Meta-analysis to determine the incidence of obstetric anal sphincter damage

Abstract

Background The reported incidence of anal sphincter injury after first (11·5–35·0 per cent) and subsequent (3·4–12·1 per cent) vaginal deliveries varies widely. In addition, the reported incidence of associated faecal incontinence ranges from zero to 68·2 per cent. The aim of this study was to perform a meta-analysis of reported incidences of postpartum anal sphincter defect diagnosed by endoanal ultrasonography (EAUS) and associated incidences of faecal incontinence. Methods A Medline search yielded five studies with more than 100 subjects who underwent EAUS after childbirth for evaluation of anal sphincter disruption and who were questioned about symptoms of faecal incontinence, defined as any impairment in flatus and stool control but not including urgency of defaecation. A Bayesian meta-analysis was performed to produce one inference while accounting for potential heterogeneity among the five study populations. Results Meta-analysis of 717 vaginal deliveries revealed a 26·9 per cent incidence of anal sphincter defect in primiparous women and an 8·5 per cent incidence of new sphincter defects in multiparous women. Overall, 29·7 per cent of anal sphincter defects were symptomatic. Some 3·4 per cent of women experienced postpartum faecal incontinence without an anal sphincter defect. In a Bayesian calculation, the probability of postpartum faecal incontinence due to a sphincter defect was 76·8–82·8 per cent. Conclusion The incidence of occult anal sphincter disruption following vaginal delivery is much higher than commonly estimated. However, at least two-thirds of occult defects are asymptomatic postpartum. The probability of faecal incontinence associated with an anal sphincter defect was 76·8–82·8 per cent.