Predictors and outcome of surgical repair of obstetric fistula at a regional referral hospital, Mbarara, western Uganda

BACKGROUND Obstetric fistula although virtually eliminated in high income countries, still remains a prevalent and debilitating condition in many parts of the developing world. It occurs in areas where access to care at childbirth is limited, or of poor quality and where few hospitals offer the necessary corrective surgery. METHODS This was a prospective observational study where all women who attended Mbarara Regional Referral Hospital in western Uganda with obstetric fistula during the study period were assessed pre-operatively for social demographics, fistula characteristics, classification and outcomes after surgery. Assessment for fistula closure and stress incontinence after surgery was done using a dye test before discharge RESULTS Of the 77 women who were recruited in this study, 60 (77.9%) had successful closure of their fistulae. Unsuccessful fistula closure was significantly associated with large fistula size (Odds Ratio 6 95% Confidential interval 1.46-24.63), circumferential fistulae (Odds ratio 9.33 95% Confidential interval 2.23-39.12) and moderate to severe vaginal scarring (Odds ratio 12.24 95% Confidential interval 1.52-98.30). Vaginal scarring was the only factor independently associated with unsuccessful fistula repair (Odds ratio 10 95% confidential interval 1.12-100.57). Residual stress incontinence after successful fistula closure was associated with type IIb fistulae (Odds ratio 5.56 95% Confidential interval 1.34-23.02), circumferential fistulae (Odds ratio 10.5 95% Confidential interval 1.39-79.13) and previous unsuccessful fistula repair (Odds ratio 4.8 95% Confidential interval 1.27-18.11). Independent predictors for residual stress incontinence after successful fistula closure were urethral involvement (Odds Ratio 4.024 95% Confidential interval 2.77-5.83) and previous unsuccessful fistula repair (Odds ratio 38.69 95% Confidential interval 2.13-703.88). CONCLUSIONS This study demonstrated that large fistula size, circumferential fistulae and marked vaginal scarring are predictors for unsuccessful fistula repair while predictors for residual stress incontinence after successful fistula closure were urethral involvement, circumferential fistulae and previous unsuccessful fistula repair.