Background
Cervical cerclage is used in an attempt to reduce recurrence risk of preterm birth, but evidence for use is limited.

Aims
To compare pregnancy outcomes among women with a single previous midtrimester delivery when managed with or without a cervical cerclage.

Materials and methods
Population‐based cohort study of all women in New South Wales, Australia with a singleton pregnancy ending in birth/miscarriage ≥14 and <28 weeks, between 2003 and 2011. Modified Poisson regression was used to compare outcomes in the next subsequent pregnancy, for women with a cerclage inserted <14 weeks, and those without cerclage. The primary outcome was gestational age <37 weeks at birth/miscarriage in the next pregnancy. Secondary outcomes included: maternal morbidity, preterm prelabour rupture of membranes (PPROM), stillbirth/neonatal death and composite neonatal morbidity for liveborn infants ≥28 weeks. Adjusted risk ratios (ARR) and 95% confidence intervals (CI) were determined. Results Five thousand, six hundred and ninety‐eight births/miscarriages were potential index deliveries. Of these, 2175 women had an eligible subsequent pregnancy: 108 received cerclage at <14 weeks gestation, 2067 did not. Women with cerclage were significantly more likely to deliver <37 weeks than those without (39.8% vs 19.3%, ARR 1.92, 95% CI 1.48–2.48), and had increased risks of PPROM (ARR 4.38, 95% CI 2.62–7.32) and stillbirth/neonatal death (ARR 2.20, 95% CI 1.02–4.73). Following cerclage, liveborn infants ≥28 weeks had double the risk of severe morbidity (ARR 2.54, 95% CI 1.55–4.16). Conclusions In women with a single previous midtrimester delivery, cervical cerclage <14 weeks gestation in subsequent pregnancy was associated with worse pregnancy outcomes.