Limitations on the prognostic value of predischarge data after myocardial infarction.

Clinical variables and those obtained by non-invasive techniques were studied prospectively in a series of 306 patients discharged from hospital after an acute myocardial infarction. The predictive value of the data at two and 12 months was assessed by univariate and multivariate analyses. The best correlation was found for age, hypertension, bundle branch block, early and late heart failure, x ray cardiothoracic ratio, digoxin use, the number of metabolic equivalents reached during the stress test, echocardiographic wall motion score index, left ventricular end diastolic diameter, left ventricular ejection fraction, and the presence of an aneurysm. The prognostic value of the same data at 12 months was studied in those surviving for two months. There was a noticeable decline in the relative risk of all but two of the factors (number of metabolic equivalents, ventricular arrhythmias). All of the predictive variables except the x ray cardiothoracic ratio, number of metabolic equivalents, and the presence of an aneurysm lost their discriminant power. The explanation for this is the strength of statistical relations of these variables with the outcome at two months. They continued to influence the score at 12 months even when the entire patient series was considered. In conclusion, the study shows that the predictive value of most of the predischarge variables usually taken into account in the assessment of risk in patients one year after infarction does not extend beyond the first two months.