The following chart summarizes the Symptom Reduction Rates reported in the studies listed. Additional details about the studies charted are provided after the chart. Following the tradition of Burgio et al, symptom reduction rates are calculated on the basis of "before" vs. "after" leakage reports, or, where those reports are not provided, pad weights. Only studies which include enough data to calculate the reduction rate have been included here; i.e., studies which report only "patient satisfaction" without supporting data were not included. Where possible, only "Stress Incontinent" patients are included; see individual notes.
What these studies have in common is the daily use of a home training instrument, based on either an EMG (Taylor and Perry) or Pressure sensor (all others), throughout the entire course of treatment. Patients typically received 100 to 200 biofeedback sessions. Weekly clinical office visits to assess patient progress were also required. Perry is the only one of these five that also used a "clinical" office instrument in addition to the home trainer with all patients.
100% - Taylor, K & Henderson, J. 1986. Effects of biofeedback on urinary stress incontinence in older women. J. Gerontological Nursing, 12(9): 25-30.
N=4. This was a pilot study in preparation for a full-scale project that was never begun due to the untimely death of Dr. Taylor. All patients used the "Personal Perineometer" home trainer.
99% - Perry, J, Hullett, L & Bollinger, J. 1988. EMG biofeedback treatment of incontinence (and other disorders of the pelvic musculature). Paper presented at the Biofeedback Society of America, Colorado Springs, March 26, 1988. (Abstract, Biofeedback and Self Regulation, 1989, p. Revised version available on this website as <effective.htm>.
The aim of this study determine how effective EMG biofeedback could be if ALL of the manufacturer's recommended parameters were followed precisely. All patients used "Personal Perineometer" for daily at-home practice. The protocol is further described in The Perry Protocol.
93% - Kegel, A. 1950. Active exercise of the pubococcygeus muscle. Chapter 16 in Progress in Gynecology (NY: Grune & Stratten, p. 789).
Based on his first 300 cases, Kegel claimed 93% obtained "complete relief of urinary stress incontinence". He did not actually provide symptom reduction data, which would be much higher than 93%. Before he retired about 1975, he claimed to have treated over 3,000 cases.
87% - Sussett, JG, Galea, G, & Read, L. 1990. Biofeedback therapy for female incontinence due to low urethral resistance. The Journal of Urology, 143, June. p. 1205-1208.
An investigation of the Contimed pneumatic home trainer (Hollister/InCare). Sussett did not actually calculate the symptom reduction rate; one outlier, which Sussett classified as a "failure", dragged the rate down to 77%, but apparently the AHCPR dropped that data to better represent the work, so we'll use their 87% figure (1996, p. 40)
83% - Shepherd, AM, et al, 1983. Treatment of genuine stress incontinence with a new perineometer. Physiotherapy 69(4): 113.
N=11 in experimental group. They used a new pressure-based perineometer for daily at-home practice with weekly office checkups.
It is not surprising that these studies produced such uniform results; all are based on the participation of Dr. Kathy Burgio, a brilliant young psychologist who was brought into Bernie Engel's lab in 1981 to translate their fecal incontinence work into the urinary incontinence field. Because ano-rectal manometry is so painful for patients, only a limited number of office visits (1-5, typically 3 or 4) are utilized. Although manometric home trainers are available (BioSearch, for example), the NIA typically did not use them. The NIA-Burgio method is well illustrated in a 1984 videotape available from the NIA.
82% - Burgio, KL, Whitehead, WE, and Engel, BT. 1985. Urinary incontinence in the elderly; Bladder/sphincter biofeedback and toileting skills training. Ann Internal Med, 103: 507-515.
19 patients with stress incontinence achieved 82% symptom reduction after 3.3 sessions of ano-rectal manometry. (And 8 Urge patients achieved 94% reduction.)
82% - McDowell, BJ, Burgio, LK, Dombrowski, M, Locher, JL, Rodgiruquez. 1992. An interdisciplinary appoach to the assessment and behavioral treatment of urinary incontinence in geriatric outpatients. J Am Geriatr Soc 40:370-374.
N=29. 59% of the 29 patients received only a single biofeedback traning session! This study used both ano-rectal manometry and Perry brand EMG sensors, but comparative data is not presented. None of the subjects used home trainers. Curiously, the 16.9 to 2.5 mean leaks that they report works out to an 85%. However, a single patient with 140 leaks/week (11 times the sample average) accounted three percentage points; with her data removed (quite appropriate), the remaining average is 82%, which they reported.
79% - Burton, JR, Pearce, KL, Burgio, KL, Engel, BT, Whitehead, WE. 1988. Behavioral training for urinary incontinence in elderly ambulatory patients. J Am Geriatr Soc 36:693-98.
N=4 stress incontinence patients who received biofeedback. Based on interpolation from the graph, it appears they went from about 102 to 12 leaks per week, which would yield an 88% symptom reducation rate for S.I. Poorer results for the 9 urge incontinence patients lowered the biofeedback group's average by 9 percentage points.
This is one of the studies currently being used to discredit biofeedback, since the "education" group reduced their accidents by 82%, vs. 79% reported for biofeedback. However, this is an artifact of the mis-matched assignment; the education group had 33% more accidents to start with, and therefore, MUCH more room to improve. Both groups ended up with about 3 accidents per week.
The fact that the manometric method cannot evaluate pelvic resting tension levels, which often play a part in sensory urge incontinence, may account for the poorer performance with urge patients.
78% - Burgio, KL, Stutzman, RE, and Engel, BT. 1989 Behavioral training for post-prostatectomy urinary incontience. J Urol 141: 303-306
Eight patients with stress incontinence decreased their accidents by 78 percent. (And 8 patients with Urge incontinence reduced their accidents by 81%.)
76% - Burgio, KL, Robinson, JC, and Engel, BT. 1986. The role of biofeedback in Kegel's exercise training for stress incontinence. Am J Obstet Gynecol 154:5-64.
Thirteen patients decreased their accidents by 76% after four biweekly training sessions of 25 exercises each. Homework of 51 contractions per day was assigned, without instrumentation. (A control group performed the same biweekly office exercises with a therapist's fingers inserted in the vagina, and achieved 52% reduction under verbal feedback.)
87% - Williams, MI & Kaczmarek, P. 1995. EMG biofeedback is an effective treatment for pelvic floor dysfunctions. Paper presented at the Association for Applied Psychophysiology and Biofeedback, Albuquerque, March, 1996.
N=173 patients with urinary incontinence. Home trainers were not used except in a few cases; most patients were instructed to practice exercises at home using their personal vaginal or rectal sensor and an audio tape.
76% - Baigis-Smith, J, 1989. Managing urinary incontinence in community-residing elderly persons. The Gerontologist 29(2): 229-233.
It is unfortunate that this study is so often cited in research on the effectiveness of biofeedback, since effectiveness was not a consideration in the research design. The project, run by the University of Pennsylvania for the State of New Jersey, was aimed at discovering "how much senior citizens could be helped with a limited expenditure of funds"! To that end, the authors developed what they called "Focused Treatment" -- the patient is treated only so long as she is making good weekly progress; once the weekly improvement slowed, the patient was dropped, regardless of the level of symptom reduction.
Therefore, one would expect the results to be lower than they were. One reason for such a high symptom reduction rate is that they actually used eight home trainers (Personal Perineometer) with all new patients -- but only for the first one, two or three weeks of their therapy -- until the trainer was needed by another incoming patient. Sporadic record keeping led the authors to overlook mention of the home trainers in their published report.
67% - Woolner, BF & Ouslander, JG. 1992. Experience with biofeedback for incontinence and related urinary symptoms in the old old. Proceedings fo the International Continence Society, Halifax, September 1992.
The first report on treating residential geriatric patients (mean age, 87 years). In the first 25 patients, incontinence episodes were reduced 67% after 5.9 weekly therapy visits.
61% - Woolner, BF. 1994. [reference missing]
Based on 113 patients, incontinence episodes dropped 61%.
61% - Burns, PA, Pranifoff, K, Nochasjski, TH, Hadley, EC, Levy, KJ, & Ory, MG. 1993. A comparison of effectiveness of biofeedback and pelvic muscle exercise treatment of stress incontinence in older community-dwelling woman. Journal of Gerontology, 48:4, M167-M174. See also, Burns et al, Treatment of stress incontinence in community-residing elderly persons. J Am Geriatr Soc, 1990, 38(3): 341-344, which covers the same research program.
This study produced a 61% decrease in accidents for the biofeedback group (n=40), compared with a 54% decrease for a so-called pelvic muscle exercise group (n=43), and a 6% decrease for a control group (n=39). Biofeedback therapy consisted of eight twenty minute weekly training sessions.
Because this study is being used by Aetna and Medicare to DENY benefits for biofeedback, it is or will soon be the subject of a separate "critical review" on this website. Suffice it to say that Burns obtained the WORST results of any project ever using intravaginal EMG sensors. Non-standard data collection, contamination of the "exercise only" control group, and use of a fixed number of treatment sessions (regardless of patient progress), are among the most obvious problems with this study. Burgio (1986), for example, in a similar study, obtained statistically significant differences of 76% vs. 52%. Burns' biofeedback results were 15% worse than Burgio, 22% worse than Williams, and 38% worse than Perry and Taylor (above).
61% - O'Donnell, PD, Doyle, R, 1991. Biofeedback Therapy technique for treatment of urinary incontinence. Urology, 37(5): 432-436.
This is the only study to utilize the Dantec "sponge" electrode, originally developed for static use in urodynamics, as a biofeedback sensor. The "sugar cube" electrode was taped over the anus of male patients. The sensor was so unstable that the authors were forced to monitor electrode impedance constantly with an oscilliscope to detect poor contact.
20 subjects were patients were elderly inpatients at a VA facility. They received twice-weekly therapy for five weeks. Urine loss was "improved" from 525 cc/day to 209 cc/day on average.
Electrical stimulation typically achieves far poorer results over a much longer period of time than biofeedback therapies. Authors typically report non-specific results, such as "patient satisfaction" or "improvement", rather objective improvement rates. The studies below provided sufficient data (usually pad weights) to be included and compared.
44% - Fall, M, Erlandson, B-E, Ahlstrom, K, Frankenberg, S, Carlsson, C-A, Mattiasson, A, and Ek, A. 1986. Contelle: Pelvic Floor Stimulator for Female Stress-Urge Incontinence: A multi-center study. Urology, 27(3): 282-287.
The combined pad weights for the 8 "cured", 25 "improved", and 7 "no effect" patients were weighted and averaged to produce the 44% symptom reduction value. The authors admit that the best results were obtained with detrusor instability patients, with or without stress incontinence.
38% - Richardson, DA, Miller, KL, Siegel, SW, Karram, MM, Blackwood, NB, and Staskin, DR. 1996. Pelvic floor electrical stimulation: A comparison of daily and every-other-day therapy for genuine stress incontinence. Urology 48:110-118.
28 subjects had 20 weeks of estim treatment on twice-daily or twice every-other-day (QOD) basis (280 or 140 estim sessions). 38% reduction refers to average pad weight. The QOD group did better on most measures: 57% vs. 35% total leak reduction, 46% vs. 34% stress leak reduction, 49% vs. 25% reduction in pad weight, and 69% vs. 46% reduction in pad count. However, they started with 25% fewer leaks, and initially leaked 75% less urine. Authors conclude that "a long-term maintenance program of approximately 3 days/week of treatment [for the rest of the patient's life] is necessary to maintain treatment outcomes." (p. 118). Subjects received their device or $200 cash payment for participation.
30% - Sand, PK, Richardson, DA, Staskin, DR, Swift, SE, Appell, RA, Whitmore, KE, and Ostergard, DR. 1995. Pelvic floor electrical stimulation in the treatment of genuine stress incontinence: A multi-center, placebo-controlled study. Am J Obstet Gynecol, 173:72-79.
28 women received 12 weeks of twice daily estim (EMPI Innova brand) (168 estim sessions), and reduced their accident rate by 30% (14.2 leaks/week before, 10 after). In contrast, 16 women in a control group went from 20.1 to 27.0 leaks (34% WORSE!). The improvement was statistically signficant, but only because the control group got so much worse.
27% - Peattie, AB, Pelvnik, S, Stanton, SL. 1988. Vaginal Cones: a conservative method of treating genuine stress incontinence. British Journal of Obstet and Gynecol 95(1049-1053).
N=30. The authors conclude that "70% felt they were improved or cured". However, 37% still requried surgery after their trial with vaginal cones. The authors include the before-after urine loss data, but do not calculate the change. When that is done, we observe that 3 (10%) were cured, but 8 (27%) actually got worse after using cones. The overall symptom reduction rate for cones was only 27%, the least effective of any method.
This version dated: Wednesday, December 21, 2005
Copyright 1997 by IncontiNet
Click Here to return to our Critical Reviews menu.
Click Here to return to our Home Page