Koh, C.E., Young, C.J., Young, J.M., & Solomon, M.J. (2008). Systematic review of randomized controlled trials of the effectiveness of biofeedback for pelvic floor dysfunction. British Journal of Surgery, 95, 1079–1087.

To determine whether biofeedback improves outcomes for patients with pelvic floor dysfunction (PFD), and to assess the relative effectiveness of different types of biofeedback therapy.

Databases searched were CINAHL, Embase, Medline, PsycINFO, Evidence-Based Medicine Reviews (EBMR), and the Cochrane Database. References of retrieved articles also were hand searched.

Search keywords were constipation, anismus, dyssynergia, obstructive defecation, rectocele, rectal intussusception, rectal prolapse, ​and biofeedback.

Studies were included in the review if they

• Were randomized controlled trials (RCTs)
• Reported on adults with PFD
• Included biofeedback as at least one of the treatments studied.

Studies were excluded from the review if they reported on pediatric cases.

The initial searching provided 5,028 references. Study selection and screening for inclusion criteria provided a final set of seven studies. Study quality was evaluated by two reviewers.

• The final sample of seven RCTs included 413 patients.
• Sample sizes ranged from 26 to 109.
• Medical diagnostic information across studies was not provided.

• Three trials compared biofeedback with other interventions such as laxatives, diazepam, and placebo. Meta-analysis showed an odds ratio (OR) of 5.861 (95% confidence interval [CI] [2.175, 15.794], p < 0.001) in favor of biofeedback.
• Four trials compared differences in biofeedback techniques. A meta-analysis of the most widely used technique, electromyogram biofeedback, showed an OR of 6.738 (95% CI [2.194, 15.58], p < 0.001) in favor of biofeedback.

• Results suggest biofeedback is associated with symptomatic relief of constipation; however, the number of studies included was very small and substantial heterogeneity existed across studies.
• Study quality overall was poor.
• Findings are not specific to the oncology population, and whether any of the studies included patients with cancer is not known.

Additional, better-designed studies are needed in this area to determine efficacy. Future studies should compare different biofeedback modalities to identify the most effective approaches.