2011 Theses Doctoral

Toward a better understanding of urinary fistula repair prognosis: Results from a multi-country prospective cohort study

Frajzyngier, Veronica Maya

This dissertation addresses several critical gaps in the evidence-base with regard to urinary fistula care and treatment in developing countries. First, I systematically reviewed and synthesized the small but growing body of literature examining the patient, fistula and facility-level factors that influence repair outcomes in developing countries. There was insufficient evidence to support a role of patient characteristics in influencing repair outcomes. In contrast, the weight of evidence suggested that some fistula characteristics, particularly scarring and urethral involvement, may influence the risk of failure to close the fistula, residual incontinence following closure and any incontinence. Results from randomized controlled trials examining prophylactic antibiotic use and repair outcomes were inconclusive, and observational studies examining the influence of peri-operative procedures were limited by small sample sizes and lack of statistical adjustment for potential confounding factors. Secondly, using data from a multi-country facility-based prospective cohort study, I examined the prognostic value of five existing classification systems - those developed by Lawson, Tafesse, Goh, the World Health Organization (WHO) and Waaldijk - for predicting fistula closure, and evaluated the prognostic value of a score derived empirically from the data from this study. The scoring systems representing the Tafesse, Goh and WHO and empirically-derived classification systems were similar, and had the highest predictive values. However, none of the scores evaluated achieved good discriminatory ability (AUC > 0.70), suggesting that other factors unrelated to fistula characteristics may be equally or more important in predicting repair outcomes. Finally, I examined several issues surrounding two peri-operative procedures related to fistula surgery: abdominal versus vaginal route of repair, and catheterization duration greater than 14 days (compared to 14 days or less). Specifically, I explored the factors influencing the choice of these procedures, the influence of each of these procedures on repair outcomes independent of indication for repair or repair prognosis, and whether indication for the procedure or fistula prognosis moderates the influence of each of these procedures on repair outcomes. Abdominal route of repair was independently associated with site, parity > 3, and having a fistula that met indications for an abdominal route of repair (limited vaginal access due to extensive scarring or tissue loss, genital infibulation, ureteric involvement, or trigonal, supra-trigonal, vesico-uterine or intracervical location, or other abdominal pathology). Surgeon experience conducting complex repairs and mid-vaginal location were inversely associated with abdominal route of repair. Increased prognostic score was independently associated with catheterization > 14 days, as were site and surgeon experience doing complex repairs. Vaginal route of repair was independently associated with increased risk of failure to close the fistula, relative to abdominal route of repair; however, stratified analyses suggested that the risk of failed repair among those repaired vaginally may be particularly elevated among women who met common indications for abdominal route of repair. Duration of catheterization > 14 days was associated with failure to close the fistula, after adjusting for repair prognosis and surgeon experience; however, residual confounding by indication and reverse causation cannot be excluded as explanations for this finding. Additional research is needed to confirm our findings regarding the discriminatory value of the classification systems evaluated. Further, since the value of a classification system lies not only in its discriminatory ability but also its reliability and ease of use, tests of inter- and intra-rater reliability of these systems are priority area for future research. Given the cost and health implications associated with abdominal route of repair and longer duration catheterization, additional studies examining the influence of these procedures on repair outcomes are warranted. Such studies must ensure adequate control of confounding by indication and prognosis of repair.