Objective: In a systematic overview of individual data from 100,000 patients with Myocardial Infarction (MI), two Prognostic Indexes (PI) were calculated and compared to test the efficacy of the ACE-Inhibitor therapy in patients at different risk of death.

Methods: The PIs were calculated on the basis of the main risk factors (age, heart rate, previous MI, killip class and site of MI) collected for the overview. The univariate PI was merely computed by counting the number of risk factors and patients were grouped into four categories of growing risk (1='O-1', 2='2', 3='3', 4='>3' risk factors). Tbe multivariate PI was based on the predicted probability of death obtained from a logistic regression as a combination of the risk factors. As for the univariate PI, four categories of growing risk were generated (1='1ow risk', 2='medium risk', 3='high risk', 4='very high risk') on the basis of the predicted probability of death.

Results: The efficacy of the treatment in term of mortality was tested on both the univariate and multivariate PIs and the results are shown in the following table:

Univariate PI Multivariate PI
Category OR C.I.(99%) OR C.I.(99%)
1 1.00 0.85-1.18 0.86 0.76-0.98
2 0.93 0.84-1.04 1.02 0.90-1.16
3 0.90 0.79-1.02 0.90 0.79-1.02
4 0.81 0.66-1.00 0.94 0.82-1.08
2p(heterog. ?2)=0.22 2p(heterog. ?2)=0.1
2p(trend ?2)=0.02 2p(trend ?2)=0.52

Discussion: A more sophisticated multivariate PI did not confirm the trend showed by univariate PI of an Lncreasiiig efficacy of ACE-Inhibitor with increased risk.