High-risk patients with ST-elevation myocardial infarction derive greatest absolute benefit from primary percutaneous coronary intervention: Results from the Primary Coronary Angioplasty Trialist versus Thrombolysis (PCAT)-2 Collaboration
Background: Meta-analyses of randomized trials show that primary percutaneous coronary intervention (PPCI) results in lower mortality than fibrinolytic therapy in patients with myocardial infarction. We investigated which categories of patients with myocardial infarction would benefit most from the strategy of PPCI and, thus, have lowest numbers needed to treat to prevent a death. Methods: Individual patient data were obtained from 22 (n = 6,763) randomized trials evaluating efficacy and safety of PPCI versus fibrinolysis. A risk score was developed and validated to estimate the probability of 30-day death in individuals. Patients were then divided in quartiles according to risk. Subsequent analyses were performed to evaluate if the treatment effect was modified by estimated risk. Results: Overall, 446 patients (6.6%) died within 30 days after randomization. The mortality risk score contained clinical characteristics, including the time from symptom onset to randomization. The c-index was 0.76, and the Hosmer-Lemeshow test was nonsignificant, reflecting adequate discrimination and calibration. Patients randomized to PPCI had lower mortality than did patients randomized to fibrinolysis (5.3% vs 7.9%, adjusted odds ratio 0.63, 95% CI 0.42-0.84, P < .001). The interaction between risk score and allocated treatment interaction term had no significant contribution (P = .52) to the model, indicating that the relative mortality reduction by PPCI was similar at all levels of estimated risk. In contrast, the absolute risk reduction was strongly related to estimated risk at baseline: the numbers needed to treat to prevent a death by PPCI versus fibrinolysis was 516 in the lowest quartile of estimated risk compared with only 17 in the highest quartile. Conclusion: Primary percutaneous coronary intervention is consistently associated with a strong relative reduction in 30-day mortality, irrespective of patient baseline risk, and should therefore be considered as the first choice reperfusion strategy whenever feasible. If access to percutaneous coronary intervention is >2 hours, fibrinolysis remains a legitimate option in low-risk patients because of the small absolute risk reduction by PPCI in this particular cohort.