Are there predictors of outcome following rectovaginal fistula repair?

BACKGROUND: Rectovaginal fistula is a distressing condition for patients and for physicians who are continuously challenged in providing durable treatment options. The aim of this study is to assess the results of rectovaginal fistula repair and identify predictive factors for poor outcome.
METHODS: Retrospective analysis of patients who underwent rectovaginal fistula repair from 1988 to 2008 was performed. chi tests and logistical regression analysis were used to study treatment outcomes according to the following fistula characteristics: etiology, size, location, and number of prior attempts at fistula repair. In addition, patient factors such as age, body mass index, smoking history, comorbid condition of diabetes, use of steroid and immunosuppressive medications, number of prior vaginal deliveries, and presence of a diverting stoma were analyzed.
RESULTS: A total of 184 procedures were performed in 125 patients. Inflammatory bowel disease was the most common indication for surgery (45.6%), followed by obstetric injury (24%) and surgical trauma (16%). The mean duration of fistula presence was 31.2 months. The procedures performed included endorectal advancement flap (35.3%), gracilis muscle interposition (13.6%), seton placement (13.6%), and transperineal (8.7%) and transvaginal repair (8.1%). The overall success rate per procedure was 60%, with no difference in recurrence rates based on the type of repair. Patients with Crohn's disease had more recurrent fistulas (44.2% success per procedure; P < .01), although 78% eventually healed after an average of 1.8 procedures. Patients with obstetric injuries had an 89% success rate after an average of 1.3 procedures per patient, which is similar to the success rate for traumatic fistulas. Pouch vaginal fistulas had a 91% success rate after an average of 1.6 procedures per patient. The overall success rate per patient was 88% after multiple procedures with a mean follow-up of 16.3 months. Age, body mass index, diabetes, use of steroids and immunosuppressive agents, size and location of the fistula, number of vaginal deliveries, time interval between a recurrent episode and subsequent repair, and the presence of fecal diversion did not affect outcomes. The presence of Crohn's disease and a smoking history are strongly associated with rectovaginal fistula recurrence (P = .02).
CONCLUSIONS: Despite a relatively low initial success rate (60%), most rectovaginal fistulas can be successfully repaired with subsequent operations. Crohn's disease and smoking are associated with adverse outcomes.