To the Editor: The articles by Meier et al.1 and Carroll et al.2 and the corresponding editorial by Messe and Kent3 (March 21 issue) illustrate a major problem in clinical trials. When it is not obvious which of two therapies is better, sufficient numbers of events are essential to reach a conclusion. Performing a prospective, randomized trial is not enough. In the PC Trial (Clinical Trial Comparing Percutaneous Closure of Patent Foramen Ovale [PFO] Using the Amplatzer PFO Occluder with Medical Treatment in Patients with Cryptogenic Embolism),1 primary-end-point events occurred in only 18 patients in the two groups, even though 414 patients and 29 international sites participated. Any difference may have been due to chance. The investigators in the RESPECT (Randomized Evaluation of Recurrent Stroke Comparing PFO Closure to Established Current Standard of Care Treatment)2 trial enrolled 980 patients at 69 sites, yet there were only 25 primaryend-point events. The proper conclusion of these studies, even though they were well designed and carefully performed, is that too few events were observed to draw any conclusion. When the Early Breast Cancer Trialists’ Collaborative Group4 performed the first metaanalysis of the value of adjuvant tamoxifen — one of the most effective medications in our pharmacopoeia — for breast cancer, only 6 of 42 randomized trials had shown significant benefit. Only trials with many patients and many events are likely to overcome the play of chance.