The role of luteal phase support in infertility treatment: a meta-analysis of randomized trials**Presented at the annual meeting of the Society of Obstetricians and Gynaecologists of Canada in Ottawa, Canada, June 22 to 26, 1993.

OBJECTIVE: To determine whether the use of luteal phase support improves pregnancy rate (PR) in infertility. DESIGN: A meta-analysis of randomized trials of luteal phase support. Search of the National Library of Medicine MEDLINE data base from 1971 using the words luteal, pregnancy, human, and comparative. Bibliography of relevant articles, reviews, and abstracts of scientific meetings were hand searched. All randomized controlled trials of luteal phase support in infertility were included. Luteal phase support for recurrent abortion and nonrandomized trials were excluded. The common odds ratio was calculated for each intervention using the Mantel-Haentzel test. Homogeneity of treatment effect was evaluated using the Breslow-Day test. MAIN OUTCOME MEASURES: Pregnancy per cycle, rate of spontaneous abortion, and ovarian hyperstimulation syndrome rate. RESULTS: Eighteen trials met the above criteria. Human chorionic gonadotropin improved PRs in IVF when GnRH agonist (GnRH-a) was used (n = 151) and was superior to P (n = 352). Its benefit in all IVF cycles, however, was not established because of significant heterogeneity of treatment effect. Progesterone improved the PR in all IVF cycles (n = 457). No significant reduction in spontaneous abortion was noted with luteal support (n = 200). Ovarian hyperstimulation syndrome occurred in 5% of patients with hCG. Combination of data from trials of luteal support with other infertility therapies was not possible because of the differences in patient populations. CONCLUSIONS: The meta-analysis supports the routine use of hCG in IVF cycles using a GnRH-a. Progesterone was also beneficial for luteal phase support in IVF. For other infertility therapy, however, further research is needed to evaluate the role of luteal phase support.