Comparison of clinical variables and variables derived from a limited predischarge exercise test as predictors of early and late mortality after myocardial infarction.

An exercise test limited to 5 METS or 70% of age-predicted maximal heart rate was performed 1 day before hospital discharge by 225 survivors of acute myocardial infarction, all of whom were subsequently followed up for at least 5 years. The mortality rate was 11.1% during the first year, but averaged only 2.9% per year from the second to fifth year. Over the entire follow-up period, the five variables that predicted mortality by multivariate analysis were QRS score, an exercise-induced ST segment shift, previous infarction, failure to achieve target heart rate or work load and ventricular arrhythmia during the exercise test. Because mortality differed markedly before and after 1 year, Cox regression analyses were performed separately for both of these periods. The factors that were predictive of mortality during the first year were an exercise-induced ST shift (p less than 0.0001, relative risk 7.8), failure to increase systolic blood pressure by 10 mm Hg or more during exercise (p = 0.0039, relative risk 4.3) and angina in hospital 48 hours or longer after admission (p = 0.0046, relative risk 3.4). None of these three variables was predictive of mortality after 1 year. Previous infarction (p = 0.0007), QRS score (p = 0.0042) and ventricular arrhythmia during the exercise test (p = 0.016) were predictive of mortality after the first year. Thus, clinical and exercise test variables are complementary predictors of mortality after myocardial infarction. An abnormal ST segment response during an early limited exercise test and angina in the hospital are common strong predictors of mortality to 1 year, but not thereafter. Late mortality correlates with markers of poor left ventricular function.