POUCH VAGINAL FISTULA

Between January 1976 and June 1991, 161 women underwent restorative pr octocolectomy. Seventeen (10.6 per cent) developed pouch-vaginal fistu la. There was no difference in the incidence of inflammatory (15 of 14 1; 10.6 per cent) and non-inflammatory (two of 20; 10 per cent) diseas e, the severity of colitis or the number of stages in the operation. I n 15 patients the fistula involved the ileoanal anastomosis; in two it had possibly occurred at the dentate line. There was no significant d ifference in the proportion of hand-sewn (ten of 120; 8.3 per cent) an d stapled (seven of 41; 17 per cent) ileoanal anastomosis. Of the 17 w omen, six had pelvic sepsis in the immediate postoperative period and five had an anastomotic complication. There was no case of Crohn's dis ease. Three patients developed a pouch-vaginal.fistula before closure of the ileostomy. The remaining 14 fistulas occurred a median of 7 (ra nge 1-144) months after closure. Five further patients were referred f rom elsewhere to give a total of 22 for analysis of treatment and outc ome. The fistula developed before ileostomy closure in five patients ( group 1) and after closure in 17 (group 2). Following treatment, six f istulas had healed with five patients still undergoing treatment. In g roup 1, three of the five fistulas healed, whereas in group 2 only thr ee of the 17 healed. Eight patients required excision of the pouch. Th e prognosis appears to be worse when pouch-vaginal fistula occurs afte r ileostomy closure. The optimal management is not yet established.