Statistical analysis
Assuming a mean (SD) Vaizey score at M6 of 5 (6) in the control group8,18, 86 women/group would provide 90% power at a 2-sided α-level of 0.05 to detect a clinically meaningful difference of mean Vaizey score of 3 between groups. The target for enrollment was increased to account for potential loss to follow-up..
Baseline characteristics are reported by trial group (CS and VD) as numbers (%) for categorical variables and means (± standard deviations, SD) or medians [interquartile range, IQR] for continuous variables, as appropriate.
All analyses were performed according to the intention-to-treat (ITT) principle. Missing data were handled using multiple imputations (5 datasets) on principal and secondary endpoints except maternal and neonatal outcomes. The Vaizey score at M6 post-partum was compared between CS and VD groups using a permutation test, as this variable was not normally distributed and showed a floor effect. A post-hoc subgroup analysis of the primary outcome was conducted in the 27 women with Vaizey scores ≥5 (a cut-off usually defining anal incontinence22) at the prenatal visit, after testing positive for interaction with trial arm. Secondary outcomes were compared between VD and CS groups using Chi-square or Fisher exact test, Student, Wilcoxon or permutation test as appropriate.
All statistical analyses were performed using SAS V.9.4 (SAS Institute Inc., Cary, North Carolina, USA).