Risk factors for preterm birth and prolonged pregnancy following cervical cerclage in women with cervical insufficiency: a retrospective analysis and predictive model.

backgroundCervical insufficiency is one of the major risk factors for preterm birth. The efficacy of cervical cerclage (CC) in improving pregnancy outcomes remains controversial. This study aimed to compare obstetric outcomes between patients with cervical insufficiency who underwent CC and those who did not to determine the optimal cutoff point for preterm birth on the basis of fetal survival outcomes and to identify risk factors influencing preterm birth. Furthermore, we sought to analyze the determinants of pregnancy prolongation after CC and to develop predictive models using pregnancy prolongation as the endpoint.

methodsPatients who underwent CC intervention and those who declined it, were included. Following 1:1 propensity score matching, the duration from cervical insufficiency diagnosis to delivery and fetal survival outcomes were compared. On the basis of fetal survival status, receiver operating characteristic (ROC) curve analysis identified 30.5 weeks as an exploratory cutoff for discriminating fetal survival within this cohort. Using pregnancy prolongation after CC as a continuous endpoint, linear regression analysis was conducted, and nomograms were constructed to assess predictive performance.

results133 pregnant women with cervical insufficiency who underwent CC were included and matched with 133 women who did not receive the procedure. CC significantly prolonged the gestational duration (P < 0.01) but did not improve fetal survival (P = 0.45). The neutrophil percentage was significantly greater in mothers who delivered preterm at the 30.5-week cutoff (P = 0.022), and this difference persisted at 32 and 37 weeks. Multivariate analysis indicated that both neutrophil and C-reactive protein (CRP) levels were significantly associated with prolonged pregnancy (P < 0.05 for both). The nomogram model, which uses prolonged pregnancy as the endpoint, demonstrated good predictive performance (R²=0.49).

conclusionCerclage performance was associated with a longer gestational duration but was not significantly related to fetal survival outcomes. Inflammatory markers such as the neutrophil count and CRP level may help predict preterm labor and pregnancy prolongation following CC. The nomogram model based on preoperative characteristics demonstrated accurate predictive ability for fetal retention time. However, the ROC-derived gestational threshold and predictive models should be regarded strictly as exploratory, proof-of-concept findings requiring mandatory external validation before any clinical implementation.