Ima 1

8122019 Ima 1

httpslidepdfcomreaderfullima-1 19

Articles

wwwthelancetcom Published online January 23 2014 httpdxdoiorg101016S0140-6736(13)62070-X 1

Acute myocardial infarction a comparison of short-term

survival in national outcome registries in Sweden and the UKSheng-Chia Chung Rolf Gedeborg Owen Nicholas Stefan James Anders Jeppsson Charles Wolfe Peter Heuschmann Lars Wallentin

John Deanfield Adam Timmis Tomas Jernberg Harry Hemingway

SummaryBackground International research for acute myocardial infarction lacks comparisons of whole health systems Weassessed time trends for care and outcomes in Sweden and the UK

Methods We used data from national registries on consecutive patients registered between 2004 and 2010 in allhospitals providing care for acute coronary syndrome in Sweden and the UK The primary outcome was all-causemortality 30 days after admission We compared effectiveness of treatment by indirect casemix standardisation Thisstudy is registered with ClinicalTrialsgov number NCT01359033

Findings We assessed data for 119 786 patients in Sweden and 391 077 in the UK 30-day mortality was 7middot6 (95 CI7middot4ndash7middot7) in Sweden and 10middot5 (10middot4ndash10middot6) in the UK Mortality was higher in the UK in clinically relevant subgroupsdefined by troponin concentration ST-segment elevation age sex heart rate systolic blood pressure diabetesmellitus status and smoking status In Sweden compared with the UK there was earlier and more extensive uptakeof primary percutaneous coronary intervention (59 vs 22) and more frequent use of β blockers at discharge (89vs 78) After casemix standardisation the 30-day mortality ratio for UK versus Sweden was 1middot37 (95 CI 1middot30ndash1middot45)which corresponds to 11 263 (95 CI 9620ndash12 827) excess deaths but did decline over time (from 1middot47 95 CI1middot38ndash1middot58 in 2004 to 1middot20 1middot12ndash1middot29 in 2010 p=0middot01)

Interpretation We found clinically important differences between countries in acute myocardial infarction care andoutcomes International comparisons research might help to improve health systems and prevent deaths

Funding Seventh Framework Programme for Research National Institute for Health Research Wellcome Trust (UK)Swedish Association of Local Authorities and Regions Swedish Heart-Lung Foundation

IntroductionRecognition is growing of the need for comparativeeffectiveness research to improve the quality andoutcomes of health care International comparisons ofcancer survival1 and years of life lost to ischaemic heartdisease2 from 1990ndash2010 suggest that the performanceof the UK health system needs to be improved Howeverstudies that simultaneously examine care and outcomesare lacking The Institute of Medicine identified health-care delivery systems and cardiovascular care as amongthe highest priorities for comparative effectiveness

research3 The effi cacy of treatments for acutemyocardial infarction has been extensively studied inrandomised trials45 but uptake and use of thesetreatments vary within and between the UK (Englandand Wales) and Sweden6ndash10 A study in patients with ST-segment-elevation myocardial infarction (STEMI) in12 European countries reported increasing use ofprimary percutaneous coronary intervention (PCI)but showed striking differences betweencountries 11 Attributes of care systems includingorganisational culture care pathways and programmesto improve quality are not assessed in trials but mightbe associated with outcome12 30-day mortality for acutemyocardial infarction is an important indicator ofhospital performance and delivery of care has more

immediate potential to improve outcomes thantreatment innovations13 International comparisons ofwhole health-care delivery systems therefore mightyield important actionable insights to guidedevelopment of policies and clinical practice14

International comparative effectiveness research foracute myocardial infarction has had three mainlimitations First is a lack of comparison of whole healthsystems Existing studies lack population coveragebecause they are based on selected samples of hospitalpatients reported in voluntary registries8ndash10 one-off

surveys15 or trials16 that are known to differ from thenational population in treatments and outcomes17 Secondinternational studies have compared only care8 oroutcomes18 or have been restricted to patients with eitherSTEMI7ndash9 or non-STEMI10 Third attempts have not beenmade to standardise the mortality of patients in onecountry by the casemix in another As a result there arefew studies between health systems from which to setbenchmark outcome goals

A crucial feature of the health systems in Sweden andthe UK is that they are the only two countries worldwidethat have continuous national clinical registries for acutecoronary syndrome with mandated participation for allhospitals1920 Comparison of these two countries isfacilitated by the similarity of their health systems

Published Online

January 23 2014

httpdxdoiorg101016

S0140-6736(13)62070-X

See OnlineComment

httpdxdoiorg101016

S0140-6736(13)62367-3

Copyright copy Chung et al Open

Access article distributed underthe terms of CC BY-NC-ND

Farr Institute of Health

Informatics Research at UCL

Partners (S-C Chung PhD

Prof H Hemingway FRCP)

National Institute for Clinical

Outcomes Research

(O Nicholas PhD) and Centre

for Cardiovascular Prevention

and Outcomes

(Prof J Deanfield FRCP)

University College London

London UK Uppsala Clinical

Research Centre

(R Gedeborg MD

S James MD Prof L Wallentin MD)

and Department of Medical

Sciences Cardiology (S James

Prof L Wallentin) Uppsala

University Uppsala Sweden

Department of Molecular and

Clinical Medicine Institute of

Medicine Sahlgrenska

Academy University of

Gothenburg Gothenburg

Sweden (Prof A Jeppsson MD)

Department of Cardiothoracic

Surgery Sahlgrenska University

Hospital Gothenburg Sweden

(Prof A Jeppsson) Division of

Health and Social Care Research

Kingrsquos College London London

UK (Prof C Wolfe MD) National

Institute of Health Research

Comprehensive Biomedical

Research Centre Guyrsquos amp

St Thomasrsquo NHS Foundation

Trust and Kingrsquos College

London London UK

(Prof C Wolfe) Institute for

Clinical Epidemiology and

Biometry and Comprehensive

Heart Failure Centre University

of Wuumlrzburg Wuumlrzburg

Germany (P Heuschmann MD)

National Institute for Health

Research Biomedical Research

Unit Bartrsquos Health London

London UK

(Prof A Timmis FRCP) and

Department of Medicine

8122019 Ima 1


Articles

2 wwwthelancetcom Published online January 23 2014 httpdxdoiorg101016S0140-6736(13)62070-X

Huddinge Section of

Cardiology Karolinska

Institute Stockholm Sweden

(T Jernberg MD)

Correspondence to

Dr Tomas Jernberg Department

of Cardiology Karolinska

University Hospital Karolinska

Institute Huddinge 141 86

Stockholm Sweden

tomasjernbergkarolinskase

or

Prof Harry Hemingway Farr

Institute of Health Informatics

Research at UCL Partners


222 Euston Road

London NW1 2DA UK

hhemingwayuclacuk

(universal funded from taxation and free at the point of

use) proportion of gross domestic product spent onhealth and national policy guidance provided for theevidence-based management of acute myocardialinfarction2122 Differences are that Sweden has morerapid diffusion of some new technologies23 morecomplete use of evidence-based practice9 and a moreestablished system for evaluating and reporting thequality and outcomes of care24 than the UK

In the absence of previous international comparisonsour objectives in this study were as follows first to assessthe validity of comparing data from the two nationwideclinical registries second to compare time trends forproportions of patients in receipt of effective interventionswhile in hospital and at discharge between 2004 and 2010

third to compare crude and casemix-standardised 30-daymortality between the two countries and between clinicallyimportant subgroups fourth to estimate time trends incasemix-standardised mortality and compare thembetween Sweden and the UK and fifth to explore thecontribution of clinical care to any difference in mortality

MethodsStudy populationAll hospitals providing care for acute myocardialinfarction in Sweden and the UK contribute data onconsecutive patients to the Swedish Web-System forEnhancement and Development of Evidence-BasedCare in Heart Disease Evaluated According toRecommended Therapies (SWEDEHEART)Register ofInformation and Knowledge about Swedish HeartIntensive care Admissions (RIKS-HIA) and the UKMyocardial Ischaemia National Audit Project (MINAP)

respectively RIKS-HIA and MINAP comply with the

Cardiology Audit and Registration Data Standards (foracute coronary syndrome in Europe72526 We obtaineddata for all patients who were admitted to hospitalbecause of acute myocardial infarction between Jan 12004 and Dec 31 2010 For patients with multipleadmissions we used the earliest record Acutemyocardial infarction diagnosis was based on guidelinesfrom the European Society of CardiologyAmericanCollege of CardiologyAmerican Heart Association27 Thestudy was approved by the MINAP Academic Group andthe steering group of SWEDEHEART

Casemix and treatment measuresThe definitions of casemix evidence-based hospital

treatment and discharge medications in SWEDEHEARTRIKS-HIA and MINAP were compared (appendixpp 2ndash9) The 17 casemix characteristics were demographicfactors (age sex year of admission) risk factors (smokinghistory of diabetes mellitus and hypertension) severity ofacute myocardial infarction (troponin concentrationssystolic blood pressure at admission heart rate atadmission) history of heart failure cardiac arrest atadmission history of cerebrovascular disease or historyof acute myocardial infarction and previous proceduresor use of medication (antiplatelet treatment with aspirinclopidogrel or both PCI coronary artery bypass graftsurgery) Features of hospital treatment for STEMI werereperfusion therapy (primary PCI or fibrinolytic therapybefore or during hospital stay) delay between symptomonset and primary PCI or fibrinolysis coronaryintervention other than primary PCI intravenousglycoprotein IIbIIIa inhibitors and use of anticoagulantsDischarge medications assessed were antiplatelet therapy(aspirin clopidogrel or both) β blockers angiotensin-converting-enzyme (ACE) inhibitors or angiotensin-receptor blockers (ARBs) and statins Validation ofSWEDEHEARTRIKS-HIA (each year a trained monitorcompares data with a chart review in 30 randomly selectedpatients within each hospital) showed a 96middot1agreement20 Validation of MINAP data (compared withreaudit data generated from each hospital annually by

the re-entering of 20 data items on 20 randomly selectedpatients) showed a median agreement of 89middot519

MortalityThe primary clinical outcome was all-cause mortalitywithin 30 days after hospital admission National uniqueidentifiers were used to link patients with the NationalDeath Registry in Sweden or the Offi ce for NationalStatistics in the UK We accessed these registries toascertain vital status or date of death at 30 days

Statistical analysisWe compared troponin I and T concentrations casemixhospital treatment and medication at discharge betweenSwedish and UK patients Data are shown as proportion

See Online for appendix

All hospitals providing AMI careSweden 86 UK 242

Admission records from all hospitals

Sweden 414831 UK 739 828

Study population of AMI patientsSweden 119786 UK 391077

Deaths at 30 daysSweden 9173 UK 41 509

SwedenUK excluded 276 147330175

ineligible records51626998 age lt30 years or missing

266 78773 031 non-AMI diagnosis 0126916 admitted before or after

inclusion dates 4198123 230 did not meet other

demographic criteria1889818576 subsequent admissions

Figure 983089 Study populationAMI=acute myocardial infarction ID=identifier

8122019 Ima 1


Articles


(95 CI) for categorical variables and mean (SD) or

median (IQR) for continuous variables To furtherinvestigate the comparability of acute myocardialinfarction diagnosis in the two registries we comparedthe propensity of acute myocardial infarction diagnosisbetween UK and Sweden patients (appendix pp 12ndash13)

Treatments for acute myocardial infarction were also

compared by year of hospital admissionWe compared mortality outcomes with Kaplan-Meieranalysis in clinically important subgroups (STEMI ornon-STEMI troponin concentration systolic bloodpressure at admission heart rate at admission sex age

Sweden (n=119 786) UK (n=391 077)

Casemix

STEMI 38 432 (32middot1 31middot8ndash32middot3) 157 418 (40middot3 40middot1ndash40middot4)

Mean (SD) age (years) 71middot2 (12middot3) 69middot5 (13middot6)

Female 43 512 (36middot3 36middot1ndash36middot6) 135 664 (34middot8 34middot7ndash34middot9)

Median (IQR) AMI severity

Systolic blood pressure (mm Hg) 145 (125ndash165) 139 (120ndash158)

Heart rate (beats per min) 78 (65ndash93) 79 (66ndash94)

Troponin I (microgL) 4middot2 (0middot8ndash18middot0) 4middot4 (0middot8ndash21middot7)

Troponin T (microgL) 0middot7 (0middot2ndash2middot3) 0middot65 (0middot2ndash2middot3)

Risk factor

Current smoker 25 085 (23middot3 23middot0ndash23middot5) 104 522 (29middot5 29middot3ndash29middot6)

History of diabetes 26 992 (22middot7 22middot4ndash22middot9) 65 458 (17middot6 17middot4ndash17middot7)

History of hypertension 53 155 (45middot2 44middot9ndash45middot5) 173 342 (47middot3 47middot2ndash47middot5)

Cardiovascular disease history

Heart failure 10 859 (9middot7 9middot5ndash9middot8) 18 944 (5middot3 5middot2ndash5middot4)

Cardiac arrest before admission 1578 (1middot3 1middot3ndash1middot4) 8478 (2middot3 2middot2ndash2middot3)

Cerebrovascular disease 9816 (10middot1 9middot9ndash10middot3) 30 091 (8middot5 8middot4ndash8middot5)

Myocardial infarction 26 526 (22middot4 22middot1ndash22middot6) 67 346 (18middot3 18middot1ndash18middot4)

Prehospital treatment

Antiplatelet monotherapy 43 485 (36middot6 36middot3ndash36middot9) 98 247 (26middot4 26middot3ndash26middot6)Antiplatelet dual therapy 4788 (4middot0 3middot9ndash4middot1) 10 931 (2middot9 2middot9ndash3middot0)

PCI 9475 (8middot0 7middot8ndash8middot2) 19 473 (5middot4 5middot4ndash5middot5)

CABG 9192 (7middot7 7middot6ndash7middot9) 17 383 (4middot8 4middot8ndash4middot9)

Hospital treatment

STEMI patients

Total reperfusion treatment 27 354 (71middot2 70middot7ndash71middot6) 118 880 (76middot6 76middot4ndash76middot8)

Prehospital fibrinolysis 1533 (4middot1 3middot9ndash4middot4) 13 903 (9middot3 9middot2ndash9middot5)

In-hospital fibrinolysis 4539 (11middot8 11middot5ndash12middot1) 84 112 (54middot2 54middot0ndash54middot5)

Primary PCI 22 773 (59middot3 58middot8ndash59middot8) 34 695 (22middot4 22middot2ndash22middot6)

Median (IQR) delay from symptom to fibrinolysis (min) 177 (108ndash322) 150 (94ndash285)

Median (IQR) delay from symptom to primary PCI (min) 198 (129ndash365) 199 (140ndash328)

Non-STEMI and STEMI patients

Coronary intervention other than primary PCI 34 288 (28middot6 28middot4ndash28middot9) 58 492 (17middot3 17middot2ndash17middot5)IV glycoprotein IIbIIIa receptor inhibitors 24 993 (21middot0 20middot8ndash21middot2) 28 389 (8middot6 8middot5ndash8middot7)

Anticoagulants 87 271 (73middot2 73middot0ndash73middot5) 28 3344 (83middot0 82middot9ndash83middot1)

Median (IQR) duration of hospital stay (days) 5 (3ndash7) 6 (3ndash10)

Discharge medication

Antiplatelet monotherapy 30 409 (27 26middot8ndash27middot3) 103 218 (34middot2 34middot1ndash34middot4)

Antiplatelet dual therapy 76 099 (67middot6 67middot3ndash67middot9) 183 753 (60middot9 60middot8ndash61middot1)

β blocker 99 779 (88middot7 88middot5ndash88middot9) 231 505 (78middot2 78ndash78middot3)

ACE inhibitor or ARB 63 102 (56middot2 55middot9ndash56middot5) 242 300 (82middot3 82middot2ndash82middot5)

Statin 89 767 (79middot7 79middot5ndash79middot9) 276 335 (92middot8 92middot7ndash92middot9)

Values are number ( 95 CI) unless stated otherwise STEMI=ST-segment-elevation myocardial infarction AMI=acute myocardial infarction PCI=percutaneous coronary

intervention CABG=coronary artery bypass graft surgery IV=intravenous ACE=angiotensin-converting enzyme ARB=angiotensin-receptor blocker Prehospital and

in-hospital fibrinolysis are not mutually exclusive

Table 983089 Casemix and treatment for patients with AMI in Sweden and the UK

8122019 Ima 1


Articles


year of admission diabetes status and smoking) aftercasemix standardisation and by propensity-scorematching In casemix standardisation we modelled30-day mortality for the two countries with the 17 casemixvariables then applied the Sweden model to the UK acutemyocardial infarction population to estimate the casemix-standardised relative risk of observed UK 30-day mortalityfor each study year and overall In propensity matchingwe estimated the propensity of being a STEMI patient

and the propensity of being a non-STEMI patient inSweden and the UK according to logistic regression andthen matched patients on the basis of propensity scores(appendix pp 12ndash14)

Casemix models incorporated a random effect forparticipant hospital The extent of missing values wasassessed and then managed by multiple imputation(appendix pp 16ndash19) Analyses were performed with SAS(version 93) R (version 292) and Matlab (version 714)This study is registered with ClinicalTrialsgov numberNCT01359033

Role of the funding sourceThe sponsors of the study had no role in study design

data collection data analysis data interpretation orwriting of the report S-CC had full access to all the datain the study TJ and HH had final responsibility for thedecision to submit for publication

ResultsThe study population was drawn from 86 hospitals inSweden and 242 in the UK 119 786 patients were eligiblein Sweden and 391 077 in the UK (figure 1) and data on30-day mortality were available for 119 786 (100) and390 951 (99middot97) of these respectively

Median concentrations (IQR) maximum troponin Iand troponin T were similar in Sweden and the UKoverall (table 1) and for STEMI and non-STEMI patientsby year (appendix pp 10ndash11)

Primary PCI SwedenPrimary PCI UK

0

20

40

60

80

100

P r o p o r t i o n ( )

Any fibrinolysis Sweden

Any fibrinolysis UK

Reperfusion

Any antiplatelet SwedenStatin Sweden

2004 2005 2006 2007 2008 2009 20100

20

40

60

80

100


Years2004 2005 2006 2007 2008 2009 2010

Years

Any antiplatelet UKStatin UK

β blocker SwedenACEIARB Sweden

β blocker UKACEIARB UK

Discharge medications

A B

C D

Figure 983090 Use of reperfusion or fibrinolysis to treat STEMI and medication at discharge among all patients by year

(A) Primary PCI and (B) fibrinolysis including any given before admission or in hospital (C) Any antiplatelet therapy and statin and (D) use of β blockers and ACEI or

ARBs among all acute myocardial infarction patients who survived to discharge STEMI=ST-segment-elevation myocardial infarction PCI=percutaneous coronary

intervention ACEI=angiotensin-converting-enzyme inhibitor ARB=angiotensin-receptor blocker

Figure 983091 Kaplan-Meier curves for cumulative mortality at 30 days after admission with acute myocardial

infarction in Sweden and the UK

Time of censoring or vital status at 30 days missing for 129 patients in the UK

Number at risk

SwedenUK

0 5 10 15 20 25

UK 30-day risk of death 10middot5(95 CI 10middot4ndash10middot6)

Sweden 30-day risk of death 7middot6(95 CI 7middot4ndash7middot7)

30

119786390948

115 113370883

113 364362 830

112 345357954

111 646354475

111 136351919

110693349845

Time since admission (days)

0

2

4

6

8

10

C u m u l a t i v e m o r t a l i t y f r o m a

n y c a u s e ( )

8122019 Ima 1


Articles


The proportion of patients with STEMI was lower in

Sweden than in the UK (32 vs 40) The distributionsof age and sex were similar in the two countries (table 1)Swedish patients had more favourable risk profiles thanUK patients for some factors (eg lower prevalence ofcurrent smoking and higher systolic blood pressure atadmission) but worse risk profiles for other factors (eghigher prevalence of diabetes heart failure andcerebrovascular disease) Previous use of antiplatelet andβ-blocker therapy was greater in Sweden but use ofstatins on admission was lower (table 1)

Total reperfusion for STEMI was more common in theUK than in Sweden (77 vs 71) Fibrinolysis was alsomore common in the UK (54 vs 12) but primary PCIwas notably more common in Sweden (59 vs 22

table 1) The median delay from symptom onset tohospital admission was similar in the two countries Themedian delay from symptom onset to primary PCI inSTEMI patients was similar but for fibrinolysis the delaywas longer in Sweden (table 1) Overall coronaryinterventions (other than primary PCI) and use ofintravenous glycoprotein IIbIIIa agents were higher inSweden than in the UK For patients who survived tohospital discharge those in Sweden were more likelythan those in the UK to be prescribed dual antiplatelettherapy or β blockers but less likely to be prescribed ACEinhibitors or ARBs and statins (table 1)

In both countries the use of primary PCI to treatSTEMI increased over time and use of fibrinolysisdecreased (figure 2 appendix p 20) In 2004 almost allSTEMI patients in the UK received fibrinolysis but thisdecreased substantially over time and use of primary PCIreached the Swedish 2004 rate in 2009 Over time theuse of any antiplatelet medication at discharge wassimilar in the two countries Use of dual antiplatelettherapy at discharge was initially low in the UK butincreased over time and had exceeded that in Sweden by2009 (appendix p 20) Use of β blockers at discharge wasconstantly higher in Sweden whereas use of ACEinhibitors or ARBs and use of statins were higher in theUK (figure 2)

Cumulative 30-day mortality was higher in the UK

than in Sweden (figure 3) When assessed according totroponin concentrations acute infarct severity (heartrate and blood pressure at admission) age sex year ofadmission smoking and diabetes status 30-daymortality in the UK remained consistently higher thanthat in Sweden (table 2) The strength and directionof casemix-adjusted associations between 30-daymortality and age year of admission blood pressureheart rate diabetes status history of cardiovasculardisease and previous revascularisation were similar inSweden and the UK (appendix p 18) In-hospitalmortality was also higher in the UK (8middot8 8middot7ndash8middot9)than in Sweden (5middot8 5middot7ndash5middot9) In clinically importantsubgroups in-hospital mortality was consistentlyhigher in the UK than in Sweden (appendix pp 21)

After standardisation with the Swedish casemix

model UK mortality was lower than the unadjustedcrude estimates (appendix p 22) After standardisationthe forecast mean 30-day mortality between 2004 and

Numbe r of deathspatients 3 0-d ay mortality ( 95 CI )

Sweden

(n=119 786)

UK

(n=391 077)

Sweden

(n=119 786)

UK

(n=391 077)

ST-segment elevation

STEMI 324837 937 17 68115 7365 8middot6 (8middot3ndash8middot8) 11middot2 (11middot1ndash11middot4)

Non-STEMI 592380 724 23 82823 3586 7middot3 (7middot2ndash7middot5) 10middot2 (10middot1ndash10middot3)

Severity of myocardial infarction

Troponin I (microgL)

lt5middot0 148729 793 599480 991 5middot0 (4middot7ndash5middot2) 7middot4 (7middot2ndash7middot6)

5middot0ndash9middot9 4846725 167017 524 7middot2 (6middot6ndash7middot8) 9middot5 (9middot1ndash10middot0)10middot0ndash20middot0 4876256 156016 342 7middot8 (7middot1ndash8middot5) 9middot5 (9middot1ndash10middot0)

ge20middot0 140413 255 463040 765 10middot6 (10middot1ndash11middot1) 11middot4 (11middot1ndash11middot7)

Troponin T (microgL)


0middot2ndash0middot5 59810 263 206023 937 5middot8 (5middot4ndash6middot3) 8middot6 (8middot3ndash9middot0)

0middot5ndash1middot0 5597963 167417 774 7middot0 (6middot5ndash7middot6) 9middot4 (9middot0ndash9middot9)


Admission systolic blood pressure (mm Hg)

lt110 22799729 10 53247 796 23middot4 (22middot6ndash24middot3) 22middot0 (21middot7ndash22middot4)

110ndash140 288231 750 13 331123 072 9middot1 (8middot8ndash9middot4) 10middot8 (10middot7ndash11)

ge140 273262 743 11 406166 274 4middot4 (4middot2ndash4middot5) 6middot9 (6middot7ndash7middot0)

Admission heart rate (beats per min)

lt90 414373 584 18 307230 648 5middot6 (5middot5ndash5middot8) 7middot9 (7middot8ndash8middot0)90ndash120 276024 235 1269482674 11middot4 (11ndash11middot8) 15middot4 (15middot1ndash15middot6)

ge120 11037933 441824 525 13middot9 (13middot2ndash14middot7) 18middot0 (17middot5ndash18middot5)

Demographic characteristics

Male 506975 406 22 54925 4110 6middot7 (6middot5ndash6middot9) 8middot9 (8middot8ndash9middot0)

Female 410243 255 18 823135 608 9middot5 (9middot2ndash9middot8) 13middot9 (13middot7ndash14middot1)

Age

lt65 years 71735 262 4487142 392 2middot0 (1middot9ndash2middot2) 3middot2 (3middot1ndash3middot2)

65ndash75 years 138529 859 765194 442 4middot6 (4middot4ndash4middot9) 8middot1 (7middot9ndash8middot3)

75ndash85 years 373136 324 16 288101 943 10middot3 (10ndash10middot6) 16middot0 (15middot8ndash16middot2)

ge85 years 333817 216 13 08352 174 19middot4 (18middot8ndash20) 25middot1 (24middot7ndash25middot4)

Year of admission

2004 180318 294 730055 447 9middot9 (9middot4ndash10middot3) 13middot2 (12middot9ndash13middot4)


2006 136816 865 596652 757 8middot1 (7middot7ndash8middot5) 11middot3 (11ndash11middot6)



2009 101015 796 523758 479 6middot4 (6ndash6middot8) 9middot0 (8middot7ndash9middot2)


Risk factors

Diabetes 265726 769 844465 424 9middot9 (9middot6ndash10middot3) 12middot9 (12middot7ndash13middot2)

No diabetes 637191 275 29 647307 409 7middot0 (6middot8ndash7middot1) 9middot6 (9middot5ndash9middot7)

Current smoker 108624 651 5765104 503 4middot4 (4middot2ndash4middot7) 5middot5 (5middot4ndash5middot7)

Non-smoker 593982 205 26 463249 814 7middot2 (7ndash7middot4) 10middot6 (10middot5ndash10middot7)

STEMI=ST-segment-elevation myocardial infarction

Table 983090 30-day mortality in patients with acute myocardial infarction in Sweden and the UK by clinically

relevant subgroups

8122019 Ima 1


Articles


2010 was 7middot7 (95 CI 7middot3ndash8middot2) If the same casemixwas assumed in Sweden and the UK the standardisedmortality ratio was 1middot37 (1middot30ndash1middot45) whichcorresponded to an estimated 11 263 (9620ndash12 827) moredeaths in the UK between 2004 and 2010 than inSweden The greatest annual difference betweencountries in mortality was seen in 2004 when thestandardised mortality ratio was 1middot47 (1middot38ndash1middot58) butdecreased significantly over time to 1middot20 (1middot12ndash1middot29) in2010 (figure 4) The propensity-matched analyses gavesimilar findings (appendix p 15) We explored the extentto which mortality differences between countries mightbe explained by differences in medical care by casemixstandardisation and treatment in hospital and also byestimating what might have differed if the UK had thesame level of use of primary PCI and β blockersas Sweden from 2004 onwards assuming treatmentbenefits reported in randomised clinical trials If thelevel of use of primary PCI and β blockers had been thesame in the UK as in Sweden we estimate thestandardised mortality ratio would have reduced from1middot37 to 1middot31 (95 CI 1middot30ndash1middot33 appendix pp 22 24)When in-hospital treatments were included in additionto casemix the standardised mortality ratio decreased

from 1middot37 to 1middot21 (1middot15ndash1middot29 appendix pp 22ndash23)

DiscussionWe found greater mortality among patients with acutemyocardial infarction in the UK than similar patients inSweden The differences in the care and outcomes ofacute myocardial infarction are a cause for concernUptake of effective treatments especially primary PCI totreat STEMI and β blockers at discharge was slower inthe UK than in Sweden The greater cumulative 30-daymortality in the UK was much improved afterstandardisation with the Swedish casemix This approachsuggests that more than 10 000 deaths at 30 days wouldhave been prevented or delayed had UK patientsexperienced the care of their Swedish counterparts

Several lines of evidence support the validity of

comparing care and outcomes across these registriesFirst by design we captured data on the whole system inboth countries (all hospitals and consecutive patients)and have used common data definitions That these dataare comparable is supported by our finding that theassociations between casemix variables and mortalitywere similar in the two countries Second the diagnosisof acute myocardial infarction was comparable withtroponin values and propensity to make a diagnosis beingsimilar Third differences in 30-day mortality betweencountries were consistent within strata defined bytroponin values and other clinically important subgroups

The use of primary PCI to treat STEMI in the UKlagged behind that in Sweden the rate in 2010 in the UK

(53) was similar to that in Sweden in 2005 (50)Primary PCI was more effective than fibrinolysis in ameta-analysis28 and this finding was reflected inguideline recommendations in the USA in 200429 and inEurope in 200530 The UK did not have a national policyfor primary PCI until October 200831 which couldexplain the rapid increase in the use of this treatmentfrom 2008 onwards but it took until 2011ndash12 for rates toexceed 9032

The use of evidence-based secondary preventionshowed a mixed pattern with statin therapy and ACEinhibitors or ARBs being more commonly prescribed inthe UK than in Sweden whereas use of β blockers atdischarge was more common in Sweden β blockers havebeen reported to be effi cacious for secondary preventionof acute myocardial infarction in randomised controlledtrials and have been included in guidelinerecommendations for acute myocardial infarction inEurope33 and the USA34 since 1996 but their use was notrecommended in the UK until 200135

Our findings suggest that these differences in clinicalcare contributed to international differences in patientsrsquooutcomes The mortality gap between Sweden and theUK decreased over time which is consistent with thenarrowing gap between treatments Application ofSwedish rates of primary PCI and β blocker use to the UKpopulation or inclusion of in-hospital treatments in the

standardisation were associated separately with reduceddifferences in mortality between the two countries

Other features of health care are not measured in theseregistries (eg why particular patients do or do not receiveparticular types of care) and these features probablycontribute to explaining the mortality gap We show thatimproved understanding of selection processes in the twocountries is needed (appendix p 23) Mortality might beaffected by multiple unmeasured features of careincluding doses timing adherence to drugs differencesin operator experience shared and specialty care pathwaysuse of decision-support tools and organisational culture12

In all countries patients payers and policymakerscould ask how outcomes in their health systems comparewith our results Policy initiatives are required to identify

Figure 983092 30-day mortality of UK patients admitted in each study y ear standardised according to the Swedish

casemix model

p=0middot01 for linear trend across years for the relative risks AMI=acute myocardial infarction

Casemix-

standardisedrisk ratio (95 CI)

Number of

UK AMIpatients

Observed

UK 30-daymortality ()

Year

2004

2005

2006

2007

2008

2009

2010

55447

54421

52757

53276

55 851

58479

60720

13middot2

12middot6

11middot3

10middot4

9middot9

9middot0

8middot4

1middot47 (1middot38ndash1middot58)







1middot00middot8 1middot61middot1 1middot2 1middot3 1middot4 1middot50middot9

Mortality Sweden higher

Standardised relative risk

Mortality UK higher

8122019 Ima 1


Articles


understand and reduce gaps between treatment use and

outcomes in different health systems We suggest thatprogress towards this goal is achievable because thenarrowing of the gap between mortality in Sweden andthe UK indicates that differences are reversible Between-country comparisons of nationwide care and outcomes isa novel approach most quality-outcome initiatives so farhave been concentrated on within-country and within-system metrics For Sweden our results highlight thevalue of quality of sustained system-wide initiatives toimprove quality including the public reporting ofoutcomes at hospital level For the UK our results suggestthe usefulness of learning from systems that seem to beperforming better12

Our study has several limitations First we cannot

exclude the possibility that unmeasured features ofcasemix contributed to the differences in mortalityHowever standardisation stratification and analysis ofpropensity scores all reported similar results whichsupports an actual and significant difference betweencountries Our casemix model is as comprehensive as theregistries allow and included 17 variables in demographyacute myocardial infarction severity risk factorcomorbidity and prehospital treatment Second we couldnot assess the care received by patients who died beforereaching hospital although we believe this is unlikely toexplain the higher mortality in the UK because the timefrom symptom onset to admission was similar to that inSweden and differences in mortality became apparentafter the first day in hospital Third the registries do notcapture all patients admitted with acute myocardialinfarction and in the UK missed patients are likely to beolder and less likely to be under the care of a cardiologistthan patients recorded in the registry36 Because thenationwide registries in Sweden have been establishedlonger than in the UK missed cases might be less frequentin Sweden which suggests the actual difference inmortality could be wider Fourth the quality andcompleteness of data (in themselves markers of quality ofcare) might introduce bias in our casemix model Insensitivity analyses however estimates for the associationsbetween casemix variables and 30-day mortality based on

complete case analysis verified the results from multipleimputed data (appendix pp 18ndash19)

Our comparison of international outcomes suggests anovel research agenda (panel)37 First additional patient-level health-care factors that are not measured mightexplain differences between countries Internationalharmonisation of detailed measures of quality of care inclinical registries such as pathways of care is neededSecond at the national level whether initiatives such asnational policies financial incentives organisation andleadership affect the delivery of care is unclear Thirdoutcomes should be compared with those in othercountries such as France Hungary Poland and the USAthrough assessment of national albeit voluntary registriesFourth with linkage to national electronic health records

there are opportunities to extend comparisons to includeambulatory care before and after admission for heartattack non-fatal events and longer-term outcomes38

We found clinically important differences in the careand outcomes of patients with acute myocardialinfarction in Sweden and the UK Internationalcomparisons of care and outcome registries mightinform new research and policy initiatives to improve thequality of health systems

Contributors

S-CC RG and ON analysed the data All authors contributed to thepreparation of the manuscript TJ and HH oversaw the study and providedimportant scientific input leadership and collaboration in the study

Conflicts of interest

We declare that we have no conflicts of interest

AcknowledgmentsThis study was done on behalf of the Swedish web-system forEnhancement and Development of Evidence-Based Care in HeartDisease Evaluated According to Recommended TherapiesTheSwedish Register of Information and Knowledge about SwedishHealth Intensive Care Admissions the National Institute forCardiovascular Outcomes ResearchMyocardial Ischaemia NationalAudit Project the Clinical Research Using Linked Bespoke Studiesand Electronic Health Records programme and the EuropeanImplementation Score project This study was funded by the EuropeanUnion Seventh Framework Programme for Research (CW PH andHH grant 223153) National Institute for Health Research ([NIHR] ATand HH grant RP-PG-0407ndash10314) Wellcome Trust (AT and HHgrant 086091Z08Z) and the Medical Research Council PrognosisResearch Strategy (PROGRESS) Partnership (AT and HH grantG090239399558) and by awards to establish the Farr Institute ofHealth Informatics Research at UCL Partners from the Medical

Research Council Arthritis Research UK British Heart Foundation

Panel Research in context

Systematic review

We searched Medline via PubMed with the medical subject headings ldquomyocardial

infarctionrdquo ldquoOutcome Assessment (Health Care)rdquo and ldquointernationalityrdquo and identified

studies published in English from January 1990 to September 2013 Additional

references from identified studies reviews or relevant citations provided by experts were

manually checked to supplement the literature searches We identified six closely relevant

studies reporting substantial variation in the use and uptake of evidence-based medicine

for care of acute myocardial infarction between countries8ndash111516 These studies however

were based on selected samples of hospital patients in voluntary registries8ndash10 cross-

sectional surveys15 a clinical trial16 or with only ST-segment-elevation acute myocardial

infarction (STEMI)8ndash911 or non-STEMI10 One study reported important between-countries

differences in acute myocardial infarction mortality18 and the uptake of some evidence-

based medicine Our search revealed no previous studies comparing quality of care

between two countries with nationwide coverage and taking into account heterogeneityin patientsrsquo casemix

Interpretation

We found evidence of clinically important international differences in the uptake of effective

treatments for and outcomes from AMI 30-day mortality after AMI was higher among UK

patients than among Swedish patients Differences in mortality were not explained by

differences in casemix but were partly attributable to recorded differences in clinical care

International comparisons of care and outcome registries might yield important actionable

insights to guide health-care policy and clinical practice to improve the quality of health

systems and prevent avoidable deaths from acute myocardial infarction

8122019 Ima 1


Articles


Cancer Research UK Chief Scientist Offi ce Economic and SocialResearch Council Engineering and Physical Sciences Research

Council NIHR National Institute for Social Care and HealthResearch and Wellcome Trust CW was supported by the NIHRBiomedical Research Centre at Guyrsquos and St Thomasrsquo National HealthService Foundation Trust and Kingrsquos College London funded by theNIHR AT was supported by Barts and The London NIHRCardiovascular Biomedical Research Unit funded by the NationalInstitute for Health Research SJ AJ LW and TJ were supported bythe Swedish Heart and Lung Foundation RG is employed at theMedical Products Agency in Sweden The views expressed in thispaper do not necessarily represent the views of the funding bodies

References1 Coleman MP Forman D Bryant H et al for the ICBP Module

1 Working Group Cancer survival in Australia Canada DenmarkNorway Sweden and the UK 1995-2007 (the International CancerBenchmarking Partnership) an analysis of population-based cancerregistry data Lancet 2011 377 127ndash38

2 Murray CJ Richards MA Newton JN et al UK health performancefindings of the Global Burden of Disease Study 2010 Lancet 2013381 997ndash1020

3 Iglehart JK Prioritizing comparative-effectiveness researchmdashIOMrecommendations N Engl J Med 2009 361 325ndash28

4 Wright RS Anderson JL Adams CD et al for the American Collegeof Cardiology FoundationAmerican Heart Association Task Forceon Practice Guidelines 2011 ACCFAHA focused updateincorporated into the ACCAHA 2007 guidelines for themanagement of patients with unstable anginanon-ST-elevationmyocardial infarction a report of the American College ofCardiology FoundationAmerican Heart Association Task Force onPractice Guidelines developed in collaboration with the AmericanAcademy of Family Physicians Society for CardiovascularAngiography and Interventions and the Society of ThoracicSurgeons J Am Coll Cardiol 2011 57 e215ndash367

5 Kushner FG Hand M Smith SC Jr et al 2009 focused updatesACCAHA guidelines for the management of patientswith ST-elevation myocardial infarction (updating the2004 guideline and 2007 focused update) and ACCAHASCAIguidelines on percutaneous coronary intervention (updating the2005 guideline and 2007 focused update) a report of the AmericanCollege of Cardiology FoundationAmerican Heart Association TaskForce on Practice Guidelines J Am Coll Cardiol 2009 54 2205ndash41

6 Gavalova L Myocardial Ischaemia National Audit Project (MINAP)2011 report London National Institute for CardiovascularOutcomes Research 2012

7 Jernberg T Johanson P Held C Svennblad B Lindbaumlck J Wallentin Lfor the SWEDEHEARTRIKS-HIA Association between adoption ofevidence-based treatment and survival for patients with ST-elevationmyocardial infarction JAMA 2011 305 1677ndash84

8 Eagle KA Goodman SG Avezum A Budaj A Sullivan CMLoacutepez-Sendoacuten J for the GRACE Investigators Practice variationand missed opportunities for reperfusion in ST-segment-elevationmyocardial infarction findings from the Global Registry of AcuteCoronary Events (GRACE) Lancet 2002 359 373ndash77

9 Widimsky P Wijns W Fajadet J et al for the European Associationfor Percutaneous Cardiovascular Interventions Reperfusiontherapy for ST elevation acute myocardial infarction in Europedescription of the current situation in 30 countries Eur Heart J 2010 31 943ndash57

10 Yusuf S Flather M Pogue J et al for the OASIS (Organisation toAssess Strategies for Ischaemic Syndromes) RegistryInvestigators Variations between countries in invasive cardiacprocedures and outcomes in patients with suspected unstableangina or myocardial infarction without initial ST elevationLancet 1998 352 507ndash14

11 Laut KG Gale CP Lash TL Kristensen SD Determinants andpatterns of utilization of primary percutaneous coronaryintervention across 12 European countries 2003ndash2008 Int J Cardiol 2013168 2745ndash53

12 Curry LA Spatz E Cherlin E et al What distinguishes top-performing hospitals in acute myocardial infarction mortality ratesA qualitative study Ann Intern Med 2011 154 384ndash90

13 White HD Chew DP Acute myocardial infarction Lancet 2008372 570ndash84

14 Levy AR Mitton C Johnston KM Harrigan B Briggs AHInternational comparison of comparative effectiveness research in five

jurisdictions insights for the US Pharmacoeconomics 2010 28 813ndash3015 Schiele F Hochadel M Tubaro M et al Reperfusion strategy in

Europe temporal trends in performance measures for reperfusiontherapy in ST-elevation myocardial infarction Eur Heart J 201031 2614ndash24

16 Kociol RD Lopes RD Clare R et al International variation in andfactors associated with hospital readmission after myocardialinfarction JAMA 2012 307 66ndash74

17 Terkelsen CJ Lassen JF Noslashrgaard BL et al Mortality rates in patientswith ST-elevation vs non-ST-elevation acute myocardial infarctionobservations from an unselected cohort Eur Heart J 2005 26 18ndash26

18 Abildstrom SZ Rasmussen S Roseacuten M Madsen M Trends inincidence and case fatality rates of acute myocardial infarction inDenmark and Sweden Heart 2003 89 507ndash11

19 Herrett E Smeeth L Walker L Weston C and the MINAPAcademic Group The Myocardial Ischaemia National Audit Project(MINAP) Heart 2010 96 1264ndash67

20 Jernberg T Attebring MF Hambraeus K et al The SwedishWeb-system for enhancement and development of evidence-basedcare in heart disease evaluated according to recommendedtherapies (SWEDEHEART) Heart 2010 96 1617ndash21

21 Sweden National Board of Health and Welfare Nationalguidelines for cardiac care httpwww socialstyrelsen senationellariktlinjerforhjartsjukvard (accessed Nov 12 2013in Swedish)

22 The National Archives Evaluation of the Coronary Heart DiseaseNational Service Framework httpwebarchivenationalarchivesgovuk+wwwdhgovukenFreedomOfInformationFreedomofinformationpublicationschemefeedbackFOIreleasesDH_126679 (accessed Nov 12 2013)

23 Packer C Simpson S Stevens A for the EuroScan the EuropeanInformation Network on New and Changing Health TechnologiesInternational diffusion of new health technologies a ten-countryanalysis of six health technologies Int J Technol Assess Health Care 2006 22 419ndash28

24 Carlsson P Health technology assessment and priority setting forhealth policy in Sweden Int J Technol Assess Health Care 200420 44ndash54

25 Flynn MR Barrett C Cosiacuteo FG et al The Cardiology Audit andRegistration Data Standards (CARDS) European data standards forclinical cardiology practice Eur Heart J 2005 26 308ndash13

26 Birkhead JS Walker L Pearson M Weston C Cunningham ADRickards AF for the National Audit of Myocardial Infarction ProjectSteering Group Improving care for patients with acute coronarysyndromes initial results from the National Audit of MyocardialInfarction Project (MINAP) Heart 2004 90 1004ndash09

27 Thygesen K Alpert JS White HD for the Joint ESCACCFAHAWHF Task Force for the Redefinition of Myocardial InfarctionUniversal definition of myocardial infarction Eur Heart J 200728 2525ndash38

28 Keeley EC Boura JA Grines CL Primary angioplasty versusintravenous thrombolytic therapy for acute myocardial infarctiona quantitative review of 23 randomised trials Lancet 2003 361 13ndash20

29 Antman EM Anbe DT Armstrong PW et al for the AmericanCollege of Cardiology and the American Heart Association TaskForce on Practice Guidelines and the Canadian CardiovascularSociety ACCAHA guidelines for the management of patients withST-elevation myocardial infarction a report of the American Collegeof CardiologyAmerican Heart Association Task Force on PracticeGuidelines (Committee to Revise the 1999 Guidelines for theManagement of Patients with Acute Myocardial Infarction)Circulation 2004 110 e82ndash292

30 Silber S Albertsson P Avileacutes FF et al and the Task Force forPercutaneous Coronary Interventions of the European Society ofCardiology Guidelines for percutaneous coronary interventionsEur Heart J 2005 26 804ndash47

31 Improvement NHS National roll-out of Primary PCI for patientswith ST segment elevation myocardial infarction an interim report(2010) httpsystemimprovementnhsukImprovementSystemViewDocumentaspxpath=Cardiac2FNational2FWebsite2FReperfusion2FInterim20Report20text20finaldocpdf

(acccessed Jan 13 2014)

8122019 Ima 1


Articles


32 National Health Service Growth of primary PCI for the treatmentof heart attack patients in England 2008ndash2011 the role of NHS

Improvement and the cardiac network 2012 httpwebarchivenationalarchivesgovuk20130221101407httpwwwimprovementnhsukLinkClickaspxfileticket=PWttejHG45M3Damptabid=63(accessed Jan 13 2014)

33 The Task Force on the Management of Acute Myocardial Infarctionof the European Society of Cardiology Acute myocardial infarctionpre-hospital and in-hospital management The Task Force on theManagement of Acute Myocardial Infarction of the EuropeanSociety of Cardiology Eur Heart J 1996 17 43ndash63

34 Ryan TJ Anderson JL Antman EM et al ACCAHA guidelines forthe management of patients with acute myocardial infarctionA report of the American College of CardiologyAmerican HeartAssociation Task Force on Practice Guidelines (Committee onManagement of Acute Myocardial Infarction) J Am Coll Cardiol 1996 28 1328ndash428

35 NICE A prophylaxis for patients who have experienced amyocardial infarction London National Institute for Clinical

Excellence 200136 Herrett E Shah AD Boggon R et al Completeness and diagnostic

validity of recording acute myocardial infarction events in primarycare hospital care disease registry and national mortality recordscohort study BMJ 2013 346 f2350

37 Hemingway H Croft P Perel P et al and the PROGRESS GroupPrognosis research strategy (PROGRESS) 1 a framework forresearching clinical outcomes BMJ 2013 346 e5595

38 Denaxas SC George J Herrett E et al Data resource profilecardiovascular disease research using linked bespoke studies andelectronic health records (CALIBER) Int J Epidemiol 201241 1625ndash38

8122019 Ima 1


Articles


Huddinge Section of

Cardiology Karolinska

Institute Stockholm Sweden

(T Jernberg MD)

Correspondence to

Dr Tomas Jernberg Department

of Cardiology Karolinska

University Hospital Karolinska

Institute Huddinge 141 86

Stockholm Sweden

tomasjernbergkarolinskase

or

Prof Harry Hemingway Farr

Institute of Health Informatics

Research at UCL Partners


222 Euston Road

London NW1 2DA UK

hhemingwayuclacuk

(universal funded from taxation and free at the point of

use) proportion of gross domestic product spent onhealth and national policy guidance provided for theevidence-based management of acute myocardialinfarction2122 Differences are that Sweden has morerapid diffusion of some new technologies23 morecomplete use of evidence-based practice9 and a moreestablished system for evaluating and reporting thequality and outcomes of care24 than the UK

In the absence of previous international comparisonsour objectives in this study were as follows first to assessthe validity of comparing data from the two nationwideclinical registries second to compare time trends forproportions of patients in receipt of effective interventionswhile in hospital and at discharge between 2004 and 2010

third to compare crude and casemix-standardised 30-daymortality between the two countries and between clinicallyimportant subgroups fourth to estimate time trends incasemix-standardised mortality and compare thembetween Sweden and the UK and fifth to explore thecontribution of clinical care to any difference in mortality

MethodsStudy populationAll hospitals providing care for acute myocardialinfarction in Sweden and the UK contribute data onconsecutive patients to the Swedish Web-System forEnhancement and Development of Evidence-BasedCare in Heart Disease Evaluated According toRecommended Therapies (SWEDEHEART)Register ofInformation and Knowledge about Swedish HeartIntensive care Admissions (RIKS-HIA) and the UKMyocardial Ischaemia National Audit Project (MINAP)

respectively RIKS-HIA and MINAP comply with the

Cardiology Audit and Registration Data Standards (foracute coronary syndrome in Europe72526 We obtaineddata for all patients who were admitted to hospitalbecause of acute myocardial infarction between Jan 12004 and Dec 31 2010 For patients with multipleadmissions we used the earliest record Acutemyocardial infarction diagnosis was based on guidelinesfrom the European Society of CardiologyAmericanCollege of CardiologyAmerican Heart Association27 Thestudy was approved by the MINAP Academic Group andthe steering group of SWEDEHEART

Casemix and treatment measuresThe definitions of casemix evidence-based hospital

treatment and discharge medications in SWEDEHEARTRIKS-HIA and MINAP were compared (appendixpp 2ndash9) The 17 casemix characteristics were demographicfactors (age sex year of admission) risk factors (smokinghistory of diabetes mellitus and hypertension) severity ofacute myocardial infarction (troponin concentrationssystolic blood pressure at admission heart rate atadmission) history of heart failure cardiac arrest atadmission history of cerebrovascular disease or historyof acute myocardial infarction and previous proceduresor use of medication (antiplatelet treatment with aspirinclopidogrel or both PCI coronary artery bypass graftsurgery) Features of hospital treatment for STEMI werereperfusion therapy (primary PCI or fibrinolytic therapybefore or during hospital stay) delay between symptomonset and primary PCI or fibrinolysis coronaryintervention other than primary PCI intravenousglycoprotein IIbIIIa inhibitors and use of anticoagulantsDischarge medications assessed were antiplatelet therapy(aspirin clopidogrel or both) β blockers angiotensin-converting-enzyme (ACE) inhibitors or angiotensin-receptor blockers (ARBs) and statins Validation ofSWEDEHEARTRIKS-HIA (each year a trained monitorcompares data with a chart review in 30 randomly selectedpatients within each hospital) showed a 96middot1agreement20 Validation of MINAP data (compared withreaudit data generated from each hospital annually by

the re-entering of 20 data items on 20 randomly selectedpatients) showed a median agreement of 89middot519

MortalityThe primary clinical outcome was all-cause mortalitywithin 30 days after hospital admission National uniqueidentifiers were used to link patients with the NationalDeath Registry in Sweden or the Offi ce for NationalStatistics in the UK We accessed these registries toascertain vital status or date of death at 30 days

Statistical analysisWe compared troponin I and T concentrations casemixhospital treatment and medication at discharge betweenSwedish and UK patients Data are shown as proportion

See Online for appendix

All hospitals providing AMI careSweden 86 UK 242

Admission records from all hospitals

Sweden 414831 UK 739 828

Study population of AMI patientsSweden 119786 UK 391077

Deaths at 30 daysSweden 9173 UK 41 509

SwedenUK excluded 276 147330175

ineligible records51626998 age lt30 years or missing

266 78773 031 non-AMI diagnosis 0126916 admitted before or after

inclusion dates 4198123 230 did not meet other

demographic criteria1889818576 subsequent admissions

Figure 983089 Study populationAMI=acute myocardial infarction ID=identifier

8122019 Ima 1


Articles






Sweden (n=119 786) UK (n=391 077)

Casemix









Risk factor













Hospital treatment

STEMI patients





















8122019 Ima 1


Articles










0

20

40

60

80

100



Any fibrinolysis UK

Reperfusion


2004 2005 2006 2007 2008 2009 20100

20

40

60

80

100


Years2004 2005 2006 2007 2008 2009 2010

Years





A B

C D








Number at risk

SwedenUK

0 5 10 15 20 25



30

119786390948

115 113370883

113 364362 830

112 345357954

111 646354475

111 136351919

110693349845


0

2

4

6

8

10


n y c a u s e ( )

8122019 Ima 1


Articles












Sweden

(n=119 786)

UK

(n=391 077)

Sweden

(n=119 786)

UK

(n=391 077)
























Age





Year of admission








Risk factors







relevant subgroups

8122019 Ima 1


Articles















casemix model


Casemix-


Number of

UK AMIpatients

Observed


Year

2004

2005

2006

2007

2008

2009

2010

55447

54421

52757

53276

55 851

58479

60720

13middot2

12middot6

11middot3

10middot4

9middot9

9middot0

8middot4











Mortality UK higher

8122019 Ima 1


Articles










Contributors







Systematic review














Interpretation








8122019 Ima 1


Articles






































8122019 Ima 1


Articles











8122019 Ima 1


Articles






Sweden (n=119 786) UK (n=391 077)

Casemix









Risk factor













Hospital treatment

STEMI patients





















8122019 Ima 1


Articles










0

20

40

60

80

100



Any fibrinolysis UK

Reperfusion


2004 2005 2006 2007 2008 2009 20100

20

40

60

80

100


Years2004 2005 2006 2007 2008 2009 2010

Years





A B

C D








Number at risk

SwedenUK

0 5 10 15 20 25



30

119786390948

115 113370883

113 364362 830

112 345357954

111 646354475

111 136351919

110693349845


0

2

4

6

8

10


n y c a u s e ( )

8122019 Ima 1


Articles












Sweden

(n=119 786)

UK

(n=391 077)

Sweden

(n=119 786)

UK

(n=391 077)
























Age





Year of admission








Risk factors







relevant subgroups

8122019 Ima 1


Articles















casemix model


Casemix-


Number of

UK AMIpatients

Observed


Year

2004

2005

2006

2007

2008

2009

2010

55447

54421

52757

53276

55 851

58479

60720

13middot2

12middot6

11middot3

10middot4

9middot9

9middot0

8middot4











Mortality UK higher

8122019 Ima 1


Articles










Contributors







Systematic review














Interpretation








8122019 Ima 1


Articles






































8122019 Ima 1


Articles











8122019 Ima 1


Articles










0

20

40

60

80

100



Any fibrinolysis UK

Reperfusion


2004 2005 2006 2007 2008 2009 20100

20

40

60

80

100


Years2004 2005 2006 2007 2008 2009 2010

Years





A B

C D








Number at risk

SwedenUK

0 5 10 15 20 25



30

119786390948

115 113370883

113 364362 830

112 345357954

111 646354475

111 136351919

110693349845


0

2

4

6

8

10


n y c a u s e ( )

8122019 Ima 1


Articles












Sweden

(n=119 786)

UK

(n=391 077)

Sweden

(n=119 786)

UK

(n=391 077)
























Age





Year of admission








Risk factors







relevant subgroups

8122019 Ima 1


Articles















casemix model


Casemix-


Number of

UK AMIpatients

Observed


Year

2004

2005

2006

2007

2008

2009

2010

55447

54421

52757

53276

55 851

58479

60720

13middot2

12middot6

11middot3

10middot4

9middot9

9middot0

8middot4











Mortality UK higher

8122019 Ima 1


Articles










Contributors







Systematic review














Interpretation








8122019 Ima 1


Articles






































8122019 Ima 1


Articles











8122019 Ima 1


Articles












Sweden

(n=119 786)

UK

(n=391 077)

Sweden

(n=119 786)

UK

(n=391 077)
























Age





Year of admission








Risk factors







relevant subgroups

8122019 Ima 1


Articles















casemix model


Casemix-


Number of

UK AMIpatients

Observed


Year

2004

2005

2006

2007

2008

2009

2010

55447

54421

52757

53276

55 851

58479

60720

13middot2

12middot6

11middot3

10middot4

9middot9

9middot0

8middot4











Mortality UK higher

8122019 Ima 1


Articles










Contributors







Systematic review














Interpretation








8122019 Ima 1


Articles






































8122019 Ima 1


Articles











8122019 Ima 1


Articles















casemix model


Casemix-


Number of

UK AMIpatients

Observed


Year

2004

2005

2006

2007

2008

2009

2010

55447

54421

52757

53276

55 851

58479

60720

13middot2

12middot6

11middot3

10middot4

9middot9

9middot0

8middot4











Mortality UK higher

8122019 Ima 1


Articles










Contributors







Systematic review














Interpretation








8122019 Ima 1


Articles






































8122019 Ima 1


Articles











8122019 Ima 1


Articles










Contributors







Systematic review














Interpretation








8122019 Ima 1


Articles






































8122019 Ima 1


Articles











8122019 Ima 1


Articles






































8122019 Ima 1


Articles











8122019 Ima 1


Articles











Documents

Ima 1