Objective versus subjective outcome measures of biofeedback: what really matters?

OBJECTIVE: Clinical epidemiologic studies suggest that once established, voiding dysfunction can become a lifelong condition if not treated correctly early on in life. Biofeedback is one component of a voiding retraining program to help children with voiding dysfunction. Our goal was to compare objective non-invasive urodynamic data obtained during office biofeedback sessions with patient reported voiding symptom scores.

METHODS: Charts of 55 children referred in 2010 for pelvic floor muscle biofeedback therapy for urinary incontinence were retrospectively reviewed. Patients with any anatomic diagnoses were excluded. Forty-seven (86%) females and eight males (14%) with a mean age of 8.2 years made up the cohort. Uroflow curves, voided volumes, and post-void residuals were recorded at each visit and served as objective data. Volumes were normalized as a percentage of expected bladder capacity according to age. The patient reported symptom score and patient reported outcome (improved, no change or worse) served as subjective measures of intervention.

RESULTS: The primary referral diagnoses were day and night wetting in 37 (67%) and daytime incontinence in 18 (33%) children. A history of urinary tract infection (UTI) was noted in 32 (64%) patients, and 25% were maintained on antibiotic prophylaxis during the study period. Twenty-nine percent were maintained on anticholinergic medication. Patients attended an average of 2.5 biofeedback sessions. Voided volumes and post void residual volumes were unchanged, 50% of the abnormal uroflow curves normalized over the course of treatment (p < 0.05). Patient reported symptom score decreased from 12.8 ± 5.6 to 8.0 ± 6.5 (p < 0.002) over an average follow-up time of 276 days reflecting fewer daytime voiding symptoms. There was no significant change in the patient symptom score component for the night-time wetting. Patient-reported outcomes at the final session of biofeedback were rated an improved in 26 (47%), no change in 15 (27%), worse in three (5%) patients, and not rated in 11 patients (21%).

CONCLUSIONS: Pelvic floor muscle biofeedback is associated with patient-reported improvement in symptoms, reduction in voiding symptom score, and normalization of uroflow curves, but these improvements are not correlated with objective parameters of voided volumes and post-void residual urine obtained during office visits for biofeedback. It is important to identify the most relevant outcome measures for BFB, as insurance coverage for medical interventions that cannot offer outcomes analysis that demonstrates a benefit for the patient will eventually be eliminated.