Poor long-term outcome in acute coronary syndrome in a real-life setting : Ten-year outcome of the TACOS study

ONLINE FIRST This is a p o isional PDF only. Copyedi ed and ully o ma ed e sion will be made a ailable soon. ISSN: 1897-5593 e-ISSN: 1898-018X Poo long- e m ou come in acu e co ona y synd ome in a eal- li e se ing: Ten-yea ou come o he TACOS s udy Au ho s: Kaa i K. Kon ila, Kimmo Koi ula, Ma kku J. Eskola, Mika Ma iskainen, Heini Huh ala, Vesa K. Vi anen, Jussi Mikkelsson, Ka i Jä elä, Ka i O. Niemelä, Pekka J. Ka hunen, Kjell C. Nikus DOI: 10.5603/CJ.a2019.0037 A icle ype: O iginal a icles Submi ed: 2019-02-02 Accep ed: 2019-03-28 Published online: 2019-04-11 This a icle has been pee e iewed and published immedia ely upon accep ance. I is an open access a icle, which means ha i can be downloaded, p in ed, and dis ibu ed eely, p o ided he wo k is p ope ly ci ed. A icles in "Ca diology Jou nal" a e lis ed in PubMed. Powe ed by TCPDF (www. cpd .o g) Poo long- e m ou come in acu e co ona y synd ome in a eal-li e se ing: Ten-yea ou come o he TACOS s udy Long- e m ou come in ACS Kaa i K. Kon ila1, Kimmo Koi ula1, 2, Ma kku J. Eskola3, Mika Ma iskainen1, Heini Huh ala5, Vesa K. Vi anen3, Jussi Mikkelsson6, Ka i Jä elä7, Ka i O. Niemelä3, Pekka J. Ka hunen1, 4, Kjell C. Nikus1, 3 1Facul y o Medicine and Heal h Technology, Tampe e Uni e si y, Finland 2Sou h-Ka elia Cen al Hospi al, Finland 3Hea Cen e , Depa men o Ca diology, Tampe e Uni e si y Hospi al, Finland 4Fimlab Labo a o ies Tampe e Uni e si y Hospi al, Tampe e, Finland 5Facul y o Social Sciences, Uni e si y o Tampe e, Finland 6Hea Cen e , Sa akun a Cen al Hospi al, Po i, Finland 7Hea Cen e , Tampe e Uni e si y Hospi al, Finland Add ess o co espondence: D . Kjell Nikus, MD, PhD, Hea Cen e , Depa men o Ca diology, Tampe e Uni e si y Hospi al, Ensi ie 4, 33520 Tampe e, Finland, el: +358 50 5575 396, e-mail: kjell.nikus@sydänsai aala. i Abs ac Backg ound: Long- e m ou come o he h ee ca ego ies o acu e co ona y synd ome (ACS) in eal-li e pa ien coho s is no well known. The objec i e o his s udy was o su ey he 10-yea ou come o an ACS pa ien coho admi ed o a uni e si y hospi al and o explo e ac o s a ec ing he ou come. Me hods: A o al o 1188 consecu i e pa ien s (median age 73 yea s) wi h ST-ele a ion myoca dial in a c ion (STEMI), non-ST-ele a ion myoca dial in a c ion (NSTEMI) o uns able angina pec o is (UA) in 2002–2003 we e included and ollowed up o ≥ 10 yea s. Resul s: Mo ali y o STEMI, NSTEMI and UA pa ien s du ing he ollow-up pe iod was 52.5%, 69.9% and 41.0% (p < 0.001), espec i ely. In mul i a iable Cox eg ession analysis, only age and c ea inine le el a admission we e independen ly associa ed wi h pa ien ou come in all he h ee ACS ca ego ies when analyzed sepa a ely. Conclusions: All he h ee ACS ca ego ies p o ed o ha e high mo ali y a es du ing long- e m ollow-up in a eal-li e pa ien coho . NSTEMI pa ien s had wo se ou come han STEMI and UA pa ien s du ing he whole ollow-up pe iod. Ou s udy esul s indica e clea di e ences in he p ognos ic signi icance o a ious demog aphic and he apeu ic pa ame e s wi hin he h ee ACS ca ego ies. Key wo ds: acu e co ona y synd ome, myoca dial in a c ion, p ognosis, uns able angina INTRODUCTION Acu e co ona y synd omes (ACS) ep esen a spec um o clinical e en s anging om uns able angina pec o is (UA) o non-ST ele a ion (NSTEMI) and ST-ele a ion myoca dial in a c ion (STEMI). Despi e he ac ha ischemic hea disease emains he leading cause o dea h globally [1], da a on long- e m mo ali y, especially beyond he i s ew yea s, is sca ce. Elde ly pa ien s a e unde ep esen ed o e en excluded in clinical ials. As many as 50% o eal-wo ld acu e myoca dial in a c ion (MI) pa ien s may no be ep esen ed in andomized clinical ials [2]. On he o he hand, he gene al popula ion is aging, elde ly indi iduals comp ise he as es g owing segmen o he popula ion wo ldwide, and co ona y a e y disease is common in he elde ly [3, 4]. Olde MI pa ien s a e less likely o ecei e e idence-based ca e han younge pa ien s (5). S udies ha e shown ha UA pa ien s ha e be e sho - e m ou come han pa ien s wi h acu e MI, bu long- e m ou come may no di e g ea ly [6]. Acco ding o andomized clinical ials, NSTEMI pa ien s ha e be e ou come han STEMI pa ien s du ing he i s ew weeks a e he acu e e en , bu hey a e a highe isk o ad e se ou come o e he long- e m [7]. In a p ospec i e obse a ional s udy, we p e iously epo ed 10-mon h ou come da a o consecu i e ACS pa ien s (n = 1188) ea ed in a uni e si y hospi al [8]. The aim o he p esen s udy was o es ablish he 10-yea ou come da a o all he h ee clinical en i ies o ACS in he same pa ien coho . We also s udied he e ec o baseline clinical ac o s and da a collec ed du ing he ini ial hospi al s ay on pa ien ou come. METHODS S udy popula ion De ails o he pa ien selec ion ha e been desc ibed elsewhe e [8]. B ie ly, he Tampe e Acu e CO ona y S udy (TACOS) s udy coho consis ed o 1188 ACS pa ien s admi ed o Tampe e Uni e si y hospi al om he ci y o Tampe e and 11 neighbo ing municipali ies, a egion o 340,000 inhabi an s. F om Janua y 1s 2002 o Ma ch 31s 2003 all pa ien s admi ed o he eme gency depa men p esen ing wi h acu e MI as e i ied by an ele a ed blood oponin I (cTnI > 0.2 μg/L) alue we e ec ui ed. In addi ion, om Sep embe 1s 2002 o Ma ch 31s 2003 all consecu i e oponin-nega i e pa ien s wi h UA we e also ec ui ed. Pa ien s who died in o we e discha ged om he eme gency depa men we e no included. The comple e s udy popula ion consis ed o 343 (29%) pa ien s wi h STEMI, 655 (55%) wi h NSTEMI and 190 (16%) wi h UA. The s udy complies wi h he Decla a ion o Helsinki. The E hics Commi ee o he Pi kanmaa Hospi al Dis ic app o ed he s udy p o ocol (Pe mission R02100). All subjec s ga e hei w i en in o med consen o pa icipa ion. ACS ca ego ies All pa ien s had symp oms and/o clinical signs sugges i e o ACS. Pa ien s wi h STEMI had ele a ed oponin le els (> 0.2 μg/L) and hei elec oca diog am (ECG) ul illed he p ede ined c i e ia o STEMI: ST-segmen ele a ion in ≥ 2 adjacen leads, in leads V1–V6 ≥ 1.5 mm (≥ 2 mm in a leas one lead), in leads II, III, aVF, and I and aVL ≥ 1 mm. Also, in NSTEMI pa ien s, he oponin alues we e ele a ed, bu he ECG did no ul il he c i e ia o STEMI. UA pa ien s showed no ele a ion in a minimum o wo cTnI le els 6– 12 h apa . Follow-up Da a was collec ed by a s udy nu se and wo o he in es iga o s (ME and KJN). The ollow-up was se o begin a he momen o he ECG eco ding used o analysis, and i ended a dea h o a he end o ollow-up — Ma ch 31s 2013. Mo ali y was ga he ed by linking he pe sonal iden i y code om he TACOS s udy o he Causes o Dea h egis e , main ained by S a is ics Finland, which eco ds 100% o dea hs o Finnish ci izens a home and nea ly 100% ab oad. Follow-up was comple e wi h 716 dea hs and 472 pa ien s ali e a he end o he ollow up. When compa ing mo ali y o li e a u e, exac 10-yea mo ali y was used. S a is ical analysis Ca ego ical a iables we e exp essed as numbe s o pa ien s o pe cen ages and con inuous a iables as means o medians ollowed by qua iles (Q1–Q3). Fishe ’s exac es was used o ca ego ical a iables and he Mann-Whi ney U o K uskal-Wallis es o nume ical a iables. A wo- ailed p- alue o < 0.05 was conside ed s a is ically signi ican . Kaplan-Meie cu es we e used o p esen he unadjus ed su i al da a. Cox eg ession analysis was used o iden i y he baseline and in-hospi al p ognos ic a iables conce ning mo ali y a ollow-up. Cox uni a ia e and mul i a iable eg ession analyses including all he a iables we e p esen ed. T oponin I alues we e used only o he STEMI and NSTEMI ca ego ies due o immeasu able low (< 0.2 μg/L) alues in UA pa ien s. To u ilize he powe o he wide s udy popula ion, he a iables p e ious smoking and co ona y angiog aphy we e no included in he inal model because o lack o da a in a signi ican p opo ion o pa ien s. Mo ali y a es a p e-speci ied poin s in ime we e calcula ed by di iding he amoun o cumula i e e en s be o e he ime poin by he numbe o pa ien s a isk a he beginning o he ollow-up. All calcula ions we e pe o med wi h he SPSS 22.0 s a is ical package. RESULTS Baseline cha ac e is ics and in-hospi al da a o he s udy pa ien s we e epo ed p e iously [9]. The median age o pa ien s a s udy inclusion was 73 yea s (63–80 yea s) and he male/ emale a io was 58%/42%. The NSTEMI pa ien s we e olde (median age 75 yea s) han he STEMI (69 yea s) and UA (68 yea s) pa ien s. The ela i e p opo ion o emale pa ien s was highe in he NSTEMI han in he STEMI and UA ca ego ies (46%, 36%, and 37%, espec i ely; p = 0.003). The e we e no signi ican di e ences in he a e o hype ension (50–55%, p = 0.297) o diabe es (22–29%, p = 0.065) be ween he h ee g oups. The a e o diu e ic usage a admission was highes in he NSTEMI ca ego y (42%, 19%, and 32%, espec i ely; p < 0.001). The median su i al imes o he STEMI and NSTEMI ca ego ies we e 9.7 yea s and 4.7 yea s. The mean su i al imes we e 7.3 (95% con idence in e al [CI] 6.8–7.7), 5.4 (95% CI 5.0–5.7) and 7.7 (95% CI 7.2–8.3) o STEMI, NSTEMI and UA ca ego ies, espec i ely (p < 0.001). The 5-yea mo ali y a es we e 32.4%, 51.3%, and 25.3% (p < 0.001), while he 10-yea mo ali y a es we e 52.5%, 69.9%, and 41.0% (p < 0.001) o he STEMI, NSTEMI and UA ca ego ies, espec i ely (Fig. 1). Among all dea hs, 73.9%, 72.5% and 57.7% we e due o ca dio ascula causes o he STEMI, NSTEMI and UA pa ien ca ego ies, espec i ely (p = 0.019). Va iables p edic ing ou come a ollow-up acco ding o Cox uni a ia e and mul i a iable eg ession analyses a e p esen ed in Table 1. Age, male gende , ac i e smoking, diabe es, highe c ea inine le el, STEMI and NSTEMI ACS ca ego ies we e independen p edic o s o wo se ou come, while bypass su ge y and hype ension we e associa ed wi h be e ou come. Diu e ic use bo h a hospi al a i al and discha ge was associa ed wi h wo se ou come, while s a in use a discha ge was associa ed wi h be e ou come (Table 2). When mul i a iable Cox eg ession analysis was pe o med sepa a ely o he ACS ca ego ies, only age and c ea inine le el a admission p o ed o be independen ou come p edic o s o all h ee ca ego ies (Table 3). Ac i e smoking was an indica o o wo se ou come in bo h STEMI and NSTEMI ca ego ies. Diu e ic use a discha ge had a s ong nega i e impac on ou come bo h in NSTEMI and UA pa ien s (Table 2). In NSTEMI, which was he la ges pa ien ca ego y, in asi e ea men and be ablocke use a discha ge we e associa ed wi h be e ou come. DISCUSSION The p esen all-come s’ s udy showed ha : 1) all 3 pa ien ca ego ies o ACS ha e poo long- e m ou come, 2) NSTEMI pa ien s ha e he wo s ou come, 3) he su i al cu es o STEMI and NSTEMI pa ien s s ay clea ly sepa a ed o a ollow-up pe iod o ≥ 10 yea s, 4) UA pa ien s ha e be e ou come han MI pa ien s also in he long e m, and 5) ac o s a ec ing ou come di e be ween he h ee ACS ca ego ies. Randomized clinical ials and he eal-li e se ing in ACS: “Two di e en wo lds” In gene al, he e is limi ed da a on pa ien ou come in ACS beyond he i s ew yea s [9]. Especially, he e is e y li le long- e m mo ali y da a om comple e ACS coho s, which include STEMI, NSTEMI and UA pa ien s. Exis ing da a shows wide a ia ion in mo ali y e lec ing dis inc di e ences be ween andomized con olled ials wi h p e-speci ied exclusion c i e ia and “ eal-li e” popula ions, which include consecu i e pa ien s independen ly o co- mo bidi ies, e hnici y, age and gende . In andomized con olled ials o in asi ely ea ed STEMI pa ien s, he 5-yea mo ali y a e in STEMI may be as low as 10% [10]. The Global Regis y o Acu e Co ona y E en s (GRACE) s udy is widely acknowledged and has had signi ican impac on isk s a i ica ion in ACS [11]. In he “long- e m” GRACE s udy (GRACE UK-Belgian), 5-yea mo ali y o STEMI and NSTEMI pa ien s was 19% and 22%, espec i ely [9]. These igu es a e in s ong con as wi h he co esponding mo ali y igu es o 32.4%, and 51.3% in he p esen s udy. The 2002 New Zealand ACS Audi G oup ca ied ou a comp ehensi e collec ion o da a om all ACS pa ien s admi ed o a New Zealand hospi al o e a 14-day pe iod in May 2002, and ound mo ali y a es close o hose o he p esen s udy in STEMI pa ien s (34%), while he mo ali y a e (33%) o NSTEMI pa ien s was be ween ha epo ed in he GRACE UK-Belgian s udy and he p esen s udy [12]. Di e ences in pa ien age is p obably an impo an explana o y ac o o he obse ed a ia ion in mo ali y a es; age a s udy inclusion was 65/72/69 yea s o STEMI and 67/73/75 yea s o NSTEMI in GRACE, New Zealand ACS and TACOS, espec i ely. Also, a e ospec i e “ eal li e” analysis o 2,763 consecu i e ACS pa ien s ound much highe mo ali y a long- e m (median 8.2 yea s) in pa ien s > 65 yea s (69.7%) compa ed wi h hose ≤ 65 yea s (18.6%) [13]. When compa ing longe ou come in STEMI pa ien s, he 10-yea mo ali y a es in he New Zealand ACS audi s udy (48%) and he p esen s udy (52.5%) a e compa able. In NSTEMI pa ien s, highe 10-yea mo ali y a es we e ound: 51% and 69.9%, p obably no en i ely explained by he 2-yea age di e ence a s udy inclusion. A ecen me a-analysis o 8 andomized non-ST-ele a ion ACS (NSTE-ACS; NSTEMI and UA oge he ) ials included 6,657 pa ien s [14]. A a mean o 10.3 yea ollow-up, he isk o all-cause mo ali y was 28.5%. Again, his is ce ainly much lowe han in bo h NSTEMI (69.9%) and UA (41%) in he p esen s udy. Howe e , he mean age o he NSTE-ACS pa ien s in he me a-analysis was ~76 a he end o 10.3-yea ollow-up, while in he p esen s udy, he median age a s udy inclusion in NSTEMI pa ien s was 75 yea s (68 yea s o UA) [8]. STEMI/NSTEMI compa ison Clinical ial e idence is limi ed wi h ega d o he e icacy and haza ds o pha macological and in asi e managemen o NSTE-ACS in he elde ly. Acco ding o Alexande e al. [15], he age gap be ween ials and communi y popula ions begins a age 75 and widens wi h age. S udies ha e shown ha long- e m ou come in NSTEMI pa ien s is no imp o ing, and his has been a ibu ed o he ac ha hey ha e a mo e complex pheno ype [16]. Compa ed wi h STEMI pa ien s, hose wi h NSTEMI end o be olde and ha e mo e como bidi y. In he Wo ces e Hea A ack S udy (WHAS) wi h a popula ion 3,762 pa ien s, pos -discha ge dea h a es in a sub-coho wi h longe ollow-up, 5-yea dea h a es o STEMI (mean age o all pa ien s 65.5 yea s) and NSTEMI (mean age o all pa ien s 72.6 yea s) we e 30.2% and 52.4%, which a e in he same ange as in he p esen s udy: 32.4% o STEMI, 51.3% o NSTEMI [17]. Rega ding STEMI, he in oduc ion o p ima y pe cu aneous co ona y in e en ion (PCI) p og ams and imp o emen s in co ona y in e en ions and medical he apy ha e esul ed in de ini e imp o emen in pa ien ou come [18, 19]. Howe e , pa ien s > 75 yea s o age a e unde ep esen ed in andomized clinical STEMI ials; age o e 75 o 80 yea s was a ypicalexclusion c i e ia in many ials [20]. The e o e, limi ed da a is a ailable o guidance on he bes managemen o his g owing subse o pa ien s, al hough egis y da a seems o suppo he supe io i y o p ima y PCI o e conse a i e ea men also in he elde ly [21]. The Flo ence Acu e Myoca dial In a c ion Regis y (AMI-Flo ence) was a popula ion-based p ospec i e obse a ional egis y, whe e he baseline da a we e collec ed in 2000–2001 (2002–2003 in ou s udy) [21]. In STEMI pa ien s (n = 875), he 8-yea mo ali y a e was 49%, compa able o 42.3% in he p esen s udy. In AMI-Flo ence, p ima y PCI was pe o med in 50% o he STEMI pa ien s admi ed wi hin 24 h, whe eas in he cu en s udy 24% had PCI du ing he index hospi al admission, while 57% ecei ed ib inoly ic he apy [8]. Uns able angina pec o is Exis ing da a on he long- e m ou come o UA is sca ce mainly due o he ac ha esea che s end o combine NSTEMI and UA in o NSTE-ACS [22]. I was p e iously epo ed ha UA pa ien s (median age a s udy inclusion 68 yea s) had low in-hospi al mo ali y (2.6%), bu a 10 mon hs, he mo ali y a e had inc eased o 12% [8]. Wi h longe ollow-up, 5- and 10-yea mo ali y a es o UA pa ien s clea ly inc eased o 25.3% and 41%, espec i ely. The co esponding mo ali y a e a 10 yea s in he New Zealand ACS Audi ial was 32% [12]. In he GRACE UK- Belgian s udy, 5-yea mo ali y a e in UA was 18% [9]. Wi h he in oduc ion o he sensi i e oponins o de ec myoca dial inju y, i is p obable ha a conside able p opo ion o he UA pa ien s in he p esen s udy would be classi ied as NSTEMI using oday´s diagnos ic me hods [23]. P edic o s o mo ali y When analyzing all pa ien s oge he in he p esen s udy, he well-es ablished ca dio ascula isk ac o s e ained hei s a is ical signi icance as independen ou come p edic o s in he mul i a iable analyses. Howe e , only age and enal dys unc ion (highe c ea inine le els), which a e well documen ed isk ac o s, showed nega i e p ognos ic impac uni o mly in all h ee ACS ca ego ies. Fo example, ac i e smoking a ec ed ou come only in STEMI and NSTEMI pa ien s, while male gende was associa ed wi h in e io ou come only in UA pa ien s. P e ious s udy e idence o a gende di e ence in mo ali y in ACS pa ien s is con lic ing. In a majo sys ema ic e iew, Buchholz e al. [24] ound conside able he e ogenei y o s udy esul s when analyzing 26 s udies epo ing mo ali y a 5 o 9 yea s. Mos s udies epo ed clea a enua ion o s udy esul s a e co a ia es o he han age we e in oduced in he analyses. The ac ha diu e ic use had he s onges impac on he ou come o pa ien s in he NSTEMI ca ego y is no su p ising, as hese pa ien s we e olde and p obably had mo e co- mo bidi y, such as hea ailu e. In he PRAIS-UK egis y, which deal wi h NSTEMI pa ien s ea ed in he la e 1990’s, his o y o hea ailu e was a p edic o o in e io ou come du ing 10-yea ollow-up [25]. He ein, he e is no de ini e explana ion o he p o ec i e e ec on ou come o hype ension in he NSTEMI pa ien s, o he han possible posi i e e ec s on use o hype ensi e medica ion on en icula emodeling. Hype ension could also main ain ci cula ion o he kidneys longe in he se e ely ill, hypo olemic pa ien s and hence, a delayed p og ession o kidney ailu e. Limi a ions o he s udy This s udy has clea limi a ions; hose ela ed o da a collec ion and pa ien classi ica ion we e desc ibed p e iously [8]. The ollow-up o UA pa ien s was sho e han in he STEMI and NSTEMI g oups. The ca ego iza ion o hose wi h le bundle b anch block as NSTEMI o UA pa ien s could inc ease he isk o andom e o . Howe e , only nine pe cen o le bundle b anch block pa ien s we e ea ed wi h ib inoly ic he apy, which suppo s he decision o his classi ica ion. The e a e wo addi ional limi a ions ypical o ou come s udies wi h long ollow-up in pa ien s wi h ca dio ascula diseases. The i s limi a ion is he low a e o in asi e p ocedu es [17]. Especially in STEMI, he a e o in asi e p ocedu es du ing he index hospi al s ay in he p esen s udy was lowe han wha is ypical o Wes e n coun ies oday. Ye , mos (55%) pa ien s in he examined coho had NSTE-ACS, whe e he a e o in asi e p ocedu es did no inc ease as much as in he ea men o STEMI [26]. In addi ion, in he NSTEMI ca ego y, he median age a s udy inclusion was 75 yea s, and olde pa ien s end o ha e lowe a es o in asi e p ocedu es [27]. Also, he use o medical he apy is known o imp o e ou come, such as s a ins, we e no a he le el ha is expec ed in pa ien ca e oday. Because o hese limi a ions, he s udy esul s do no necessa ily e lec he ou come o ACS pa ien s ea ed acco ding o a mode n s anda d. Ano he gene al limi a ion o s udies wi h long- e m ollow-up is he ac ha changes in pa ien medica ion and new co ona y in e en ions a e di icul o impossible o con ol o . Conclusions All h ee ACS ca ego ies he ein p o ed o ha e high mo ali y a es du ing long- e m ollow-up in a eal-li e pa ien coho . NSTEMI pa ien s had wo se ou comes han STEMI and UA pa ien s du ing he whole ollow-up pe iod. The p esen s udy esul s also indica es conside able di e ences in he p ognos ic signi icance o a ious demog aphic and he apeu ic pa ame e s wi hin he h ee ACS ca ego ies. Age Male gende Ac i e smoking Hype ension Diabe es: No diabe es Diabe es melli us ype 1 Diabe es melli us ype 2 P e ious MI Plasma c ea inine [/10 µmol/L] C- eac i e p o ein [/10 mg/L] cTnI [/10µmol/L] Medica ion a admission: Diu e ic ACE-inhibi o Wa a in 1.044 1.121 1.537 0.753 1.774 1.144 1.066 1.037 0.999 1.035 1.827 1.104 1.370 1.029–1.060 0.892–1.410 1.091–2.165 0.593–0.955 0.637–4.939 0.911–1.436 0.842–1.351 1.020–1.055 0.985–1.013 1.014–1.058 1.411–2.366 0.864–1.412 1.003–1.870 < 0.001 0.328 0.014 0.019 0.272 0.247 0.595 < 0.001 0.874 0.001 < 0.001 0.429 0.048 PTCA CABG Medica ion a discha ge: Be a-blocke Diu e ic S a in Digi alis 0.569 0.456 0.554 2.104 0.795 1.250 0.374–0.864 0.310–0.673 0.352–0.872 1.547–2.862 0.629–1.005 0.951–1.642 0.008 < 0.001 0.011 < 0.001 0.055 0.109 UAP ca ego y Age 1.117 1.073–1.164 < 0.001 Male gende 3.400 1.625–7.113 0.001 Ac i e smoking 1.995 0.614–6.481 0.251 Hype ension 1.003 0.558–1.805 0.992 Diabe es: No diabe es Diabe es melli us ype 1 131.881 0.882–19712.989 0.056 Diabe es melli us ype 2 2.103 1.173–3.770 0.013 P e ious MI 0.696 0.361–1.345 0.281 Plasma c ea inine [/10 µmol/L] 0.946 0.905–0.989 0.015 C- eac i e p o ein [/10 mg/L] 1.221 1.102–1.352 < 0.001 Medica ion a admission: Diu e ic 0.683 0.296–1.577 0.372 ACE-inhibi o 1.354 0.704–2.606 0.364 Wa a in 0.700 0.342–1.429 0.327 PTCA 0.028 0.000–4.118 0.160 CABG 0.222 0.047–1.039 0.056 Medica ion a discha ge: Be a-blocke 1.281 0.571–2.874 0.548 Diu e ic 4.807 1.937–11.931 0.001 S a in 1.131 0.610–2.099 0.695 Digi alis 0.907 0.432–1.900 0.795 Figu e 1. Kaplan-Meie es ima es o su i al and he numbe a isk a di e en ime poin s in he h ee acu e co ona y synd ome ca ego ies. The y axis shows he p opo ion o pa ien s ali e a di e en ime poin s (1.0 = 100%).