T1348 Predictors of Outcome of Biofeedback Therapy in Patients With Constipation

Abstract

Background: Ano-rectal biofeedback therapy (BFT) is a safe and effective treatment in patients with constipation. Given the high prevalence of constipation, there is a need to prioritise patients. Therefore, the aim of this study was to further explore factors which predict success or failure of BFT. Several previous studies have examined this in part, with the only consistent finding being that the patient's willingness to participate predicts success of therapy. Methods: 102 consecutive patients (mean age 48±2 yrs, 88 females) with constipation referred for anorectal BFT were evaluated. All patients completed the Rome questionnaire, a validated constipation questionnaire, and 10 cm visual analogue scales for willingness to participate in BFT, satisfaction with bowel habit and impact on quality of life. All patients underwent comprehensive anorectal manometry and balloon expulsion test. Patients were enrolled in a 6 weekly-visit BFT program. Independent predictors were identified using a bootstrapped backward elimination procedure based on linear regression with change in bowel satisfaction as the outcome. Only predictors selected in ≥50% of bootstrap samples were considered. Results: 96 patients completed the full course of BFT. Stool consistency (p=0.009), willingness to participate (p<0.001), balloon expulsion time (p=0.02) and resting anal sphincter pressure (p=0.04) were positively correlated with an improvement in bowel satisfaction scores. On the other hand, more laxative use (p=0.049) and a greater satisfaction with bowel habit at baseline (p<0.001) correlated with no improvement in bowel satisfaction scores. The outcome of BFTwas not influenced by age, duration of symptoms or compliance with therapy. Multiple linear regression indicated that stool consistency (p=0.02) and willingness to participate (p= 0.003) independently predicted an improvement in bowel satisfaction scores after BFT, and total constipation score (p=0.006) and baseline satisfaction with bowel habit (p<0.001) independently predicted worsening in bowel satisfaction scores after BFT. The model including these 4 variables explained 52% of the variance in treatment outcome. Conclusion: In agreement with previous studies, success of BFT is predicted by a greater willingness to participate. New findings are that (1) harder stools predict a successful outcome, while more prolonged balloon expulsion and anal sphincter hypertonia at baseline are associated with greater bowel satisfaction after BFT; and (2) the presence of more severe constipation predicts failure of BFT. This information should be taken into account when prioritising patients for BFT.