Comparison of cesarean versus vaginal delivery of extremely preterm gestations in breech presentation: retrospective cohort study

Objective: As survival increases at earlier gestational ages, the optimal mode of delivery, especially in cases of breech presentation, is of increasing importance. The objective of this study was to compare outcomes of vaginal delivery (VD) and cesarean section (CS) births for infants in breech presentation at borderline viability.

Study design: A retrospective chart review of live breech births between 23?+?0 and 25?+?6 weeks gestation at a tertiary university center from 2003 to 2013 was conducted. Those delivered vaginally were compared with those delivered by CS. Stillbirths and deliveries where no resuscitation was intended were removed from the analysis. Variables were compared using a Student t-test (continuous), Mann–Whitney U test (categorical), or a Chi-squared test (count). Logistic regression analysis was performed to further evaluate the results. Results with p?Results: One hundred seventy-six births were included, 36 VD and 140 CS. Baseline characteristics were similar between groups. Gestational age at delivery was significantly higher in CS deliveries (24.9?±?0.6 versus 24.5?±?0.7, p?=?.0007). The rate of neonatal death (23.6% versus 44.4%, p?=?.0127) was significantly lower in those born by CS. All other neonatal outcomes including Apgar scores at one and 5?min, cord gases, birth weight, length of stay in NICU, incidence of respiratory complications, and incidence of high-grade IVH demonstrated no significant differences. Logistic regression suggested that male sex, lower birth weight, and earlier gestational age are significantly associated with neonatal mortality. Thirty percent of uterine incisions were of the classical, high transverse or inverted-T types. The estimated blood loss was significantly higher in CS births (706.6?±?226.4 versus 327.4?±?174.1?mL, p?Conclusion: CS delivery of breech infants at borderline viability had a protective effect on neonatal mortality compared to VD depending on the regression model utilized. Infant sex, birth weight, and gestational age also contribute significantly to neonatal mortality. A prospective study of planned method of delivery is recommended to further explore this finding.