Biofeedback therapy for children with BBD: what is the maximum number of sessions we should offer?
Arighno Das, MD, Fardod O'Kelly, MD, Josiah Wolf, BS, Gabriella Hermes, BS, Muen Wang, BS, Clara Nemr, BS, Jennika L. Finup, NP, Walid A. Farhat, MD, Ruthie R. Su, MD.
University of Wisconsin, Madison, WI, USA.
BACKGROUND: Biofeedback therapy is an effective treatment for pediatric bladder bowel dysfunction (BBD), however, there is no consensus on the number of sessions needed to achieve clinical improvement. We sought to determine the relationship between symptoms resolution and number of sessions by evaluating our center’s experience with children referred for biofeedback therapy.
METHODS: We reviewed all pediatric patients undergoing biofeedback from 2013 to 2021 and included all children with >6 sessions. At each session, patients and their parents documented their urinary symptoms (urgency/frequency, straining, or pain), incontinence, medications, and stool pattern. This longitudinal data was abstracted, and the outcome was defined as resolution of urinary symptoms or incontinence. Descriptive statistics were done in R. The log odds of symptoms was modeled with number of sessions as a predictor using generalized estimating equations and robust standard errors in SAS v9.4.
RESULTS: During this period, 837 patients received biofeedback sessions for BBD, while 490 patients underwent ≥ 6 sessions. There were 166/490 males (34%) and 324 females (66%), with average age at the time of the first session of 8.9 years (range 4-22). Median number of sessions per patient was 10 (IQR 8-12, range 6-33). At first session, 91% patients reported urinary symptoms, 84% urinary incontinence, 37% constipation, and 28% encopresis. The relationship between session number and the log odds of having symptoms was curvilinear therefore a cubic regression function was fitted. A knot at session 5 improved model fit and identified a significant cutoff suggesting a decreasing trend in symptoms from session 1 to 5 there was (X2=86,df=3, p<0.0001). After session 5, however, there was a rise in the log odds of symptoms with sessions (b=0.125, p=0.03) (Figure). 231 patients had >10 sessions of biofeedback and of these, 34 patients had a lapse of more than 6 months after session 10 which suggests that 197/231 (85%) had sessions beyond #10 with intent rather than due to relapse.
CONCLUSIONS: The benefits of biofeedback should be evident by session 5 after which there is a plateau in the probability of symptom resolution. Beyond session 10, sessions were associated with an increasing probability of observing symptomatic patients. It is more likely that later sessions became a marker for biofeedback resistance rather than therapy causing worsening of BBD. Further studies are needed to adequately select patients and maximize the chances of success with biofeedback, hence improving both patient and provider investment in this resource. 
Back to 2021 Abstracts