Sub-grouping of post myocardial infarction patients according to their one-year death risk

This study was undertaken to investigate whether sub-grouping post myocardial infarction (MI) patients with various risks of death was possible. Data on 6900 consecutive patients with clinical symptoms of acute MI were collected prospectively within 48 h of their admission. Vital status at one year was obtained for all but 264 (3.8%). A total of 476 (7.2%) patients died within the first 7 days. Correlation of baseline historical, demographic, biological and ECG variables with total mortality was first investigated with univariate analysis. Fifty-three secondary risk factors (SRF) were identified, based on: (1) a P value for comparison between dead and alive lower than 0.02; (2) a minimum number of 100 subjects in each cell of the four-fold table. Based on this set of SRFs, two approaches were used to investigate the distribution of patients still alive at 8 days regarding the one-year mortality risk, and to assign them in three sub-groups of equal size of increasing risk of death at one year. One approach consisted in merely distributing patients by the number of SRFs; the other consisted in distributing patients according to an individual expected risk computed with a logistic regression. Both approaches gave consistent findings. One third of the patients with the lowest risk of death within 12 months had an individual risk between 0% and 8%; the average mortality was 5-6% which represented 10.7% of the total number of deaths. The third with the highest individual risks had an individual risk ranging from 17-19% to 100%, with an average mortality of 35-37%; they accounted for 66-67% of the total number of deaths.(ABSTRACT TRUNCATED AT 250 WORDS)