Progesterone supplementation for preventing preterm birth: a systematic review and meta-analysis.

AIM The aim of this study is to assess the role of progesterone in preterm birth prevention. METHODS A MEDLINE search (from 1966 to the present; date of last search January 2005) was performed - using the key words progesterone, pregnancy, preterm birth, preterm labor, and randomized, controlled trial - in order to identify randomized, controlled trials in which progesterone (either intramuscular or vaginal administration) was compared with placebo or no treatment. Data were extracted and a meta-analysis was performed. RESULTS Seven randomized, controlled trials were identified. Women who received progesterone were statistically significantly less likely to give birth before 37 weeks (seven studies, 1020 women, RR = 0.58, 95% CI = 0.48-0.70), to have an infant with birth weight of < or =2.5 kg (six studies, 872 infants, RR = 0.62, 95% CI = 0.49-0.78), or to have an infant diagnosed with intraventricular hemorrhage (one study, 458 infants, RR = 0.25, 95% CI = 0.08-0.82). CONCLUSIONS For progesterone supplementation to be advocated for women at the risk of preterm birth, the prolongation of gestation demonstrated in this meta-analysis must translate into improved infant outcomes, including a reduction in mortality. There is currently insufficient information to allow recommendations regarding the optimal dose, route, and timing of administration of progesterone supplementation.