Short cervix, no prior preterm birth: start with vaginal progesterone alone

Design

Systematic review and bayesian network meta-analysis of 26 RCTs

Population

8,561 asymptomatic singleton pregnancies with short cervix, no prior spontaneous preterm birth

Primary outcome

Preterm birth before 34 and 37 weeks

Effect

Vaginal progesterone vs control, <34 weeks: RR 0.56 (0.44–0.72)

This bayesian network meta-analysis took 26 randomised trials (8,561 women) of asymptomatic singleton pregnancies with a sonographic short cervix and no previous spontaneous preterm birth, comparing vaginal progesterone, other progestogens, cerclage, pessary and their combinations.

In direct comparisons, vaginal progesterone alone reduced birth before 34 weeks (RR 0.56, 95% CI 0.44–0.72), on high-certainty evidence. In the network, pessary plus progesterone ranked first, with vaginal progesterone alone close behind. But combinations rested on lower-certainty evidence and carried more maternal adverse events: progesterone alone had about half the adverse-event risk of combination regimens (RR 0.56, 0.42–0.76). Rankings held across sensitivity analyses.

The authors' reading, which is reasonable, is that progesterone alone remains the standard and that combinations belong in trials for now. A ranking is not an effect size: a top rank on sparse data is less informative than a firm pairwise estimate. A pessary may suit selected women who cannot use vaginal progesterone.

• For a short cervix with no prior spontaneous preterm birth, offer vaginal progesterone alone as first-line.
• Avoid routinely adding a pessary or cerclage to progesterone outside a trial.
• Consider a pessary for women who cannot or will not use vaginal progesterone.
• Tell women combinations carried more maternal side-effects in this analysis.
• This applies only to women without a previous spontaneous preterm birth.

Why it matters

It stops 'more is better' escalation — adding pessary or cerclage to progesterone buys side-effects without firm extra benefit.

Don't overread it

Combination rankings rest on low-certainty evidence; they suggest a question to test, not a reason to combine.

The statistics, in plain English

A risk ratio of 0.56 means about 44% fewer births before 34 weeks with progesterone. 'Probability RR below 1 = 96%' means the model is very confident the effect is beneficial but says nothing about its size — which is why a high rank on thin data should not trump a precise direct estimate.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.