Cervical cerclage for short cervix at 24 to 26 weeks of gestation: systematic review and meta-analysis of randomized controlled trials using individual patient-level data.

Objective: This study aimed to determine whether cervical cerclage for a transvaginal ultrasound-detected short cervical length after 24 weeks of gestation in singleton pregnancies reduces the risk for preterm birth.

Data Sources: Ovid MEDLINE, Scopus, and the Cochrane Central Register of Controlled Trials were searched using the following terms: "cerclage, cervical," "uterine cervical incompetence," "obstetrical surgical procedures," "cervix uteri," "randomized controlled trial," and "controlled clinical trial."

Study Eligibility Criteria: All randomized controlled trials comparing cerclage placement with no cerclage in singleton gestations with a transvaginal ultrasound-detected short cervical length ≤25 mm between 24+0/7 and 29+6/7 weeks of gestation were eligible for inclusion.

Methods: Individual patient-level data from each trial were collected. If an eligible trial included patients with both multiple and singleton gestations with a short cervical length detected either before or after 24+0/7 weeks of gestation, only singletons who presented at or after 24+0/7 weeks were included. The primary outcome was preterm birth <37 weeks' gestation. Secondary outcomes included preterm birth <34, <32, and <28 weeks' gestation, gestational age at delivery, latency, preterm prelabor rupture of membranes, chorioamnionitis, and adverse neonatal outcomes. Individual patient-level data from each trial were analyzed using a 2-stage approach. Pooled relative risks or mean differences with 95% confidence intervals were calculated as appropriate.

Results: Data from the 4 eligible randomized controlled trials were included. A total of 131 singletons presented at 24+0/7 to 26+6/7 weeks of gestation and were further analyzed; there were no data on patients with a cerclage at 27+0/7 weeks' gestation or later. Of those included, 66 (50.4%) were in the cerclage group and 65 (49.6%) were in the no cerclage group. The rate of preterm birth <37 weeks' gestation was similar between patients who were randomized to the cerclage group and those who were randomized to the no cerclage group (27.3% vs 38.5%; relative risk, 0.78; 95% confidence interval, 0.37-1.28). Secondary outcomes including preterm birth <34, <32, and <28 weeks' gestation, gestational age at delivery, time interval from randomization to delivery, preterm prelabor rupture of membranes, and adverse neonatal outcomes such as low birthweight, very low birthweight, and perinatal death were similar between the 2 groups. Planned subgroup analyses revealed no statistically significant differences in the rate of preterm birth <37 weeks' gestation between the 2 groups when compared based on cervical length measurement (≤15 mm or ≤10 mm), gestational age at randomization (24+0/7 to 24+6/7 weeks or 25+0/7 to 26+6/7 weeks), or history of preterm birth.

Conclusion: Cervical cerclage did not reduce or increase the rate of preterm birth among singleton pregnancies with a short cervical length detected after 24 weeks of gestation. Because there was a 22% nonsignificant decrease in preterm birth associated with cerclage, which is a similar amount of risk reduction often associated with ultrasound-indicated cerclage before 24 weeks' gestation, further randomized controlled trials in this patient population are warranted.