeintraubWS, BraunwaldE;TACTICS-TIMI18Investigators.Abilityofminorelevationsof troponinsIandTtopredictbene tfromanearlyinvasivestrategyinpatients withunstableanginaandnon-STelevationmyocardialinfarction:resultsfrom arandomizedtrial. JAMA .2001;286:2405 2412. 29.SoritaA,AhmedA,StarrSR,ThompsonKM,ReedDA,ProkopL,ShahND, MuradMH,TingHH.Off-hourpresentationandoutcomesinpatientswith acutemyocardialinfarction:systematicreviewandmeta-analysis. BMJ 2014;348:f7393. 30.Rof M,PatronoC,ColletJP,MuellerC,ValgimigliM,AndreottiF,BaxJJ, BorgerMA,BrotonsC,ChewDP,GencerB,HasenfussG,KjeldsenK, LancellottiP,LandmesserU,MehilliJ,MukherjeeD,StoreyRF,WindeckerS, BaumgartnerH,GaemperliO,AchenbachS,AgewallS,BadimonL,BaigentC, BuenoH,BugiardiniR,CarerjS,CasselmanF,CuissetT,ErolC ,FitzsimonsD, HalleM,HammC,Hildick-SmithD,HuberK,IliodromitisE,JamesS,LewisBS, LipGY,PiepoliMF,RichterD,RosemannT,SechtemU,StegPG,VrintsC,Luis ZamoranoJ;ManagementofAcuteCoronarySyndromesinPatientsPresentingwithoutPersistentST-SegmentElevationoftheEuropeanSocietyof Cardiology.2015ESCguidelinesforthemanagementofacutecoronary syndromesinpatientspresentingwithoutpersistentST-segmentelevation: TaskForcefortheManagementofAcuteCoronarySyndromesinPatients PresentingwithoutPersistentST-SegmentElevationoftheEuropeanSociety ofCardiology(ESC). EurHeartJ.2016;37:267 315. 31.CannonCP,WeintraubWS,DemopoulosLA,VicariR,FreyMJ,LakkisN, NeumannFJ,RobertsonDH,DeLuccaPT,DiBattistePM,GibsonCM, BraunwaldE;TACTICS(TreatAnginawithAggrastatandDetermineCostof TherapywithanInvasiveorConservativeStrategy) Thrombolysisin MyocardialInfarction18Investigators.Comparisonofearlyinvasiveand conservativestrategiesinpatientswithunstablecoronarysyndromestreated withtheglycoproteinIIb/IIIainhibitortiro ban. NEnglJMed 2001;344:1879 1887. 32.MehtaSR,GrangerCB,BodenWE,StegPG,BassandJP,FaxonDP,AfzalR, ChrolaviciusS,JollySS,WidimskyP,AvezumA,RupprechtHJ,ZhuJ,ColJ, NatarajanMK,HorsmanC,FoxKA,YusufS;TIMACSInvestigators.Early versusdelayedinvasiveinterventioninacutecoronarysyndromes. NEnglJ Med .2009;360:2165 2175. 33.KheraS,KolteD,PalaniswamyC,MujibM,AronowWS,SinghT,GotsisW, SilvermanG,FrishmanWH.ST-elevationmyocardialinfarctionintheelderly temporaltrendsinincidence,utilizationofpercutaneouscoronary interventionandoutcomesintheUnitedStates. IntJCardiol .2013; 168:3683 3690. DOI:10.1161/JAHA.116.005369 JournaloftheAmericanHeartAssociation10 EarlyInvasiveStrategyinDiabeticsWithNSTE-ACS MahmoudetalORIGINALRESEARCH by guest on May 12, 2017 http://jaha.ahajournals.org/ Downloaded from
PAGE 11
SUPPLEMENTAL MATERIAL by guest on May 12, 2017 http://jaha.ahajournals.org/ Downloaded from
PAGE 12
T able S1 Patient and hospital characteristics of the post hoc propensity matched cohort of patients with lower match tolerance Variable (%) Propensity score matched P value Early invasive (%) Initial conservative (%) Number of patients 12,363 (100) 1 2,367 (100) Patient demographics Age, mean years (SD) 67.6(11) 67.7(12) 0.47 Female sex 5,122(41 ) 5,271(41 ) 0. 0 6 Race 0. 60 White 8,523(69 ) 8,543 (69) Black 1,652(14 ) 1,690(13 ) Hispanic 1,321(11 ) 1,284 (10) Asian or Pacific Islander 350 (3) 333 (3) Other 427 (4) 419 (3) Primary expected payer 0. 13 Medicare 8,076 (65) 8,173(66 ) Medicaid 960(8 ) 985 (8) Private insurance 2,278(18 ) 2,237 (18) Uninsured 662 (5) 586 (5) Other 326 (3) 333 (3) Weekend admission 3,064(25 ) 3,026 (25) 0. 58 Household income (median) 0. 66 0 25 th percentile 4,229 (34) 4,229 (34) 26 50 th percentile 3,248(26 ) 3,239 (26) 51 75 th percentile 2,812 (23) 2,761(22 ) 76 100 th percentile 2,074 (17) 2,138(17 ) Patient characteristics Smoking 3,450 (28) 3,366 (27) 0.23 Dyslipidemia 8,919(72 ) 8,905 (72) 0. 81 Obesity 3,182(26 ) 3,103 (25) 0.24 Known history of CAD 1 0,365(84 ) 1 0,527 (85) 0. 01 Family history of CAD 631 (5) 624 (5) 0. 8 4 Prior myocardial infarction 1,668 (14) 1,719 (14) 0. 57 Prior PCI 2,021 (16) 2,034 (16) 0. 83 Prior CABG 1,251 (10) 1,251(10 ) 0.99 Prior stroke or TIA 750 (6) 730 (6) 0. 73 Carotid artery disease 325 (3) 328 (3) 0.9 1 Peripheral vascular disease 2,161 (18) 2,231 (18) 0. 20 Pulmonary circulation disease 15 (<1) 13 (<1) 0. 71 Dementia 353 (3) 304 ( 3 ) 0. 05 Atrial fibrillation 2,027 (16) 2,034 (1 6 ) 0. 91 Alcohol abuse 259 (2) 260 (2) 0. 97 Deficiency anemia 2,936(24 ) 2,976 (24) 0.5 6 Collagen vascular disease 254 (2) 2 71(2) 0. 46 Chronic blood loss anemia 1 16 (1) 126 (1) 0. 52 Congestive heart failure 68 (<1) 81 (<1) 0.2 9 Valvular disease 15 (<1) 20 (<1) 0. 40 Chronic pulmonary disease 3,013 (24) 3,044 (25) 0. 66 Coagulopathy 781 (6) 711 (6) 0. 43 by guest on May 12, 2017 http://jaha.ahajournals.org/ Downloaded from
PAGE 13
Liver disease 236 (2) 243 (2) 0. 75 Renal disease (chronic) 4,006 (32) 4,044(33 ) 0. 62 Electrolytes abnormalities 3,006 (24) 3,030(25 ) 0.73 AIDS 1 1 (<1) 9 (<1) 0. 65 Drug abuse 232 (2) 225 (2) 0. 74 Depression 1, 156 (9) 1,154 (9) 0. 96 Hypertension 1 0 373(84 ) 1 0 290 (83) 0. 14 Hypothyroidism 1,579 (13) 1,618 (13) 0. 47 Lymphoma 60 (<1) 67 (<1) 0. 54 Metastatic cancer 60 (<1) 74 (<1) 0. 23 Solid tumor without metastasis 165 (1) 148 (1) 0. 33 Other neurological disorder 698(6 ) 679 (6) 0. 59 Paralysis 236 (2) 250 (2) 0. 52 Psychoses 359(3) 334(3) 0. 33 Peptic ulcer (non bleeding) 1 (<1) 1 (<1) 1 00 Weight loss 317 (3) 327(3 ) 0. 69 Cardiogenic shock 400 (3) 374 (3) 0. 34 Intra cranial hemorrhage 9 (<1) 15 (<1) 0. 22 Acute ischemic stroke 194 (2) 182 (2) 0. 5 3 Gastrointestinal bleeding 245 (2) 238 (2) 0. 75 Hospital characteristics Hospital bed size 0. 25 Small 1,161(9 ) 1,238 (10) Medium 3,178 (26) 3,176 (26) Large 8,024 (65) 7,953 (64) Hospital location 0.21 Urban teaching 6,309(49 ) 6,453 (5 0 ) Urban non teaching 5,033 (41) 4,877(39 ) Rural 1, 021 (8) 1, 037 ( 8 ) In hospital mortality 303(2.5) 462(3.7) <0.0 00 1 All percentages are approximated to the nearest integer. CAD= coronary artery disease, PCI= percutaneous coronary intervention, CABG= coronary artery bypass graft surgery, TIA= transient ischemic attack, AIDS= acquired immune deficiency syndrome. by guest on May 12, 2017 http://jaha.ahajournals.org/ Downloaded from
PAGE 14
Anthony A. Bavry and R. David Anderson Ahmed N. Mahmoud, Islam Y. Elgendy, Hend Mansoor, Xuerong Wen, Mohammad K. Mojadidi, Acute Coronary Syndromes: A Contemporary National Insight Elevation ST Hospital Survival Among Diabetics With Non Early Invasive Strategy and In Online ISSN: 2047-9980 Dallas, TX 75231 is published by the American Heart Association, 7272 Greenville Avenue, Journal of the American Heart Association The doi: 10.1161/JAHA.116.005369 2017;6:e005369; originally published March 18, 2017; J Am Heart Assoc. http://jaha.ahajournals.org/content/6/3/e005369 World Wide Web at: The online version of this article, along with updated information and services, is located on the for more information. http://jaha.ahajournals.org Access publication. Visit the Journal at is an online only Open Journal of the American Heart Association Subscriptions, Permissions, and Reprints: The by guest on May 12, 2017 http://jaha.ahajournals.org/ Downloaded from
|