Perspective:

Several recent randomized trials have failed to demonstrate superiority of PFO closure over medical therapy for patients with cryptogenic stroke. The results of these studies have been questioned because of concerns for adequacy of sample size, varying inclusion criteria, heterogeneity of medical management, and possible differences between devices. This study uses the method of network meta-analysis, which allows for comparison of differing treatment approaches across multiple studies. A limitation of the study is heterogeneity in the medical management group. The study suggests superiority of the Amplatzer device over the other devices as well as medical therapy. The number needed to treat for the Amplatzer was 29, with an accompanying cost analysis placing the cost-effectiveness ratio just at the ‘willingness to pay’ threshold for countries with national health care programs. An interesting accompanying editorial by Benkhadra, et al. discusses the methodology of network meta-analysis, and questions the authors’ conclusions of this study over concerns for the relatively high rate of lost-to-follow-up leading to possible bias and the low event rate leading to imprecision.