Follicle-stimulating hormone versus human menopausal gonadotropin for in vitro fertilization cycles: a meta-analysis**Presented at the annual meeting of the Canadian Fertility and Andrology Society in St. Andrews-by-the-Sea, New Brunswick, Canada, September 7 to 10, 1994.

OBJECTIVE: To conduct a systematic overview of available data comparing FSH with hMG in IVF treatment cycles. DESIGN: A meta-analysis of randomized trials of FSH versus hMG use in ovarian stimulation protocols, with or without GnRH agonists, in IVF treatment cycles. Search strategies included on-line searching of the National Library of Medicine MEDLINE data base from 1975 to 1993 and hand searching of bibliographies of relevant publications and reviews and abstracts of scientific meetings. Common odds ratio (OR) and overall absolute treatment effect were calculated after demonstrating homogeneity of treatment effect across all trials. MAIN OUTCOME MEASURES: Clinical pregnancy rates per cycle start, per cycle reaching oocyte retrieval, and per cycle reaching ET. RESULTS: Eight trials met the inclusion criteria. The overall OR in favor of FSH for cycle start, oocyte retrieval, and ET were 1.71, 1.69 and 1.70, respectively, and represented an overall absolute treatment effect of 8.5%, 8.0% and 8.3%, respectively. CONCLUSIONS: This meta-analysis demonstrates that in IVF cycles the use of FSH is associated with a significantly higher clinical pregnancy rate than hMG.