Progesterone for early-pregnancy bleeding: who the trials say benefits

Progesterone given for bleeding in early pregnancy did not increase the overall rate of live births in the largest trial to test it, but it appeared to help one group: women who were bleeding and had already had three or more miscarriages [s1][s2]. In that subgroup the live-birth rate was 72% with progesterone against 57% with placebo, a 15-percentage-point difference [s2].

Bleeding in the first weeks of pregnancy is common and is strongly associated with miscarriage, and progesterone is a hormone essential to maintaining a pregnancy [s1]. Several small studies had hinted that giving supplemental progesterone might reduce loss, but they were small and methodologically weak [s2]. Two large, high-quality trials were built to settle the question, and the more precise reading of them is not "progesterone works" or "progesterone fails" but "it depends on who is treated."

What the main trial found

The PRISM trial randomly assigned 4,153 women with vaginal bleeding in early pregnancy, recruited at 48 UK hospitals, to vaginal suppositories containing either 400 mg of progesterone or a matching placebo twice daily, from the time they presented with bleeding through 16 weeks of gestation [s1]. The primary outcome was the birth of a live baby after at least 34 weeks [s1].

Overall, the result was a near-miss that did not reach significance. Live births occurred in 75% of the progesterone group (1,513 of 2,025 women) and 72% of the placebo group (1,459 of 2,013) — a relative rate of 1.03 (95% CI, 1.00 to 1.07; P = 0.08) [s1]. The incidence of adverse events did not differ significantly between the groups [s1]. Read on its own, PRISM is a null trial: a three-percentage-point gap that could be chance.

The signal is in a subgroup

A later critical evaluation by the trial team pooled PRISM with PROMISE, a companion trial of 836 women with recurrent miscarriage but no requirement for bleeding, which had also found a 3% higher live-birth rate with progesterone amid substantial statistical uncertainty [s2]. Across both trials, one pattern held: live-birth rates rose with the number of a woman's previous miscarriages [s2].

In PRISM's prespecified subgroup of women with both risk factors — at least one previous miscarriage and current bleeding — the live-birth rate was 75% (689 of 914) with progesterone versus 70% (619 of 886) with placebo (risk ratio 1.09; 95% CI, 1.03 to 1.15; P = 0.003) [s2]. The effect was larger still in women with three or more previous miscarriages and current bleeding: 72% (98 of 137) versus 57% (85 of 148), a 15-percentage-point gap (risk ratio 1.28; 95% CI, 1.08 to 1.51; P = 0.004) [s2]. No short-term safety concerns emerged in either trial [s2].

How much weight the finding bears

Subgroup results deserve caution, because carving a trial into pieces produces apparent effects by chance. This one is stronger than most: it was prespecified, it followed a biologically coherent dose-response with miscarriage history, and it was first seen in PROMISE and then replicated in PRISM [s2]. The trial team also report that the subgroup analysis met all 11 conditions on a standard checklist for judging whether a subgroup effect is credible [s2]. But the strongest subgroup rests on small numbers — 137 and 148 women — and the overall trial did not cross the significance line [s1][s2]. This is a credible signal in a defined group, not proof of a general benefit.

The trial team's own conclusion is measured: women with a history of miscarriage who present with bleeding in early pregnancy may benefit from vaginal micronized progesterone 400 mg twice daily, and the decision is one for shared decision-making with a clinician [s2].

For related evidence on treating and preventing pregnancy complications, see our coverage of the aspirin dose debate in preeclampsia prevention, and of the WHO's first global guidelines on diabetes in pregnancy.

What to watch

Whether guidelines and future trials continue to focus progesterone on the group where the benefit concentrates — women bleeding in early pregnancy who have miscarried before — rather than offering it to all women with early bleeding, most of whom the trials suggest gain little.

This article describes trial evidence about a treatment and is not medical advice. Decisions about progesterone in pregnancy belong with a clinician who knows the individual case.

Sources

1. A Randomized Trial of Progesterone in Women with Bleeding in Early Pregnancy — New England Journal of Medicine, 8 May 2019
2. Micronized vaginal progesterone to prevent miscarriage: a critical evaluation of randomized evidence — American Journal of Obstetrics and Gynecology, 30 January 2020