Randomized, controlled trial shows biofeedback to be superior to alternative treatments for patients with pelvic floor dyssynergia-type constipation.

Heymen S, Scarlett Y, Jones K, Ringel Y, Drossman D, Whitehead W. Randomized, controlled trial shows biofeedback to be superior to alternative treatments for patients with pelvic floor dyssynergia-type constipation. Dis Colon Rectum. 2007;50:428-441.

Appraised by: Beth Northrop PT, DPT

January 11, 2012

Clinical Bottom Line: Incorporating biofeedback training with behavioral training is effective for improving constipation symptoms and quality of life in patients with pelvic floor dyssynergia.

Study design: single-blinded randomized controlled trial with intention to treat

The Study Patients:  

·         All patients fulfilled the Rome II diagnostic criteria for pelvic floor dyssynergia, with 2 or more symptoms of functional constipation for at least 12 weeks in the past year, evidence for inappropriate contraction or failure to relax the pelvic floor muscles during attempts to defecate, evidence for in adequate propulsive forces during attempts to defecate, and evidence of incomplete evacuation.

·         There were 3 phases in the study: run in phase, training phase, and follow up. 117 patients participated in the run-in phase which consisted of 4 weeks of education and medical management. Education included: physiology of defecation, pelvic floor anatomy, dietary recommendations, general physical exercise recommendation, instruction to sit on toilet for 10 minutes after dinner and not to strain excessively, instruction to keep diary recording straining, feeling of incomplete evacuation, unassisted and assisted bowel movements. The average increase in unassisted bowel movements during run-in for all subjects combined was 3.8 per week. Average decrease in assisted bowel movements (defined as any bowel movement that did occur within 12 hours of taking a laxative or suppository) during run-in for all patients combined was 4.35. Patients were excluded from the training phase if they reported adequate relief of constipation at the end of the run- in phase (18 with adequate relief, 15 withdrew).

·         Training phase included 3 groups; placebo, diazepam, EMG biofeedback. The patients in the 3 groups had an average age of 50 years, 85 percent female. Patients in all groups received more intensive educational intervention, pelvic floor muscle exercises, and they were instructed to use of stool softeners and diet manipulations to modify stool consistency.

Control groups (Placebo N = 24; 20 analysed, Diazepam N = 30; 23 analysed):      

·         Patients took a placebo tablet or 5 mg of diazepam, which is a skeletal muscle relaxer, 1-2 hours before attempting to defecate each evening. Contraindications of diazepam and side effects were discussed with patients in both groups, and both the patients and the therapist were blinded to which pill the patients received. Patients had 6 biweekly visits lasting 50 minutes. Patients were shown videos of defecogram studies (normal and pelvic floor dyssenergia), verbally trained to perform pelvic floor exercises, and asked to practice exercises several times daily. Neither patients nor treating professional knew which pill was being taken.

Experimental group (N = 30; 23 analysed):

·         Patients received instrumented EMG biofeedback. An internal rectal sensor and a second channel with electrodes over rectus abdominis were used. 50-minute training sessions were performed with patient sitting in a chair to simulate defecation postures. Resting EMG from pelvic floor and rectus abdominis was recorded for 3 minutes, then patient was instructed to relax, squeeze, or strain gently for a series of 10-second trials.

·         The experimental group received the same training strategies as the control group. 

Results/Comments:   

·         Patients in the biofeedback group had significantly lower pelvic floor EMG activity when straining to defecate at 3-month follow-up. Adequate relief of symptoms was reported by 70% of patients in biofeedback group, 23% in diazepam group, 38% in placebo group. The biofeedback group also had more unassisted bowel movements.

·         The follow-up rate of the biofeedback and diazepam groups was 76.67%, which makes it difficult to confirm their statistics. Patients who did not report adequate relief at three months` follow-up or who failed to return for follow-up evaluation at three months were labeled treatment failures. Prior to beginning the study, they estimated that they needed 50 patients per group for 80% power to detect between group differences. However, after approximately 30 patients per group had completed the three-month follow-up, an interim intent-to-treat analysis was performed. At that point, the analysis showed the biofeedback group to be significantly better than the two pill groups combined, and they stopped enrolling patients in the study.

·         Patients with slow transit constipation were included in the study, though it is reported that these patients are less responsive to biofeedback.

·         The authors concluded that this study provides definitive support for the efficacy of biofeedback treatment for patients with pelvic floor dyssynergia constipation, and that instrumented biofeedback is an essential element of successful training, though the study did not have a comparison group that received non-instrumented biofeedback.

·         It is not clear whether the biofeedback group received 50 minutes of treatment consisting of both biofeedback and training strategies, or if they had separate sessions of biofeedback and training strategies.