Age, clinical presentation, and outcome of acute coronary syndromes in the Euroheart acute coronary syndrome survey

Abstract

Aims Age is one of the most powerful determinants of prognosis in myocardial infarction, but there is comparatively little recent data across the whole spectrum of acute coronary syndromes (ACS). We examined the impact of increasing age on clinical presentation and hospital outcome in a large sample of patients with ACS.

Methods and results Patients (n=10 253) from the Euroheart ACS survey in 103 hospitals in 25 countries were investigated. There was a significant inverse association between the age and the likelihood of presenting with ST-elevation. For each decade of life, the odds of presenting with ST-elevation decreased by 0.82 [95% confidence interval (CI) 0.79–0.84]; P<0.0001. Elderly patients were considerably less often treated by cardiologists, less extensively investigated, and, when presenting with ST-elevation ACS, less likely to be treated with reperfusion. Compared with patients <55 years, the odds ratios of hospital mortality were 1.87 (1.21–2.88) at age 55–64, 3.70 (2.51–5.44) at age 65–74, 6.23 (4.25–9.14) at age 75–84, and 14.5 (9.47–22.1) among patients ≥85 years, with no major differences across different types of admission or discharge diagnoses.

Conclusion Elderly ACS patients were less likely to present with ST-elevation but had substantial in-hospital mortality, yet they were markedly less intensively treated and investigated.

Introduction

In recent years, there has been a shift in the clinical presentation of acute coronary syndromes (ACS) to milder forms, with evidence to suggest that case severity in acute myocardial infarction (AMI) may be decreasing^1–3 and that hospitalized patients with AMI have smaller infarcts⁴ with lower case fatality.⁵ Concomitantly, there is an increase in the rates of unstable angina,⁶ which is a milder form of the ACS. Old age is a powerful predictor of mortality of patients with AMI.^7–15 Even though the impact of age on clinical presentation and outcome in AMI is well characterized, the effects of age with these emerging characteristics of ACS have not been well studied. With increasing life expectancy, the mean age of ACS patients is growing steadily, emphasizing the need to define the impact of age across the whole spectrum of ACS. Toward this end, we investigated the impact of increasing age on clinical presentation and hospital outcome in a large population of patients with different manifestations of ACS.

Methods

The details of the Euroheart ACS survey have been previously described in detail.¹⁶ The survey was performed in clusters composed of academic and non-academic hospitals and hospitals with and without cardiac catheterization laboratories, and cardiac surgery facilities. During the enrolment period lasting from 4 September 2000 to 15 May 2001, 14 271 patients in 25 countries admitted with suspected ACS were registered, of whom 10 484 were finally diagnosed with either AMI or unstable angina. In 231 patients, there were missing data on either age or sex, leaving 10 253 who form the study population for the present analysis.

For all patients, the tentative initial diagnosis made by the attending physicians was recorded on the basis of the initial electrocardiographic pattern: ACS with ST-elevation, ACS without ST-elevation, and ACS with an undetermined electrocardiographic pattern. The full case report form was filled out for patients with a confirmed diagnosis of unstable angina or AMI and categorized according to the discharge diagnosis as either unstable angina, non-Q-wave AMI, or Q-wave AMI. The case report form included details regarding the demographic, clinical, and electrocardiographic characteristics of the patient, the diagnostic and treatment modalities, the in-hospital complications, and the discharge status. Body mass index was calculated as weight in kilograms divided by height in meters squared. Hypertension was defined as previously diagnosed by a physician, receiving medication to lower blood pressure, or known blood pressure values of ≥140 mmHg systolic or ≥90 mm diastolic on two or more occasions. Diabetes was defined as previously diagnosed by a physician. Current smoking was defined as smoking cigarettes up to 1 month before admission. Angina pectoris of recent or prior onset was recorded and defined as chronic if present at least 30 days before admission, or of unknown duration. Hospital mortality was defined as any death occurring before discharge from any cause.

Coronary angiography was done in 5437 patients (53%). Among patients ≥85 years, only 63 patients (13%) underwent coronary angiography, and because they were so few and, in all probability, heavily selected, we excluded them from the analyses pertaining to angiographic findings. The presence of ≥50% stenosis in any of the three main vessels or left coronary main stem was recorded.

Statistical methods

All analyses were performed using the SAS software version 8e. Baseline characteristics of the patients in the study were summarized in terms of frequencies and percentages. Correlations were tested by linear regression, with age as a continuous variable. With respect to angiographic findings, a summary score was created for the number of vessels with significant (>50%) stenosis, ranging from 0 to 3 and tested against age as a continuous variable. To determine whether the age differences in clinical presentation or outcome were due to differences in baseline clinical and demographic characteristics, multiple logistic regression models were fitted with hospital mortality as the dependent variables while controlling for potentially confounding factors. Of the factors present on admission, we considered gender, prior AMI, hypertension, diabetes, chronic angina, prior revascularization, and known prior heart failure. A similar analysis was conducted with ST-elevation ACS as the dependent variable, which also included an interaction term for sex and age. The clustering of observations within countries was taken into account by using multilevel techniques (university/non-university hospital and country). This method adjusts the estimation of model coefficients for the correlation among the observations, which is due to clustering within countries. Because of the large number of statistical comparisons in this report, P<0.01 was considered to provide borderline evidence for an association and P<0.001 was considered significant.

Results

Of the 10 253 patients enrolled in the Euroheart ACS survey, 2321 (23%) were <55 years and 2544 (25%) were ≥75 years. The proportion of women increased from 17% among patients aged <55 to 56% among patients aged ≥85 (Table 1). The distribution of coronary risk factors varied strongly with age, with few exceptions. Younger patients were more often obese and smokers, and they had a higher prevalence of a positive family history, whereas older patients had more hypertension and diabetes. Prior angina, prior AMI, cerebrovascular events, and heart failure prior to admission were more prevalent with increasing age, as were use of aspirin, beta-blockers, and diuretics.

In the youngest age group, <55 years, 56% of men and 44% of women presented with ST-elevation ACS (Table 2). This proportion decreased with age for the men, whereas there was no significant decrease with age among women. The proportion of ACS with undetermined ECG pattern increased with age. With respect to diagnosis at discharge, the proportion of ACS patients <55 years with Q-wave MI was 44% for men and 29% for women, decreasing to 33% for men 85 years and older, whereas there was no decrease with age in women. The majority of patients in all age groups presented with typical anginal pain, but less frequently so with increasing age. Conversely, the proportion presenting with heart failure (Killip class 3–4) increased markedly with age.

The proportion of patients undergoing a coronary angiogram decreased with age (Table 3). The proportion with a normal coronary angiogram was low overall and became less common with age. Almost one in five had a left main stenosis among patients aged 75–84, and approximately half had either left main stenosis ≥50% or three-vessel disease.

Complications such as heart failure or pulmonary oedema or shock were increasingly more common with age, particularly pulmonary oedema, which occurred in 18% of men and 16% of women 85 years or older (Table 4). Overall, over 40% of ACS patients ≥85 years displayed some form of heart failure, including shock, during their hospital stay. Atrial fibrillation occurred in few of the youngest patients, but in 16% of men and 9% of women ≥85 years. Recurrent ischaemia and re-infarction increased only very slightly with age in men and not at all in women.

Of patients who were <55 years, 91% of men and 85% of women were treated in a coronary care unit or cardiology ward from the start (Table 5). One in five of patients aged 75–84, and one in four among those aged ≥85, were initially treated in non-cardiology units. Among patients presenting with ST-elevation and presenting within 12 h, only 31 and 21% among men and women, respectively, ≥85 years were treated with primary reperfusion, compared with 70% of men and 76% of women <55 years. Use of angiography and evaluation of left ventricular function decreased markedly with age. Angiotensin converting enzyme inhibitors prescribed on discharge were slightly higher in the higher age groups, whereas prescription of beta-blockers and statins were lower.

The proportion of ST-elevation ACS decreased with age, also after considering potential confounders (Table 6). For each decade of life, the odds of presenting with ST-elevation decreased by 0.82 (95% CI 0.79–0.84); P<0.0001 (data not shown). Overall, in-hospital death increased from 1% among patients <55 years to 17% in patients ≥85 years. There were no significant interactions between age and type of ACS, or type of AMI, and hospital mortality (all P>0.2). The overall worst prognosis was found among patients with an undetermined electrocardiogram at admission, but this was largely due to the high average age in this category. Introducing potential confounders into the models including random effects in a multilevel analysis (university/non-university hospital and country) did not influence the odds ratios associated with age to any great extent. Slightly more than half of the patients were ≥65 years, but 81% of the deaths occurred in this group.

Discussion

Major findings

Despite improvement in treatment in ACS, the short-term outcome in elderly patients is still very poor. In this international cohort of comparatively unselected ACS patients, we found that overall in-hospital mortality was <5%, but this varied widely across the different ACS and across age groups. For example, in-hospital mortality was <1% in patients aged <55 without ST-elevation, whereas almost one in four of patients ≥85 years with ST-elevation ACS died. Older patients had less ST-elevation ACS and Q-wave AMI but had more ACS with undetermined ECG pattern, which, in turn, had a poor prognosis, as described in a prior publication.¹⁷ The risk factor pattern also varied with age, in that younger patients more often had a positive family history, were smokers, or obese, whereas older patients more often had a history of diabetes, hypertension or prior cardiovascular disease, and more medical treatment. Heart failure and atrial fibrillation were common complications during the hospital stay and were significantly more common with higher age, whereas recurrent ischaemia and re-infarction displayed only weak or non-existent associations with age.

Risk factors and treatments

Manifestations of coronary disease in populations undergo rapid changes over time. In many European countries, as well as in the USA, there have been decreases in the incidence of AMIs^18,19 and in mortality from coronary disease.²⁰ Although these changes have not been universal, most of the participating centres in the Euroheart ACS survey represent countries with these characteristics. Another feature of recent developments is that the severity of ACS seems to be decreasing,^1–3 with better prognosis.^4,5 To a large extent, this will be due to better treatment.¹⁹ Out-of-hospital mortality, however, has also decreased,²¹ indicating that other factors, such as changes in risk factor pattern may play a role.²² One recent study found that modest reductions in major risk factors led to gains in life-years four times higher than did the various treatments.²³ The different risk factor patterns in younger and older patients might reflect variations related to aging and differences between birth cohorts. To what extent elderly patients would benefit from better implementation of proven therapies or improvement of risk factors remains undetermined.

Prior studies

Relative odds of 1.44 and 1.82 for hospital mortality for a 10 year increment in age in women and men, respectively, have been found in AMI patients.²⁴ Other studies, however, in non-ST-elevation ACS have found higher increments in short-term mortality with age.²⁵ Among Spanish patients with AMI, and investigated from 1995 to 2001, hospital mortality almost doubled from each age group from 2.6% in patients <55 years to 25.8% in the >84 years group, figures similar to that of the present study,²⁶ even though, unlike the Euroheart survey, only patients treated in coronary care units or intensive care were included. Improvement in hospital outcome seems to have been greater in younger age groups. In the Worcester study, improvement in outcome over two decades was demonstrated to be less in older patients.⁸

Some of the improvement in younger patients could be due to the fact that proven therapies are better implemented among the younger patients. A substantial proportion of patients who are eligible for reperfusion therapy still do not receive this treatment, particularly among the elderly^27,28 although a benefit has been demonstrated also in the elderly,²⁹ and other proven therapies are also underused,¹³ even in the absence of contraindications.³⁰ Despite having more severe and extensive coronary disease, elderly patients have been shown to receive less aggressive anti-ischaemic therapy and to be less likely to undergo coronary angiography, with fewer revascularization procedures than their younger counterparts.³¹ Consistent with previous studies,³² older patients had less typical symptoms which partly might explain their different treatment. However, although older patients have an increased risk for major bleeding, a routine early invasive strategy has been shown to significantly improve ischaemic outcomes in elderly patients with non-ST-segment elevation ACS.³³ This study supports the notion that older patients may not be as well treated as they should be, and consequently, there may be considerable scope for improvement with respect to hospital outcome in elderly patients.

Limitations

In the Euroheart ACS survey, there was no strict validation of the AMI or unstable angina diagnoses. The discharge diagnosis of the attending physician was accepted. About a quarter of patients diagnosed with AMI had normal creatine kinase status, indicating either that the test was not done within the optimal time frame or that the traditional definition was used, with symptoms and electrocardiographic changes sufficient to diagnose infarction.³⁴ Likewise, 23% of patients diagnosed with unstable angina had elevated creatine kinase, although for the most part only moderately so. The new definition of AMI was published on the same day as this survey started,³⁴ and these inconsistencies reflect clinical practice during this transitional period. Even so, the differences in mortality between different diagnoses were considerable. The age gradient with respect to mortality, however, was similar across all ACS diagnoses.

Conclusions

In this large survey of ACS patients from 25 countries in Europe and the Mediterranean basin, we found that old age remains a strong predictor of increased hospital mortality across all types of ACS. Old patients with ACS have their hospital course frequently complicated, in particular, with heart failure. Young patients had comparatively low mortality, four out of five deaths were in patients ≥65years, indicating that improvement in hospital care for older patients may potentially save a substantial number of lives. However, more detailed data with respect to comorbid conditions and contraindications are needed to estimate the size of this potential benefit.

Acknowledgement

The Euroheart ACS survey was sponsored by Schering-Plough and Centocor. The Swedish participation was supported by the Swedish Heart and Lung Foundation. We thank Georg Lappas for help with the statistical analyses.

Conflict of interest: none declared.

References