OUP user menu

Resolving inequalities in care? Reduced mortality in the elderly after acute coronary syndromes. The Myocardial Ischaemia National Audit Project 2003–2010

Chris P. Gale, B.A. Cattle, A. Woolston, P.D. Baxter, T.H. West, A.D. Simms, J. Blaxill, D.C. Greenwood, K.A.A. Fox, R.M. West
DOI: http://dx.doi.org/10.1093/eurheartj/ehr381 630-639 First published online: 18 October 2011

Abstract

Aims To examine age-dependent in-hospital mortality for hospitalization with acute coronary syndromes (ACS) in England and Wales.

Methods and results Mixed-effects regression analysis using data from 616 011 ACS events at 255 hospitals as recorded in the Myocardial Ischemia National Audit Project (MINAP) 2003–2010; 102 415 (16.7%) patients were aged <55 years and 72 721 (11.9%) ≥85 years. Patients ≥85 years with ST-elevation myocardial infarction (STEMI) were less likely to receive emergency reperfusion therapy than those <55 years (RR = 0.27, 95% CI: 0.25–0.28). Older patients had greater lengths of stay (P< 0.001) and higher in-hospital mortality (P< 0.001). For STEMI and non-ST-elevation myocardial infarction (NSTEMI), there were reductions in in-hospital mortality from 2003 to 2010 across all age groups including the very elderly. For STEMI ≥ 85 years, in-hospital mortality reduced from 30.1% in 2003 to 19.4% in 2010 (RR = 0.54, 95% CI: 0.38–0.75, P< 0.001), and for NSTEMI ≥ 85 years, from 31.5% in 2003 to 20.4% in 2010 (RR = 0.56, 95% CI: 0.42–0.73, P< 0.001). Findings were upheld after multi-level adjustment (base = 2003): male STEMI 2010 OR = 0.60, 95% CI: 0.48–0.75; female STEMI 2010 OR = 0.55, 95% CI: 0.42–0.71; male NSTEMI OR = 0.50, 95% CI: 0.42–0.60; female NSTEMI OR = 0.49, 95% CI: 0.40–0.59.

Conclusion For patients hospitalized with ACS in England and Wales, there have been substantial reductions in in-hospital mortality rates from 2003 to 2010 across all age groups. The temporal improvements in mortality were similar for sex and type of acute myocardial infarction. Age-dependent inequalities in the management of ACS were apparent.

  • Acute coronary syndrome
  • STEMI
  • NSTEMI
  • Age
  • In-hospital mortality
  • MINAP
  • Quality of care

See page 562 for the editorial comment on this article (doi:10.1093/eurheartj/ehr364)

Introduction

Data from international studies suggest that elderly patients who are admitted to hospital with an acute coronary syndrome (ACS) are less likely to receive evidence-based care and that they have higher mortality rates than their younger counterparts.17 Recently, however, there have been substantial improvements in the treatment and outcome of ACS among a range of developed and developing countries.8,9 In part, this has been attributed to increased use of evidence-based ACS therapies.811 For the elderly, it has been advocated that improvements in hospital care may translate into a reduction in mortality and research has highlighted the need for quality-of-care programmes that reinforce the use of evidence-based therapies among this group.2

With the advent of the new definition of acute myocardial infarction (AMI) and a greater emphasis placed on the results of the cardiac troponin concentration,12 a contemporary analysis of patients admitted to hospital with an ACS may reveal a changing burden of disease and early outcomes. Moreover, it is not known whether the effects of reported improvements in ACS care have occurred equally across the spectrum of ACS ages. This study, therefore, aimed to (i) establish whether, in light of international recommendations,13,14 age-dependent inequalities in care continue to exist in a modern national healthcare system and (ii) quantify and compare temporal effects in in-hospital mortality by age for patients who present to hospital with an ACS.

Methods

Study design

The analyses were based on data from the Myocardial Ischaemia National Audit Project (MINAP) whose national database was established in 1999 to examine the quality of management of AMI in England and Wales and to meet the audit requirements of the National Service Framework (NSF) for Coronary Heart Disease.1518 MINAP data collection and management has previously been described.19,20

Data for patients admitted with an ACS are collected prospectively at each acute hospital by a secure electronic system, developed by the Central Cardiac Audit Database (CCAD), electronically encrypted and transferred online to a central database.21 CCAD is part of the National Institute for Cardiovascular Outcomes Research (NICOR) based at University College London. MINAP is overseen by a multi-professional steering group representing the stakeholders.16

Each patient entry offers details of the patient journey, including the method and timing of admission, inpatient investigations, treatment, and date of all-cause death [from linkage to the National Health Service Central Register (NHSCR) using a unique National Health Service (NHS) number]. Data entry is subject to routine online error checking. There is a mandatory annual data validation exercise for each hospital.22

Cohort description

The investigators had access to data in which patient identity was protected. The MINAP cohort comprised 616 011 index patient events admitted to 238 acute hospitals in England and 17 acute hospitals in Wales between 1 January 2003 and 2 October 2010.

Ages on admission were categorized into five groups: <55 years, 55–64 years, 65–74 years, 75–84 years, and ≥85 years of age. The initial diagnosis was based on the working diagnosis generated by a paramedic or first attending physician who was in a position to provide definitive treatment. The final diagnosis was formed from the patients' presenting history, clinical examination and the results of inpatient investigations, and made by a senior member of the medical staff. The consensus document of the Joint European Society of Cardiology/American College of Cardiology23 was used as the diagnostic standard for AMI and provided the basis for categorization into ST-elevation myocardial infarction (STEMI). Non-ST-elevation acute coronary syndrome (NSTEACS) was defined as a troponin positive [non-ST-elevation myocardial infarction (NSTEMI)] or troponin negative (unstable angina) ACS.

Statistical methods

The population was described without adjustment and by percentages with respect to discrete data, and by medians and interquartile range or mean and 95% range for continuous variables. Pearson's χ2 test was used to determine whether there was a significant difference between the expected frequencies and the observed frequencies in one or more categories. The Kruskall–Wallis rank test was used to test the difference in distributions across groups. The analysis of variance test was use to ascertain whether the means of several groups were all equal.

Given that there was a significant interaction between age, in-hospital mortality, and sex for STEMI (P< 0.001) and NSTEMI (P< 0.001), models were fitted by sex. To account for variations at the hospital level, a linear mixed-effects regression model was used to quantify the relationship between age category and ACS final diagnosis at discharge from hospital, and between age category and in-hospital all-cause mortality. The model fitted included age, history of diabetes, hypertension, previous AMI, angina, history of heart failure, previous revascularization [percutaneous coronary intervention (PCI) and/or coronary artery bypass grafting (CABG)], admitting consultant, admission ward, and emergency reperfusion (primary PCI and/or thrombolysis). The temporal risk of in-hospital mortality was quantified by STEMI and NSTEMI after adjustment for age category and consideration of hospital random effects. Finally, the risk of in-hospital mortality was estimated for each age category after adjustment for the final diagnosis and consideration of hospital random effects. We used STATA IC version 11.0 (Stat Corp LP, TX, USA) for the analyses.

Results

Of the 616 011 patients, 102 415 (16.7%) were <55 years and 72 721 (11.9%) were ≥85 years of age. Data for age were missing for 4.2% of men and 4.5% of women, and in-hospital status was missing for 5.6% of the cohort. The proportion of men reduced from 79.4% among patients aged <55 years to 41.9% among patients aged ≥85 years (Table 1). There were 208 358 (33.8%) patients with a final diagnosis of STEMI, 325 299 (52.8%) NSTEACS, 24 320 (3.9%) unconfirmed ACS, 35 783 (5.8%) non-ACS/other, and 19 217 (3.1%) with a missing final diagnosis.

View this table:
Table 1

Baseline clinical characteristics Myocardial Ischemia National Audit Project patients (all acute coronary syndrome diagnoses combined) by age group

Age group (years)<55, % (n)55–64, % (n)65–74, % (n)75–84, % (n)≥85, % (n)Totals, % (n)
Men81 31086 66094 66885 74130 464395 464
Women21 10527 81648 16468 39142 257217 513
Diabetes
 Men10.1 (8224)14.3 (12 412)19.9 (18 878)19.8 (16 953)15.7 (4771)16.1 (63 592)
 Women13.6 (2867)16.2 (4508)20.7 (9962)19.0 (12 986)13.6 (5742)17.2 (37 507)
Hypertension
 Men27.2 (22 086)37.3 (32 315)43.7 (41 378)45.9 (39 394)44.0 (13 403)38.9 (153 969)
 Women31.0 (6534)42.2 (11 751)48.9 (23 535)52.3 (35 768)49.2 (20 777)46.9 (102 094)
Current smoker
 Men52.6 (42 791)36.6 (31 748)20.9 (19 762)10.9 (9361)6.2 (1880)27.7 (109 362)
 Women47.4 (10 009)36.1 (10 032)21.5 (10 349)10.9 (7452)3.6 (1534)18.8 (40 955)
Prior AMI
 Men15.2 (12 324)20.9 (18 135)28.1 (26 591)33.3 (28 525)35.0 (10 668)25.3 (100 186)
 Women11.9 (2504)15.5 (4307)21.6 (10 420)26.4 (18 078)27.1 (11 440)22.3 (48 541)
Angina
 Men15.4 (12 498)23.2 (20 123)32.2 (30 460)39.1 (33 525)42.2 (12 860)28.8 (113 935)
 Women17.0 (3592)23.2 (6462)30.3 (14 594)35.5 (24 272)35.9 (15 150)30.7 (66 725)
Prior revascularization
 Men10.4 (8429)14.1 (12 246)17.2 (16 313)14.5 (12 390)6.7 (2053)13.5 (53 520)
 Women7.8 (1641)10.0 (2786)11.4 (5505)8.2 (5642)3.0 (1288)8.0 (17 506)
Chronic heart failure
 Men1.1 (890)2.3 (1961)4.8 (4543)8.4 (7202)12.1 (3686)4.8 (19 060)
 Women1.3 (273)2.4 (665)5.0 (2418)8.9 (6053)12.4 (5237)7.0 (15 252)
Chronic renal failure
 Men1.1 (869)1.7 (1514)3.7 (3456)6.8 (5795)9.5 (2901)3.8 (15 216)
 Women1.5 (325)2.0 (544)3.2 (1559)4.8 (3307)5.8 (2450)3.9 (8558)

Cardiovascular risk factors

The distribution of baseline risk factors varied by age groups and sex (Table 1). Older patients were less often current smokers and more often had hypertension, prior AMI, angina, chronic heart failure and chronic renal failure. Compared with men ≥85 years, women ≥85 years of age were less often current smokers. They less often had diabetes, previous AMI, angina, previous revacularization (PCI or CABG), and chronic renal failure.

Diagnoses, presentation, and provision of care

Table 2 shows the distribution of initial and final diagnoses, method of presentation, and provision of care by age group. Younger patients more often had an initial diagnosis of STEMI. Older patients more often had a final diagnosis of NSTEACS. Compared with men ≥85 years of age, women ≥85 years of age less often had an initial diagnosis of NSTEACS. Older patients were less likely to call the emergency services or make their own way to the hospital, and more likely to have an ACS in hospital than their younger counterparts. Also, older patients were less likely to be admitted to the Cardiac Care Unit, a Cardiology ward, and be under the care of a Consultant Cardiologist. For STEMI and NSTEMI, the proportion of patients ≥85 years of age with cardiogenic shock was higher than that for patients <55 years of age: STEMI: <55 years = 2.1%, ≥85 years = 5.0%, P< 0.001; NSTEMI: <55 years = 1.2%, ≥85 years = 3.1%, P< 0.001.

View this table:
Table 2

Initial and final acute coronary syndrome diagnoses, method of presentation and provision of care by age group

Age group<55, % (n)55–64, % (n)65–74, % (n)75–84, % (n)≥85, % (n)Missing age, % (n)Totals, % (n)
Men81 31086 66094 66885 74130 46416 621395 464
Women21 10527 81648 16468 39142 2579780217 513
Initial diagnosis
 STEMI
  Men37.8 (30 709)35.6 (30 841)27.7 (26 265)20.1 (17 250)13.9 (4229)19.1 (3182)28.4 (112 476)
  Women25.7 (5433)26.8 (7458)23.9 (11 494)19.0 (13 013)15.0 (6324)13.8 (1345)20.7 (45 067)
 NSTEACS
  Men39.5 (32 152)42.8 (37 123)45.1 (42 687)44.8 (38 403)43.5 (13 244)40.5 (6734)43.1 (170 343)
  Women41.5 (8767)44.3 (12 331)43.8 (21 082)42.1 (28 824)38.6 (16 318)37.7 (3687)41.8 (91 009)
Final diagnosis
 STEMI
  Men45.9 (37 313)44.3 (38 376)36.4 (34 449)28.3 (24 244)21.8 (6633)24.7 (4113)36.7 (145 128)
  Women32.9 (6952)34.7 (9663)32.1 (15 473)27.9 (19 106)23.9 (10 081)20.0 (1955)29.1 (63 230)
 NSTEACS
  Men37.7 (30 690)45.1 (39 063)52.7 (49 886)61.3 (52 541)68.7 (20 926)51.0 (8473)51.0 (201 579)
  Women42.3 (8931)49.3 (13 706)54.7 (26 338)60.7 (41 541)66.2 (27 973)53.5 (5231)56.9 (123 720)
Presentation and provision of care—all ACS
 Called emergency services
  Men50.7 (41 228)54.7 (47 383)59.7 (56 475)64.8 (55 594)68.6 (208 99)32.7 (5436)57.4 (227 015)
  Women49.7 (10 488)54.5 (15 150)59.3 (28 557)64.8 (44 317)68.1 (28 784)34.6 (3381)60.1 (130 677)
 Made own way to hospital
  Men25.7 (20 913)19.7 (17 065)13.3 (12 621)7.6 (6518)4.1 (1239)11.5 (1914)15.2 (60 270)
  Women22.7 (4785)17.6 (4897)11.7 (5628)6.2 (4268)3.4 (1434)7.9 (774)10.0 (21 786)
 Already in hospital
  Men1.7 (1403)2.5 (2184)4.1 (3838)6.0 (5104)7.1 (2175)12.6 (2096)4.2 (16 800)
  Women2.9 (615)3.4 (952)4.9 (2368)6.6 (4544)7.7 (3244)16.1 (1571)6.1 (13 294)
 First ward: Cardiac Care Unit
  Men66.1 (53 780)63.1 (54 709)57.2 (54 318)49.0 (42 054)38.9 (11 858)43.3 (7203)56.6 (223 922)
  Women59.1 (12 471)56.8 (15 793)52.8 (25 423)45.1 (30 816)35.1 (14 823)36.9 (3610)47.3 (102 936)
 First ward: Cardiology
  Men4.8 (3901)5.5 (4779)5.8 (5472)5.8 (4958)5.8 (1781)4.7 (784)5.5 (21 675)
  Women5.0 (1056)5.7 (1579)5.8 (2780)5.5 (3771)5.2 (2193)3.9 (375)5.4 (11 754)
 Admitting consultant: Cardiologist
  Men50.1 (40 729)48.0 (41 627)44.0 (41 637)36.8 (31 561)28.5 (8689)38.3 (6365)43.1 (170 608)
  Women47.1 (9950)44.5 (12 372)42.0 (20 229)34.0 (23 285)25.1 (10 609)33.1 (3238)36.6 (79 683)

Management and in-hospital mortality

Table 3 shows the distribution of evidence-based management and outcomes by age category. For all ACS combined, older patients had greater lengths of stay (df = 5, P< 0.001) and higher in-hospital mortality rates (P< 0.001). For those with an initial diagnosis of STEMI, older patients were less likely to receive primary PCI, pre-hospital thrombolysis, and to a lesser extent in-hospital thrombolysis. They too had greater lengths of stay (df = 5, P< 0.001) and higher in-hospital mortality rates (df = 5, P< 0.001). Patients ≥85 years of age with an initial diagnosis of STEMI were up to 75% less likely to be reperfused (by either primary PCI or thrombolysis) compared with those <55 years of age with STEMI: RR = 0.27, 95% CI: 0.25–0.28.

View this table:
Table 3

Management and outcome by age group and acute coronary syndrome subgroups

Age group<55, % (n)55–64, % (n)65–74, % (n)75–84, % (n)≥85, % (n)Missing age, % (n)Totals
Men81 31086 66094 66885 74130 46416 621395 464
Women21 10527 81648 16468 39142 2579780217 513
All ACS
 Median length of hospital stay (IQR)
  Men4 (5)5 (5)6 (7)7 (8)7 (10)5 (5)5 (6)
  Women4 (5)5 (5)6 (6)7 (9)8 (11)6 (8)6 (8)
 In-hospital mortality
  Men1.3 (1070)2.5 (2161)5.6 (5296)11.8 (10 141)19.7 (6004)7.4 (1224)6.6 (25 996)
  Women1.7 (360)3.2 (888)6.5 (3151)12.7 (8660)20.3 (8570)10.8 (1053)10.4 (22 682)
Patients with an initial diagnosis is STEMI
 Primary PCI
  Men24.3 (7470)21.1 (6509)19.5 (5123)17.3 (2992)15.1 (640)8.8 (280)20.5 (23 014)
  Women23.7 (1287)20.6 (1534)19.3 (2223)16.7 (2174)13.4 (849)6.5 (88)18.1 (8155)
 Pre-hospital thrombolysis
  Men11.8 (3644)12.9 (3973)11.7 (3072)7.1 (1221)2.4 (102)1.6 (50)10.7 (12 062)
  Women9.5 (516)11.3 (840)9.6 (1102)5.6 (734)1.7 (106)1.3 (17)7.4 (3315)
 In-hospital thrombolysis
  Men52.6 (16 175)53.8 (16 607)54.3 (14 253)54.7 (9442)52.8 (2233)49.6 (1580)53.6 (60 290)
  Women53.7 (2920)55.1 (4112)54.4 (6256)54.5 (7091)50.2 (3172)44.8 (603)53.6 (24 154)
 Median length of hospital stay (IQR)
  Men4 (3)5 (4)5 (5)6 (7)6 (9)5 (5)5 (5)
  Women5 (4)5 (4)6 (5)6 (7)7 (10)6 (6)6 (7)
 In-hospital mortality
  Men1.5 (467)2.6 (815)6.2 (1625)14.0 (2423)24.5 (1035)5.5 (174)5.8 (6539)
  Women2.3 (126)4.2 (313)8.2 (944)17.2 (2232)28.9 (1827)14.1 (189)12.5 (5631)
Patients with a final diagnosis is AMI (NSTEMI + STEMI)
 Aspirin on discharge
  Men79.1 (48 675)77.9 (55 083)74.6 (57 505)70.2 (50 171)66.5 (17 508)73.5 (8210)74.5 (237 152)
  Women77.0 (10 690)77.2 (16 011)74.0 (27 850)69.6 (38 934)65.5 (23 562)68.3 (4373)77.0 (214 905)
 Clopidogrel on discharge
  Men36.7 (22 566)35.7 (25 230)32.5 (25 031)29.9 (21 343)28.5 (7509)22.9 (2557)32.7 (104 236)
  Women36.0 (4992)35.3 (7331)32.0 (12 048)29.4 (16 463)25.7 (9255)19.4 (1241)30.1 (51 330)
 β-Blocker on discharge
  Men71.9 (44 265)68.3 (48 266)61.3 (47 285)53.6 (38 327)46.0 (12 108)60.1 (6709)61.9 (196 960)
  Women64.3 (8925)63.1 (13 084)58.3 (21 917)53.3 (29 843)46.3 (16 642)50.7 (3249)54.9 (93 660)
 ACE inhibitor on discharge
  Men70.1 (43 168)69.6 (49 228)66.2 (51 016)59.4 (42 504)49.4 (13 021)61.7 (6886)64.7 (205 823)
  Women63.4 (8792)65.4 (13 565)63.7 (23 979)58.5 (32 743)48.1 (17 286)53.4 (3421)58.5 (99 786)
 Statin on discharge
  Men78.4 (48 261)78.0 (55 116)75.5 (58 154)70.4 (50 344)61.3 (16 131)72.3 (8072)74.2 (236 078)
  Women76.0 (10 545)77.1 (15 986)74.9 (28 181)69.6 (38 920)56.2 (20 224)63.9 (4093)69.2 (117 949)
 Coronary angiography
  Men56.7 (34 883)51.1 (36 119)43.6 (33 577)27.5 (19 654)12.1 (3196)34.3 (3831)41.2 (131 260)
  Women53.9 (7485)47.5 (9850)38.7 (14 567)22.5 (12 592)8.8 (3153)24.0 (1535)28.8 (49 182)
 Echocardiography
  Men48.7 (29 987)48.9 (34 557)49.6 (38 257)48.6 (34 781)41.3 (10 872)46.4 (5182)48.3 (153 636)
  Women49.2 (6823)48.4 (10 036)49.5 (18 642)48.2 (26 941)38.2 (13 728)42.9 (2745)46.3 (78 915)
 Median length of hospital stay (IQR)
  Men5 (4)5 (5)6 (6)7 (8)7 (9)6 (5)6 (6)
  Women5 (4)5 (5)6 (7)7 (9)8 (11)7 (8)7 (8)
 In-hospital mortality
  Men1.3 (824)2.6 (1872)5.6 (4506)12.0 (8894)19.7 (5329)8.0 (940)6.8 (22 365)
  Women1.9 (276)3.5 (747)6.9 (2704)13.2 (7674)20.6 (7700)12.8 (861)11.3 (19 962)

For patients with AMI (STEMI or NSTEMI), the risk (RR, 95% CI) of being prescribed aspirin (0.54, 0.53–0.56), clopidogrel (0.59, 0.57–0.62), β-blockers (0.38, 0.37–0.39), statins (0.41, 0.40–0.42), or ACE inhibitors (0.50, 0.49–0.51) was considerably lower for those ≥85 years of age with AMI compared with those <55 year of age with AMI. The elderly were less likely to undergo coronary angiography (df = 5, P< 0.001) and echocardiography (df = 5, P< 0.001), and had greater lengths of stay (df = 5, P< 0.001) and higher in-hospital mortality rates (P< 0.001) than their younger counterparts.

View this table:
Table 4

Association of age with risk of in-hospital all-cause mortality for ST-elevation myocardial infarction and non-ST-elevation myocardial infarction, by sex

Risk of in-hospital mortalityMaleFemale
Odds ratioa (95% CI)Odds ratiob (95% CI)Odds ratioa (95% CI)Odds ratiob (95% CI)
STEMI
 <55 years1.001.001.001.00
 55–64 years1.95 (1.76–2.15)1.97 (1.75–2.23)1.79 (1.51–2.14)1.89 (1.54–2.33)
 65–74 years4.55 (4.14–4.99)4.41 (3.94–4.94)3.75 (3.21–4.40)3.80 (3.14–4.59)
 75–84 years11.50 (10.51–12.59)10.62 (9.51–11.86)8.19 (7.03–9.54)8.29 (6.91–9.95)
 ≥85 years23.30 (21.09–25.74)20.31 (17.97–22.95)15.28 (13.10–17.83)14.98 (12.44–18.03)
NSTEMI
 <55 years1.001.001.001.00
 55–64 years2.00 (1.77–2.19)2.27 (1.99–2.59)1.78 (1.49–2.12)1.82 (1.48–2.23)
 65–74 years4.17 (3.79–4.59)4.94 (4.38–5.56)3.81 (3.26–4.46)4.10 (3.42–4.92)
 75–84 years8.94 (8.14–9.81)10.46 (9.31–11.75)7.53 (6.47–8.77)8.08 (6.77–9.64)
 ≥85 years15.71 (14.27–17.28)18.10 (16.05–20.41)12.66 (10.88–14.74)13.47 (11.27–16.08)
  • aUnadjusted odds ratio for age, with hospital random intercept effects.

  • bAdjusted odds ratio for age, diabetes, hypertension, previous AMI, angina, previous revascularization, chronic heart failure, reperfusion (primary PCI or thrombolysis) during admission, admitting ward, admitting consultant, with hospital random intercept effects.

Risk of ST-elevation myocardial infarction and in-hospital mortality

For males and less so females, increasing age predicted a lower risk of a final diagnosis of STEMI (males ≥85 years: OR = 0.33, 95% CI: 0.32–0.34; females ≥85 years: OR = 0.62, 95% CI: 0.60–0.65). For both sexes, the risk of in-hospital mortality increased with age for STEMI (males: OR = 20.31, 95% CI: 17.97–22.95; females: OR = 14.98, 95% CI: 12.44–18.03) and NSTEMI (males: OR = 18.10, 95% CI: 16.05–20.41; females: OR = 13.47, 95% CI: 11.27–16.08). The highest risk of death occurred in males ≥85 years with STEMI (OR = 20.31, 95% CI: 17.97–22.95) (Table 4).

Year of admission and acute coronary syndrome care

For patients with AMI, the proportion with cardiogenic shock increased between 2003 and 2010: <55 years of age: 0.6% (2003) to 1.7% (2010), RR = 2.80, 95% CI: 1.79–4.32, P< 0.001; ≥85 years of age: 1.6% (2003) to 3.1% (2010), RR = 2.00, 95% CI: 1.36–2.89, P< 0.001. For patients with an admission diagnosis of STEMI, rates of primary PCI increased from 1.6% in 2003 to 60.9% in 2010 (RR = 92.41, 95% CI: 71.11–120.99, P< 0.001) for patients aged <55 years and from 0.2% to 48.5% (RR = 376.79, 95% CI: 140.18–1412.26, P< 0.001) for patients aged ≥85 years. For patients with AMI, there were significant increases in the rates of use of evidence-based pharmacological therapy rates from 2003 to 2010 (Table 5).

View this table:
Table 5

Acute coronary syndrome provision of care and in-hospital mortality by year of admission

Age groupYearsRelative risk (95% CI)
Provision of care
2003–20042009–2010
Primary PCI for STEMI<55 years3.3%52.1%31.57 (28.22–35.33)
>85 years0.5%32.2%82.52 (55.28–128.55)
Aspirin on admission for AMI<55 years86.5%90.2%1.43 (1.33–1.53)
>85 years76.6%86.6%1.97 (1.85–2.11)
GP IIb–IIIa for AMI<55 years10.2%15.2%1.59 (1.47–1.73)
>85 years2.0%2.6%1.31 (1.08–1.59)
Aspirin on discharge for AMI<55 years95.8%82.5%0.20 (0.19–0.22)
>85 years81.1%71.6%0.59 (0.55–0.63)
ACE inhibitor on discharge for AMI<55 years81.4%76.5%1.35 (1.27–1.42)
>85 years57.4%55.9%1.06 (1.01–1.12)
β-Blocker on discharge for AMI<55 years85.5%75.3%0.52 (0.49–0.55)
>85 years49.1%56.7%1.35 (1.29–1.43)
Clopidogrel on discharge for AMI<55 years56.1%97.3%28.48 (20.64–39.69)
>85 years28.1%89.1%81.31 (59.06–112.26)
Statin on discharge for AMI<55 years94.2%82.4%0.29 (0.26–0.31)
>85 years61.3%68.6%1.38 (1.31–1.46)
In-hospital mortalitya
20032010
STEMI<55 years2.0%1.5%0.72 (0.39–1.25)
>85 years30.1%19.4%0.54 (0.38–0.75)
NSTEMI<55 years1.9%0.9%0.89 (0.48–1.34)
>85 years31.5%20.4%0.56 (0.42–0.73)
  • aUnadjusted rates

Year of admission and in-hospital mortality

For patients with STEMI, in-hospital mortality reduced from 2.0% in 2003 to 1.5% in 2010 (RR = 0.72, 95% CI: 0.39–1.25, P= 0.24) for patients aged <55 years, from 4.0% to 1.6% (RR = 0.28, 95% CI: 0.14–0.52, P< 0.001) for patients aged 55–64 years, from 19.6% to 10.6% (RR = 0.47, 95% CI: 0.36–0.60, P< 0.001) for patients aged 75–84 years, and from 30.1% to 19.4% (RR = 0.54, 95% CI: 0.38–0.75, P< 0.001) for patients aged ≥85 years (Table 5). For patients with NSTEMI, in-hospital mortality reduced from 1.9% in 2003 to 0.9% in 2010 (RR = 0.89, 95% CI: 0.48–1.34, P= 0.43) for patients aged <55 years, from 3.5% to 1.8% (RR = 0.40, 95% CI: 0.23–0.65, P= 0.001) for patients aged 55–64 years, from 19.6% to 10.6% (RR = 0.49, 95% CI: 0.39–0.61, P= 0.001) for patients aged 75–84, and from 31.5% to 20.4% (RR = 0.56, 95% CI: 0.42–0.73, P< 0.001) for patients aged ≥85 years (Table 5). These findings were upheld after multi-level adjustment (base = 2003, adjusted odds ratio for age group, diabetes, hypertension, previous AMI, angina, previous revascularization, chronic heart failure, reperfusion (primary PCI or thrombolysis) during admission, admitting ward, admitting consultant, with hospital random intercept effects): male STEMI 2010: OR = 0.60, 95% CI: 0.48–0.75; female STEMI 2010: OR = 0.55, 95% CI: 0.42–0.71; male NSTEMI: OR = 0.50, 95% CI: 0.42–0.60; female NSTEMI: OR = 0.49, 95% CI: 0.40–0.59. After adjustment for final diagnosis and hospital-level effects, there was a reduction in inpatient mortality from 2003 to 2010 across all age groups including patients ≥85 years of age: OR, 95% CI: 2004: 0.94, 0.88–1.01; 2010: 0.52, 0.44–0.61; 75–84 years of age: 2004: 0.98, 0.93–1.03; 2010: 0.52, 0.45–0.60, and patients <55 years of age: 2004: 0.94, 0.79–1.13; 2010: 0.64, 0.44–0.93 (Figure 1).

Figure 1

Odds ratios by year for in-hospital all-cause mortality, stratified by age category. 2003 = base, adjustment for final diagnosis and hospital-level random effects.

Discussion

Despite earlier research from Europe which has highlighted the need to address age-dependent inequalities in ACS quality of care,2 when compared with their younger counterparts the elderly hospitalized with an ACS continue to be disadvantaged. This is important when the elderly comprise up to a third of the ACS admissions in England and Wales. Yet, there was good evidence to suggest that all age groups including the old and very old have benefited from improvements in ACS management—for AMI, there were substantial year-on-year reductions in in-hospital mortality. Notably, the temporal improvements in the risk of in-hospital mortality were similar for males and females and for STEMI and NSTEMI.

To date, many studies have described the differential presentation, management, and outcome of elderly vs. young ACS patients.2,2428 This research corroborates these findings; revealing that the profile of the elderly hospitalized with an ACS has not changed greatly. What has changed is the reduction in in-hospital mortality. We refute findings from a recent single centre observational study which suggested that no temporal improvements in mortality rates were evident for the elderly who underwent primary PCI.29 Our research readily highlights that although age-dependent biases in quality of care exist, in England and Wales significant improvements in ACS care have occurred. From 2003 to 2010, improvements in the application of evidence-based ACS care were evident across all age groups—this is despite the increased proportion of patients presenting with cardiogenic shock. The unadjusted risk of in-hospital mortality after an ACS admission in 2010 was half that of 2003 (RR = 0.50, 95% CI: 0.45–0.54).

Several studies have suggested that improvements in hospital care for the elderly would reduce elderly ACS mortality rates.2,2428 Our analyses using contemporary MINAP data demonstrate this association. The reductions in in-hospital mortality over time were unlikely to be due to reduction in lengths of hospital stay. For STEMI and NSTEMI, we found no significant relationships between the length of hospital stay and in-hospital mortality and there was no significant interaction between the length of hospital stay and in-hospital mortality by year of hospital admission. Furthermore, from 2003 to 2010, 30-day mortality rates fell for STEMI (RR = 0.43, 95% CI: 0.34–0.54, P< 0.001) and NSTEMI (RR = 0.66, 95% CI: 0.55–0.78, P< 0.001) suggesting that the reduction in in-hospital mortality rates was unlikely to be related to hasty (or inappropriate) discharge from hospital care. It is possible, however, that some of the improvements in NSTEMI mortality rates related to a lower risk profile in later years: the median (IQR) troponin concentration for NSTEMI decreased from 0.57 (2.80) in 2003 to 0.48 (2.56), P< 0.001.

While the adjusted risk of the temporal decline in in-hospital mortality for STEMI and NSTEMI <55 years of age were statistically significant, we found there was only a non-significant trend in the decline of the absolute risk (20% and 47%, respectively) in this group. In 2003, mortality rates in the young were already low (2.0%) and it is possible that in-hospital mortality rates lower than 1.6% (2010) are now reaching a ‘plateau of achievable care’,30,31 and that a statistically significant association would require much greater numbers of patients or evaluation of survival beyond the hospital stay.

Overall our findings are in keeping with international advances in the provision of evidence-based acute cardiac care. They herald the accomplishment in England and Wales of the NSF for Coronary Heart disease (2000–2010).15 This was a nationwide implementation strategy of changes to the delivery of care for patients with coronary heart disease and encouraged the adoption of the translation of contemporary evidence into best practice.20,32 Nevertheless, our research continues to support a notion of age-dependent inequality in ACS care and, moreover, highlights gaps in key aspects of the management of elderly patients with ACS who benefit equally as much as their younger counterparts from an early invasive strategy.28,33,34

This study provides evidence for opportunities for improvements in the quality of clinical care. For example, despite high frequencies of previous AMI in the very elderly, they had previously less often undergone revascularization when advanced age alone must not be considered a contraindication to performing coronary angiography and PCI.34 Overall, rates of emergency reperfusion (primary PCI and thrombolysis) for STEMI in those aged <55 years were nearly three times higher than those aged ≥85 years. For those with a final diagnosis of AMI, older patients were less likely to be discharged on aspirin, clopidogrel, β-blockers, ACE inhibitors, and statins. In light of our evidence for increased risk of early mortality and greater lengths of hospital stay, the application of evidence-base ACS therapies to appropriate patients regardless of age may further reduce overall cost and improve early outcomes.2

The causes for discrepancies in quality care for the elderly are multifactorial. In part, the shortfalls in treatment may be due to the lack of appropriate specialist care and inappropriate placement within the hospital.35 Although the MINAP database includes data relating to the indication, contraindication, refusal of treatments (all taken into account in the analyses), we were unable to evaluate the appropriateness of ACS management.36 Reductions in risk of inpatient death may be related to improved primary37 and secondary prevention;38 however, we specifically considered in-hospital mortality (rather than longer term survival) because this more clearly reflects acute care associated with the index admission. Nonetheless, improvements in mortality are associated with the application of evidence-based medicine,38 and it is likely that the implementation of strategic networks of care (such as the national primary PCI service in England and Wales20) has contributed to the greater application of ACS treatments and hence better outcomes.39

Not withstanding age-dependent inequalities in care, the elderly are more likely to present differently and less likely to have the same diagnosis on discharge from hospital as that which they were given on admission. In our study, the risk of a change in diagnosis from that on admission to a different one on discharge in patients ≥85 years of age was over 10% greater than that for patients <55 years of age: RR, 95% CI 1.12, 1.09–1.16. Multi-level adjustment made little difference to the risk of in-hospital mortality and suggests that the ‘diagnosis’ per se is a stronger predictor of outcome than the covariates modelled. As such, mechanisms to improve the early and accurate diagnosis of specific ACS subgroups in the elderly are needed so that timely risk-evaluated ACS interventions may be implemented. It is plausible that physicians already know that the likelihood of an elderly patient presenting with STEMI is much lower than that of a younger patient and that this influences their perception of a diagnosis of STEMI in an older patient. Finally, age-dependent inequalities in treatments may be the legacy of a risk-adverse strategy to ACS care4 through lack of accurate estimation of ACS risk.40

Limitations

MINAP does not collect data on all patients in England and Wales, and it is possible that patients entered into the MINAP database differ from those not recorded. We noted that data missingness for age was 4.3% and for final diagnosis 3.2%. Although this may introduce systematic bias, we have previously noted that while being statistically significant the inclusion of missing data does not alter regional standardized mortality ratios.41 As with all observational data, the modelling of diagnosis, in-hospital mortality, and effect of year considered hospital-level and patient-specific influences and the use of alternative covariates may change the effect sizes demonstrated. Finally, this research reveals important associations but cannot prove causation.

Conclusion

The elderly comprise a substantial proportion of ACS admissions. They have a different risk factor and ACS diagnosis profile to younger patients. Biases in elderly ACS care remain and the elderly have significantly longer hospital lengths of stay and higher in-hospital mortality rates. Despite this, improvements in the application of evidence-based ACS care were evident across all age groups from 2003 to 2010. There were significant year-on-year reductions in in-hospital mortality equally across all age groups, both sexes, and for STEMI and NSTEMI.

Ethical approval

The National Institute for Cardiovascular Research (NICOR) which includes the Myocardial Ischaemia National Audit Project (MINAP) (Ref: NIGB: ECC 1-06 (d)/2011) has support under section 251 of the National Health Service (NHS) Act 2006. On seeking advice from Leeds (West) Research Ethics Committee, formal ethical approval was not required under NHS research governance arrangements for the project.

Funding

MINAP is funded by the Health Quality Improvement Partnership (HQIP). This study was funded by the British Heart Foundation. C.P.G. is funded by the National Institute for Health Research as a Clinician Scientist and Honorary Consultant Cardiologist

Conflicts of interest: none declared.

Acknowledgements

All of the authors gratefully acknowledge funding from the British Heart Foundation PG/07/057/23215. C.P.G. is funded through the National Institute for Health Research (NIHR). The extract from the MINAP database was provided through the MINAP Academic Group. We acknowledge all the hospitals in England and Wales for their contribution of data to MINAP. There are no competing interests and the authors have nothing to declare.

References

View Abstract