Outcomes
Primary and secondary outcomes are reported in Table 3. At W6-8 after delivery, anal incontinence was not statistically different between trial arms, nor was post-partum transient anal incontinence (11.7% in the CS arm vs 25.0% in the VD arm (absolute risk difference [95% CI]: -13.3 [-25.1 to 0.0]).
At M6, the median [IQR] Vaizey score for anal incontinence was 1/24 [0-4] in the CS arm vs 1/24 [0-3] in the VD arm (p=0.34) (figure 2). This primary outcome was actually measured at a median [IQR] time of 8.0 [6.8 – 11.2] months post-partum. When comparing Vaizey scores at inclusion and at the M6 visit, the results did not differ between the CS and VD groups (median (IQR) differences 0.0 [-1.5-2.0] and 0.0 [0.0-1.0], respectively, p = 0.9825). The effect of trial arm on Vaizey score at M6 differed between women with Vaizey score at inclusion <5 and women with Vaizey score at inclusion ≥5 (significant interaction, p=0.008). Post-hoc subgroup analyses showed that in the subgroup of 27 women with a Vaizey score before delivery ≥5, Vaizey score at M6 were significantly lower in the CS than in the VD arm (median 3 IQR [0-7] vs 6 [3.5-8.5], p=0.026).
At M6, there was no statistically significant difference between groups for urinary incontinence (MUH score), sexual function (FSFI) and physical and mental quality of life assessed with SF12.
Regarding maternal morbidity, 4 (4.9%) patients had at least one minor complication in the VD arm and 8 (8.8%) in the CS arm, including 3 (3.3%) anesthetic complications (headaches) in the CS arm and none in the VD arm. For neonatal outcomes, 5 (6.1%) had at least one complication in the VD arm including 4 transfers to neonatal care units (2 for respiratory distress and 2 for infection) versus none in the CS arm.
Among the 222 randomized women, 125 (56.3%) had post-partum endosonography at the M6 visit, 61 (54.5%) in the VD arm and 64 (58.2%) in the CS arm. Baseline characteristics of these women did not differ from those without endosonography (see supplementary appendix). External sphincter lesions deteriorated more frequently in the VD arm than in the CS arm (11 (22.4%) women vs 1 (2.2%), absolute risk difference -20.2 [95% CI: -31.7 to -7.6]), but no additional internal sphincter lesions were observed.