Several methodologic difficulties arise in assessing biofeedback. For example, most interventions that include biofeedback are multimodal and include relaxation and behavioral instruction, which may have effects separate from those that may occur due to biofeedback. While studies may report a beneficial effect of multimodality treatment, without appropriate control conditions, it is impossible to isolate the specific contribution of biofeedback to the overall treatment effect. For example, relaxation, attention, or suggestion may account for successful results that have been attributed to biofeedback. These effects are nonspecific therapeutic factors, some of which can be considered placebo effects.

 

A 1995 TEC Assessment evaluated the use of biofeedback in the treatment of 9 different conditions: anxiety disorders, headaches, hypertension, movement disorders, incontinence, pain, asthma, Raynaud’s disease, and insomnia.  The Assessment concluded that while a substantial number of studies reported improvement in the biofeedback group relative to the no-treatment group, there were generally no differences when the isolated effect of biofeedback was compared with relaxation or behavioral therapy alone. The Assessment also concluded that while there was evidence that feedback on physiological processes provides patients with an enhanced ability to control these processes, there was nevertheless no consistent evidence of any relationship between a patient’s ability to exert control over the targeted physiological process and any health benefits of the intervention. This finding underscores the importance of seeking controlled studies showing whether use of biofeedback improves disease-related health outcomes, as opposed to physiological, intermediate outcomes. Studies also failed to consistently address the durability of effects beyond the initial short-term biofeedback training period. The findings that were reviewed suggest that the outcomes of biofeedback relative to no treatment are due to the other components of therapy or to the nonspecific effects of the therapeutic setting and are not a result of the biofeedback training per se.

 

The specific indication of biofeedback as a treatment of urinary incontinence was re-evaluated in 1997 and 2000 TEC Assessments.  These Assessments focused on the independent contribution of biofeedback as an adjunct to pelvic floor muscular exercises. The 1997 TEC Assessment concluded that while the controlled trials that isolated the contribution of biofeedback reported conflicting results, the weight of the evidence suggested no additional benefit for biofeedback above that obtained with pelvic floor muscle exercises alone.  All of the trials had low power to detect a small difference in outcomes; therefore, the possibility exists that larger trials with improved statistical power could demonstrate a beneficial effect of biofeedback. However, the TEC Assessment concluded that based on the available data, any such benefit, if present, is likely to be small and may not be clinically significant.

 

The conclusions of the 2000 TEC Assessment were similar to the 1997 Assessment, i.e., that the evidence is not sufficient to demonstrate an additional benefit for biofeedback above that obtained with pelvic floor muscle exercises (PME) alone:

 

Two papers published since the 2000 Assessment have supported the findings of this earlier study. A randomized study of 42 post-prostatectomy patients found no significant difference in outcome between biofeedback plus PME and verbal feedback plus PME.  Both groups improved significantly over time. A second randomized trial of 100 patients evaluated a preoperative training program of biofeedback plus PME versus PME alone.  Again, both groups improved, but the continence rates as defined by pad test and the rate of continence return were not significantly different between groups. A third study, a randomized controlled trial of 102 post-prostatectomy patients, reported that improvement in both duration and degree of incontinence was significantly better in a group receiving PME plus biofeedback versus placebo. This trial, however, did not evaluate the additive effect of biofeedback.

 

In summary, the literature update did not identify evidence sufficient to alter the 2000 TEC Assessment findings on biofeedback for urge, stress, or post-prostatectomy incontinence.

 

Some trials have reported favorable outcomes with biofeedback plus pelvic floor exercises, but these studies have not examined the incremental effect of adding biofeedback to PME. For example, Burgio and colleagues reported on a study that randomized 197 women to receive either biofeedback-assisted pelvic muscle exercises or drug treatment or placebo.  While the biofeedback group reported improved results compared to placebo or drug treatment alone, this study did not isolate the independent contribution of the biofeedback to the pelvic muscle exercises. In a more recent study, Burgio and coworkers used a modified crossover design to evaluate the effects of combining biofeedback-assisted PME and drug treatment. Subjects with urge incontinence were randomized to biofeedback, drug, or placebo groups. Subjects not totally continent or satisfied after 8 weeks of treatment were offered the opportunity to cross over into combined therapy. Although the results showed that combining drug and biofeedback with PME in a stepped program produces added benefits beyond either single therapy, the incremental benefit of adding biofeedback to PME was not addressed. Similarly, McDowell and colleagues reported that outcomes with biofeedback-assisted pelvic muscle exercises were superior to a waiting-list control group.  Again, the independent impact of biofeedback on outcomes was not studied.

 

The focus of both the 1997 and 2000 TEC Assessments contrasts with the 1996 assessment on treatment of incontinence published by the Agency for Healthcare Research and Quality (formerly the Agency for Health Care Policy and Research, AHCPR).  While the AHCPR assessment endorsed the use of behavioral therapy as a first-line treatment of incontinence, and identified biofeedback as a component of behavioral therapy, the AHCPR did not specifically evaluate the independent contribution of biofeedback to an overall behavioral approach.

 

Other evidence-based reviews on the effectiveness of biofeedback for urinary incontinence have generally found limited evidence or no evidence supporting a beneficial effect beyond that offered by PME alone. A Cochrane review conducted in 2000 concluded that formal comparisons of biofeedback-assisted PME versus PME alone consistently suggest that there is no added benefit in women with stress or mixed incontinence.  Berghmans and colleagues published a systematic review with qualitative data synthesis.  These authors concluded that there was strong evidence to support that the addition of biofeedback to pelvic floor muscle exercises (PME) does not offer additional benefits over PME alone. Weatherall performed a quantitative meta-analysis of the data included in the Berghmans report.  This analysis revealed a pooled odds ratio of 2.1 in favor of biofeedback, a result that reached marginal statistical significance.

 

Aksac and colleagues reported the results of a trial that randomized 50 patients with stress incontinence to 1 of 3 groups: self-directed PME, biofeedback-directed PME, or no treatment.  Outcomes were assessed with pad tests, perineometry, and pelvic floor muscle strength as assessed by digital palpation. The first 2 groups had a significant improvement in outcomes compared to the control (no treatment) group. The biofeedback group had increased strength in the pelvic floor muscles compared to those with self-directed PME, but the clinical significance of this difference is unclear. Two trials investigated the role of biofeedback in post-prostatectomy patients. In a randomized trial, Wille and colleagues reported that the addition of biofeedback did not improve the outcomes compared to PME alone.  Parekh and colleagues examined the role of biofeedback-enhanced PME offered both pre- and postoperatively to post-prostatectomy patients. (19) Since this trial did not include PME alone, the contribution of biofeedback cannot be isolated. Aukee and colleagues reported on a study that randomized 35 women with stress incontinence to receive pelvic floor training with or without additional biofeedback.  Patients were evaluated after a year; however, during this time 14 of the 35 underwent an incontinence operation, limiting interpretation of the study. Wang and colleagues conducted a study that randomized 103 women with overactive bladder to 1 of 3 arms: pelvic floor exercises, biofeedback-assisted pelvic floor exercises, or electrical stimulation.  There was a 14% dropout rate in this 12-week study; these results were not included in the statistical analysis. There was no significant difference in reduction in symptoms between the pelvic floor exercise group with or without biofeedback assistance.

 

A systematic review of pelvic floor muscle training to improve urinary incontinence after radical prostatectomy discussed 3 studies (281 men) that focused on the incremental value of biofeedback over written/verbal PME.  Although PME appeared to reduce the time to recover continence compared to no training, there was no evidence for an advantage of training with biofeedback over written/verbal instructions. An incremental improvement with biofeedback-PME or a combination of biofeedback-PME plus neuromuscular electrical stimulation was reported in a small study (10 subjects per group) of patients with multiple sclerosis; interpretation of the study results is limited due to the small group size and an imbalance in baseline measures following randomization.  A subsequent randomized study by the same group (74 patients with multiple sclerosis) reported that the addition of neuromuscular electrical stimulation (with biofeedback training) resulted in 85% incontinence reduction, compared to a 47% incontinence reduction in the control group trained only with biofeedback.  No evidence was identified that would alter the conclusions reached above; the policy statement is unchanged.

  

In October 2006, National Institute for Health and Clinical Excellence (NICE) issued their guideline on the management of urinary incontinence in women. NICE states that “perineometry or pelvic floor electromyography as biofeedback should not be used as a routine part of pelvic floor muscle training”, but that “electrical stimulation and/or biofeedback should be considered in women who cannot actively contract pelvic floor muscles in order to aid motivation and adherence to therapy.” This conclusion regarding use of biofeedback is based on expert opinion.

 

An Evidence Report/Technology Assessment, Prevention of Urinary and FecalIncontinence in Adults, based on research conducted by the Minnesota Evidence-based Practice Center (EPC) under contract to the Agency for Healthcare Research and Quality (AHRQ), Rockville was issued in December 2007.

 

The research objectives were to assess the prevalence of and risk factors for urinary incontinence (UI) and fecal incontinence (FI) in adults in long-term care (LTC) settings and in the community, the effectiveness of diagnostic methods to identify adults at risk and patients with incontinence, and to review the effectiveness of clinical interventions to reduce the risk of incontinence. The Report executive summary section titled “Effects of Clinical Interventions on UI”, includes the authors’ overview of the evidence on the clinical effectiveness of interventions including biofeedback for various UI conditions in adult females and males. The following are excerpts from the report that include biofeedback;

The authors in discussing evidence limitations concluded that the applicability of clinical interventions that demonstrated significant improvement in incontinence was restricted to the sampled population groups. Whether these effective interventions would result in the same clinical benefit in other populations requires future research.

 

A systematic review of randomized, controlled trials of non-surgical treatments for urinary incontinence in women was conducted for the AHRQ evidence report.  This review examined 96 RCTs and 3 systematic reviews to synthesize evidence on the management of urinary incontinence in women. The studies included women with stress and urge urinary incontinence, overactive bladder with urge incontinence, or minimal urinary incontinence. The studies did not explicitly exclude women with other types of urinary incontinence; therefore, the effects of the treatments can be applied to women with mixed urinary incontinence. Study quality was analyzed using the following criteria: participant selection; length and loss of follow-up; use of intention-to treat principle; masking of the treatment status; randomization scheme; adequacy of randomization and allocation concealment; and justification of sample sizes. The efficacy of the clinical interventions was analyzed from the trials that compared active treatment to placebo, regular care, or no active treatments. The comparative effectiveness of the interventions was analyzed from the trials with active controls, long-term follow-up, adequate sample size and intention-to treat analysis. The evidence reviewed included 22 clinical intervention comparative trials with biofeedback in conjunction with pelvic floor muscle training. The overall strength of the evidence in these 22 studies was deemed low.

 

Nineteen RCTs (n=2,441) included pelvic floor muscle training with biofeedback compared with active interventions (education, bladder training, medical device, medications). The authors concluded continence and improvement rates did not differ when two active treatments were compared. Three RCTs (n=179) included pelvic floor muscle training with biofeedback and bladder training compared with regular care. The authors concluded that pelvic floor training with biofeedback and bladder training improved but did not resolve UI. In conclusion, the studies of behavioral interventions including pelvic floor muscle training and biofeedback relied largely on convenience samples that involved recruiting participants in the clinics while few studies reported population-based recruitment. Selection criteria varied for the same interventions and some clinical trials reported statistically significant differences at baseline among the treatment groups despite randomization. Pooling analysis was questionable due to clinical and methodologic differences across the studies and previous systematic reviews. Despite extensive efforts to standardize outcomes, assessment for urinary incontinence in the included studies measured a variety of outcomes.

 

In conclusion, despite both CMS and NICE guidance supporting biofeedback-assisted PME for women unable to voluntarily contract their pelvic floor muscles, no published evidence supports these hypotheses. Future research is needed to determine the comparative effectiveness of available treatments, including biofeedback.

 

No published evidence supports the unsupervised home use of biofeedback for treatment of urinary incontinence.

 

Therefore, the policy remains unchanged; biofeedback as a treatment for urinary incontinence is considered investigational due to insufficient evidence to permit conclusions concerning the impact of this procedure on net health outcomes. Specifically the value of adding biofeedback to a program of pelvic muscle exercises has not been demonstrated.

 

In December 2007, NIH convened a Consensus Development Conference; Prevention of Fecal and Urinary Incontinence and subsequently released a statement.  Included in this statement was the following regarding pelvic floor muscle training and biofeedback: …”Pelvic floor muscle training and biofeedback are effective in preventing and reversing some pregnancy-related fecal and urinary incontinence for the first year after delivery. There is insufficient research on the sustained long-term benefits of pelvic floor muscle training or biofeedback on preventing fecal or urinary incontinence.”

 

 

A review of the literature has been conducted through August 2010.  There was no new literature identified that would prompt a change in the coverage statement.

 

In 2009, Mariotti and colleagues conducted a randomized controlled trial in Italy comparing a program of pelvic floor electrical stimulation and electromyographic biofeedback (2 sessions weekly for 6 weeks) to written/verbal instructions for pelvic muscle exercises (Mariotti, 2009).   All 60 patients (30 per group) completed the study through the 6-month follow-up. The mean time to regain continence was significantly shorter in the treatment group (8.0 weeks) than the control group (13.9 weeks, p=0.003). The continence rate was significantly higher in the treatment group beginning at the 4-week visit and continuing through the 20-week visit at which time 29 of 30 (96.7%) in the treatment group and 18 of 30 (60%) in the control group were continent. The difference in the rate of continence was not statistically significantly different at the final, 6-month visit at which time 29 patients in the treatment group continued to be continent compared to 20 of 30 (66.7%) in the control group. This is one study suggesting that biofeedback in combination with pelvic electrical stimulation may shorten the time to continence after prostatectomy; however, the effect of biofeedback without electrical stimulation compared to written/verbal instructions to perform pelvic floor muscle exercises was not evaluated.

 

In 2010, the Cochrane Collaboration published a systematic review of trials on pelvic floor muscle training for urinary incontinence in women (Dumoulin, 2010).   Several of the trials included biofeedback as part of the intervention. However, the Cochrane review did not include an analysis that specifically focused on the efficacy of pelvic floor muscle training with and without biofeedback and did not comment on the efficacy of biofeedback in their conclusions.

 

 

 

 

 

 

 

 

 

 

 

 

 

In 2014, the American College of Physicians published a clinical practice guideline on nonsurgical management of urinary incontinence in women (Qaseem, 2014). The guideline was based on literature published through December 2013. The authors concluded that low-quality evidence showed that pelvic floor muscle training (PFMT) with biofeedback using a vaginal electromyography probe increased continent compared to no active treatment and that high-quality evidence showed that this combination of treatments improved urinary incontinence symptoms compared to no active treatment. The guideline did not address the comparison of PFMT alone and PFMT plus biofeedback.

 

 

Moroni and colleagues published a systematic review of 37 randomized controlled trials (RCTs) on conservative treatment of stress urinary incontinence in women (Moroni, 2016). Five trials were identified that compared pelvic floor muscle training (PFMT) plus biofeedback with biofeedback alone; the studies included a total of 250 women. A pooled analysis of 4 studies found significantly more urine loss as measured by a posttreatment pad test with PFMT alone than with PFMT plus biofeedback (mean difference [MD], 0.90; 95% confidence interval [CI], 0.71 to 1.10). Reviewers noted that the difference between groups was likely not clinically significant because there was only about a 1-gram difference. Moreover, the finding was largely due to the effect of 1 study. Results on other outcomes (eg, quality of life, number of incontinence episodes) could not be pooled due to imprecision of the estimates.

 

Hsu and colleagues published a systematic review of PFMT with biofeedback in men who had radical prostatectomy (Hsu, 2016). Thirteen trials met reviewers’ inclusion criteria. However, on closer inspection, not all trials included a biofeedback intervention, and other trials did not compare PFMT alone to PFMT plus biofeedback. Thus, conclusions about the added efficacy of biofeedback cannot be determined from the results of this meta-analysis.

 

A 2003 randomized trial by Willie and colleagues randomized 139 men prior to radical prostatectomy to 1 of 3 groups (Willie, 2003). Group 1 received verbal and written instructions about PFMT from a physical therapist. Group 2 received PFMT instruction and instruction on using an electrical stimulation device. Group 3 received the previous 2 intervention components and training on using biofeedback with the electrical stimulation device. Patients had regular contact with a health care provider for the first 5 weeks after surgery. In the immediate postsurgical period, 20.5% in group 1, 22.9% in group 2, and 20.7% in group 3 were continent (p=0.815). After 6 and 12 months, continence rates remained similar among the groups. Twelve-month continence rates were 88% in group 1, 81% in group 2, and 88.6% in group 3 (p=0.524).