Comparative Effectiveness of Pessary Placement, Cervical Cerclage, or Expectant Management in Preventing Preterm Delivery in Twin Pregnancies

<b>Objective:</b> The objective of this study was to evaluate the association between cervical management strategies, specifically pessary placement, cervical cerclage, or expectant management, and gestational age at delivery in twin pregnancies and to assess the prognostic value of cervical characteristics for early preterm birth (<33 weeks). <b>Methods</b>: We conducted a retrospective cohort study including 120 twin pregnancies managed at a tertiary referral center between 2019 and 2024. Pregnancies with positive vaginal or cervical microbiological cultures or abnormal cervical cytology were excluded. The management strategy was selected based on cervical characteristics and clinical judgment. Gestational age at delivery was compared across intervention groups using descriptive statistics, kernel density plots, boxplots, and Kaplan-Meier survival analysis. Multivariable Cox proportional hazards regression was performed to estimate adjusted hazard ratios (HRs) for early delivery, including intervention type and cervical parameters (length, diameter, and funneling). <b>Results</b>: Overall, 26 of 120 pregnancies (21.6%) resulted in delivery before 33 weeks. Pessary placement was associated with longer gestational duration compared with cerclage or expectant management. Kaplan-Meier analysis demonstrated a clear separation of survival curves by intervention group, with the pessary group maintaining pregnancy to later gestational ages (log-rank <i>p</i> < 0.001). In multivariable Cox regression analysis, pessary use was associated with a significantly lower hazard of early delivery compared with cerclage (HR = 0.088, 95% CI: 0.035-0.220; <i>p</i> < 0.001). Expectant management showed a trend toward an increased risk of early delivery (HR = 2.44; <i>p</i> = 0.067). Cervical length and diameter were not independently associated with early delivery after adjustment for intervention type. Funneling was associated with a lower hazard of early delivery, a finding that warrants cautious interpretation. <b>Conclusions</b>: In this retrospective cohort of twin pregnancies without microbiological evidence of infection, pessary placement was associated with prolonged gestation and a lower hazard of early preterm delivery compared with cerclage or expectant management. These findings support a personalized, risk-adapted approach to cervical intervention selection in twin pregnancies. Prospective, randomized studies incorporating etiologic stratification are needed to confirm these associations and guide clinical practice.