Biofeedback training is useful in fecal incontinence but disappointing in constipation

BACKGROUND: Successful biofeedback therapy has been reported in the treatment of fecal incontinence and constipation. It is uncertain which groups of incontinent patients benefit from biofeedback, and our impression has been that biofeedback is more successful for incontinence than for constipation. PURPOSE: This study was designed to review the results of biofeedback therapy at the Lahey Clinic. METHODS: Biofeedback was performed using an eightchannel, water-perfused manometry system. Patients saw anal canal pressures as a color bar graph on a computer screen. Assessment after biofeedback was by manometry and by telephone interview with an independent researcher. RESULTS: Fifteen patients (13 women and 2 men) with incontinence underwent a mean of three (range, 1-7) biofeedback sessions. The cause was obstetric (four patients), postsurgical (five patients), and idiopathic (six patients). Complete resolution of symptoms was reported in four patients, considerable improvement in four patients, and some improvement in three patients. Manometry showed a mean increase of 15.3 (range, −3-30) mmHg in resting pressure and 35.7 (range, 13-57) mmHg in squeezing pressure after biofeedback. A successful outcome could not be predicted on the basis of cause, severity of incontinence, or initial manometry. Twelve patients (10 women and 2 men) with constipation underwent a mean of three (range, 1-14) biofeedback sessions. Each had manometric evidence of paradoxic nonrelaxing external sphincter or puborectalis muscle confirmed by defography or electromyography. All patients could be taught to relax their sphincter in response to bearing down. Despite this, only one patient reported resolution of symptoms, three patients had reduced straining, and three patients had some gain in insight. CONCLUSIONS: Biofeedback helped 73 percent of patients with fecal incontinence, and its use should be considered regardless of the cause or severity of incontinence or of results on initial manometry. In contrast, biofeedback directed at correcting paradoxic external sphincter contraction has been disappointing.