Introduction: Studies on the role of pelvimetry in predicting the mode of delivery have yielded mixed results. The aim of this study is to determine utility of the mid-pelvic measurement in predicting the mode of delivery in a cohort of women attempting vaginal delivery.
 Study Design and Methods: With this prospective cohort study, we evaluated the use of x-ray in predicting the delivery outcome of women attempting to deliver vaginally. Women were recruited before labor at >36 weeks gestational age and X-ray pelvimetry was performed after delivery. The exposures of interest were mid pelvic measures including anteroposterior diameters, transverse diameters and circumferences. The outcome measures were whether the women delivered via vaginal route or had cesarean delivery. We estimated the distribution and calculated measures of central tendency and spread of each pelvic dimension. Area under the receiver-operating characteristics curve (AUC) was used to estimate the overall predictive ability for each pelvic dimension and the optimal cut-point was estimated using the method of Liu. Logistic regression analysis was used to identify independent predictors for mode of delivery. The Hosmer – Lemeshow goodness-of-fit test was used to estimate the overall fit while the AUC was used to estimate the overall prediction of the final model.
 Results: A total of 426 women met the inclusion criteria. The mean gestational age at delivery was 40 (±6.0) weeks and the majority were black parturient (62.6%0. A slight majority were nulliparous (52.1%). In all, 127 women (29.8%) were delivered by cesarean delivery. All the pelvic inlet and mid pelvic dimensions were approximately normally distributed. The AUC ranged from 0.62 to 0.86. While the pelvic inlet and mid pelvic diameters had equivalent AUC, the optimal cut-point of the mid pelvic anteroposterior diameter (10.8cm) had both higher sensitivity and specificity than the pelvic inlet anteroposterior diameter (10.2cm) (95% and 85% versus 90% and 80%, respectively). There was an inverse relationship between the mid pelvic anteroposterior diameter and mode of delivery by cesarean section, with a 100% risk among women with a diameter of 9cm or less. However, the multivariable model developed had an AUC of 0.90, indicating overall good and higher predictive ability than the mid-pelvic anteroposterior diameter alone.
 Conclusions: In this large cohort study, the mid pelvic anteroposterior diameter best predicted mode of delivery by cesarean section.
Read full abstract