Relationship between time to treatment and mortality among patients undergoing primary percutaneous coronary intervention according to Korea Acute Myocardial Infarction Registry

Originalarticle

Relationship between time to treatment and mortality among patients undergoing primary percutaneous coronary intervention according to Korea Acute Myocardial Infarction Registry

Hyun KukKim(MD)^a,Myung HoJeong(MD, PhD,FACC, FAHA,FESC, FSCAI)^b,*,

YoungkeunAhn(MD)^b,ShungChullChae(MD)^c,YoungJoKim(MD)^d,SeungHoHur(MD)^e, InWhanSeong (MD)^f,Taek Jong Hong (MD)^g,Dong HoonChoi(MD)^h,

MyeongChanCho(MD)ⁱ,ChongJinKim(MD)^j,KiBaeSeung(MD)^k,YangSooJang(MD)^h, SeungWoonRha(MD)^l,JangHoBae (MD)^m, Sung SooKim(MD)ⁿ,

SeungJung Park(MD)^o otherKoreaAcute MyocardialInfarction RegistryInvestigators¹

aChosunUniversityHospital,Gwangju,RepublicofKorea

bChonnamNationalUniversityHospital,Gwangju,RepublicofKorea

cKyungpookNationalUniversityHosptial,Daegu,RepublicofKorea

dYeungnamUniversityHospital,Daegu,RepublicofKorea

eKeimyungUniversityHospital,Daegu,RepublicofKorea

fChungnamNationalUniversityHospital,Daejon,RepublicofKorea

gBusanNationalUniversityHospital,Busan,RepublicofKorea

hYonseiUniversitySeveransHospital,Seoul,RepublicofKorea

iChungbukNationalUniversityHospital,Cheonju,RepublicofKorea

jKyungheeUniversityHospital,Seoul,RepublicofKorea

kCatholicUniversityHospital,Seoul,RepublicofKorea

lKoreaUniversityHospital,Seoul,RepublicofKorea

mKonyangUniversityHospital,Seoul,RepublicofKorea

nKwangjuChristianHospital,Gwangju,RepublicofKorea

oUlsanUniversityHospital,Seoul,RepublicofKorea

ARTICLE INFO

Articlehistory:

Received12August2016

Receivedinrevisedform4September2016 Accepted13September2016

Availableonline6October2016

Keywords:

Myocardialinfarction Door-to-balloontime Primarypercutaneouscoronary intervention

ABSTRACT

Background: Despitelargereductionsindoor-to-balloontimesovertheperiod,severalstudiesfrom regionalandnationaldatashowedthatannualmortalityrateswerenotdecreasedamongpatientswho underwentprimarypercutaneouscoronaryintervention(PCI).However,thesestudiesmostlyfocused ondoor-to-balloontime,andtherewasnoconsiderationoftotalischemictimeinatrendofmortality.

Theaimofthisstudywastoassesstheannualtrendbetweentimetotreatmentand1-monthmortality amongpatientsundergoingprimaryPCI.

Methodsandresults: Thestudypopulationconsistedof8040patientswhounderwentprimaryPCIat hospitals participating in the nationwide prospective Korea Acute Myocardial Infarction Registry (KAMIR)betweenJanuary2008andDecember2011.Theprimaryendpointofthisstudywas1-month all-causemortality, andtimeto treatment(door-to-balloontime, symptom-to-balloon time).One- monthdeathoccurredin452patients(5.6%)from2008to 2011.Additionalreductionsindoor-to- balloontimewerenottranslatedintoparallelreductionsinmortalityrateandtotalischemictime.After adjustmentusingclinicalrisk,shortertotalischemictimewasanindependentpredictorof1-month mortality [adjusted oddsratio (OR) 0.78, 95% confidential interval (CI) 0.62–0.99, p=0.04]. Total ischemictimecouldbereducedbyusingemergencymedicalservices.

Conclusion:Despiteimprovementsindoor-to-balloontime,noparallelreductionsinmortalityrateand totalischemictimewereobserved.Totalischemictimewasassociatedwithmortality.Thepresentstudy

* Correspondingauthorat:PrincipalInvestigatorofKoreaAcuteMyocardialInfarctionRegistry,ChonnamNationalUniversityHospital,671Jaebongro,DongKu, Gwangju501-757,SouthKorea.Fax:+82622287174.

E-mailaddress:myungho@chollian.net(M.H.Jeong).

1SeeAppendixA.

ContentslistsavailableatScienceDirect

Journal of Cardiology

j our na l ho me pa g e : w ww . e l se v i e r . com / l oca t e / j j cc

http://dx.doi.org/10.1016/j.jjcc.2016.09.002

0914-5087/ß2016PublishedbyElsevierLtdonbehalfofJapaneseCollegeofCardiology.

Introduction

Door-to-balloon time, time interval from hospital arrival to ballooninflation,isassociated withbetterin-hospitalandlong- termsurvivalforpatientswithST-segmentelevationmyocardial infarction(STEMI)whoreceivedprimarypercutaneouscoronary intervention(PCI)[1–3]. Based onthesedata,current guideline recommendedgoals of90minfordoor-to-balloontime [4].Be- causeofthisrecommendation,door-to-balloontimehasbeena keyqualitymetricforprimaryPCIcenters[5].

ThepreviousreportinKoreashowedthatonly51%ofpatients achievedthis goalforreperfusionbetween November2005and January2007[6].Thus,therehavebeeninstitutionalandnational qualityimprovementinitiativestoachievedoor-to-balloontime within 90min in Korea. But, it is unknown whether these movements are associated withbetterprognosis. Despite large reductions in door-to-balloon times over the period, several studies from regional and national data showed that annual mortalityrateswerenotdecreasedamong patientswhounder- wentprimaryPCI[7,8].However,thesestudiesmostlyfocusedon door-to-balloontime inWesternpopulations,and therewasno considerationoftotalischemictimeinatrendofmortality.

In thepresentstudy,weusedlargenationwideregistrydata fromKoreatoassesstheannualtrendbetweentimetotreatment and1-monthmortalityamongpatientswhounderwentprimary PCI.

Methods Studypopulation

Thestudypopulationconsistedof8040patientswhounder- went primary PCI at hospitals participating in the nationwide prospectiveKoreaAcuteMyocardialInfarctionRegistry(KAMIR) betweenJanuary2008andDecember2011.TheKAMIR,launched inNovember2005,isaprospectivemulti-center datacollection registryreflectingreal-worldtreatmentpracticesandoutcomesin Asian patients diagnosed with AMI. The registry consists of 50communityandteachinghospitalswithfacilitiesforprimary PCIandon-sitecardiacsurgery.Includedpatientswererepresen- tativeoftheKoreanSTEMIpopulation.ThefirstdiagnosisofSTEMI wasestablishedintheparticipatingcenters.Datawerecollectedby atrainedstudycoordinatorusingastandardizedcasereportform and protocol. The study protocol was approved by the ethics committeeateachparticipatinginstitution.Detailsoftheregistry havebeenpreviouslypublishedinmanyinternationaljournals[9].

Inclusion criteria for the present study were consecutive patientsaged 18years presenting within12hafter symptom onset,hadasuggestivehistorywithST-segmentelevationabove 2mmintwoormoreprecordialleads,orabove1mmintwoor morelimbleads,ornewonsetofleftbundlebranchblockonthe 12-leadelectrocardiogram(ECG) withincreasedcardiacenzyme activity (troponin or myocardial fraction of creatine kinase), treatedwithprimaryPCI,andhadcompleteda1-monthclinical follow-up.Weexcludedpatientsforwhomtimetabledatawere notcompleted(symptomonset,hospitalarrival,andballoontime), andthedoor-to-balloontimeswerenotavailableduetoanerror value.

Interventionalproceduresandmedicaltreatment

Loading doses of aspirin and clopidogrel were administered afterpatientsconsentedtoreceivingPCI,andaminimumdoseof 100mgaspirinanda300–600mgloadingdoseofclopidogreland unfractionatedheparin(50–70U/kg)wereadministeredinorder to maintain an activated clotting time above 250–300s. The maintenancedosewas100mg/dayforaspirinand75mg/dayfor clopidogrel. Coronary arterystenting was performed using the standardtechnique.Performanceofpre-dilation,directstenting, post-adjunctive balloon inflation, mode of revascularization (staged or ad-hoc PCI), using intravascular ultrasound (IVUS), usingthrombusaspirationdevice,stenttype,andadministration ofaglycoproteinIIb/IIIareceptorblockerwereatthediscretionof theoperator.

Statisticalanalysis

AllanalyseswereperformedusingStatisticalPackageforthe SocialSciences(SPSS)softwareversion21.0(IBMsoftwaregroup, Chicago,IL,USA).Baselineclinicalandproceduralcharacteristics were compared with theuse of chi-square test for categorical variables and the analysis-of-variance F-test for continuous variables.Continuousvariablesarepresentedasmeanstandard standarddeviation(SD),andcategoricalvariablesaspercentages.The primaryendpointofthisstudywas1-monthall-causemortality,and timetotreatment(door-to-balloontime,symptom-to-balloontime).

Multi-variatelogisticregressionanalysiswasperformedtoassessthe relation between time to treatment (door-to-balloon time, total ischemic time) and 1-month mortality. Global Registry of Acute CoronaryEvents(GRACE)scoreforin-hospitalmortalitywasusedasa co-variabletoadjusttherelativerisk[10].Allstatisticaltestswere two-tailed,andap-valuelessthan0.05wasconsideredsignificant.

Results

Baselineandproceduralcharacteristics

Atotalof9154patientswithtimetopresentationwithin12h whounderwentprimaryPCIbetweenJanuary2008andDecember 2011wereincludedinthisstudy.Patientswereexcludedifmissing datawerepresentintimetable(n=668),iftheyhadanerrorvalue ofdoor-to-balloontime(n=393),or1-monthmortalitydatawere not available (n=53). Thus, thefinal studycohort consisted of 8040patients(Fig.1).TheincidenceratesofSTEMIinKoreawere decreasedcomparedtoNSTEMIasinapreviousreport[11].

Themeanagewas63.9yearswith74%ofmalepatients.Table1 shows baseline clinical and procedural characteristics for each year.Exceptdyslipidemia,theprevalenceofriskfactorssuchas hypertension, diabetes, and current smoking was relatively constant. The proportion of patients with a prior MI, prior revascularization, and heart failurewas remarkablylowerthan previousWesternstudy(MI2.3%vs.18.5%,PCI4.8%vs.20.5%,heart failure0.9% vs.4.0%).Left ventricularejectionfractionwasalso more preserved maybedue to differences of these parameters [8]. Location of culprit artery and prevalence of multi-vessel diseasewere similarin each year. However, procedural factors showed some tendencies over the course of the study period.

suggeststhatadditionaleffortsareneededtoshortentotalischemictimeincludingpatientandpre- hospitalsystemicdelayforbetterprognosisafterprimaryPCI.

ß2016PublishedbyElsevierLtdonbehalfofJapaneseCollegeofCardiology.

Fig.1.Studypopulationdiagram.PCI,percutaneouscoronaryintervention.

Table1

Baselineclinicalandproceduralcharacteristics.

Characteristic Total(n=8040) 2008(n=2135) 2009(n=2162) 2010(n=1890) 2011(n=1853) p-Value

Age(years)^a 63.913.0 63.913.1 64.012.8 63.613.1 64.013.0 0.767

Age>75(%) 23.3 22.9 23.3 22.9 24.1 0.783

Male(%) 74.3 73.4 74.4 73.4 76.0 0.219

Clinicalhistory(%)

Hypertension 46.8 46.0 46.8 46.8 47.9 0.696

Diabetesmellitus 24.4 23.8 24.0 24.3 25.6 0.546

Dyslipidemia 11.2 10.7 9.2 13.5 11.7 <0.001

Currentsmoker 60.9 60.7 60.9 60.1 62.0 0.676

Heartfailure 0.9 0.7 0.9 0.9 0.8 0.901

Stroke 5.2 5.5 5.4 4.9 5.1 0.818

Familyhistory 7.8 6.3 8.7 8.6 7.5 0.014

PriorMI 2.3 2.4 2.3 1.9 2.8 0.320

PriorPCI 4.8 4.0 4.9 5.2 5.2 0.237

PriorCABG 0.2 0.2 0.3 0.3 0.2 0.837

Cardiogenicshock 8.4 8.1 9.7 7.4 8.5 0.078

Ejectionfraction(%)^b 52.020.9 52.115.4 51.820.8 51.924.9 51.920.9 0.961

Proceduralvariables(%) Door-to-balloontime

Median(min) 65 72 69 64 59 <0.001

<90min(%) 79.9 70.3 75.4 85.7 90.2 <0.001

Totalischemictime(min) 240 257 245 227 224 0.380

Useofstent 89.9 90.8 86.8 90.7 91.6 <0.001

Drug-elutingstent 83.7 82.2 80.3 85.2 87.7 <0.001

GPIIb/IIIainhibitor 23.3 29.2 22.0 20.6 20.9 <0.001

Thrombectomy 19.1 15.8 17.9 20.1 23.2 <0.001

Radialapproach 17.4 16.8 16.6 17.6 18.8 <0.001

IVUS-guidedPCI 17.4 17.9 18.5 18.4 14.4 <0.001

Targetcoronaryartery(%)

Leftmain 1.4 1.6 1.5 1.4 1.3 0.808

Leftanteriordescending 51.0 50.9 50.2 52.1 51.0 0.704

Leftcircumflex 9.3 10.2 10.1 8.4 9.3 0.253

Rightcoronary 38.3 37.3 39.2 38.2 38.5 0.644

Multi-vesseldisease(%) 51.2 52.8 51.8 49.7 50.1 0.167

Ventriculararrhythmia(%) 3.6 4.0 3.7 3.2 3.6 0.562

MI,myocardialinfarction;PCI,percutaneouscoronaryintervention;CABG,coronaryarterybypassgraft;GP,glycoprotein;IVUS,intravascularultrasound.

a Plusminusvaluesaremeansstandarddeviation.

b Availablein1884patients(88.2%)in2008,1936patients(89.1%)in2009,1684patients(89.1%)in2010,and1709patients(90.1%)in2011.

Drug-elutingstentimplantationtendedtoincreasefrom82.2%in 2008to87.7%in2011.Thrombectomyandtrans-radialapproach graduallyincreasedoverthestudyperiod(thrombectomy15.8–

23.2%, radial approach 16.8–18.8%). The percentage of patients receiving IVUS-guided PCI and glycoprotein IIb/IIIa inhibitor declinedannually.

Clinicaloutcomes

Themediandoor-to-balloontimedecreasedsignificantlyfrom 72minin2008to59minin2011.Theachievedrateofgoalof door-to-balloontime <90min wassignificantly increasedfrom 70.3%in2008to90.2%in2011.Similarly,nosignificantchangein thistrendwasobservedinanyofthehigh-risksubgroups(Table2).

However,symptom-to-balloontime(totalischemictime)wasnot significantlydifferentdespitedoor-to-balloontimeimprovement (p=0.380), and the rate of 1-month mortality was relatively constantovertheperiodfrom5.8%in2008to5.8%in2011(Fig.2).

One-monthdeathoccurred in452patients(5.6%) from2008to 2011.AfteradjustmentbyGRACEscore,totalischemictimebelow 180min was an independent predictor of 1-month mortality [adjustedoddsratio(OR)0.78,95%confidentialinterval(CI)0.62–

0.99, p=0.04]. However, lowered only door-to-balloon time

(adjustedOR0.86,95% CI0.66–1.12, p=0.258)orsymptom-to- doortimewasnotsignificantlyassociatedwith1-monthmortality (adjustedOR0.82,95%CI0.63–1.06,p=0.136)(Table3).

Emergencymedicalservices(EMS)wereusedinonly36.1%of patients (2902/8040). Median time of total ischemic time was significantly decreased in EMS users (202min vs. 264min, p=0.011), but not door-to-balloon time (64min vs. 66min, p=0.223).

Discussion

Themajorfindingsofthepresentstudyareasfollows.First, door-to-balloontimehadatendencytodeclineduringthe4years.

Second,additional reductionsindoor-to-balloon time werenot translated into parallel reductions in mortality rate and total ischemic time.A lackofmortalitybenefitwasalsoobservedin high-risk subgroups. Third,total ischemic time was associated with1-monthmortalityrates.

Differenttomanyothercountries,allhealthinsurancesocieties wereintegratedintoasingleinsurer,theNationalHealthInsurance Program in Korea[12]. All medical institutionsof Koreashould claim almost all procedural costs to government (National Health Insurance Corporation). Then, government will pay the Table2

Door-to-balloontimeand1-monthmortalityinthehigh-risksubgroups.

Characteristic Total 2008 2009 2010 2011 p-Value

Age>75years(%) 23.3 22.9 23.3 22.9 24.1 0.783

Door-to-balloontime

Median(min) 68 74 70 66 63 <0.001

<90min(%) 77.4 68.6 72.7 81.4 88.0 <0.001

Mortality(%) 12.5 11.3 12.1 13.8 12.5 0.669

Multi-vesseldisease(%) 51.2 52.8 51.8 49.7 50.1 0.783

Door-to-balloontime

Median(min) 66 72 68 64 60 <0.001

<90min(%) 79.4 70.7 74.9 85.8 88.6 <0.001

Mortality(%) 7.3 7.2 7.0 7.3 7.6 0.970

Cardiogenicshock(%) 8.4 8.1 9.7 7.4 8.5 0.078

Door-to-balloontime

Median(min) 69 72 73 61 65 <0.001

<90min(%) 76.3 69.9 67.9 90.8 80.6 <0.001

Mortality(%) 32.0 32.2 31.8 29.0 34.5 0.770

GRACEhighrisk(%) 66.9 67.8 68.5 64.9 66.4 0.115

Door-to-balloontime

Median(min) 68 74 70 66 63 <0.001

<90min(%) 79.5 71.3 74.3 85.3 89.4 <0.001

Mortality(%) 7.7 7.0 7.7 7.9 8.1 0.772

GRACE,GlobalRegistryofAcuteCoronaryEvents.

Fig.2.Mortalityratesaccordingtothetimetotreatment,2008–2011.Themediantimetotreatmentand1-monthmortalityamongpatientswithST-segmentelevation myocardialinfarctionwhounderwentprimarypercutaneouscoronaryinterventionareshownbetweenJanuary2008andDecember2011.Resultsbetweendoor-to-balloon timeand1-monthmortality(A),andbetweentotalischemictimeand1-monthmortality(B).

reimbursement to hospitals after reviewing appropriateness of medicalfeesandhealthcaremanagement.Thegovernmentgraded thequalityofcareofalltertiarycentersregardlessofPCI-capability basedon in-hospital reperfusion time (door-to-needle<30min, door-to-balloon <90min), and have provided incentives or reductionof reimbursement for procedural costs depending on achievingtheratesofthereperfusiongoal.Moreover,resultsare madepublic throughthe variousmass mediaand Internetweb sites.Asaresult,theachievedrateofgoalofdoor-to-balloontime

<90minwasdramaticallyincreasedfrom70.3%in2008to90.2%in 2011.One-monthmortalityratewasnotdecreaseddespitealarge reductionindoor-to-balloontime.Theresultsweresimilarwith previousstudies[7,8].But,itdoesnotmeanthatdoor-to-balloon timeis a uselessmetricfor quality-improvementofPCI-capable centers[13].

Totalischemictimeisrelatedtolong-termclinicaloutcomes especially in high-risk subgroups [14,15]. Clinical study using cardiovascularmagneticresonanceshowedthatshorterischemic time in patients withprimary PCI wasassociated withsmaller infarct size and micro-vascular obstruction, and larger salvage myocardium[16].Asmentioned above,shorter door-to-balloon timeisassociatedwithreducedmortality.But,theeffectofdoor- to-balloontimereductioncouldbevariedacrossthesubgroupsof patients.Earlypresenter(3h)aftersymptomonsetandhigh-risk patientsappeartobenefitthemostafterachieveddoor-to-balloon timegoal[17].Pooledanalysisfromtworandomizedtrials also showed that patients with later presentation (>3h) had no significant 1-year mortality reduction despite short door-to- balloontime (90min) [18]. Also in Asian data,thebenefit of shortdoor-to-balloontimewaslimitedtopatientswhopresented early[14].Arecentstudyshowedthattotalischemictimewasa betterpredictorthandoor-to-balloontimefor1-monthmortality andinfarctsize[19].

However,therehavebeennospecialattentionsorregulations exceptdoor-to-reperfusiontime in Korea. In-hospitalreperfu- siontimeshouldbecontinuouslyusedasaperformancemeasure atthePCIcenter.But,itmaynotbeanidealgoalforthegeneral healthcare system. Total ischemic time includingpatient and systemic delay might be a more suitable target for national quality improvement program. Forreduction of patient delay, educationaboutSTEMI symptomsandrapidcontactwithEMS areneeded[20].Inthepresentstudy,EMSwasassociatedwith reducingtotalischemictime.SeveralreportsshowedthatEMS- based pre-hospital network to hospitals or interventional cardiologist reduced systemic delay and improved outcomes [21–24]. Fornon-PCI capablecenters,telemedicineandestab- lishment referral system to PCI centers could be helpful to decreasesystemicdelay[25,26].Several subgroupsof patients such as older age, female gender, hypertension, diabetes mellitus,andstroke should befocused tominimizetreatment delay[27].

There wereseveral limitations tothis study. First, the non- randomizednatureof theregistrydatacouldresultinselection bias.Therewereseveralclinicalandproceduraldifferencesover thestudyperiod.Unmeasuredchangesinthepatientpopulation couldaffectthe1-monthmortality.Itisalsopossiblethatlow-risk patientsarebeingtreatedmorequickly,andhigher-riskpatients arebeingdelayed.Second,thisstudyincluded1-monthfollow-up period. It could be possible that shorter door-to-balloon time translatesintoreductionoflong-termmortalityandheartfailure.

Third,therewerenotenoughdataaboutdetailedinformationof pre-hospitalsystemdelayinKAMIR.Itisinsufficienttoevaluate therelationshipbetweenmortalityandsystemicdelayinprimary PCI.

Conclusion

Thepresentstudyshowedthattherewasatrendofsignificant decline of door-to-balloon time during 4 years. Despite these improvements,noparallelreductionsinmortalityrateandtotal ischemictimewereobserved.Afteradjustmentusingclinicalrisk, totalischemictimewasassociatedwith1-monthmortalityrates.

Door-to-balloon time should probably become only one of a numberofmetricsinsteadofthesolequalitymetric.Additional effortsareneededtoreducetotalischemictimeincludingpatient andpre-hospitalsystemicdelayforbetterprognosisafterprimary PCI.

Funding

This study was supported by a grant of the Korean Health TechnologyR&DProject(HI13C1527),sponsoredbytheMinistry forHealth&Welfare,RepublicofKorea.

Conflictofinterest

Theauthorsdeclarethatthereisnoconflictofinterest.

Acknowledgment

This study was performed with the support of the Korean Circulation Society (KCS) as a memorandum of the 50th AnniversaryoftheKCS.

AppendixA

KAMIRInvestigators:MyungHoJeong,YoungJoKim,ChongJin Kim,MyeongChanCho,YoungKeunAhn,JongHyunKim,Shung ChullChae,SeungHoHur,InWhanSeong,TaekJongHong,Dong HoonChoi,JeiKeonChae,JaeYoungRhew,DooIlKim,InHoChae, Table3

Door-to-balloontime,totalischemictime,and1-monthmortality.

One-monthmortality(%) UnadjustedOR(95%CI) p-Value ^aAdjustedOR(95%CI) p-Value Achievedgoal Notachievedgoal

Door-to-balloontime

<90min

5.1 7.6 0.65(0.53–0.81) <0.001 0.86(0.66–1.12) 0.258

Symptom-to-doortime

<90min

4.7 6.6 0.69(0.57–0.84) <0.001 0.82(0.63–1.06) 0.136

Totalischemictime

<180min

4.2 6.3 0.64(0.53–0.78) <0.001 0.78(0.62–0.99) 0.040

Totalischemictime=symptom-to-balloontime.

OR,oddsratio;CI,confidenceinterval.

a AdjustedbyGRACE(GlobalRegistryofAcuteCoronaryEvents)riskscore.