Predicting long-term mortality after a myocardial infarction from routine hospital data.

Among 528 patients under 67 years of age discharged alive after a myocardial infarction (MI), the cumulative survival rates after 3, 5, and 7 years were 84.1%, 75.9% and 68.6%, respectively. Compared with the "normal" population, the relative mortality risk was 4.8 for the first year, 3.1 for the second, and on average 2.1 for the next 5 years. Significant age differences were not observed for relative mortality. A multivariate Cox proportional hazards model showed long-term mortality to be independently related to higher age, a reduced working activity before the MI, previous cardiovascular disease, and a higher inhospital complication score, which was computed by summing eight defined clinical events weighted for severity. The results indicate that a reasonable prediction of long-term survival after a MI can be made from routine hospital data.