Description of the condition
World wide, urinary incontinence in men is a common problem. In western countries, estimates of prevalence rates up to 65 years of age are between 0.9% and 5% (Campbell 2012). In men over 50 years of age, after radical prostatectomy the prevalence of urinary incontinence varies between 5% and 60% (van Kampen 2000; Wolin 2009).
Male stress urinary incontinence is rare except after radical prostatectomy and transurethral resection of the prostate (TURP), which may cause both sphincter and local nerve injury leading to intrinsic sphincter deficiency or weakness of the urethra with stress urinary incontinence as a consequence (Gacci 2003; Groutz 2000; Hu 2003; Moore 1999; Palmer 2003). As well as intrinsic sphincter deficiency, detrusor overactivity and poor compliance of the bladder neck may be factors related to urinary incontinence (Carlson 2001; Groutz 2000; Gudziak 1996; Moore 1999). In general, persistent stress urinary incontinence tends to occur only in a minority of men, mostly because of surgery of the prostate (Buckley 2008). In the case of (persistent) urinary incontinence, this health problem has a major impact on their daily life activities and quality of life (Coyne 2012).
Male detrusor overactivity, including urgency urinary incontinence, may be due to bladder outlet obstruction such as with benign prostatic hyperplasia (De Nunzio 2003; Dmochowski 2002) or neurogenic disease (Steers 2002). Urgency urinary incontinence might also be a symptom related to many different health problems of the lower urinary tract, such as inflammation, infections, kidney stones and tumours.
For urinary incontinence, detrusor overactivity or symptoms of urgency and frequency, treatment options vary between several surgical and conservative or non-surgical interventions, pharmaceuticals and non-pharmaceuticals (Abrams 2009). In the algorithms of the International Consultation of Incontinence, electrical stimulation has been suggested as one of the first-line treatment options (Abrams 2009).
Description of the intervention
In the context of conservative therapy, electrical stimulation can be applied using surface electrodes (Appell 1998; Brubaker 2000; Goldberg 2000; Govier 2001; Hasan 1998; Jabs 2001; Siegel 1992; van Kerrebroeck 1998). Surface electrodes include:
(1) transcutaneous electrical stimulation (Berghmans 2002; Brubaker 2000; Jabs 2001) or transcutaneous electrical nerve stimulation (TENS) via suprapubic, sacral or penile attachment of electrodes, anal plug electrodes, plantar or thigh and similar stimulation, and other placement of surface electrodes such as for interferential or maximum electrical stimulation;
(2) percutaneous electrical stimulation (Govier 2001; Janknegt 1997; van Balken 2001), e.g. posterior tibial nerve stimulation, electro-acupuncture, in general via insertion of a percutaneous needle electrode with another stick-on surface electrode placed near to the needle as the reference electrode.
There are two main types of electrical stimulation:
- long-term or chronic electrical stimulation delivered below the sensory threshold aiming at detrusor inhibition by afferent pudendal nerve stimulation. The electrically evoked activity is suggested to result in reflex activation of hypogastric efferents and central inhibition of pelvic efferent mechanisms sensitive to low-frequency stimulation (Fall 1994). The device is used six to 12 hours a day for several months (Eriksen 1989);
- maximal electrical stimulation using a high-intensity stimulus (just below the pain threshold). It aims to improve urethral closure. Fall (Fall 1991) suggested a direct and reflexogenic contraction of striated peri-urethral musculature. Detrusor inhibition by afferent pudendal nerve stimulation has also been suggested (Berghmans 2002). Maximal electrical stimulation is applied with short duration (15 to 30 minutes) several times a week (or one to two times daily using portable devices at home) (Yamanishi 1997; Yamanishi 1998; Yamanishi 2000a).
Parameters used in previous studies, that is current source, pulse width and duration, current intensity (range), stimulus frequency, pulse shape, time and total number of sessions and rest to work ratio, vary according to type of urinary incontinence and type of electrical stimulation. Berghmans (Berghmans 2002) reported that frequencies of 5 to 20 Hz are usually used for urgency urinary incontinence, 20 to 50 Hz for stress (post-prostatectomy) incontinence, and for mixed urinary incontinence around 20 Hz or high and low frequency alternately (Smith 2009). Pulse durations of 200 (Smith 2009), 300 (Yamanishi 2010), 400 to 600 (Everaert 1999) and 1000 μsec (Moore 1999) have been reported for stress urinary incontinence, for detrusor overactivity 200 to 500 μsec, and for mixed urinary incontinence depending on the dominant factor of the urinary incontinence (Berghmans 2002; Smith 2009). The pulse shape is generally rectangular, and biphasic pulses are preferred (Smith 2009).
In the literature, authors suggest that intermittent, short-term stimulation (maximal electrical stimulation) using a portable stimulation device for home-use should usually be used. In men, rectal or anal or surface electrodes can be used to apply electrical stimulation (ES). Surface electrodes can be positioned:
- over the dorsal nerve of the penis (Yamanishi 2000a);
- at the peri-anal region;
- at the S3, S4 dermatome (Walsh 1999);
- at the tibial nerve (Amarenco 2003); or
- over the quadriceps and hamstring muscles (Okada 1998).
For the treatment of post-prostatectomy incontinence, ES has been used in combination with pelvic floor muscle training (PFMT) but also on its own (Yokoyama 2004). Normally, an anal or rectal electrode is used and the stimulation artificially stimulates the pudendal nerve and its branches to cause direct reflex responses of the urethral and peri-urethral striated muscles (Moore 1999). Frequencies of 5 to 20 Hz are recommended for urgency urinary incontinence, for example Yamanishi (Yamanishi 2000a) used 10 Hz. For stress (post-prostatectomy) urinary incontinence (Moore 1999; Wille 2003; Yamanishi 1998a; Yokoyama 2004) 20 to 50 Hz is used.
Although a wide range of parameters has been claimed to be successful, the optimal set of parameters for each type of urinary incontinence has not been determined (Smith 2009). Duration of the stimulation varies according to the investigators, from 15 minutes/day (Amarenco 2003) to 15 minutes/twice daily (Wille 2003; Yokoyama 2004), to 20 minutes three times/week (Hoffmann 2005) to twice weekly (Moore 1999). As for the number of sessions needed, some authors have recommended at least 10 treatment sessions before the clinical effect was assessed (Fall 1998; Primus 1996). Others used a treatment episode of one month (electrical stimulator also used at home) (Yokoyama 2004), four to six weeks (Amarenco 2003; Bent 1993; Plevnik 1986), 12 weeks (Moore 1999; Soomro 2001) or three months (Wille 2003).
So far, no studies reporting on a comparison of ES protocols or the parameters specified above have been identified, so up to now it is completely unclear what the most appropriate ES protocol and parameters might be, and for which type of urinary incontinence they might be most suitable.
How the intervention might work
ES of the pelvic floor aims at stimulating motor fibres of the pudendal nerve, which may elicit a direct contraction of the pelvic floor muscles or the striated peri-urethral musculature to support the intrinsic part of the urethral sphincter closing mechanism (Fall 1991; Scheepens 2003). As such, ES might contribute to compensation of a weak intrinsic sphincter but it is questionable whether or not ESwould be the first choice treatment option in such cases or would have any additional value to functional training (Berghmans 1998b; Smith 2009).
ES might also be used as a kind of biofeedback. As such, by stimulating awareness, it might help male patients with stress urinary incontinence and who lack conscious control how and where exactly to contract and relax the pelvic floor muscles to regain this control (Berghmans 1998a). Although it is suggested that ES might be useful to train or strengthen the pelvic floor muscles (Sand 1995), or to strengthen the structural support of the urethra and the bladder neck (Plevnik 1991), these hypotheses are not well supported by scientific evidence so far.
In patients with detrusor overactivity or symptoms of urgency and urgency urinary incontinence, ES can elicit direct contractions of the pelvic floor muscles. The contractions stimulate afferent fibres of the pudendal nerve going to the sacral spinal cord, which reflexively decrease the feeling or sensation of urgency and inhibit parasympathetic activity at the level of the sacral micturition centre in the sacral cord in order to reduce involuntary detrusor contractions and reflexively activate the striated peri-urethral musculature. Several authors have suggested that ES of these afferent nerve fibres influences remodelling of neuronal reflex loops, such as the detrusor inhibition reflex (Berghmans 2002; Fall 1994; Vodusek 1986; Weil 2000).
Why it is important to do this review
It has been postulated that by using ES in combination with PFMT, continence is regained more rapidly (Hirakawa 1993; Salinas 1993) and the duration of the application of ES is reduced when PFMT is augmented with ES (Campbell 2012).
The aim of this systematic review was to assess the effects of ES in adult men with stress, urgency and mixed urinary incontinence, and combinations of types of urinary incontinence.
ES for urinary incontinence is the subject of more than one Cochrane review. In men, a Cochrane review (Campbell 2012) reviewed a combination of PFMT and rectal ES for urinary incontinence after radical prostatectomy. ES with the use of implanted stimulation is the subject of a further Cochrane review (Herbison 2009). Because we only deal with non-implanted electrodes in this review, there is no overlap with that review.
Definitions and classifications used in this review are according to the Standardization Committees of the International Continence Society (ICS) and the International Urogynecological Association (Abrams 2002; Haylen 2010; Messelink 2005).
The following comparisons will be considered, in men with stress urinary incontinence; detrusor overactivity with urgency urinary incontinence or men with symptoms of urgency, frequency and urgency urinary incontinence; or mixed urinary incontinence.
1. Electrical stimulation with non-implanted devices versus no treatment.
2. Electrical stimulation with non-implanted devices versus placebo treatments.
3. Electrical stimulation with non-implanted devices versus other 'single' treatments (e.g. a physical therapy such as pelvic floor exercises, a pharmaceutical such as an anticholinergic, or a type of surgery).
4. Electrical stimulation with non-implanted devices in combination with another therapy versus the other therapy alone (e.g. pelvic floor muscle training).
5. One method of providing electrical stimulation with non-implanted devices (e.g. anal plug electrodes) versus another method (e.g. transcutaneous electrical nerve stimulation (TENS).
Criteria for considering studies for this review
Types of studies
Randomized and quasi-randomized controlled trials of surface electrical stimulation (ES) for the management of urinary incontinence were included. Other forms of clinical trials were excluded. As well as trials reported as full-text, we also included trials of which only abstracts were available.
Types of participants
Adult men with: stress urinary incontinence; detrusor overactivity with urgency urinary incontinence or symptoms of urgency, frequency and urgency urinary incontinence; or mixed urinary incontinence.
Types of interventions
One arm of the study must have used ES with non-implanted devices to provide stimulation to the nerves or muscles of the pelvic floor or the bladder, or both, according to a standardised protocol. Any type of non-implanted device was included.
Setting (hospital, office, at home), intensity (both level of electrical current and duration and frequency of stimulation) and method (surface, anal, percutaneous) of stimulation were taken into account.
Comparator interventions included:
- pelvic floor muscle training (PFMT),
- bladder training,
- drugs such as serotonin-noradrenaline reuptake inhibitors (SNRIs), anticholinergics,
- implantable electrical stimulation systems.
Types of outcome measures
- self report of urinary incontinence
Quantification of symptoms
- pad changes (voiding diary)
- frequency symptoms (voiding diary)
- urgency symptoms (voiding diary)
- incontinent episodes (voiding diary)
- standardised pad tests (24-hr, 1-hr, 20-min) measuring grams of involuntary loss of urine (continuous)
Health status measures
- condition-specific quality of life (QOL) e.g. Incontinence Quality of Life (IQOL) Questionnaire
- observation of urinary incontinence
- urodynamic measurements and studies
- pelvic floor muscle function and strength
Health status measures
- general health status e.g. Short Form 36 (Ware 1993)
- tissue damage
- exhaustion of stimulated muscle fibres
- pain, discomfort
- infection of the lower urinary tract
- cost of interventions
- resource implications of differences in outcome
- formal economic analysis (e.g. cost-effectiveness and cost utility)
Search methods for identification of studies
We did not impose any language or other limits on the searches described below.
Relevant trials have been identified from the Specialised Register of controlled trials of the Cochrane Incontinence Group. This Register includes trials identified from searches in the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, PreMEDLINE, CINAHL, and by handsearching of journals and conference proceedings. The methods used to derive the Register, including the search strategy, are described under the Group's module in The Cochrane Library. Date of last search: 21 January 2012.
Additional trials have been sought by the authors from a broader search of computerized bibliographic databases (EMBASE, the Excerpta Medica website, the Dutch National Institute of Allied Health Professions, and the database of the Cochrane Rehabilitation and Related Therapies Field at Maastricht University), from 1980 to 21 January 2012. The keywords used were: incontinence, urinary incontinence, detrusor instability, detrusor overactivity, bladder, overactive bladder, stress incontinence, urge incontinence, mixed incontinence, urgency, frequency, nocturia, physiotherapy, physical therapy, conservative management, conservative therapy, non-surgical stimulation, electrostimulation, neuromuscular stimulation, electrical stimulation, electrotherapy, RCTs, controlled trials, evaluation, effectiveness, efficacy and outcomes.
Searching other resources
One non-electronic bibliographic database was also searched: Physiotherapy Index (up to September 2011), using the search terms given above. In addition, published abstracts presented at the International Continence Society, the European Association of Urology, the American Urogynaecology Society, and the American Urological Association have been reviewed (from 2000 to 2012) and cross-referenced to find if a full-length report has been published. Known trialists and other experts in the field have been contacted to ask for possible relevant trials, published or unpublished. Additional trials have been sought from the reference lists of included studies.
Data collection and analysis
Selection of studies
Only randomized and quasi-randomized controlled trials were included. Two review authors independently screened the list of titles and abstracts generated by our search. Full-text articles of potentially relevant studies were retrieved. Two review authors independently assessed the full-text articles for eligibility. We contacted study investigators as required. Any differences of opinion were resolved by discussion or involvement of a third party. Studies formally considered for the review but excluded were listed with reasons given for their exclusion.
Data extraction and management
Data extraction of the included studies was performed independently by two of the review authors (BB and EH) using a standardised form. Any disagreement was resolved by discussion or by consulting a third party (RdB or MO). Where there was insufficient information regarding the primary outcome in the published reports, study authors were contacted. For data entry, performed by BB and EH, Review Manager software (RevMan 5.2) was used. Processing of the included data of trials was according to the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).
Data were grouped by type of incontinence.
Assessment of risk of bias in included studies
The risk of bias in the included studies was assessed using the Cochrane risk of bias assessment tool (Higgins 2011). This included:
- sequence generation,
- allocation concealment,
- blinding of participants or therapists,
- blinding of outcome assessors,
- completeness of outcome data,
- selective outcome reporting, and
- other potential sources of bias.
Two of the review authors (BB and EH) independently assessed these domains. Any differences of opinion were resolved by consensus or by consulting a third party (MO).
Measures of treatment effect
Analyses were based on available data from all included trials relevant to the comparisons and outcomes of interest. For trials with multiple publications, only the most up-to-date or complete data for each outcome were included. Meta-analysis was undertaken where data were available from more than one study assessing the same outcome. A fixed-effect model was used for calculations of pooled estimates and their 95% confidence intervals. For categorical outcomes we related the numbers reporting an outcome to the numbers at risk in each group to calculate a risk ratio (RR) with 95% confidence interval (CI). For continuous variables we used means and standard deviations to calculate a mean difference (MD) with 95% confidence interval. If similar outcomes were reported on different scales, we calculated the standardised mean difference (SMD). We reversed the direction of effect, if necessary, to ensure consistency across trials. If data to calculate RRs or MDs were not given, we utilised the most detailed numerical data available to calculate the actual numbers or means and standard deviations (for example test statistics, P values).
Unit of analysis issues
The primary analysis was per man randomised.
Dealing with missing data
The data were analysed on an intention-to-treat basis, as far as possible, meaning that all participants must be analysed in the groups to which they were randomised. If this was not the case, we considered whether the trial should be excluded. Attempts were made to obtain missing data from the original trialists. However, if this was not possible, data were reported as given in the studies, except if there was evidence of differential loss to follow up from the randomised groups. In that case, the use of imputation of missing data was considered. If trials reported sufficient detail to calculate mean differences but gave no information on associated standard deviations (SD), the outcome was assumed to have a standard deviation (SD) equal to the highest SD from other trials within the same analysis.
Assessment of heterogeneity
Trials were only combined if they were thought to be clinically similar. Heterogeneity between studies was assessed by visual inspection of plots of the data, the χ
Assessment of reporting biases
In view of the difficulty of detecting and correcting for publication bias and other reporting biases, the authors aimed to minimise their potential impact by ensuring a comprehensive search for eligible studies and by being alert for duplication of data.
Trials were combined if interventions were similar, based on clinical criteria. To combine trial data, a meta-analysis was conducted and a fixed effect model approach to the analysis was used unless there was evidence of heterogeneity across studies.
Subgroup analysis and investigation of heterogeneity
Data were subgrouped, if possible, by the type of underlying urinary incontinence or lower urinary tract symptoms:
- stress urinary incontinence;
- detrusor overactivity with urgency urinary incontinence or symptoms of urgency, frequency and urgency urinary incontinence; or
- mixed urinary incontinence (both stress and urgency urinary incontinence).
If heterogeneity between trials was sufficiently large, an investigation to identify its causes was conducted. The investigation of heterogeneity addressed populations and interventions in the individual trials. The investigation could also include subgroup analyses, meta-regression and sensitivity analyses. If heterogeneity remained after appropriate investigation and possible removal of outlying trials, a random-effects model was used in the meta analysis.
The effects of including or excluding trials at high risk of bias were investigated by means of sensitivity analyses.
Description of studies
Results of the search
Our search strategy yielded 20 potentially eligible studies. All 20 studies were deemed to study ES protocols. For assessment of ES effect only six trials (five full publications: Hoffmann 2005; Moore 1999; Wille 2003; Yamanishi 2010; Yokoyama 2004; 1 abstract: Ceresoli 2002) could be included, whereas 14 studies were excluded. Closer investigation demonstrated that the abstract of Yamanishi 2006 and the full publication of Yamanishi 2010 were the same trial. The randomized clinical trial (RCT) of Yokoyama (Yokoyama 2004) was included but data were not usable as they were not presented in a format usable in the tables of comparison. The literature assessment process is documented with a PRISMA flow chart (Figure 1).
|Figure 1. PRISMA study flow diagram.|
All six included trials involved patients after radical prostatectomy. In total, 544 men were included of whom 305 received some form of ES and 239 a control or comparator treatment.
The trials were conducted at different time points:
- Wille 2003 recruited candidates pre-operatively,
- Hoffmann 2005 shortly after surgery,
- Yokoyama 2004 11 to 15 days post-operatively,
- Yamanishi 2010 three weeks post-operatively,
- Moore 1999 more than eight weeks after surgery, and
- Ceresoli 2002 post-operatively without specification when exactly.
This means that the study population of Wille 2003 might have consisted of men with and without urinary incontinence after radical prostatectomy, while in the other studies only patients with urinary incontinence after radical prostatectomy were included. So, based upon status of incontinence, study populations might already have been heterogeneous.
The following interventions or management strategies were included:
- transcutaneous (perineal) electrical nerve stimulation (Hoffmann 2005);
- transcutaneous (not specified) electrical nerve stimulation (Ceresoli 2002);
- PFMT plus biofeedback (Wille 2003);
- extracorporeal magnetic innervation or stimulation (Yokoyama 2004);
- placebo treatment (sham stimulation) (Yamanishi 2010).
The specific study characteristics (participants, interventions, etc.) of the six included RCTs are presented in the table Characteristics of included studies.
Electrical stimulation (ES) protocols
As in all other ES trials, the included trials showed considerable variation in design, parameters used and intensity of ES:
As described above, the variety in the duration of each session of stimulation and the number of sessions was considerable. Frequencies used in the different studies varied between:
Also, intensity differed between trials:
- in the Wille study patients could control intensity, from 10% to 100% of maximal intensity (Wille 2003);
- whereas patients in the Moore study used ‘comfortable’ intensity (Moore 1999);
- Yokoyama (Yokoyama 2004) reported a maximal output current of 24 mA;
- Ceresoli (Ceresoli 2002) used daily TENS, with intensity below the sensory threshold.
In four trials a biphasic pulse shape was used, only Ceresoli (Ceresoli 2002) and Yokoyama (Yokoyama 2004) did not specify this. Pulse width was between 1 sec (Moore 1999; Wille 2003) and 250 msec (Hoffmann 2005). Yamanishi (Yamanishi 2010) and Yokoyama (Yokoyama 2004) reported a pulse duration of 300 μsec.
Fourteen studies needed to be excluded, mainly because in these trials both men and women were included without separate reporting of results according to gender. Reasons for exclusion can be found in the table Characteristics of excluded studies.
Risk of bias in included studies
Specific characteristics, details of assessment, methodological quality and risk of bias of the included studies are presented in the table Characteristics of included studies and in Figure 2 and Figure 3.
|Figure 2. Risk of bias summary: review authors' judgements about each risk of bias item for each included study.|
|Figure 3. Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.|
Random sequence generation (selection bias)
The method used for random sequence generation was unclear in 3/6 trials (Ceresoli 2002; Wille 2003; Yokoyama 2004). Three trials were judged to be at low risk of bias: Hoffmann 2005 and Yamanishi 2010 used computer-generated randomization; whereas Moore 1999 used a computer-generated random-number list.
Allocation concealment (selection bias)
Security of allocation concealment was unclear in 4/6 trials (Ceresoli 2002; Hoffmann 2005; Wille 2003; Yokoyama 2004). It was considered adequate in two trials: Moore 1999 reported that the "participants were assigned using a computer-generated random-number list placed in sealed envelopes at the end of the assessment visit, with patient and researcher opening the sealed envelope" whereas Yamanishi 2010 reported that "none of the patients, doctors or medical staff knew which type of stimulation had been assigned until the key code was opened".
The doctors, nurses and medical staff were blinded in the trial conducted by Yamansihi and colleagues (Yamanishi 2010). Moore 1999 reported that the physiotherapists and the outcome assessors were blinded. Blinding of participants, personnel and outcome assessors was unclear in the other four trials (Ceresoli 2002; Hoffmann 2005; Wille 2003; Yokoyama 2004).
Incomplete outcome data
There were no missing data in one trial (Yokoyama 2004). Three trials reported dropouts resulting in incomplete outcome data, but these were evenly distributed across the groups (Moore 1999; Yamanishi 2010) or the information about the group distribution was not supplied (Wille 2003). In one trial we assumed that there were no missing data but the information was not provided (Ceresoli 2002). One trial was judged to be at high risk of bias because there was differential dropout (Hoffmann 2005).
It was unclear if there was selective reporting of the outcomes in all six trials because the protocols were not available (Ceresoli 2002; Hoffmann 2005; Moore 1999; Wille 2003; Yamanishi 2010; Yokoyama 2004).
Other potential sources of bias
Information about other potential sources of bias was unclear in 3/6 trials (Ceresoli 2002; Moore 1999; Wille 2003). No other potential source of bias was found in 1/6 trials (Yokoyama 2004), whereas 2/6 trials were judged to be at high risk of bias (Hoffmann 2005; Yamanishi 2010). Support for these judgements are detailed under Characteristics of included studies and summarised in Figure 2 and Figure 3.
Outcome measures in the included studies showed a lack of consistency. With regard to the primary outcomes of the included studies, objective measures were reported by Moore (Moore 1999), Ceresoli (Ceresoli 2002), Yokoyama (Yokoyama 2004) and Yamanishi (Yamanishi 2010) using a 24-hr pad test; Wille (Wille 2003) used a 20-min provocative pad test and Hoffmann (Hoffmann 2005) a 1-hr standardised pad test. Self report of cure or improvement using a visual analogue scale (VAS) was done only by Ceresoli (Ceresoli 2002) and Hoffmann (Hoffmann 2005). Micturition charts were used in the Wille (Wille 2003), Ceresoli (Ceresoli 2002) and Moore (Moore 1999) trials.
Quality of life (QOL) as a secondary outcome measure, using the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ), was measured by Hoffmann (Hoffmann 2005) and Moore (Moore 1999) while the latter also used the Incontinence Impact Questionaire (IIQ-7). In the study of Yokoyama (Yokoyama 2004) the Incontinence Quality of Life (I-QOL) was scored. The Incontinence Consultation on Incontinence Questionnaire Short Form (ICIQ-SF) and King's Health Questionnaire (KHQ) were secondary endpoints in Yamanishi (Yamanishi 2010).
There was no reporting of economic measures in any of the included studies.
Effects of interventions
The specific characteristics and outcomes of the included studies are in the table 'Characteristics of included studies'. Data, when available, are summarised in the paragraph 'Comparisons and data'.
1. Electrical stimulation (ES) with non-implanted devices versus no active treatment
One trial addressed this comparison (Moore 1999). All the men had stress urinary incontinence at baseline. Moore compared anal electrical nerve stimulation with PFMT. The trial was too small to reliably detect differences between the groups ( Analysis 1.1; Analysis 1.2).
2. Electrical stimulation (ES) with non-implanted devices versus placebo treatments
One RCT compared ES against placebo treatment (sham stimulation) (Yamanishi 2010). All the men had stress urinary incontinence.
ES resulted in statistically significantly fewer incontinent men at 3 and 6 months (for example RR at 6 months 0.38, 95% CI 0.16 to 0.87, Analysis 2.1), but not at 12 months. Pad tests also showed less urine loss at 3, but not at 6 or 12 months ( Analysis 2.2). These findings were reflected in the ICI-Questionaire (Short Form) data: the total score (range 0 to 21) was significantly less (better) at 3 and 6 months (MD -3.9, 95% CI -7.15 to -0.65, Analysis 2.3.2) although the QOL score (range 0 to 10) was only better at 3 months (MD -1.5, 95% CI -2.9 to -0.1, Analysis 2.4.1).
Adverse effects in the active ES group occurred in 2 of 26 men, and in 4 out of 30 men in the sham ES group (discomfort or anal pain) ( Analysis 2.5).
3. Electrical stimulation with non-implanted devices versus other 'single' treatments
One small trial addressed this comparison (Yokoyama 2004) but it provided no useable data. There were no statistically significant differences between the groups but the trial was too small to detect them.
4. Electrical stimulation with non-implanted devices in combination with another therapy versus the other therapy alone
- PFMT plus biofeedback (Wille 2003).
Transcutaneous (anal) ES plus PFMT versus PFMT alone
All four trials provided information about the number of men with urinary incontinence at three months after treatment. There was no evidence of a difference in the number of men with urinary incontinence (RR 0.98, 95% CI 0.84 to 1.15, Analysis 4.1.1), nor in the pad test data ( Analysis 4.7). However, one small trial suggested that men reported better QOL scores with PFMT alone (MD 2.34, 95% CI 0.07 to 4.61, Analysis 4.4.1). Overall, men reported more adverse effects with anal ES (20/79, 25% versus 2/99, 2% with PFMT alone; RR 9.77, 95% CI 2.42 to 39.44, Analysis 4.8.1).
Transcutaneous (perineal) ES plus PFMT versus PFMT alone
One trial used perineal ES (Hoffmann 2005). There was no statistically significant difference in the incontinence rate ( Analysis 4.12). However, men reported better quality of life with PFMT (MD 6.51, 95% CI 4.70 to 8.32, Analysis 4.4.2) (Hoffmann 2005). There were too few adverse effects for meaningful analysis ( Analysis 4.8.2).
Transcutaneous (any) ES plus PFMT versus PFMT alone
Ceresoli 2002 did not specify the method of transcutaneous ES. There was no statistically significant difference in the incontinence rate ( Analysis 4.13). Men used fewer pads per day in the PFMT group (MD 1, 95% CI 0.53 to 1.47, Analysis 4.6,1) (Ceresoli 2002).
Transcutaneous (anal or perineal) ES plus PFMT versus PFMT alone
Overall, there was no evidence of a statistically significant difference in the number of men with urinary incontinence at 3 months (146/239, 61% for combined treatment versus 98/156, 63% with PFMT alone; RR 0.93, 95% CI 0.82 to 1.06, Analysis 4.1.3). However, there were more adverse effects (23/139, 17% versus 2/99, 2% with PFMT alone; RR 7.04, 95% CI 1.51 to 32.94, Analysis 4.8.3) and QOL also seemed better with PFMT alone ( Analysis 4.4).
Based on these six trials there seems to be no additional value of the addition of ES to PFMT in men with stress urinary incontinence after a radical prostatectomy.
5. One method of providing electrical stimulation (ES) with non-implanted devices versus another method
One trial compared transcutaneous (anal) ES with PFMT versus transcutaneous (perineal) ES with PFMT (Hoffmann 2005). All the men had stress urinary incontinence at baseline.
This small trial did not detect statistically significant differences between the two methods of administration of transcutaneous ES (anal versus perineal) ( Analysis 5.1; Analysis 5.2) but the QOL score was lower (better) in the anal stimulation group ( Analysis 5.3).
Value and benefits of electrical stimulation
Today, it is still very difficult to clarify the real value, potentials and benefits of electrical stimulation (ES) in the treatment of male urinary incontinence (Smith 2009). There are several reasons for this. First, the nomenclature used to describe electrical stimulation has been inconsistent, leading to confusion how to categorize ES. Stimulation has sometimes been described on the basis of the type of current being used (for example faradic stimulation, interferential therapy) but is also described on the basis of the structures being targeted (for example neuromuscular electrical stimulation), the current intensity (for example low-intensity stimulation, or maximal stimulation) and the proposed mechanism of action (for example neuromodulation). Second, in electrical stimulation studies many combinations of current types, amplitudes, types of waveforms, frequencies, intensities, electrode placements etc. are reported (Smith 2009). Third, although it has been suggested that ES as an intervention for urinary incontinence is using the natural neural pathways and micturition reflexes (Fall 1998; Yamanishi 1998a), and the understanding of both the neuro-anatomy and neurophysiology of the central and peripheral nervous systems is increasing, there is still lack of a well-substantiated biological rationale supporting the use of ES (Smith 2009).
Comparing ES in combination with another treatment modality with that treatment modality alone, there seems to be little to no additional value of the addition of ES to PFMT in men with stress urinary incontinence after a radical prostatectomy. However, this conclusion should be taken with caution because in all four included RCTs the number of included participants was relatively small. There were too few data to draw final conclusions about the effects of ES with non-implantable devices.
Summary of main results
One small trial showed that ES was more effective than sham treatment in the first few months in both urinary outcomes and quality of life (QOL), but this did not persist after six months (Yamanishi 2010).
There was not enough evidence from four small trials to suggest that supplementing ES with PFMT resulted in better urinary outcomes than with PFMT alone ( Analysis 4.1), but there were more adverse effects ( Analysis 4.4). However, the QOL score was lower (better) in the anal stimulation group than in the perineal group ( Analysis 5.3).
Of the included trials, Ceresoli (Ceresoli 2002) and Wille (Wille 2003) did not report any adverse event or side effect. It remains unclear whether these were not reported or that there were no adverse events or side effects in these trials. The other four trials reported that a minority of ES patients suffered from discomfort from the electrode, pain or an increase in incontinence symptoms in such way that they dropped out. Most reported side effects were pain, soreness or local irritation and psychological distress.
Therefore there was not enough evidence for or against ES, partly due to the variability in the interventions of the included trials and their small size.
Overall completeness and applicability of evidence
Patient selection and electrical stimulation protocols
There were many differences in the populations, the intervention protocols and the timing of the outcome measures in the included trials, as summarised above. As all the trials included men with stress urinary incontinence, it was not possible to determine the effect of ES on other types of incontinence such as urgency. The trials were also confined to men who had undergone a prostatectomy.
Five of the six included studies used surface anal electrodes, Hoffmann et al (Hoffmann 2005) also used transcutaneous (anal and perineal) ES. Ceresoli (Ceresoli 2002) used a transcutaneous electrode. Although both anal and surface electrodes have been reported to be safe (Yamanishi 2000a), sometimes patients do not tolerate anal plug electrode because of pain, discomfort, mucosal injury or exacerbation of haemorrhoids (Moore 1999).
Most often intensity of stimulation is used up to the maximum tolerable level, because many authors report that the stronger the intensity the better the outcome (Geirsson 1997; Sand 1995; Yamanishi 1997a; Yamanishi 1998a; Yamanishi 2000a; Yashuda 1999). In the included studies frequencies and duration of the stimulation and the number of sessions varied depending on the investigators. Hoffmann (Hoffmann 2005) and Yamanishi (Yamanishi 2010) used a maximal amplitude of 70 mA, Moore (Moore 1999) reported the use of intensities "adequate to induce a visual lifting of the levator ani and pubococcygeus muscle, considering the level of comfort of the patients", Wille (Wille 2003) instructed patients to control intensity of current ‘from 10% to 100%’. Although it is not clear yet whether or not maximal tolerable intensity of current really is necessary to get an optimal result of ES therapy, it might be that in the included studies intensity of current was not adequate or sufficient to reach an optimal result.
All six trials reported some of the primary outcome measures of interest, that is patient perceived urinary incontinence, pad tests and condition specific QOL. However, none of the trials used the same subjective outcome measures. Regarding pad tests, Moore (Moore 1999), Ceresoli (Ceresoli 2002), Yamanishi (Yamanishi 2010) and Yokoyama (Yokoyama 2004) used a 24-hr pad test, while Hoffmann (Hoffmann 2005) used a 1-hr pad test and Wille (Wille 2003) a 20-min pad test. As a result, the amount of data available for meta-analysis was very limited, making interpretation of results difficult.
In all but one (Yokoyama 2004) of the six RCTs, ES was combined with PFMT. Because of this, it is difficult to assess any specific effect of ES (as could have been the case if ES was provided as a stand-alone therapy). Especially when ES is combined with PFMT, any effect may be partly due to the attention and support given to patients during their clinic sessions (Moore 1999). Both Moore et al (Moore 1999) and Hoffman et al (Hoffmann 2005) discussed that these non-specific effects might have contributed to the effect of ES.
Quality of the evidence
Study quality and methodological assessment
Various conservative treatment modalities for the management of stress, mixed, and urgency urinary incontinence are available. ES is used alone or as an adjunct to one of other eligible physiotherapeutic treatment modalities such as PFMT. All but one of the included trials in this review used ES as an adjunct to PFMT.
Only a small number of RCTs investigating the effect of ES could be identified.
When assessing methodological quality, the primary goal is to assess the reported effects of interventions and the influence, risk and direction of bias on the results of these treatments as methodological quality plays an important role in weighing the conclusions of different trials. In order to explore the possibility that an increasing likelihood of bias was reflected in the results of the trials (through an increasing number of methodological shortcomings), the 'Risk of bias' criteria presented in The Cochrane Handbook for Systematic Reviews of Intervention, were assessed.
The risk of bias in the six included studies appeared to range from low to high risk of bias. Allocation concealment and blinding were unclear in more than half of the studies while incomplete outcome reporting and selective reporting were identified in about 40% and 14% respectively, and other potential bias in 40% of the studies. So, bias with respect to measurement, intervention or allocation might have been introduced. None of the studies reported whether or not they had performed an intention-to-treat analysis. A power calculation was only performed by Moore et al (Moore 1999) and Yamanishi et al (Yamanishi 2010). It might be that studies were underpowered to find any statistical differences between study groups.
Because of small numbers per group in the included RCTs, occurrence of type II-errors was more likely than would have been the case in RCTs with larger groups.
Apart from Ceresoli et al (Ceresoli 2002), all included RCTs had sufficient descriptions of the content, the duration and the intensity of the interventions.
Potential biases in the review process
Implications for practice
The six small included trials were of disparate populations and interventions, limiting our ability to combine the data. There was some evidence that electrical stimulation (ES) enhanced the effect of PFMT in the short term but not after six months. There were, however, more adverse effects with ES.
There was a marked lack of consistency in the ES protocols, which implies a lack of consensus about the physiological principles of rehabilitating urinary incontinence in men through ES in clinical practice.
Since no RCTs on ES were available for any type of urinary incontinence other than post-prostatectomy stress urinary incontinence, at the moment it is not possible to reach any conclusion on effects, side effects or adverse events of ES as a-stand alone therapy or in combination with other treatments on men with other types of urinary incontinence.
In men with stress urinary incontinence after radical prostatectomy, up till now there is no convincing evidence from RCTs that ES is a useful treatment. So far, it is impossible to recommend the optimal ES regimen and protocol.
Implications for research
Because of the scanty evidence of an effect of ES on incontinence in men, and the lack of agreement about the correct mode of administration, duration and frequency of treatment, there is a need to design rigorous trials to determine whether there is a place for ES in the management of men with stress urinary incontinence. Any trialists should use the recommendations and principles set out in the CONSORT statement when designing and reporting their trials.
There is a need for more research to determine the best ES protocol(s) and important standardised outcome measures for these men. Future trials should follow the recommendations and principles of the CONSORT statement and should include primary and secondary outcomes which are most suitable (with respect to sensitivity, responsiveness, validity, and reliability).
The authors would like to acknowledge the team of the Cochrane Incontinence Review Group for their help.
Data and analyses
- Top of page
- Authors' conclusions
- Data and analyses
- Contributions of authors
- Declarations of interest
- Sources of support
- Index terms
Contributions of authors
All review authors contributed in the writing of the protocol. Authors Bary Berghmans, Erik Hendriks and Rob de Bie independently assessed the pertinence and quality of eligible studies and selected which to include in the review. The first two review authors independently extracted data from trial reports of identified studies and interpreted the results. Muhammad Imran Omar re-checked extracted data and risk of bias assessments and offered methodological advice and help. In the case of disagreement Rob de Bie and Muhammad Imran Omar acted as consultants and helped to make final decisions. All review authors contributed in the writing of the final version of the review.
Declarations of interest
Sources of support
- No sources of support supplied
- NIHR, UK.The National Institute for Health Research (NIHR) is the largest single funder of the Cochrane Incontinence Group.
Medical Subject Headings (MeSH)
*Electrodes; Biofeedback, Psychology [methods]; Electric Stimulation Therapy [adverse effects; instrumentation; *methods]; Exercise Therapy [*methods]; Muscle Contraction; Muscle Strength; Pelvic Floor; Randomized Controlled Trials as Topic; Transcutaneous Electric Nerve Stimulation [adverse effects; instrumentation; methods]; Urinary Incontinence, Stress [*therapy]; Urinary Incontinence, Urge [*therapy]
MeSH check words
* Indicates the major publication for the study