Is primary angioplasty for some as good as primary angioplasty for all? Modeling across trials and individual patients

OJECTIVES: To investigate whether proper patient selection might allow most of the benefits of population-wide primary coronary angioplasty to be captured in a subgroup of high-risk patients. BACKGROUND: Despite growing evidence that angioplasty yields better outcomes, thrombolytic therapy remains the most common form of reperfusion therapy in acute myocardial infarction (AMI) because of limited capacity for primary coronary angioplasty at most hospitals. METHODS: We used a validated logistic regression model, based on individual patient characteristics, to estimate the distribution of mortality risk in a community-based sample of 1,058 patients who received reperfusion therapy for AMI. To estimate the benefits across different baseline risks, we examined the results of 10 randomized controlled trials using meta-regression techniques. RESULTS: Assuming a constant relative risk reduction, 68% of all mortality benefits in our community-based patient sample could be captured by treating only those patients in the highest quartile of mortality risk and 87% of the benefit could be captured by treating those in the highest half. Moreover, meta-regression of the results from the 10 clinical trials suggests that patients with a mortality risk of less than 2% may be unlikely to receive any mortality benefit. With this risk-benefit relationship, treatment of only the 39% of patients with the highest risk would yield equivalent mortality outcomes to population-wide angioplasty. CONCLUSIONS: Most of the incremental benefits of primary angioplasty can be achieved by treating high-risk patients. For these patients, thrombolytic therapy may be difficult to justify if nearby primary angioplasty is available. For most patients, however, thrombolytic therapy appears to be an effective alternative.