Perinatal Outcome in Women Treated With Progesterone for the Prevention of Preterm Birth: A Meta-Analysis

Sotiriadis, A.; Papatheodorou, S.; Makrydimas, G.

Obstetrical & Gynecological Survey:
doi: 10.1097/01.ogx.0000427632.80511.25

Screening of pregnant women for preterm birth, based on their obstetric history and sonographic measurement of the cervical length, can identify greater than 50% of those at risk. Administration of progesterone to at-risk women with a singleton pregnancy can significantly reduce rates of preterm birth. However, effects of this agent on the actual perinatal and long-term consequences of prematurity are difficult to assess. Existing randomized controlled trials (RCTs) and systematic reviews have focused mainly on the primary outcome of reduction of preterm birth rates. The aim of this meta-analysis was to systematically review published evidence and pool data on the perinatal outcome in women treated with progesterone for the prevention of preterm birth.

MEDLINE and SCOPUS databases were searched for clinical trials in which progesterone was given to prevent preterm birth in pregnant women at risk compared with placebo. Randomized controlled trials that compared progesterone versus placebo in women with singleton or multiple pregnancies at risk for preterm birth based on previous history or short cervix were selected. The CONSORT statement was used to address the reporting quality of the RCTs. The risk of bias in the RCTs was assessed with the “risk-of-bias” tool from the Cochrane Collaboration. The primary outcomes were the rates of neonatal and perinatal mortality. Secondary outcomes were the rates of perinatal complications, including respiratory distress syndrome (RDS), intraventricular hemorrhage (IVH) grade 3–4, sepsis, necrotizing enterocolitis (NEC), sepsis, retinopathy, and admission to the neonatal intensive care unit (NICU); a composite adverse outcome was also determined and defined as the presence of any perinatal morbidity or mortality.

Of 628 retrieved items, 16 RCTs reporting on the use of progesterone in asymptomatic women to prevent preterm birth were included in the meta-analysis. Pooled data indicated that progesterone administration in these women significantly decreased the risk for composite adverse outcome (RR, 0.576; 95% confidence interval [CI], 0.373–0.891), neonatal death (relative risk, 0.487; 95% CI, 0.290–0.818), RDS (RR, 0.677; 95% CI, 0.490–0.935), and admission to the NICU (RR, 0.410; 95% CI, 0.204–0.823). No significant differences were found in rates of perinatal death, grade 3–4 IVH, NEC, retinopathy, or sepsis. Pooled data from 3 studies that used vaginal progesterone in women with a short cervix showed that progesterone significantly decreased the rate of composite adverse outcome (RR, 0.576; 95% CI, 0.373–0.891) and RDS (RR, 0.464; 95% CI, 0.275–0.786), but results did not reach statistical significance for rates of neonatal death, perinatal death, grade 3–4 IVH, NEC, sepsis, or admission to the NICU. Three studies tested systemic progesterone in women with a singleton pregnancy and a history of preterm birth; pooled results found that progesterone significantly decreased rates of neonatal death (RR, 0.412; 95% CI, 0.201–0.842) and NICU admission (RR, 0.277; 95% CI, 0.160–0.479). In 7 RCTs reporting on women with twin pregnancies, progesterone administration did not significantly affect the rates of neonatal death, grade 3–4 IVH, NEC, retinopathy, sepsis, or NICU admission. Progesterone significantly increased the rates of composite adverse outcome (RR, 1.211; 95% CI, 1.029–1.425), perinatal death (RR, 1.551; 95% CI, 1.014–2.372), and RDS (RR, 1.218; 95% CI, 1.038–1.428). The pooled data from 2 RCTs on women with triplets did not show significant differences in the rates of composite adverse outcome, neonatal death, RDS, grade 3–4 IVH, NEC, or sepsis.

Preterm birth is a major cause of perinatal mortality and morbidity with long-term consequences. Results of this meta-analysis indicate that prophylactic progesterone administration in singleton pregnancies at risk can lower the rates of neonatal mortality, RDS, admission to the NICU, and a composite adverse outcome. Data, however, also indicate that use of progesterone in multiple pregnancies may lead to increased rates of perinatal death, RDS, and a composite adverse outcome.

Author Information

Second Department of Obstetrics and Gynecology, Aristotle University of Thessaloniki, Thessaloniki (A.S.); and Department of Obstetrics and Gynecology, Ioannina University Hospital, Ioannina (S.P., G.M.), Greece

Article Outline
Back to Top | Article Outline


(After decades of failed attempts to decrease the rate of preterm birth, recent studies have demonstrated progesterone treatment to be effective in reducing preterm delivery rates in some high-risk women, including those with a prior preterm birth, as well as those with a short cervix detected by midtrimester ultrasound. Although the effectiveness of progesterone in decreasing preterm birth is well described, it is more difficult to prove that this leads to lower rates of adverse perinatal outcomes, as these are less common, and most studies of progesterone are too small to demonstrate improved neonatal outcomes.

This abstracted article is a meta-analysis done to evaluate the effect on perinatal outcome of progesterone in women treated for the prevention of preterm birth. The authors acknowledge that progesterone has been demonstrated in multiple RCTs to decrease the rate of preterm birth in women with a previous preterm birth or found to have a short cervix by ultrasound. However, they point out that the primary reason that we worry about preterm birth is because of the increased risk of adverse perinatal outcomes, such as RDS, NEC, and IVH. Therefore, by performing a meta-analysis of 16 RCTs of progesterone that included perinatal outcomes, they were able to assess whether the demonstrated decrease in preterm birth has led to improved perinatal outcomes.

The authors found that, in singleton pregnancies, progesterone did in fact decrease rates of neonatal death, RDS, NICU admission, and composite adverse outcomes. Although decreased rates of other adverse outcomes such as IVH, NEC, retinopathy, and sepsis were not significant, in general there was a trend toward decreased rates of most of these important outcomes. The authors were also interested in whether there was a difference in the effectiveness of vaginal versus systemic progesterone for women with a short cervix but were unable to reach conclusions because of insufficient data.

This study also again confirmed the lack of benefit of progesterone in twin pregnancies that has been reported in several other studies and reviews (Ultrasound Obstet Gynecol 2011;38:272–280). In this meta-analysis, the results showed that progesterone did not affect the rates of neonatal death, IVH, NEC, retinopathy, sepsis, or NICU admission in twins. They did find, though, that progesterone increased the rates of composite adverse outcome, perinatal death, and RDS in twin pregnancies.

Why there is a difference in the effectiveness of progesterone in twins is unknown, although hypotheses include inadequate dosing given the larger volume of distribution and different mechanisms of preterm birth in multiple gestations. However, theories regarding a lack of effectiveness on preterm delivery rates in multiple gestations do not explain why adverse perinatal outcomes would be increased. If the only concern with multiples was that progesterone might be less effective, one could not fault providers for nevertheless attempting this treatment as “better than nothing.” However, it appears, at least from this analysis, that progesterone might in fact be worse than nothing. For the present, until more data are available, it appears that progesterone should be avoided in multiple gestations, as treatment led to a number needed to harm of 71 for perinatal death, 39 for RDS, and 31 for composite adverse outcome. Interestingly, these adverse outcomes appeared to be independent of the rates of preterm birth, which were comparable in twins between progesterone-treated women and controls. So the bottom line is that evidence clearly supports the benefit to progesterone treatment for women at risk for preterm birth due to either a short cervix or previous preterm birth, and I’d say that progesterone is now standard of care in these circumstances. The situation is less clear for multiples, with the potential that there is a risk to such treatment for women with twin gestations. As the authors indicate, the next important outcome is longer-term neurodevelopmental outcomes. In addition, we look forward to additional data on twin gestations and adverse outcomes.—MEN)

© 2013 Lippincott Williams & Wilkins, Inc.