Abstract
INTRODUCTION: Cervical cerclage is an acceptable procedure in women with cervical insufficiency and is known to be effective in the prevention of preterm delivery. However, limited data exist regarding long-term health outcomes among offspring exposed to cerclage during pregnancy. Since the presence of a foreign body during pregnancy may change the vaginal microbiome, we aimed to study whether a cervical cerclage is associated with long-term infectious morbidity of the offspring. STUDY DESIGN: A retrospective population-based cohort study was performed at a tertiary medical center, including all singleton deliveries between the years 1991-2021. Long-term infectious morbidity was compared among offspring after pregnancies with and without cervical cerclage. The diagnoses of infectious morbidities were defined based on ICD-9 codes as recorded in community clinics and hospitalization files. A Kaplan-Meier survival curve was utilized to evaluate the cumulative incidence. A Cox proportional hazards model was used to control for confounders. RESULTS: Out of 356,356 offspring included in the analysis, 0.4% (n = 1416) were following pregnancies with cervical cerclage. Unadjusted analyses demonstrated no significant difference in total infectious morbidity between the groups (OR 1.0, 95% CI 0.9-1.1; p = 0.369, Table 1). Kaplan-Meier analysis showed no difference in cumulative incidence (log-rank test P-value = 0.19, Fig. 1). In the primary analysis, cerclage was not associated with long-term infectious morbidity. However, in a secondary model, after adjustment for confounders including gestational age, obesity and diabetes, cerclage exposure was associated with a modest reduction in the risk of long-term infectious morbidity (adjusted HR 0.9, 95% CI 0.87-0.99, p = 0.036). CONCLUSION: In this large population-based cohort, cervical cerclage was not associated with increased long-term infectious morbidity in offspring. A modest association with reduced infectious morbidity was observed after adjustment for confounding factors. These findings should be interpreted cautiously given the observational design and potential residual confounding.