Predictors of response to biofeedback treatment in anal incontinence.

PURPOSE: Biofeedback is considered an effective treatment for anal incontinence, but a substantial proportion of patients fails to improve. The purpose of this study was to identify the key predictors of outcome.

METHODS

We retrospectively analyzed the clinical and physiologic data of 145 patients consecutively treated in our unit for anal incontinence by biofeedback. Clinical evaluation was performed by means of a structured questionnaire that included previous history, symptoms of incontinence, and bowel habit. Anorectal evaluation measured anal pressure profiles, neural reflexes, defecatory dynamics, rectal compliance, and rectal sensitivity. Biofeedback treatment was performed by a manometric technique with reinforcement sessions scheduled every three months and daily exercising at home. Six months after the onset of biofeedback treatment the clinical response was evaluated as good (improvement of incontinence) or poor (no improvement or worsening).

RESULTS

Of 126 patients (104 female; age range, 17-82 years) with at least six-month follow-up, 84 percent had a good response to treatment. By univariate analysis, several factors, such as age, history of constipation, abnormal defecatory maneuver, and rectal compliance, were significantly related to treatment response, but by multivariate logistic regression only age and defecatory maneuver were independent predictors of the response. The association of both factors provided the best sensitivity and specificity; 48 percent of patients younger than age 55 years and with abnormal defecatory maneuver had negative response to treatment, whereas 96 percent of patients age 55 years or older with normal defecatory maneuver had a positive response.

CONCLUSION

In patients with anal incontinence scheduled for biofeedback treatment, potential alterations of defecation should be first searched for and corrected, particularly in younger patients.