Markers of risk after acute myocardial infarction. A comparison of clinical variables, ambulatory and exercise electrocardiography, echocardiography, and stress echocardiography.

BACKGROUND: Short-term mortality after myocardial infarction has decreased continuously among members of selected populations. Nonetheless, the long-term prognosis among members of unselected populations remains bad. Further research in risk stratification is therefore needed. In the present study we tested the additive value of clinical variables, echocadiography, ambulatory electrocardiography, exercise testing, and stress echocardiography in assessing the long-term prognosis after acute myocardial infarction. METHODS: Two-dimensional echocardiography and ambulatory electrocardiography (analysis of ST-segment changes and of heart rate variability) were performed for 74 patients aged < 75 years who had had an acute myocardial infarction. Before their discharge from hospital, 70 patients were subjected to a combined exercise test and stress echocardiography. The time of follow-up was > or = 3 years. RESULTS: During follow-up 18 patients died, and 38 suffered cardiac events defined as death, nonfatal reinfarction and the need for revascularization. We first tested 31 covariates in a univariate regression analysis. A subsequent multivariate analysis was performed in two stages. During the first of these, clinical variables (a history of systemic hypertension, infarct localization, and diabetes mellitus) and variables derived from noninvasive tests (new-onset wall-motion abnormality during stress echocardiography, ST-segment depression and heart-rate variability during ambulatory electrocardiography, the ejection fraction by echocardiography at rest, and the double product during exercise tests) predicted mortality. After the second stage, however, the only remaining independent predictors of mortality were the presence of a new-onset wall-motion abnormality (P < 0.0001, relative risk 13.5, 95% confidence interval 3.6-51.3), ST-segment depression during ambulatory electrocardiography (P = 0.003, relative risk 5.0, 95% confidence interval 1.7-15.7) and a decreased heart rate variability (P = 0.007). CONCLUSIONS: The only variables that were of independent value in assessing the long-term mortality were those expressing residual myocardial ischemia and the cardiovascular sympatho-vagal balance. It is, therefore, recommended that one should monitor these variables for patients recovering from an acute myocardial infarction.