Severity and impact of accidental bowel leakage two decades after no, one, or two sphincter injuries

Abstract

Information about the long-term severity and subjective impact of anal incontinence in women after 1 or 2 consecutive obstetrical anal sphincter injuries is still scarce and contradictory. This study aimed to describe the severity and impact of anal incontinence among women with 2 previous deliveries 2 decades after birth and to analyze the relative effect of 1 vs 2 obstetrical anal sphincter injuries in comparison with no obstetrical anal sphincter injuries and the possible influence of obstetrical anal sphincter injury on other pelvic floor disorders. We linked prospectively registered data in the Swedish Medical Birth Register with information from a postal and web-based questionnaire in 2015. Statistics Sweden identified women with 2 vaginal births from 1992 to 1998, and a simple random sample of 11,000 women was drawn from a source cohort of 64,687 women. To achieve equal-sized groups of women with 1 or 2 obstetrical anal sphincter injuries, the latter group was oversampled from 1987 to 2000. The final study cohorts consisted of 6760 women with no obstetrical anal sphincter injury, 357 women with 1 injury, and 324 women with 2 obstetrical anal sphincter injuries. Third- and fourth-degree perineal tears were grouped together for analysis. Anal incontinence was defined as either fecal or isolated gas incontinence, and fecal incontinence was defined as involuntary leakage of solid or liquid stool with or without concomitant gas. Frequencies of leakage of stool and gas were dichotomized into low frequency (less than once a month) and high frequency (several times a month or more often). Pairwise comparisons were analyzed using Fisher exact tests, Mantel-Haenszel statistics, and the Mann-Whitney U test. Trends were analyzed using Mantel-Haenszel statistics and the Spearman rank correlation test. Logistic regression models were used to obtain the age- and body-mass-index -adjusted odds ratios for outcomes. Statistical significance was set at P<.05. The response rate was 65.5% in the randomly selected cohort and 70.1% among women with 2 obstetrical anal sphincter injuries. Bothersome fecal incontinence occurred in 3.3% (212/6458) of women without obstetrical anal sphincter injury, in 10.4% (36/345) (adjusted odds ratio, 3.25; 95% confidence interval, 2.23-4.73) of those with 1 injury, and in 16.5% (52/315) (adjusted odds ratio, 5.16; 95% confidence interval, 3.69-7.22) of those with 2 obstetrical anal sphincter injuries (trend P<.0001). Fecal incontinence was perceived as bothersome in 28.2% (212/753) of women without an obstetrical anal sphincter injury compared with in 43.9% (36/82) and 46.0% (52/113) of those with 1 or 2 obstetrical anal sphincter injuries (trend P<.0001). Leakage of liquid stool occurred in 10.8% (724/6717) of those without injury, in 21.7% (77/355) of women with 1 injury, and 34.9% (113/324) of women with 2 obstetrical anal sphincter injuries (trend P<.0001). Incontinence affecting daily life was reported by 8.6% (577/6672) of women without injury and by 19.7% (69/351) and 29.6% (96/324) of women with 1 and 2 sphincter injuries, respectively (trend P<.0001). The mean Jorge-Wexner score was 2.44, 3.26, and 3.88 for women with no, 1, or 2 sphincter injuries (trend P<.0001). Among women with a Jorge-Wexner score of 6, >50% had bothersome anal incontinence. The adjusted odds ratio for the overall effect of 1 vs 2 obstetrical anal sphincter injuries on measures of anal incontinence was 2.19 (95% confidence interval, 1.68-2.85) and 3.91 (95% confidence interval, 3.06-5.00), respectively, when compared with no obstetrical anal sphincter injury (both P<.0001). Having 1 or 2 obstetrical anal sphincter injuries had no significant effect on other pelvic floor disorders or on lower urinary tract symptoms (P=.73 and P=.69). A consistent additive effect of 1 or 2 sphincter injuries on the severity and impact of anal incontinence was observed in women 2 decades after 2 vaginal births. This information is important for healthcare economics, clinical practice, and policy.