To develop a prediction model for spontaneous vaginal delivery (SVD) in women with prolonged second stage (PSS) of labor undergoing a trial of labor after cesarean (TOLAC).METHODS:
Secondary analysis of the MFMU Cesarean Registry. Inclusion criteria: term, singleton gestation, TOLAC, no prior vaginal delivery, who reached 10 cm cervical dilation. Women with PSS (≥3 hours with epidural or ≥2 hours without epidural) were compared to women with a normal second stage. Primary outcome: development of a model for predicting SVD. Univariate analysis, multivariable logistic regression were used to examine independent association of PSS with outcomes. A simple risk score was created using the regression model to calculate maximum likelihood estimates of probabilities for each subject. The association of probability quintile with incidence of SVD was examined.RESULTS:
There were 364 women with PSS with epidural and no women with PSS without epidural anesthesia; therefore analysis was conducted only for women with epidurals. SVD occurred in 96 cases (26.4%). In univariate analyses, SVD was associated with birth weight, insurance, indication for previous cesarean and male sex. The multivariate model predicting SVD, showed no independent predictors. The simple risk score was calculated. Incidence of SVD was 13.9% in the lowest risk quintile and 42.5% in the highest risk quintile. Risk quintile based on the prediction model was significantly associated with SVD (p = 0.006).CONCLUSION:
A prediction model for SVD in women with singleton gestation undergoing TOLAC with PSS and epidural may be useful to clinicians facing decisions regarding delivery.