Controversy exists regarding whether all patients with acute myocardial infarction (AMI) benefit from angiotensin-converting enzyme inhibitors (ACEIs). We examined the association between ACEI treatment and mortality in a large, unselected population of patients with AMI. The present study included 105,224 patients with AMI who were not treated with ACEIs on admission. A logistic regression analysis, including 33 variables, calculated a propensity score for each patient to estimate the probability of receiving ACEIs at discharge, given the background. The association between ACEI treatment at discharge and the 1-year outcome was evaluated in prespecified subgroups using the Cox regression analyses, adjusting for the propensity score and medications at discharge. A total of 38,395 patients (36.5%) received ACEIs at discharge. After adjustment, ACEI treatment was associated with a 24% reduction in mortality (relative risk 0.76, 95% confidence interval 0.73 to 0.80). The benefit was largest in patients with a history or present signs of heart failure. In patients without heart failure, a significant benefit of ACEI treatment was seen only in patients with renal dysfunction (relative risk 0.69, 95% confidence interval 0.54 to 0.88). In the whole group, the risk of AMI decreased by 7% (relative risk 0.93, 95% confidence interval 0.90 to 0.96), with a larger effect seen in patients with ST-segment elevation AMI or systolic left ventricular dysfunction. In conclusion, in unselected patients with AMI, ACEI treatment was associated with a reduction in 1-year mortality, mainly in patients with heart failure or renal dysfunction, and a small reduction in the risk of reinfarction, mainly in patients with ST-segment elevation AMI or systolic left ventricular dysfunction.
The clinical effects of angiotensin-converting enzyme inhibitors (ACEIs) in patients with coronary artery disease (CAD) have been extensively studied in several large clinical trials. Short-term treatment in unselected patients with acute myocardial infarction (AMI) has, in 2 mega-trials and a subsequent meta-analysis, been shown to reduce the mortality and risk of future heart failure (HF), with the largest benefit in high-risk patients. Long-term treatment in selected patients with AMI with left ventricular (LV) systolic dysfunction or signs of HF reduced both mortality and the future rate of HF in 3 trials. One of these trials also showed a reduction in the rate of reinfarction, although the other 2 did not. In patients with stable CAD and without HF or LV dysfunction, 2 large trials have shown that ACEI treatment compared to placebo reduces the risk of cardiovascular death, myocardial infarction, or stroke. Although the effects of ACEI in patients with preserved systolic LV function have been documented only in patients with stable CAD, the current guidelines now recommend ACEIs also in patients with preserved systolic LV function and acute AMI. To date, few observational studies have examined the efficacy of ACEI treatment in unselected patients with AMI. The aim of the present study was, therefore, to evaluate whether registry data can confirm an association between ACEI treatment and a reduction in mortality and a lowered risk of recurrent AMI in a broad and unselected population of patients with AMI.
Methods
The Register of Information and Knowledge about Swedish Heart Intensive Care Admissions (RIKS-HIA) was started 1995 and has registered all patients admitted to the coronary care unit at 72 of 74 hospitals treating patients with acute coronary syndrome in Sweden. Detailed information about the registry has been previously published (also available at www.riks-hia.se ). In brief, the register included >100 variables, including age, gender, risk factors of CAD, history of heart disease, medication at entry, symptoms, electrocardiogram at entry, biochemical markers, reperfusion therapy, pharmacologic treatment, results from stress testing and coronary angiography, revascularization, major complications, the outcome during the hospital stay, and medication at discharge. To ensure the correctness of the data entered into the database, a monitor visits about 20 hospitals each year and compares the data entered into the RIKS-HIA with the information in the patients’ records in 30 to 40 randomly chosen patients for each hospital. When 1,972 randomly chosen computer forms from 38 hospitals containing 161,280 variables were reviewed, the agreement was 97%. Data regarding the presence of a history of stroke, dementia, renal failure, HF, chronic pulmonary disease, or cancer were obtained by merging the RIKS-HIA database with the National Patient Registry, which includes diagnoses at discharge for all hospital stays in Sweden.
The follow-up for 1 year regarding mortality and subsequent admission because of AMI was done by merging the RIKS-HIA database with the National Cause of Death Register, which includes information about the vital status of all Swedish citizens, and the National Patient Registry. All patients were informed about their participation in the registry and follow-up and had the right to refuse participation. The National Board of Health and Welfare, the Swedish Data Inspection Board, and the local ethics committee at Uppsala University approved the registry and the merging of the registries.
The present analysis included patients who were admitted to the coronary care unit from 1995 to 2005 without ongoing ACEI treatment and who were subsequently discharged alive with a diagnosis of AMI. The patients were excluded if the medications at discharge were unknown. Angiotensin receptor blockers were introduced in Sweden 1995 and were registered as ACEIs up to (and including) 2003. Thereafter, angiotensin receptor blocker treatment was registered separately. In 2004, treatment with angiotensin receptor blocker constituted 5.9% of all treatment with inhibitors of the renin-angiotensin system (ACEIs and angiotensin receptor blockers). Thus, assuming a linear increase of angiotensin receptor blocker prescription, <3% of those registered as discharged with ACEI treatment from 1995 to 2003 were actually prescribed angiotensin receptor blocker treatment. The criteria for AMI were identical at the participating centers with the World Health Organization criteria used until 2000 and the European Society of Cardiology/American College of Cardiology/American Heart Association criteria used from 2001.
The 2 patient groups (with or without ACEI treatment at discharge) were compared using chi-square tests for categorical variables and the t test for continuous variables, all tests were 2-sided. A propensity score was calculated for each patient to estimate the probability of belonging to the group that was discharged with ACEI treatment given the background characteristics and treatments received. The purpose of adjusting for the propensity score in the subsequent analyses was to balance the groups regarding the differences in background characteristics and treatment given. A multivariate logistic regression model was performed to calculate the propensity score. This analysis included 33 covariates: demographics (age, gender, and hospitalization after 1999), risk factors (history of hypertension and diabetes), previous cardiovascular disease (previous AMI, HF, percutaneous coronary intervention, coronary artery bypass grafting, stroke, and renal insufficiency), previous other disease (chronic pulmonary disease, dementia, cancer within 3 years), medications before admission (aspirin, oral anticoagulants, β blockers, calcium antagonists, long-acting nitrates, diuretics, and statins), index event, and hospital course (circulatory arrest at arrival, ST-segment elevations or left bundle branch block on admission electrocardiogram, sinus rhythm, signs of HF during the hospital stay [Killip class >1, use of intravenous diuretics or continuous positive airway pressure], and reinfarction during the hospital stay), treatment during the hospital stay (thrombolysis, primary percutaneous coronary intervention, intravenous β blockers and nitroglycerin, and intravenous or subcutaneous heparin), and procedures before discharge (coronary angiography and revascularization). To determine the association between treatment with ACEIs at discharge and outcome, Cox regression analyses were performed adjusting for the propensity score and other medications at discharge (aspirin, oral anticoagulants, β blockers, calcium antagonists, long-acting nitrates, diuretics, and statins). The Cox regression model was then tested in predefined subgroups. The estimated glomerular filtration rate was calculated by the Modification of Diet in Renal Disease equation. The LV ejection fraction was classified as normal (>50%), slightly (40% to 50%), moderately (30% to 40%) or severely (<30%) depressed.
To exclude a possible bias of missing values, a sensitivity analysis was performed. In the present analysis, a new propensity score was calculated in which the independent dichotomous variables were coded as no, yes, or missing. The association between treatment with ACEIs at discharge and outcome were then adjusted by this new propensity score and medications at discharge. Statistical analyses were performed with the statistical program Statistical Package for Social Sciences, version 15.0 (SPSS, Chicago, Illinois).
Results
From 1995 to 2005, 107,686 patients with no previous ACEI treatment were discharged alive with a diagnosis of AMI. Whether ACEI treatment was prescribed at discharge was unknown for 2,462 patients (2.3%). Thus, 105,224 patients constituted the present study population. Of these patients, 38,395 (36.5%) received ACEI treatment at discharge. The proportion of patients who received ACEIs increased gradually from 29.9% in 1995 to 43.3% in 2005. The baseline characteristics differed between those with and without ACEI treatment in several aspects ( Table 1 ). When the baseline characteristics were adjusted for the propensity to be treated with ACEIs, no major differences remained between the groups (data not shown).
| Variable | ACEIs | p Value | |
|---|---|---|---|
| Yes (n = 38,395) | No (n = 66,829) | ||
| Age (years) | <0.001 | ||
| Median | 72 | 71 | |
| 25th–75th percentile | 62–79 | 60–80 | |
| Men | 65.7% | 64.0% | <0.001 |
| Smoking | 25.4% | 24.9% | 0.084 |
| Hypertension | 32.4% | 28.9% | <0.001 |
| Diabetes mellitus | 21.5% | 16.8% | <0.001 |
| Previous myocardial infarction | 17.1% | 17.6% | 0.038 |
| Heart failure | 7.9% | 8.6% | <0.001 |
| Percutaneous coronary intervention | 2.8% | 3.5% | <0.001 |
| Coronary artery bypass grafting | 4.2% | 4.3% | 0.271 |
| Previous stroke | 8.5% | 9.3% | <0.001 |
| Renal failure | 0.5% | 1.6% | <0.001 |
| Chronic obstructive pulmonary disease | 4.3% | 5.1% | <0.001 |
| Dementia | 0.2% | 0.4% | <0.001 |
| Cancer within 3 years | 3.0% | 3.5% | <0.001 |
| Treatment before admission | |||
| Aspirin | 30.5% | 33.0% | <0.001 |
| Oral anticoagulation | 3.4% | 3.4% | 0.886 |
| β Blockade | 29.9% | 30.9% | <0.001 |
| Calcium antagonist | 15.2% | 14.7% | 0.018 |
| Long-acting nitrates | 12.9% | 16.0% | <0.001 |
| Diuretics | 19.8% | 21.7% | <0.001 |
| Statins | 10.3% | 10.7% | 0.067 |
| Index event and hospital course | |||
| Circulatory arrest at entry | 0.7% | 0.6% | 0.003 |
| ST-segment elevation or left bundle branch block | 55.0% | 41.2% | <0.001 |
| Sinus rhythm | 86.3% | 87.0% | <0.001 |
| Signs of heart failure | 48.5% | 30.1% | <0.001 |
| Reinfarction | 2.6% | 2.0% | <0.001 |
| Treatment during hospital stay | |||
| Acute thrombolysis | 28.7% | 22.3% | <0.001 |
| Acute percutaneous coronary intervention | 12.1% | 7.4% | <0.001 |
| Intravenous β blockade | 42.4% | 32.5% | <0.001 |
| Intravenous nitroglycerin | 35.7% | 31.3% | <0.001 |
| Intravenous or subcutaneous anticoagulation | 63.2% | 57.9% | <0.001 |
| Procedures before discharge | |||
| Coronary angiography | 39.5% | 33.8% | <0.001 |
| Revascularization | 29.3% | 24.3% | <0.001 |
| Treatment at discharge | |||
| Aspirin | 86.5% | 86.8% | 0.224 |
| Oral anticoagulation | 12.0% | 6.2% | <0.001 |
| β Blockade | 86.8% | 81.7% | <0.001 |
| Calcium antagonist | 8.6% | 14.0% | <0.001 |
| Long-acting nitrates | 26.0% | 31.3% | <0.001 |
| Diuretics | 46.0% | 29.7% | <0.001 |
| Statins | 58.1% | 45.8% | <0.001 |
The unadjusted 1-year mortality rate was lower in patients treated with ACEIs than in those not treated (10.6% vs 12.1%, p <0.001). When adjusted for the propensity score and medications at discharge, ACEI treatment at discharge was associated with a 24% reduction in mortality (relative risk [RR] 0.76, 95% confidence interval [CI] 0.73 to 0.80). The effect of ACEI treatment on 1-year mortality in prespecified subgroups is shown in Figure 1 . The benefit was largest in the elderly and in patients with a history or present signs of HF. In patients without HF, no significant benefit of ACEI treatment was seen in any of the subgroups ( Figure 2 ). The estimated glomerular filtration rate and ejection fraction were known in a limited number of patients (27,038 and 28,622 patients, respectively). The benefit of ACEIs increased with a decreasing estimated glomerular filtration rate (p value for interaction = 0.007) and ejection fraction (p value for interaction = 0.006; Table 2 ). In patients without previous or current HF, a benefit of ACEIs was still seen in patients with estimated glomerular filtration rate <60 ml/min (RR 0.69, 95% CI 0.54 to 0.88), and a trend was seen toward a benefit in patients with at least a moderately depressed ejection fraction (RR 0.78, 95% CI 0.57 to 1.07).
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
