Antiplatelet agents in the secondary prevention of stroke: meta-analysis of the randomized control trials.

Randomized control trials of antiplatelet agents in the prevention of stroke following transient ischemic attacks have had conflicting results. The decision to employ aspirin instead of placebo as the control regimen in trials testing newer antiplatelet agents emphasizes the need for an accurate estimate of the efficacy of older drugs. A meta-analysis of seven randomized control trials comparing aspirin and/or sulfinpyrazone or dipyridamole with placebo was performed. For aspirin compared with placebo, a nonsignificant reduction in stroke of 15% (odds ratio 0.85, 95% confidence interval 0.60-1.19; chi 2 = 0.78, p greater than 0.30) was found. For aspirin combined with sulfinpyrazone or dipyridamole compared with placebo, a 39% reduction in stroke was observed (odds ratio 0.61, 95% confidence interval 0.39-0.95; chi 2 = 4.22, p less than 0.05); at the same time a 350% increase in gastrointestinal hemorrhage or peptic ulcer was noted (odds ratio 3.5, 95% confidence interval 1.26-9.75; chi 2 = 4.61, p less than 0.05). A trend in reduction of strokes for men was observed (odds ratio 0.58, 95% confidence interval 0.32-1.07; chi 2 = 2.52, p less than 0.15) for any regimen containing aspirin. The significant benefit of aspirin-combination therapy on stroke must be interpreted cautiously because of a number of possible biases. It is still conceivable that aspirin alone may decrease the incidence of stroke by as much as 40%, but a sample of greater than 13,000 patients would be needed to confirm the benefit observed in our analysis.