Background: The aims of the study were to evaluate the performance of the Duke Treadmill Score (DTS) and the Veteran Affairs Medical Center Score (VAMCS) in predicting 6-month death in GISSI-2 study survivors of acute myocardial infarction treated with thrombolytic agents, and to develop a simple predictive scoring system from the same database.
Methods: Patients of the GISSI-2 study (n = 6251) performed a maximal symptom-limited exercise test 1 month after myocardial infarction. We calculated for each patient the DTS and the VAMCS. Based on the coefficients of a multivariate analysis of our database, we developed a simple predictive scoring system and performed an internal validation. The prognostic value of each scoring system was assessed by multivariate analysis.
Results: Six-month mortality rates in the subgroups of each scoring system were as follows: DTS: low risk 0.6%, moderate risk 1.8%, high risk 3.4% (P < or =.0001); VAMCS: low risk 0.6%, moderate risk 1.9%, high risk 4.7% (P < or =.0001); GISSI-2 Index: low risk 0.5%, moderate risk 1.9%, high risk 6.1% (P < or =.0001). The results of multivariate analysis (relative risk [RR] and 95% CI) were as follows: DTS: moderate risk 2.50 (1.47-12.59), high risk 5.13 (3.61-15.55); VAMCS: moderate risk 2.65 (1.53-4.59), high risk 5.97 (3.10-11.49); GISSI-2 Index: moderate risk 3.16 (1.81-5.52), high risk 8.65 (4.36-17.18).
Conclusions: The use of ergometric-derived prognostic score systems in a population of survivors of acute myocardial infarction treated with thrombolytic drugs distinguishes subgroups at different risks of death and allows an appropriate recourse to more costly procedures.