Poster session II Diabetes, labor, medical-surgical-disease, obstetric quality & safety, prematurity, ultrasound-imaging: Abstracts 237 – 386| Volume 206, ISSUE 1, SUPPLEMENT , S146-S147, January 01, 2012

306: Prediction of cesarean section in women with an unfavorable cervix at term


      To identify indicators that quantify the risk of cesarean section in women with an unfavorable cervix in whom labor is induced.

      Study Design

      This was a secondary analysis of the randomized controlled trial, the PROBAAT study (NTR 1646), that compared induction of labor with a transcervical Foley catheter with induction with prostaglandins in term women with an unfavorable cervix. We prospectively collected the outcome (cesarean section) and its potential predictors (the Bishop score with its five components, maternal and prengancy characteristics). Potential predictors associated with cesarean section in univariable analysis (p≤0.50) were included in a multivariable model. Backward selection was used to develop a prediction model, predictors with a p≤0.157 remained in the final model. We used ROC analyses and calibration plots to assess the predictive accuracy of the model.


      We included 1036 women. The overall cesarean section rate was 21%. Independent predictors of the risk for cesarean section were maternal age (years) (OR 1.06, 95% CI 1.03-1.10), maternal height (cm) (OR 0.94, 95% CI 0.91-0.96), BMI (OR 1.65 to 2.97 in different categories, overall p-value 0.057), nulliparity (OR 7.14, 95% CI 11.1-4.35), indication for labor induction (Table 1, overall p-value 0.09), and no dilation (OR 1.35, 95% CI 1.89-0.98). The final model using these parameters had an area under the ROC-curve of 0.75 (95% CI 0.71-0.79), with good calibration (Figure 1).
      Tabled 1
      Table thumbnail grt34
      Figure thumbnail grr20
      Figure 1Calibration plot, showing the average observed CS rate (Y-axis) against the average calculated CS rate (X-axis), in five risk groups. In case of perfect calibration all points would be on the diagonal.


      In women with an unfavorable cervix in whom labor is induced, the risk of cesarean section can be predicted by combining maternal age, BMI, maternal height, parity, reason for labor induction, and dilation. This prediction model could be a useful tool for clinical decision making and patient information.