Cervical pessary for preventing preterm birth in singletons: A dynamic systematic review and meta-analysis

Objective To evaluate the effectiveness of cervical pessary in singleton pregnancies for the prevention of preterm birth (PTB) through an ongoing ‘dynamic’ systematic review and meta-analysis. Study Design A systematic review was completed in Medline, EMBASE, Web of Science, Cochrane CENTRAL, and ClinicalTrials.gov. All randomised controlled trials (RCTs) that studied the use of the cervical pessary compared to a control group in singleton pregnancies were included. Delivery, maternal and neonatal outcomes were evaluated comprising the core outcome set for PTB. The summary measures were reported as relative risk (RR) with 95% of confidence interval (CI) using a random effects model. A Higgins I2 of >80% was considered to be high heterogeneity and therefore the trial estimates were not pooled. The quality of evidence for pre-specified outcomes was assessed using the GRADE methodology. We plan to perform an ongoing ‘dynamic’ meta-analysis by repeating the systematic review and analysis every 6 months. Results Six qualifying trials were identified, all trials studied the use of pessary with CL <25mm (n=1,992 pessary vs 999 controls). Four trials administrated vaginal progesterone to the pessary and control group. There were no significant differences seen in the rates of spontaneous PTB (sPTB) or any PTB prior to 28, 34, or 37 weeks (Table 1). There was high heterogeneity noted for sPTB <34 weeks (I2=84%, n=1,830) (Figure 1). Three trials found no significant reduction in sPTB rate <34 weeks, while two trials demonstrated a significant reduction (6.3% vs 26.8%; 7.3% vs 15.3%, respectively). Women with pessaries were more likely to report increased vaginal discharge at follow up visits (66% vs 28%, I2=82%, n=1,830) and pelvic pain (RR 3.35, 95% CI 2.00-5.60, n=1,288). Only 5.4% requested removal of the pessary. There were no significant differences in delivery or neonatal outcomes. Conclusion Singleton gestations with short CL randomized to cervical pessary did not have a significant difference in rates of preterm birth compared to controls. However, these results demonstrate the large heterogeneity in both the statistical analysis as well as in the results of individual trials. We plan to continue to investigate using a ‘dynamic’ systematic review and meta-analysis to include forthcoming trials.