Do the Surgical Outcomes of Rectovaginal Fistula Repairs Differ for Obstetric and Nonobstetric Fistulas? A Retrospective Cohort Study.

Abstract

Rectovaginal fistulas can occur from both obstetric and nonobstetric (eg, inflammatory bowel disease, iatrogenic, or traumatic) etiologies. Current data on factors contributing to rectovaginal repair success or failure are limited, making adequate patient counseling difficult. Our objective was to compare outcomes of transperineal rectovaginal fistula repair performed in a single referral center on women with obstetric and nonobstetric causes. We performed a retrospective cohort study of women who had a transperineal rectovaginal fistula repair performed by a urogynecologist at the University of Michigan from 2005 to 2015. Data were obtained by chart review and included demographics, medical comorbidities, fistula etiology, history of a prior fistula repair, failure of current repair, time to failure, and operative details. Repair failure was defined as fistula symptoms with presence of recurrent fistula on exam or imaging in the postoperative follow-up period. Comparisons between the obstetric and nonobstetric cohorts were performed using χ, Fisher exact, and Wilcoxon rank sum tests. Relative risks were calculated to identify predictors of failure. Eighty-eight women were included-53 obstetric and 35 nonobstetric fistulas. The overall fistula repair failure rate was 22.7% (n = 20). Median follow-up was 157.0 days (range, 47.5-402.0). Of all the factors, only nonobstetric etiology was significantly associated with an increased risk of repair failure (relative risk, 3.53 [range, 1.50-8.32]; P = 0.004. Nonobstetric rectovaginal fistulas have a nearly 4-fold increased risk of repair failure compared with obstetric fistulas. Our results will help surgeons adequately counsel patients on potential outcomes of surgical repair of obstetric versus nonobstetric rectovaginal fistulas.