Anorectal biofeedback (BF) is commonly used to treat patients with fecal incontinence (FI), but demand usually exceeds availability. It is therefore important to identify patients most likely to respond to BF treatment. We aimed to identify pre-treatment clinical or physiologic factors that might be used to predict completion and success of BF in women with FI.
We analyzed data from 400 women with FI (mean age, 61±14 y) undergoing instrumental BF in a tertiary care setting from 2003 through 2016. All patients completed questionnaires before BF, including Rome and the hospital anxiety and depression scale questionnaires. Histories of medication use, surgery, medical conditions, and bowel pattern were recorded, urge was assessed, and patients kept stool diaries. Before and after treatment (6 weekly sessions with a gastroenterologist-supervised nurse specialist, 4 involving instrumented anorectal biofeedback), patients were examined by a physician and fecal incontinence severity index and visual analogue scale scores were recorded. The main outcome measure was response to therapy, defined as improvement of 50% or more in weekly FI episodes at the end of BF compared with before BF.
The BF treatment was completed by 363 women (91%); of these, 62 had low baseline symptom frequency (no FI episodes in the 2 weeks before BF). Younger age was associated with failure to complete treatment. Of the 301 patients remaining, 202 patients (67%) had a response to therapy; among these women, urge FI was associated with response at end of BF, but not at follow up (6 months after therapy). Baseline severity of symptom scores and quality of life measures were associated with greater improvement in the same variable at the end of BF and after 6 months. Patients with low baseline symptom frequency improved in all secondary outcome measures, similar to patients with higher baseline symptom frequency.
In an analysis of 363 women with FI, approximately two-thirds had a response to BF treatment. Urge FI was the only baseline variable associated with response. Baseline severity of symptoms and quality of life measures were associated with greater improvement in the same variable, but not overall response. It is therefore a challenge to select treatment for patients with FI.

Copyright © 2020 AGA Institute. Published by Elsevier Inc. All rights reserved.