Electrical stimulation with non‐implanted devices for stress urinary incontinence in women

Fiona Stewart; Bary Berghmans; Kari Bø; Cathryn MA Glazener

doi:10.1002/14651858.CD012390.pub2

. 2017 Dec 22;2017(12):CD012390. doi: 10.1002/14651858.CD012390.pub2

Electrical stimulation with non‐implanted devices for stress urinary incontinence in women

Fiona Stewart ¹, Bary Berghmans ², Kari Bø ³, Cathryn MA Glazener ^4,^✉

Editor: Cochrane Incontinence Group

PMCID: PMC6486295 PMID: 29271482

Abstract

Background

Several treatment options are available for stress urinary incontinence (SUI), including pelvic floor muscle training (PFMT), drug therapy and surgery. Problems exist such as adherence to PFMT regimens, side effects linked to drug therapy and the risks associated with surgery. We have evaluated an alternative treatment, electrical stimulation (ES) with non‐implanted devices, which aims to improve pelvic floor muscle function to reduce involuntary urine loss.

Objectives

To assess the effects of electrical stimulation with non‐implanted devices, alone or in combination with other treatment, for managing stress urinary incontinence or stress‐predominant mixed urinary incontinence in women. Among the outcomes examined were costs and cost‐effectiveness.

Search methods

We searched the Cochrane Incontinence Specialised Register, which contains trials identified from the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, MEDLINE In‐Process, MEDLINE Epub Ahead of Print, CINAHL, ClinicalTrials.gov, WHO ICTRP and handsearches of journals and conference proceedings (searched 27 February 2017). We also searched the reference lists of relevant articles and undertook separate searches to identify studies examining economic data.

Selection criteria

We included randomised or quasi‐randomised controlled trials of ES with non‐implanted devices compared with any other treatment for SUI in women. Eligible trials included adult women with SUI or stress‐predominant mixed urinary incontinence (MUI). We excluded studies of women with urgency‐predominant MUI, urgency urinary incontinence only, or incontinence associated with a neurologic condition. We would have included economic evaluations had they been conducted alongside eligible trials.

Data collection and analysis

Two review authors independently screened search results, extracted data from eligible trials and assessed risk of bias, using the Cochrane 'Risk of bias' tool. We would have performed economic evaluations using the approach recommended by Cochrane Economic Methods.

Main results

We identified 56 eligible trials (3781 randomised participants). Eighteen trials did not report the primary outcomes of subjective cure, improvement of SUI or incontinence‐specific quality of life (QoL). The risk of bias was generally unclear, as most trials provided little detail when reporting their methods. We assessed 25% of the included trials as being at high risk of bias for a variety of reasons, including industry funding and baseline differences between groups. We did not identify any economic evaluations.

For subjective cure of SUI, we found moderate‐quality evidence that ES is probably better than no active treatment (risk ratio (RR) 2.31, 95% CI 1.06 to 5.02). We found a similar result for cure or improvement of SUI (RR 1.73, 95% CI 1.41 to 2.11), but the quality of evidence was lower. We are very uncertain if there is a difference between ES and sham treatment in terms of subjective cure because of the very low quality of evidence (RR 2.21, 95% CI 0.38 to 12.73). For subjective cure or improvement, ES may be better than sham treatment (RR 2.03, 95% CI 1.02 to 4.07). The effect estimate was 660/1000 women cured/improved with ES compared to 382/1000 with no active treatment (95% CI 538 to 805 women); and for sham treatment, 402/1000 women cured/improved with ES compared to 198/1000 with sham treatment (95% CI 202 to 805 women).

Low‐quality evidence suggests that there may be no difference in cure or improvement for ES versus PFMT (RR 0.85, 95% CI 0.70 to 1.03), PFMT plus ES versus PFMT alone (RR 1.10, 95% CI 0.95 to 1.28) or ES versus vaginal cones (RR 1.09, 95% CI 0.97 to 1.21).

Electrical stimulation probably improves incontinence‐specific QoL compared to no treatment (moderate quality evidence) but there may be little or no difference between electrical stimulation and PFMT (low quality evidence). It is uncertain whether adding electrical stimulation to PFMT makes any difference in terms of quality of life, compared with PFMT alone (very low quality evidence). There may be little or no difference between electrical stimulation and vaginal cones in improving incontinence‐specific QoL (low quality evidence). The impact of electrical stimulation on subjective cure/improvement and incontinence‐specific QoL, compared with vaginal cones, PFMT plus vaginal cones, or drugs therapy, is uncertain (very low quality evidence).

In terms of subjective cure/improvement and incontinence‐specific QoL, the available evidence comparing ES versus drug therapy or PFMT plus vaginal cones was very low quality and inconclusive. Similarly, comparisons of different types of ES to each other and of ES plus surgery to surgery are also inconclusive in terms of subjective cure/improvement and incontinence‐specific QoL (very low‐quality evidence).

Adverse effects were rare: in total nine of the women treated with ES in the trials reported an adverse effect. We identified insufficient evidence to compare the risk of adverse effects in women treated with ES compared to any other treatment. We were unable to identify any economic data.

Authors' conclusions

The current evidence base indicated that electrical stimulation is probably more effective than no active or sham treatment, but it is not possible to say whether ES is similar to PFMT or other active treatments in effectiveness or not. Overall, the quality of the evidence was too low to provide reliable results. Without sufficiently powered trials measuring clinically important outcomes, such as subjective assessment of urinary incontinence, we cannot draw robust conclusions about the overall effectiveness or cost‐effectiveness of electrical stimulation for stress urinary incontinence in women.

Plain language summary

Non‐invasive electrical stimulation for stress urinary incontinence in women

Review question

We investigated whether electrical stimulation was better than no treatment at all or better than other available treatments for curing or improving stress urinary incontinence (SUI) symptoms in women. We also investigated whether SUI was cured or improved by adding electrical stimulation to other treatments, compared to other treatments and to different types of electrical stimulation. Finally, we investigated whether electrical stimulation represented value for money.

Background

About 25% to 45% of women worldwide have problems with leaking urine involuntarily. Women with SUI often leak urine with physical exertion such as coughing or sneezing. SUI can be treated with pelvic floor muscle exercises, vaginal cones, drug therapy or surgery, but there are various problems with these treatments. A possible alternative is electrical stimulation with non‐implanted devices, whereby an electrical current is delivered through vaginal electrodes.

How up‐to‐date is this review?

We searched for studies that had been published up to 27 February 2017.

Study characteristics

We found 56 trials (involving a total of 3781 women, all with stress urinary incontinence but some with urgency urinary incontinence as well) comparing electrical stimulation to no treatment or to any other available treatment.

Key results

For cure or improvement of SUI, electrical stimulation was probably better than no active or sham treatment. There was not enough evidence to say whether it was any better than pelvic floor muscle training for curing or improving SUI, or for quality of life. Adding electrical stimulation to pelvic floor muscle training may not make much difference to cure or improvement of SUI. It is uncertain whether it offers any improvement in quality of life compared with pelvic floor muscle training.

We found that few women reported adverse effects with electrical stimulation, but there was not enough reliable evidence comparing electrical stimulation to other treatments to know more about its safety.

There was not enough evidence comparing electrical stimulation to other existing treatments such as drug therapy, pelvic floor muscle training plus vaginal cones, surgery, or different forms of electrical stimulation, to provide evidence‐based guidance on which would be better, and for which women, in curing or improving SUI or in improving quality of life. There was no information from these studies to judge value for money.

Quality of the evidence

There is some evidence to support the use of electrical stimulation for stress urinary incontinence in women, but we are still very uncertain about the full potential of this treatment because of the low quality of the existing evidence. While we found evidence indicating that electrical stimulation may be better than no treatment, we did not find enough well‐designed trials with enough women to fully answer our review questions, so we do not yet know if ES is better or worse than other treatments.

Summary of findings

Background

Description of the condition

Urinary incontinence (UI) affects 25% to 45% of women worldwide (ICI 2013). UI presents in the following forms.

Stress urinary incontinence (SUI): involuntary loss of urine through physical exertion or effort, coughing or sneezing.
Urgency urinary incontinence (UUI): involuntary loss of urine associated with a sudden and compelling desire (urgency) to urinate that is difficult to delay.
Mixed urinary incontinence (MUI): involuntary loss or urine associated with both stress and urgency.

Symptomatic diagnosis of SUI is typically based on whether urine leakage occurs with physical exertion or effort, as reported by women themselves.

In addition, urodynamically proven stress incontinence (USI) is diagnosed when an observer can see urine leakage on stress such as coughing during urodynamic examination, in the absence of a detrusor contraction (ICI 2013). Symptomatic diagnosis of MUI is based on self‐report of urine leakage through both physical exertion and urgency.

This review includes women with SUI, USI and stress‐predominant MUI.

Several mechanisms are thought to contribute to stress urinary incontinence.

Suboptimal pelvic floor muscle strength.
Hypermobility or significant displacement of the urethra and bladder neck during exertion.
Intrinsic urethral sphincter deficiency (ICI 2013).

In women, these mechanisms may coexist (Kursh 1994), but few clinical trials have distinguished between them as underlying causes. We will consider women whose incontinence may be due to any of these mechanisms together in this review.

Prevalence estimates of SUI range from 3% to 25% of adult women, with older women more likely to be affected (ICI 2013). Quality of life and sexual function are often substantially impaired by the fear of leakage, resulting in avoidance of social or physical activities which might cause it, embarrassment and poor sleep (Oh 2008). SUI can severely impact the ability to carry out daily activities, resulting in debilitating embarrassment, social isolation and considerably decreased health‐related quality of life (Bartoli 2010). Women with SUI may be less likely to participate in physical activity, which in turn has a detrimental impact on overall health because inactivity is a risk factor for many diseases (Bø 2004). Other evidence has shown that up to 50% of women with UI will avoid intimacy with their partners (Roos 2014).

Furthermore, SUI is associated with a considerable economic burden for women and for healthcare providers. For instance, routine care, such as sanitary pads, can entail considerable cost for each woman affected, while conservative treatment and surgery may cost the equivalent of several thousand GBP for each woman (ICI 2013).

Description of the intervention

In Europe and the USA, conservative interventions such as pelvic floor muscle training (PFMT), with or without biofeedback, are recommended as first‐line treatment for SUI (EAU 2015; NICE 2013; Qaseem 2014); however, many women may find it difficult to adhere to these methods in the long‐term (Bø 2005; Dumoulin 2014).

Surgery is usually suggested as a second‐line option where conservative treatment has not improved a woman's symptoms or she is unwilling or unable to continue the treatment. Synthetic mid‐urethral tape, open or laparoscopic colposuspension and autologous rectus fascia sling procedures are recommended by the National Institute for Health and Care Excellence (NICE), although the use of surgery with tapes in the management of UI remains controversial in terms of safety and adverse effects (Scottish Government 2015). Several Cochrane Reviews have investigated the effects of surgical management for SUI (Dean 2017; Lapitan 2017; Nambiar 2017; Rehman 2011). Other older surgical procedures, such as anterior repair or bladder neck needle suspension, have generally fallen out of use due to lower effectiveness (Glazener 2017a; Glazener 2017b).

Other less invasive second‐line treatment options available in some countries include bulking agents, where a substance is injected into the urethral wall to increase its size and allow it to remain closed, or pharmacological therapy, typically with duloxetine. The disadvantages of these treatments are that they are likely to be less effective than surgery, and, in the case of drug therapy, long‐term adherence is usually necessary and is associated with unpleasant side effects (Alhasso 2005; Mariappan 2005). Bulking agents can cause discomfort or bleeding when urinating, and their effectiveness decreases over time, requiring retreatment. Other available treatments for SUI include artificial urinary sphincters and complementary therapies such as acupuncture.

Electrical stimulation (ES) has emerged as a first‐line alternative to PFMT in women who are unable to contract their pelvic floor muscles voluntarily or as a second‐line treatment if PFMT alone is not sufficiently effective. It may also be beneficial to combine ES with the use of vaginal cones and drug therapy.

How the intervention might work

When a nerve is stimulated, signals travel both toward the periphery and toward the central nervous system. Electrical stimulation may elicit responses to these signals, which may come from the central nervous system or the innervated tissues, or the central nervous system may be modified to reinterpret some signals (Chancellor 2002; Fall 1994).

With respect to lower urinary tract dysfunctions, electrical stimulation is applied particularly to the pelvic floor muscles, bladder and sacral nerve roots. In the context of SUI, the aim of ES is to improve pelvic floor muscle function so that the pelvic floor muscles can be used when needed to occlude (close) the urethra (such as before a cough) and to increase muscle bulk, which may help reduce urine loss by closing up the urethral walls.

Direct ES of the pelvic floor is intended to stimulate motor‐efferent fibres of the pudendal nerve, which may elicit a direct contraction of the pelvic floor muscles or the striated peri‐urethral musculature, supporting the intrinsic part of the urethral sphincter‐closing mechanism (Fall 1991; Scheepens 2003). As such, ES might contribute to compensating for a weak intrinsic sphincter, but it is questionable whether or not ES in such cases would be the first‐choice treatment option or would have any additional value to pelvic floor muscle training (Ayeleke 2015).

Different authors have suggested that ES may restore continence in women with SUI by:

strengthening the structural support of the urethra and the bladder neck by increasing muscle bulk (Plevnik 1991);
securing the resting and active closure of the proximal urethra (Erlandson 1977);
strengthening the pelvic floor muscles and hence their ability to close the urethra (Sand 1995);
inhibiting reflex bladder contractions (Berghmans 2002; Fall 1994);
modifying the vascularity (improving blood supply) of the urethral and bladder neck tissues (Fall 1991; Fall 1994; Plevnik 1991).

In the context of conservative or non‐surgical, non‐medical therapy, ES can be applied using surface electrodes in the form of transcutaneous or percutaneous ES. Transcutaneous ES is administered via suprapubic or vulval surface electrodes, or vaginal/anorectal plug electrodes. Percutaneous ES uses needle electrodes that penetrate the skin in conjunction with a surface electrode placed close to the needle to act as a reference electrode (e.g. posterior tibial nerve stimulation, percutaneous nerve evaluation). Percutaneous ES is normally used for women with overactive bladder symptoms, not SUI, so we exclude it from this review (see companion review of ES for overactive bladder; Stewart 2016a).

The frequency, dosage and duration of treatment with ES varies considerably. Although authors have claimed success for a wide range of parameters, there is no agreement on the optimal set of parameters for each type of urinary incontinence. Clinical consensus from the International Consultation on Incontinence (ICI) underlines this uncertainty:

"EStim is provided by clinic‐based mains powered machines or portable battery powered stimulators with a seemingly infinite combination of current types, waveforms, frequencies, intensities, electrode types and placements. Without a clear biological rationale it is difficult to make choices about different ways of delivering EStim. Additional confusion is created by the relatively rapid developments in the area of EStim, and a wide variety of stimulation devices and protocols have been developed even for the same condition" (ICI 2013).

Evidence from a systematic review has suggested that, in men, ES with non‐implanted devices may be more effective than sham treatment for urinary incontinence and that ES might enhance the effectiveness of pelvic floor muscle training in the short term (Berghmans 2013). Other evidence suggests that ES is more effective than sham, placebo or no active intervention for treating overactive bladder and urgency urinary incontinence, but the quality of evidence identified was generally low (Stewart 2016a). It is not yet clear whether ES has similar effects in women with SUI.

Why it is important to do this review

ES has shown promise in the treatment of UUI, but the evidence base for its use in SUI is inconclusive (Schreiner 2013). Given the adherence issues with conservative treatment, the side effects of drug therapy and the safety concerns regarding some kinds of surgical intervention, it is important to investigate alternative options for women with SUI.

Many randomised controlled trials (RCTs) have been undertaken investigating ES for SUI, compared to a variety of conservative interventions for SUI such as pelvic floor muscle exercises, drug therapy, vaginal cones, sham ES and no active treatment. Some trials have found no evidence of a difference in treatment effect, while others have found ES to be more effective than a comparator intervention. Given the heterogeneity of ES treatments, it is important to attempt to synthesise the available evidence relating to the diverse ES devices and protocols. Previous publications have synthesised some of the earlier evidence relating to ES for SUI (ICI 2013; Imamura 2010), but with a growing number of trials addressing this question, an up‐to‐date and comprehensive systematic review is needed to obtain the best possible estimate of the effectiveness of ES.

Objectives

Methods

Criteria for considering studies for this review

Types of studies

We included parallel or cross‐over RCTs, quasi‐RCTs (RCTs in which allocation to treatment is by methods such as alternate medical records, date of birth, or other predictable methods) and cluster‐randomised trials.

To critically appraise and summarise current evidence on the cost effectiveness of ES, we included relevant health economics studies conducted alongside effectiveness studies that met the eligibility criteria for the effectiveness component of the review. This includes:

full economic evaluation studies of ES compared to other treatments (i.e. cost‐effectiveness analyses, cost‐utility analyses, cost‐benefit analyses);
partial economic evaluations of ES (i.e. cost analyses, cost‐description studies, cost‐outcome descriptions);
RCTs reporting more limited information, such as estimates of resource use or costs associated with ES.

Types of participants

Eligible studies included adult women (18 years or older, or according to study authors' definitions of adult) with SUI or stress‐predominant MUI on the basis of symptoms, signs or urodynamic diagnosis. We used the trialists' definitions to classify women with SUI or stress‐predominant MUI.

We excluded studies in women with urgency‐predominant MUI, UUI only, or incontinence associated with a neurologic condition or frailty. We also excluded studies in men and women that did not report data separately by sex and studies including only men or children. We included trials of participants with MUI, UUI and SUI only if the data for women with SUI were presented separately. We included trials in women with MUI if the condition was SUI‐predominant.

Types of interventions

Eligible interventions included any method of delivering electrical stimulation with non‐implanted devices (see Table 12 and Characteristics of included studies for details of methods used). These devices could be placed in the vagina or anus or on a skin surface, but we excluded those that penetrated the skin or had to be placed surgically, which a different Cochrane Review covers (Herbison 2009). Health professionals or participants themselves could administer the treatment in any setting.

1. Description of electrical stimulation interventions.

Study	Current	Current intensity	Pulse shape & duration	Frequency (Hz)	Duty cycle	Electrodes	Treatment duration/supervision
Aaronson 1995	Unclear	Unclear	Unclear	Unclear	Unclear	Intravaginal	Unclear
Abel 1997	Unclear	Unclear	Unclear	Unclear	Unclear	Unclear	Unclear
Alves 2011	Unclear	Maximum tolerable intensity	Biphasic 2000 Hz 100 ms	50	4 s on: 8 s off	Intravaginal	Twice a week for 6 weeks (12 sessions)
Alves 2011	Unclear	Maximum tolerable intensity	Biphasic 2000 Hz 700 ms	50	4 s on: 8 s off	Intravaginal	Twice a week for 6 weeks (12 sessions)
Bernardes 2000	Unclear	10‐30 mA up to maximum tolerable intensity	Symmetrical bidirectional 1 ms	60	6 s on: 12 s off	Intravaginal	20 min daily for 10 days (10 sessions)
Beuttenmuller 2010	Unclear	Maximum tolerable intensity	0.2‐0.5 ms	50	Rest time at least twice the time of current	Intravaginal	Two 20 min sessions per week for 6 weeks (12 sessions)
Bidmead 2002	Unclear	Unclear	Unclear	Unclear	Unclear	Unclear	Unclear
Bø 1999	Unclear	0‐120 mA up to maximum tolerable intensity	0.2 ms	50	0.5‐10 s on: 0‐30 s off, adapted on basis of ability to hold voluntary contraction	Intravaginal	30 min daily
Bourcier 1994	Unclear	Unclear	Unclear	Unclear	Unclear	Unclear	Twelve 30 min sessions over 6 weeks (20 min maximal ES, 10 min EMG/pressure biofeedback)
Bridges 1988	Unclear	Maximum tolerable intensity	Unclear	0‐100	Unclear	Unclear	Three 15 min session per week for 4 weeks (12 sessions)
Brubaker 1997	Bipolar	0‐100 mA	Bipolar square wave 0.1 µs	20	2 s on: 4 s off	Intravaginal	20 min daily for 8 weeks (56 sessions)
Castro 2008	Bipolar	0‐100 mA up to maximum tolerable intensity	Bipolar square wave 0.5 milliseconds	50	5 s on: 10 s off	Intravaginal	Three 20 min session per week under supervision of trained physical therapist
Correia 2013	Unclear	Maximum tolerable intensity	700 µs	50	Unclear	4 surface electrodes: 2 in the suprapubic region and 2 medial to the ischial tuberosity	Two 20 min session per week for 3 weeks (6 sessions)
Correia 2013	Unclear	Maximum tolerable intensity	700 µs	50	Unclear	Intravaginal	Two 20 min session per week for 3 weeks (6 sessions)
Correia 2014	Unclear	Maximum tolerable intensity	700 µs	50	4 s on: 8 s off	4 surface electrodes: 2 in the suprapubic region and 2 medial to the ischial tuberosity	Two 20 min session per week for 6 weeks (12 sessions)
Correia 2014	Unclear	Maximum tolerable intensity	700 µs	50	4 s on: 8 s off	Intravaginal	Two 20 min session per week for 6 weeks (12 sessions)
Delneri 2000	Unclear	"According to the patient's sensations"	Unclear	15 min: 20 15 min: 50	4 s on: 8 s off	Unclear	12 x 30 min sessions on consecutive days, excluding Saturdays and Sundays.
Demirturk 2008	Unclear	Unclear	Unclear	0‐100 Hz	Unclear	4 vacuum electrodes: 2 in the suprapubic region, 2 near to the medial side of the ischial tuberosity, crosswise	3 x 15 min session per week for 5 weeks (15 sessions)
Edwards 2000	Unclear	Unclear	Unclear	Unclear	Unclear	Unclear	Unclear
Eyjolfsdottir 2009	Unclear	Unclear	200 µs	50	Unclear	Unclear	Unclear
Firra 2013	Unclear	Unclear current, intensity according to participant tolerance	Unclear	12.5	5 s on: 10 s off	Intravaginal	14 x 30 min sessions
Goode 2003	Biphasic	According to participant tolerance, up to 100 mA	Biphasic, 1 millisecond	20	1 s on: 1 s off1	Intravaginal	Home use, 15 min every second day for 8 weeks
Hahn 1991	Unclear	Unclear	Unclear	50	Unclear	Intravaginal	Home use, 6‐8 hours per night for 12 months
Haig 1995	Unclear	Unclear	Unclear	10‐40		Intravaginal	20 min sessions, treatment for 3 months (unclear how many sessions)
Henalla 1989	Unclear	According to participant tolerance	Unclear	0‐100	Unclear	Unclear	1 x 20 min session per week for 10 weeks (10 sessions)
Hofbauer 1990	Unclear	Intensity increased until participant felt a contraction	Unclear	Unclear	10 ms on: 15 ms off	Unclear	3 sessions per week for 6 weeks (18 sessions)
Huebner 2011	Unclear	20−80 mA	Unclear	50	8 s on: 15 s off	Intravaginal	2 x 15 min sessions per day for 12 weeks
Jeyaseelan 1999	Unclear	Up to 90 mA	Balanced, asymmetrical biphasic pulse width 250 µs	Background low frequency (to target slow twitch fibres), and intermediate frequency with initial doublet (to target fast twitch fibres)	10 s on: 50 s off	Intravaginal	Home use (portable device), 1 hour daily for 8 weeks (except when menstruating)
Jeyaseelan 2002	Unclear	Unclear	Unclear	Unclear	Unclear	Unclear	1 hour daily for 8 weeks (except when menstruating)
Jeyaseelan 2003	Unclear	Unclear	Unclear	Unclear	A range of frequencies in conjunction with a longer duty cycle than is traditionally used	Unclear	Unclear
Knight 1998	Unclear	Low intensity, barely perceptible tingling sensation	Pulse width 200 ms	Preset frequencies of 10 Hz with bursts of 35 Hz to maintain fast twitch fibre activity	5 s on: 5 s off	Intravaginal	Home use (3 hours per day) for 6 months, except during menstruation (overnight)
Knight 1998	Unclear	According to maximum participant tolerance	Pulse width 200 ms	35	5 s on: 5 s off	Intravaginal	16 x 30 min sessions in clinic
Laycock 1988	Unclear	Unclear	Unclear	Unclear	Unclear	Unclear	2‐3 30 min sessions per week for 4‐6 weeks
Laycock 1993a	Unclear	According to maximum participant tolerance	Unclear	Three different frequencies: 10 min 1 Hz, 10 min 10‐40 Hz, 10 min 40 Hz	Unclear	Transcutaneous: one medium electrode placed over perineal body and a small electrode positioned immediately inferior to the symphasis pubis	10 sessions; 1 x 15 min, 9 x 30 min
Laycock 1993b	Unclear	According to maximum participant tolerance	Unclear	Three different frequencies: 10 min 1 Hz, 10 min 10‐40 Hz, 10 min 40 Hz	Unclear	Transcutaneous: one medium electrode placed over perineal body and a small electrode positioned immediately inferior to the symphasis pubis	10 sessions; 1 x 15 min, 9 x 30 min
Lo 2003	Unclear	According to maximum participant tolerance	Unclear	0‐100	Unclear	Transcutaneous: 2 anterior flat electrodes placed over obturator foramen 1.5‐2 cm lateral to symphasis, 2 posterior electrodes placed medial to ischial tuberosities, either side of anus	One 15 min session, 11 x 30 min sessions. 3 sessions per week for 4 weeks (12 sessions)
Lopes 2014	Unclear	Unclear	Unclear	Unclear	Unclear	Unclear	3 x 30 min sessions per week at home
Luber 1997	Unclear	10−100 mA	2 ms	50	2 s on: 4 s off	Intravaginal	2 x 15 min sessions per day for 12 weeks
Maher 2009	Unclear	Unclear	Unclear	Unclear	Unclear	External electrodes	Home use, at least 4 x 30 min sessions per week for 8 weeks
Maher 2009	Unclear	Unclear	Unclear	Unclear	Unclear	Intravaginal	Home use, at least 4 x 30 min sessions per week for 8 weeks
Min 2015	Unclear	Unclear	Unclear	Unclear	Unclear	Unclear	Unclear
Olah 1990	Unclear	0−100 mA, up to maximum participant tolerance	Unclear	Unclear	Unclear	Transcutaneous: 2 electrodes placed on abdomen and 2 on inner thighs	3 x 15 min sessions per week for 4 weeks
Oldham 2013	Unclear	Pre‐programmed to increase intensity over 24 s to reach therapeutic level and switch off automatically after 30 min. All devices same level of stimulation (average intensity considered comfortable and capable of producing contractions of pelvic floor muscles)	Unclear	During the 10 s 'on time' the device delivers 10 repeats of a short high intensity burst of 50 Hz stimulation immediately preceded by a doublet (125 Hz), superimposed on continuous low frequency 2 Hz stimulation	10 s on: 10 s off	Intravaginal ‐ single use tampon‐like Pelviva device	One disposable device per day for 12 weeks except during menstruation
Parsons 2004	Unclear	Unclear	Unclear	Unclear	Unclear	Intravaginal	Home use
Patil 2010	Unclear	According to maximum participant tolerance	Unclear	0‐100	Unclear	Surface ES: 2 flat electrodes placed anteriorly over obturator foramen, 1.5‐2cm lateral to the symphysis; 2 electrodes placed posteriorly medial to ischial tuberosity on either side of the anus	1st session 15 min, if no ill effects then 30 min for all subsequent sessions. 3 times a week, for 4 weeks (12 sessions) under supervision of a physiotherapist. Participants were asked to perform 8‐12 pelvic floor contractions 3 times a day at home
Pereira 2012	Unclear	According to maximum participant tolerance	Pulse width 700 µs	50	4 s on: 8 s off	Surface ES: 2 electrodes in suprapubic region, 2 medial to the ischial tuberosity	2 x 20 min sessions per week for 6 weeks (12 sessions). "The women were not instructed to perform the contraction of the pelvic floor muscles in conjunction with electrical stimulation"
Pohl 2004	Unclear	Unclear	Unclear	Unclear	Unclear	Unclear	Unclear
Preisinger 1990	Unclear	Surging faradic‐type current	Unclear	Unclear	Unclear	Unclear	3 x 10 min sessions per week for 10‐12 weeks
Sand 1995	Unclear	Gradually adjusted amperage to 60‐80 mA or highest tolerable level	Unclear	Unclear	First 2 weeks: 5 s on: 10 s off. Weeks 3‐4: 5s: 5s; weeks 5‐6: 5 s: 10 s; weeks 7‐12: 5 s: 5 s	Vaginal electrode ( 2.6 cm diameter, 6.35 cm length) with electrode resistance 85 Ω	Women instructed to use device twice daily for 12 weeks. First 4 weeks: 15 min sessions. Weeks 5‐12: 30 min
Santos 2009	Unclear	10‐100 mA	1 ms	50	Unclear	Intravaginal: electrode: 10cm long, 3.5 cm wide with double metallic ring and cylindrical shape, positioned in medium third of the vagina	2 x 20 min sessions per week for 4 months
Schmidt 2009	Unclear	Unclear	300 μs	50	Unclear	Unclear	12 weeks (unclear how many sessions or duration of sessions)
Seo 2004	Unclear	Unclear	Unclear	"Simultaneous electrical stimulation of 35 Hz and 50 Hz for 24 secs"	Unclear	Unclear	2 x 20 min sessions per week for 6 weeks (12 sessions) (plus biofeedback)
Shepherd 1984	Unclear	Up to 40 v	Unclear	10‐50	Unclear	Maximum perineal stimulation: Scott electrode in vagina, large indifferent electrode under buttocks	Single 20 min session
Shepherd 1985	Unclear	Unclear	Unclear	10	Unclear	Intravaginal cushion attached to stimulator worn around waist	Cushion worn for 8/24 hours, day or night according to participant preference
Smith 1996	Biphasic	5 mA ‐ 10 mA, increased each month to 80 mA max (range 1‐100)	Asymmetric balanced biphasic pulse, 300 μs,	Channel 1: 50 Hz; channel 2: 12.5 Hz	5 s contraction time (range 5‐15), duty cycle 1‐2 (range 1 to 1 to 2)	Unclear	16 weeks (unclear how many sessions); increasing treatment time from 15, 30, 45, 60 minutes
Tapp 1987	Unclear	Unclear	Unclear	Unclear	Unclear	Intravaginal. Faradic stimulation using vaginal probe	2 sessions per week for 1 month (session duration not reported)
Tapp 1989	Unclear	Unclear	Unclear	Unclear	Unclear	Intravaginal. Faradic stimulation using vaginal probe	2 sessions per week for 1 month (session duration not reported)
Terlikowski 2013	Unclear	Unclear	200‐250 μs	10‐40	15 s; 30 s	Intravaginal	2 x 20 min sessions per day at home for 8 weeks
Whitmore 1995	Unclear	Unclear	Unclear	Unclear	Unclear	Unclear	Unclear
Wilson 1987	Unclear	According to maximum participant tolerance	Unclear	Unclear	Groups of 12 surges/min, 2 min rest in between each group	Faradism. Surface electrodes. Saddle shaped indifferent electrode placed over the sacrum, active electrode applied to perineum	6 weeks' treatment (session duration not reported)
Wilson 1987	Unclear	20−25 mA	Unclear	Unclear	15 pulses at pressure peak 0.25‐0.30 Pa/cm^b	Interferential. 4 suction electrodes (2 on abdomen, 2 on adductor muscles)	First treatment 10 min, if no ill effects then duration increased to 15 min
Wise 1993	Unclear	0‐90 mA, according to participant tolerance	Unclear	20	Unclear	Intravaginal	1 session per day (at home) for 6 weeks

Study	Description of sham intervention^a
Brubaker 1997	Identical device to the intervention group with disconnected wire so no electricity supplied
Hofbauer 1990	Electrodes placed in the lumbar region
Jeyaseelan 1999	One 250 μ impulse every minute for 60 min (proven to have no physiological effect on muscle)
Laycock 1993b;	Machine modified to bypass the patient circuit and divert the interferential current to a separate circuit within the machine so the participant received no current. Participants told to expect no sensation
Luber 1997	Wiring from the unit to the probe was covertly discontinuous
Sand 1995	Same system as intervention group but limited to maximum output 1 mA
Shepherd 1984	Vaginal electrode but no current
Shepherd 1985	Identical device to intervention group but not activated
Terlikowski 2013	Women were provided with a placebo set to parameters proven to have no physiological effect

Electrical stimulation versus no active treatment
Patient or population: women with stress urinary incontinence  Setting: home and/or hospitals (Brazil, France, Norway, UK)  Intervention: electrical stimulation  Comparison: no active treatment
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect  (95% CI)	№ of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Risk with no active treatment	Risk with electrical stimulation
Cure: number of women with self‐reported continence  Follow‐up: mean 6 months	Study population		RR 2.31  (1.06 to 5.02)	101  (2 RCTs)	⊕⊕⊕⊝  Moderate^a	—
	122 per 1000	283 per 1000  (130 to 615)
Improvement: number of women with self‐reported improvement in SUI (cured or improved)  Follow‐up: range 12 weeks to 9 months	Study population		RR 1.73  (1.41 to 2.11)	347  (5 RCTs)	⊕⊕⊝⊝  Low^b,c	—
	382 per 1000	660 per 1000  (538 to 805)
Incontinence‐specific quality of life (higher score = worse quality of life) assessed with: King's Health Questionnaire, Incontinence Severity Index, ICI‐Q  Follow‐up: median 6 weeks	The mean incontinence‐specific quality of life score was 0.72 standard deviations lower (0.99 lower to 0.45 lower).		—	250  (4 RCTs)	⊕⊕⊕⊝  Moderate^d	A standard deviation of 0.80  represents a large difference  between groups
Adverse effects  Follow‐up: range 6 weeks to 6 months	Study population		RR 5.96  (0.30 to 118.70)	103  (3 RCTs)	⊕⊝⊝⊝  Very low^a,d	2 trials reported no women with adverse effects in either group. 1 trial reported 2 women with adverse effects in the ES group (1 tenderness and bleeding, 1 discomfort)
	0 per 1000	0 per 1000  (0 to 0)
Cost‐effectiveness	Not reported
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; ES: electrical stimulation; ICI‐Q: International Consultation on Incontience Questionnaire; RCT: randomised controlled trial; RR: risk ratio; SUI: stress urinary incontinence.
GRADE Working Group grades of evidence  High quality: we are very confident that the true effect lies close to that of the estimate of the effect.  Moderate quality: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.  Low quality: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect  Very low quality: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Electrical stimulation versus sham treatment
Patient or population: women with stress urinary incontinence  Setting: home and/or hospital (Austria, Denmark, Poland, UK, USA)  Intervention: electrical stimulation  Comparison: sham treatment
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect  (95% CI)	№ of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Risk with sham treatment	Risk with electrical stimulation
Cure: number of women with self‐reported continence  Follow‐up: range 12 weeks to 6 months	Study population		RR 2.21  (0.38 to 12.73)	158  (3 RCTs)	⊕⊝⊝⊝  Very low^a,b	—
	95 per 1000	210 per 1000  (36 to 1000)
Improvement: number of women with self‐reported improvement in SUI (cured or improved)  Follow‐up: range 12 weeks to 6 months	Study population		RR 2.03  (1.02 to 4.07)	236  (5 RCTs)	⊕⊕⊝⊝  Low^a	—
	198 per 1000	402 per 1000  (202 to 805)
Incontinence‐specific quality of life assessed with: IIQ, UDI, I‐QoL  Follow‐up: range 8 weeks to 16 weeks	One trial found significantly better I‐QoL scores in the ES group, while another trial found no evidence of a difference between groups in IIQ or UDI scores.		—	117  (2 RCTs)	⊕⊕⊝⊝  Low^c	—
Adverse effects  Follow‐up: range 12 weeks to 6 months	Study population		RR 2.01  (0.52 to 7.67)	233  (4 RCTs)	⊕⊕⊝⊝  Low^c	2 trials reported no women with adverse effects in either group. 2 trials reported vaginal irritation, bleeding and discomfort in the ES groups.
	23 per 1000	47 per 1000  (12 to 178)
Cost‐effectiveness	Not reported
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; ES: electrical stimulation; IIQ: incontinence impact questionnaire; I‐QoL: Incontincence Quality of Life questionnaire; RCT: randomised controlled trial; RR: risk ratio; SUI: stress urinary incontinence; UDI: urogenital distress inventory.
GRADE Working Group grades of evidence  High quality: we are very confident that the true effect lies close to that of the estimate of the effect.  Moderate quality: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.  Low quality: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect  Very low quality: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Electrical stimulation versus PFMT
Patient or population: women with stress urinary incontinence  Setting: home and/or hospital (Austria, Brazil, France, Germany, Iceland, Norway, Sweden, Turkey, UK, USA)  Intervention: electrical stimulation  Comparison: pelvic floor muscle training (PFMT)
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect  (95% CI)	№ of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Risk with PFMT	Risk with Electrical stimulation
Cure: number of women with self‐reported continence  Follow‐up: median 6 months	Study population		RR 0.51  (0.16 to 1.63)	143  (4 RCTs)	⊕⊕⊝⊝  Low^a,b	—
	507 per 1000	259 per 1000  (81 to 826)
Improvement: number of women with self‐reported improvement in SUI (cured or improved)  Follow‐up: range 3 months to 4 years	Study population		RR 0.85  (0.70 to 1.03)	244  (7 RCTs)	⊕⊕⊝⊝  Low^a,b	—
	669 per 1000	569 per 1000  (469 to 690)
Incontinence‐specific quality of life assessed with: I‐QoL and unvalidated instrument  Follow‐up: range 5 weeks to 6 months	None of the trials found any evidence of a difference between groups.		—	93  (2 RCTs)	⊕⊕⊝⊝  Low^a,c	—
Adverse effects  Follow‐up: range 5 weeks to 6 months	Study population		RR 5.00  (0.25 to 99.16)	121  (3 RCTs)	⊕⊕⊝⊝  Low^d	2 trials reported no women with adverse effects in either group. 1 trial reported 2 women with adverse effects in the ES group (1 tenderness and bleeding, 1 discomfort)
	0 per 1000	0 per 1000  (0 to 0)
Cost‐effectiveness	Not reported
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; ES: electrical stimulation; I‐QoL: Incontincence Quality of Life questionnaire; PFMT: pelvic floor muscle training; RCT: randomised controlled trial; RR: risk ratio; SUI: stress urinary incontinence.
GRADE Working Group grades of evidence  High quality: we are very confident that the true effect lies close to that of the estimate of the effect.  Moderate quality: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.  Low quality: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect  Very low quality: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Electrical stimulation versus vaginal cones
Patient or population: women with stress urinary incontinence  Setting: home and/or hospital (Brazil, Italy, Korea, Norway, UK)  Intervention: electrical stimulation  Comparison: vaginal cones
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect  (95% CI)	№ of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Risk with vaginal cones	Risk with electrical stimulation
Cure: number of women with self‐reported continence  Follow‐up: median 6 months	Study population		RR 1.04 (0.70 to 1.54)	157  (3 RCTs)	⊕⊕⊝⊝  Low^a,b	—
	363 per 1000	454 per 1000  (341 to 606)
Subjective cure or improvement  Follow‐up: range 4 weeks to 6 months	Study population		RR 1.09 (0.97 to 1.21)	331  (5 RCTs)	⊕⊕⊝⊝  Low^b,c	—
	685 per 1000	768 per 1000  (678 to 863)
Incontinence‐specific quality of life  assessed with: I‐QoL (range of possible scores: 0‐100)  Follow‐up: range 4 months to 6 months	—	MD 1.59 higher  (3.72 lower to 6.9 higher)	—	96  (2 RCTs)	⊕⊕⊝⊝  Low^c,d	Minimum clinically important difference between treatments is 2.5 points
Adverse effects  Follow‐up: mean 6 months	Study population		RR 0.54  (0.11 to 2.70)	52  (1 RCT)	⊕⊕⊝⊝  Low^e,f	Adverse effects in the ES group: tenderness and bleeding discomfort Adverse effects in the vaginal cones group: abdominal pain, vaginitis bleeding
	148 per 1000	80 per 1000  (16 to 400)
Cost‐effectiveness	Not reported
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; ES: electrical stimulation; I‐QoL: Incontincence Quality of Life questionnaire; RCT: randomised controlled trial; RR: risk ratio.
GRADE Working Group grades of evidence  High quality: we are very confident that the true effect lies close to that of the estimate of the effect.  Moderate quality: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.  Low quality: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect  Very low quality: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Electrical stimulation versus PFMT plus vaginal cones
Patient or population: women with stress urinary incontinence  Setting: home (UK)  Intervention: electrical stimulation  Comparison: PFMT plus vaginal cones
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect  (95% CI)	№ of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Risk with PFMT plus vaginal cones	Risk with electrical stimulation
Cure: number of women with self‐reported continence  Follow‐up: mean 6 weeks	Study population		RR 1.45 (0.96 to 2.20)	123  (2 RCTs)	⊕⊝⊝⊝  Very low^a,b,c	—
	333 per 1000	483 per 1000  (320 to 733)
Subjective cure or improvement  Follow‐up: mean 6 weeks	Study population		RR 1.53 (1.08 to 2.18)	123  (2 RCTs)	⊕⊝⊝⊝  Very low^a,b,c	—
	426 per 1000	652 per 1000  (460 to 929)
Incontinence‐specific quality of life	Not reported
Adverse effects	Not reported
Cost‐effectiveness	Not reported
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; RCT: randomised controlled trial; RR: risk ratio.
GRADE Working Group grades of evidence  High quality: we are very confident that the true effect lies close to that of the estimate of the effect.  Moderate quality: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.  Low quality: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect  Very low quality: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Electrical stimulation versus drug therapy
Patient or population: women with stress urinary incontinence  Setting: home and hospital (UK)  Intervention: electrical stimulation  Comparison: drug therapy
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect  (95% CI)	№ of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Risk with drug therapy	Risk with electrical stimulation
Cure: number of women with self‐reported continence	Not reported
Subjective cure or improvement  Follow‐up: mean 9 months	Study population		RR 13.89  (0.84 to 230.82)	50  (1 RCT)	⊕⊝⊝⊝  Very low^a,b	—
	0 per 1000	0 per 1000  (0 to 0)
Incontinence‐specific quality of life	Not reported
Adverse effects	Not reported
Cost‐effectiveness	Not reported
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; RCT: randomised controlled trial; RR: risk ratio.
GRADE Working Group grades of evidence  High quality: we are very confident that the true effect lies close to that of the estimate of the effect.  Moderate quality: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.  Low quality: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect  Very low quality: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Electrical stimulation plus PFMT versus PFMT
Patient or population: women with stress urinary incontinence  Setting: home and/or hospital (Australia, Brazil, UK, USA)  Intervention: electrical stimulation plus PFMT  Comparison: PFMT
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect  (95% CI)	№ of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Risk with PFMT	Risk with Electrical stimulation plus PFMT
Cure: number of women with self‐reported continence  Follow‐up: range 9 weeks to 9 months	Study population		RR 0.76  (0.38 to 1.52)	99  (3 RCTs)	⊕⊕⊝⊝  Low^a,b	—
	240 per 1000	182 per 1000  (91 to 365)
Improvement: number of women with self‐reported improvement in SUI (cured or improved)  Follow‐up: range 6 weeks to 9 months	Study population		RR 1.10  (0.95 to 1.28)	308  (6 RCTs)	⊕⊕⊝⊝  Low^a,c	—
	639 per 1000	703 per 1000  (607 to 818)
Incontinence‐specific quality of life (higher score = worse quality of life)  assessed with: King's Health Questionnaire, IIQ−7, VAS for perceived effect of SUI on QoL  Follow‐up: range 4 weeks to 6 months	The mean incontinence‐specific quality of life score was 0.35 standard deviations lower (0.64 lower to 0.05 lower)		—	193  (4 RCTs)	⊕⊝⊝⊝  Very low^c,d	A standard deviation of 0.20 represents a small difference between groups. 2 other trials found no evidence of a difference between groups (data unsuitable for meta‐analysis)
Adverse effects  Follow‐up: mean 18 months	1 trial reported 4/59 women in the ES + PFMT group with adverse effects (vaginal irritation) but did not report any data for the PFMT‐only group.			133  (1 RCT)	⊕⊕⊝⊝  Low^e,f	—
Cost‐effectiveness	Not reported
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; ES: electrical stimulation; IIQ: incontinence impact questionnaire; MD: mean difference; PFMT: pelvic floor muscle training; QoL: quality of life; RR: risk ratio; SUI: stress urinary incontinence; VAS: visual analogue scale.
GRADE Working Group grades of evidence  High quality: we are very confident that the true effect lies close to that of the estimate of the effect.  Moderate quality: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.  Low quality: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect  Very low quality: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Electrical stimulation plus surgery versus surgery
Patient or population: women with stress urinary incontinence  Setting: hospital (China)  Intervention: electrical stimulation plus surgery  Comparison: surgery
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect  (95% CI)	№ of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Risk with surgery	Risk with electrical stimulation plus surgery
Cure: number of women with self‐reported continence  Follow‐up: mean 18 months	Study population		RR 1.19  (0.53 to 2.67)	120  (1 RCT)	⊕⊝⊝⊝  Very low^a,b	—
	750 per 1000	893 per 1000  (398 to 1000)
Improvement: number of women with self‐reported improvement in SUI (cured or improved)  Follow‐up: mean 18 months	Study population		RR 5.36  (0.61 to 47.36)	120  (1 RCT)	⊕⊝⊝⊝  Very low^a,b	—
	917 per 1000	1000 per 1000  (559 to 1000)
Incontinence‐specific quality of life assessed with: I‐QoL and ICIQ‐SF  Follow‐up: mean 18 months	Both I‐QoL and ICIQ‐SF scores suggested higher QoL when ES was added to surgery.		—	120  (1 RCT)	⊕⊕⊝⊝  Low^a,c	—
Adverse effects  Follow‐up: mean 18 months	Study population		RR 1.00  (0.37 to 2.72)	120  (1 RCT)	⊕⊝⊝⊝  Very low^a,b	3 women in each group reported medial thigh pain.
	150 per 1000	150 per 1000  (56 to 408)
Cost‐effectiveness	Not reported
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; ES: electrical stimulation; ICIQ‐SF: International Consultation on Incontinence Questionnaire ‐ Short Form; I‐QoL: Incontincence Quality of Life questionnaire; PFMT: pelvic floor muscle training; RR: risk ratio; SUI: stress urinary incontinence.
GRADE Working Group grades of evidence  High quality: we are very confident that the true effect lies close to that of the estimate of the effect.  Moderate quality: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.  Low quality: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect  Very low quality: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Surface ES versus intravaginal ES
Patient or population: women with stress urinary incontinence  Setting: hospital (Brazil)  Intervention: surface ES  Comparison: intravaginal ES
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect  (95% CI)	№ of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Risk with intravaginal ES	Risk with surface ES
Cure: number of women with self‐reported continence	Not reported
Improvement: number of women with self‐reported improvement in SUI (cured or improved)	Not reported
Incontinence‐specific quality of life  assessed with: King's Health Questionnaire (range of possible scores: 0‐100)  Follow‐up: mean 6 weeks	—	MD 2.9 points higher  (3.24 lower to 9.04 higher)	—	30  (1 RCT)	⊕⊕⊝⊝  Low^a	Lower score suggests greater quality of life in the intravaginal group
Adverse effects	Not reported
Cost‐effectiveness	Not reported
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; ES: electrical stimulation; MD: mean difference; PFMT: pelvic floor muscle training; SUI: stress urinary incontinence.
GRADE Working Group grades of evidence  High quality: we are very confident that the true effect lies close to that of the estimate of the effect.  Moderate quality: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.  Low quality: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect  Very low quality: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Study	Outcome	ES: mean (SD), N or n/N	No active treatment: mean (SD), N or n/N	Result
Primary outcomes
Castro 2008	I‐QoL score^a	83.4 (12.1), 27	57.6 (28.2), 24	Favours ES MD 25.80 (95% CI 13.63 to 37.97)
Castro 2008	Women with significant improvement in QoL	9/27	0/24	Favours ES RR 16.96 (95% CI 1.04 to 276.81)
Secondary outcomes
Bø 1999	Women requesting further treatment in addition to the allocated intervention	19/25	28/30	RR 0.23 (95% CI 0.04 to 1.24)
Bø 1999	Incontinence episodes per 24 hours	0.3 (0.8), 27	1.3 (0.9), 24	Favours ES MD −1.00 (95% CI 1.47 to −0.53)
Tertiary outcomes
Bø 1999	Objective cure or improvement	7/25	2/30	Favours ES Pooled RR 2.41 (95% CI 1.02 to 5.68)
Castro 2008	Objective cure or improvement	11/27	3/12	Favours ES Pooled RR 2.41 (95% CI 1.02 to 5.68)
Bø 1999	Pelvic floor muscle strength (cmH₂O)	18.6 (13.52^b), 25	16.0 (8.38^b), 30	MD 2.60 (95% CI −3.49 to 8.69) MD 4.73 (95% CI −3.92 to 13.38) Pooled MD 3.31 (95% CI −1.67 to 8.28)
Pereira 2012	Pelvic floor muscle strength (cmH₂O)	14.57 (11.55), 7	9.84 (1.71), 7
Castro 2008	Women with negative urodynamic stress test	11/27	3/12	RR 2.79 (95% CI 0.62 to 12.60)
Castro 2008	Pelvic floor muscle strength measured by PERFECT^c	2.9 (1.00), 27	2.3 (1.07), 24	Favours ES Pooled MD 0.84 (95% CI 0.55 to 1.14) MD 0.60 (95% CI 0.03 to 1.17)
Pereira 2012		1.71 (0.95), 7	1.14 (0.37), 7	MD 0.57 (95% CI −0.19 to 1.33)
Correia 2014		Surface ES: 2.53 (0.83), 15	2.25 (0.86), 15	MD 0.28 (95% CI −0.32 to 0.88)
Correia 2014		Intravaginal ES: 2.66 (0.81), 15	1.14 (0.37), 7	MD 0.41 (95% CI −0.19 to 1.01)
Henalla 1989	Maximum urethral closure pressure (cmH₂O)	26 (30.0^b), 25	20 (19.60^b), 24	MD 6.00 (95% CI −8.13, 20.13)

Study	Outcome	ES: mean (SD), N or n/N unless otherwise stated	Sham: mean (SD), N or n/N unless otherwise stated	Result
Primary outcomes
Jeyaseelan 1999	IIQ score^a	28.42 (17.22), 12	30.11 (17.94), 12	MD −1.69 (95% CI −15.76 to 12.38)
Jeyaseelan 1999	UDI score^a	34.45 (25.25), 12	38.32 (11.75), 12	MD −3.87 (95% CI −19.63 to 11.89)
Terlikowski 2013	I‐QoL score^b	80.8 (24.1), 64	50.6 (14.9), 29	Favours ES MD 30.20 (95% CI 22.18 to 38.22)
Sand 1995	Subjective assessment of SUI severity (VAS)	4.8 (NR), 28	6.3 (NR), 16	Not estimable
Laycock 1993b	Subjective assessment of SUI severity (VAS)	2.9 (NR), 15	3.5 (NR), 11	Not estimable
Whitmore 1995	Subjective assessment of improvement in SUI (VAS)	NR	NR	Significantly greater improvement in group A (ES) than group B (sham ES)
Whitmore 1995	Subjective assessment of improvement in frequency of urine loss (VAS)	NR	NR
Secondary outcomes
Jeyaseelan 1999	Incontinence episodes per week	Median (range), N: 0 (0 to 5), 7	Median (range), N: 3 (0 to 17), 10	Not estimable
Jeyaseelan 1999	1 hour pad test (g)	Median (range), N: 5.0 (1 to 91.9), 12	Median (range), N: 5.2 (0 to 75.0), 12	Not estimable
Tertiary outcomes
Laycock 1993b	Objective cure or improvement	11/13	5/9	Favours ES Pooled RR 3.32 (95% CI 1.89 to 5.84)
Luber 1997		3/20	3/24
Preisinger 1990		3/11	0/10
Terlikowski 2013		25/64	0/29
Whitmore 1995		22/35	3/17
Jeyaseelan 1999	PFM strength (cmH₂O)	24 (13), 12	19 (6), 12	MD −5.00 (95% CI −3.46 to 13.46)
Sand 1995	Vaginal muscle strength (mmHg)	15.5 (13.49^c), 28	8.9 (5.75^c), 16	Favours ES MD −6.60 (−12.34 to −0.86)
Terlikowski 2013	Oxford score^b	8 weeks: 4.2 (NR), 64 16 weeks: 4.1 (NR) 64	8 weeks: 2.6 (NR), 29 16 weeks: 2.7 (NR), 29	Favours ES

Study	Outcome	ES: mean (SD), N or n/N unless otherwise stated	PFMT: mean (SD), N or n/N unless otherwise stated	Result
Primary outcomes
Castro 2008	I‐QoL score^a	83.4 (12.1), 27	82.2 (17.6), 26	MD 1.20 (95% CI (−6.96 to 9.36)
Castro 2008	Women with significant improvement in QoL	9/27	7/26	RR 1.24 (95% CI 0.54 to 2.83)
Demirturk 2008	Quality of life questionnaire score (non‐validated instrument)^b	22.5 (17.11^c), 20	13.5 (11.41^c), 20	MD 9.00 (95% CI −0.01 to 18.01)
Pohl 2004	Subjective assessment of SUI severity (VAS)	4.81 (NR), 21	5.33 (NR), 10	Not estimable
Secondary outcomes
Castro 2008	Incontinence episodes per 24 hours	0.3 (0.8), 27	0.4 (0.5), 26	MD −0.10 (95% CI −0.46 to 0.26)
Smith 1996	Incontinence episodes per 24 hours	Mean (range), N: 1.4 (0 to 5), 9	Mean (range), N: 2.4 (0 to 6), 9	Not estimable
Bø 1999	Women requesting further treatment in addition to the allocated intervention	19/25	4/25	Favours PFMT RR 16.63 (95CI 4.06 to 68.04)
Smith 1996	Women going on to have surgery	2/9	3/9	RR 0.57 (95% CI 0.07 to 4.64)
Smith 1996	Pads per week	Mean (range), N: 4.0 (0−10), 9	Mean (range), N: 5.4 (0−10), 9	Not estimable
Pohl 2004	Pad test (g) (timescale not reported)	6.21 (NR), 21	10.00 (NR), 10	Not estimable
Tertiary outcomes
Bernardes 2000	Objective cure or improvement	2/7	5/7	Pooled RR 1.18 (95% CI 0.68 to 2.03
Bø 1999		7/25	1/25
Hahn 1991		4/10	1/10
Preisinger 1990		3/11	7/11
Smith 1996		4/9	3/9
Bø 1999	PFM strength (cmH₂O)	18.6 (13.52^c), 25	19.2 (9.95^c), 25	MD −0.60 (95% CI −7.18 to 5.98)
Bernardes 2000	Perineal contraction strength (0−5 scale; higher score = stronger contraction)	1 (0.82), 7	8 (1.83), 7	Favours PFMT MD −2.00 (95% CI −3.49, −0.51)
Castro 2008	Women with negative urodynamic stress test	11/27	10/26	RR 1.49 (95% CI 0.50 to 4.43)
Castro 2008	Oxford score^d	2.9 (1.0), 27	3.6 (0.7), 26	Favours PFMT MD −0.70 (95% CI −1.16 to −0.24)
Pohl 2004		2.55 (NR), 21	2.7 (NR), 10	Not estimable
Jeyaseelan 2002		Median (range): 13 (0−14)	Median (range): 11 (0−83	Not estimable
Henalla 1989	Maximum urethral closure pressure (cmH₂O)	26 (30.0^c), 25	32 (20.40^c), 26	MD −6.00 (95% CI −20.13, 8.13)
Preisinger 1990	Maximum urethral closure pressure (mmHg)	42.6 (8.2), 11	41.4 (14.3), 11	MD 1.20 (95% CI −8.54 to 10.94)

Study	Outcome	ES: mean (SD), N or n/N	Vaginal cones: mean (SD), N or n/N	Result
Primary outcomes
Castro 2008	Women with significant improvement in QoL	9/27	7/24	RR 1.14 (95% CI 0.50 to 2.60)
Delneri 2000	Subjective assessment of SUI severity (10 point VAS)	5/10	5/10	RR 1.00 (95% CI 0.17 to 5.77)
Secondary outcomes
Olah 1990	Women requiring continence surgery	2/30	3/24	RR 0.50 (95% CI 0.08 to 3.27)
Bø 1999	Women requesting further treatment in addition to the allocated intervention	19/25	23/27	RR 0.55 (95% CI 0.14 to 2.24)
Bø 1999	Adverse effects	2/25	4/27	RR 0.54 (95% CI 0.11 to 2.70)
Olah 1990	No leakage at 6 months	11/28	10/19	RR 0.58 (95% CI 0.18 to 1.89)
Olah 1990	Weekly leakage (g)	5.3 (9.2), 30	3.9 (9.4), 24	MD 1.40 (95% CI −3.60 to 6.40)
Tertiary outcomes
Bø 1999	Objective cure	7/25	4/27	RR 1.89 (95% CI 0.63 to 5.69)
Bø 1999	Objective cure or improvement	18/30	20/24	Pooled RR 0.93 (95% CI 0.72 to 1.20)
Bridges 1988		7/25	4/27
Wise 1993		15/30	16/21
Bø 1999	PFM strength (cmH₂O)	18.6 (12.8^a), 25	15.4 (11.40^a), 27	MD 3.20 (95% CI −3.62 to 10.02)
Castro 2008	Women with negative urodynamic stress test	11/27	9/24	RR 1.55 (95% CI 0.51 to 4.74)
Castro 2008	Oxford score^b	2.9 (1.0), 27	3.0 (0.8), 24	MD −0.10 (95% CI −0.59 to 0.39)
Seo 2004	Maximum vaginal pressure (mmHg)	33.64 (16.72), 60	27.20 (13.21), 60	Favours ES plus PFMT MD 6.44 (95% CI 1.05 to 11.83)
Seo 2004	Maximum urethral pressure (mmH₂O)	77.93 (30.96), 60	78.38 (18.30), 60	MD −0.45 (95% CI −9.55 to 8.65)

Low intensity ES plus PFMT versus maximal intensity ES plus PFMT
Patient or population: women with stress urinary incontinence  Setting: home and hospital (UK)  Intervention: low intensity ES plus PFMT  Comparison: maximal intensity ES plus PFMT
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect  (95% CI)	№ of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Risk with maximal intensity ES plus PFMT	Risk with low intensity ES plus PFMT
Cure: number of women with self‐reported continence	Not reported
Improvement: number of women with self‐reported improvement in SUI (cured or improved)  Follow‐up: mean 12 months	Study population		RR 0.28  (0.09 to 0.91)	49  (1 RCT)	⊕⊕⊝⊝  Low^a	—
	667 per 1000	187 per 1000  (60 to 607)
Incontinence‐specific quality of life	Not reported
Adverse effects ‐ not reported	Not reported
Cost‐effectiveness ‐ not reported	Not reported
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; ES: electrical stimulation; MD: mean difference; PFMT: pelvic floor muscle training; SUI: stress urinary incontinence.
GRADE Working Group grades of evidence  High quality: we are very confident that the true effect lies close to that of the estimate of the effect.  Moderate quality: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.  Low quality: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect  Very low quality: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Conventional ES plus PFMT versus dynamic ES plus PFMT
Patient or population: women with stress urinary incontinence  Setting: home (Germany)  Intervention: conventional ES plus PFMT  Comparison: dynamic ES plus PFMT
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect  (95% CI)	№ of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Risk with dynamic ES plus PFMT	Risk with conventional ES plus PFMT
Cure: number of women with self‐reported continence	Not reported
Improvement: number of women with self‐reported improvement in SUI (cured or improved) (perception of bother of UI symptoms)  assessed with: change in VAS  Scale from: 0 to 10  Follow‐up: mean 12 weeks	—	MD 0.7 higher  (0.83 lower to 2.23 higher)	—	61  (1 RCT)	⊕⊝⊝⊝  Very low^a,b,c	—
Incontinence‐specific quality of life  assessed with: change in King's Health Questionnaire scores (range of possible scores: 0‐100)  Follow‐up: mean 12 weeks	—	MD 4.1 points higher  (1.43 higher to 6.77 higher)	—	61  (1 RCT)	⊕⊝⊝⊝  Very low^a,b,c	Scores indicate greater quality of life in the dynamic ES group
Adverse effects ‐ not reported	Not reported
Cost‐effectiveness ‐ not reported	Not reported
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).  CI: confidence interval; ES: electrical stimulation; MD: mean difference; PFMT: pelvic floor muscle training; SUI: stress urinary incontinence; UI: urinary incontinence; VAS: visual analogue scale.
GRADE Working Group grades of evidence  High quality: we are very confident that the true effect lies close to that of the estimate of the effect.  Moderate quality: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.  Low quality: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect  Very low quality: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Study	Outcome	ES plus PFMT: mean (SD), N or n/N unless otherwise stated	PFMT: mean (SD), N or n/N unless otherwise stated	Result
Primary outcomes
Firra 2013	York Incontinence Perception Scale score^a	46.4 (7.2), 9	44.8 (6.3), 12	MD 0.23 (95% CI −0.64 to 1.10)
Huebner 2011	Change in perception of bother of UI symptoms (1‐10 VAS^b)	Conventional ES: −2.2 (3.2), 33	−2.5 (2.1), 27	MD 0.30 (95% CI −1.18 to 1.78)
	Change in perception of bother of UI symptoms (1‐10 VAS^b)	Dynamic ES: −2.9 (2.9), 28	−2.5 (2.1), 27	MD −0.40 (95% CI −1.73 to 0.93)
	Change in King's Health Questionnaire score^b	Conventional ES: −20.7 (5.3), 33	−20.2 (5.4), 27	MD −0.50 (95% CI −3.22 to 2.22)
	Change in King's Health Questionnaire score^b	Dynamic ES: −24.8 (5.3), 28	−20.2 (5.4), 27	Favours ES MD −4.60 (95% CI −7.43 to −1.77)
Jeyaseelan 2003	% change in UDI score	Median (range), N: −32 (−50 to −18), 6	Median (range), N: 0 (−43 to 180), 7	Not estimable
Jeyaseelan 2003	% change in IIQ score	Median (range), N: −27 (−63 to 0), 6	Median (range), N: 0 (−67 to 200), 7	Not estimable
Secondary outcomes
Goode 2003	Women 'somewhat' or 'completely' satisfied	46/47	46/47	RR 1.00 (95% CI 0.06 to 16.47)
Goode 2003	Women whose incontinence no longer restricts activities	37/47	33/47	RR 1.57 (95% CI 0.61 to 4.01)
Knight 1998	Subjective assessment of symptoms (1‐5 scale, higher score = better)	Low intensity ES at home plus PFMT: Mean (range), N: 3.3 (1 to 5), 19	Mean (range), N: 3.5 (2−5), 18	Not estimable
Knight 1998		Maximal ES in clinic plus PFMT: mean (range), N: 105.6 (−55.9 to 3.9 (3−5), 20	Mean (range), N: 3.5 (2−5), 18	Not estimable
Schmidt 2009	Incontinence episodes	Median (IQR), N: 12 weeks: 0 (0, 1), 11 6 months: 0.5 (0, 1.25), 11	Median (IQR), N: 12 weeks: 2 (0, 3), 11 6 months: 0 (0, 5.25), 11	Not estimable
Schmidt 2009	Daytime micturitions	Median (IQR), N: 12 weeks: 5 (5, 6), 11 6 months: 4.5 (4, 6), 11	Median (IQR), N: 12 weeks: 7 (5, 10), 11 6 months: 2 (1, 3), 11	Not estimable
Goode 2003	Women with adverse effects	4/59	Unclear	Not estimable
Bidmead 2002	Pad test (g)	4.3 (NR), 88	4.2 (NR), 40
Knight 1998		Low intensity ES plus PFMT: Median (range), N: 2.9 (0.0‐50.9), 19	Median (range), N: 0.8 (0.0‐88.1), 18
Knight 1998		Maximal intensity ES plus PFMT: Median (range), N: 1.5 (0.0‐28.1), 20	Median (range), N: 0.8 (0.0‐88.1), 18
Tertiary outcomes
Knight 1998	Objective cure or improvement	Low intensity ES plus PFMT: 10/25	13/21	RR 0.41 (95% CI 0.12 to 1.35)
Knight 1998	Objective cure or improvement	Maximal intensity ES plus PFMT: 16/24	13/21	RR 1.23 (95% CI 0.36 to 4.18)
Firra 2013	PFM strength (cmH₂O)	36.7 (14.1), 9	32.5 (18.5), 12	MD 0.24 (95% CI −0.63 to 1.11) MD −7.03 (95% CI −26.20 to 12.14) Pooled MD −0.04 (95% CI −0.64 to 0.57)
Schmidt 2009	PFM strength (cmH₂O)	41.85 (26.1), 11	48.88 (19.25), 11
Eyjolfsdottir 2009	Oxford score^c	4.1 (0.9), 12	3.8 (1.4), 12	MD 0.25 (95% CI −0.56 to 1.05) MD 0.48 (95% CI −0.03 to 1.00] MD 0.42 (95% CO −0.12 to 0.96) Pooled MD 0.39 (95% CI −0.01 to 0.79)
Huebner 2011		Conventional ES plus PFMT: 1.9 (0.9), 33	1.5 (0.7), 27
Huebner 2011		Dynamic ES plus PFMT: 1.8 (0.7), 27	1.5 (0.7), 27
Preisinger 1990	Maximum urethral closure pressure (mmHg)	49.9 (12.0), 11	41.4 (14.3), 11	MD 8.50 (95% CI −2.53 to 19.53)
Wilson 1987	Maximum urethral closure pressure at rest (cmH₂O)	ES (faradism): 58.0 (15.4), 14	51.5 (10.5), 9	MD 6.50 (95% CI −4.09 to 17.09)
Wilson 1987	Maximum urethral closure pressure at rest (cmH₂O)	ES (interferential) 56.0 (16.7), 12	51.5 (10.5), 9	MD 4.50 (95% CI −7.18 to16.18)
Beuttenmuller 2010	Contraction of pelvic floor at rest	30.79 (7.44), 25	32.28 (7.33), 15	MD −1.49 (95% CI −6.21 to 3.23)

Study	Outcome	ES plus surgery: mean (SD), N or n/N	Surgery: mean (SD), N or n/N	Result
Primary outcomes
Min 2015	Subjective cure	43/60	45/60	RR 1.19 (95% CI 0.53 to 2.67)
	Subjective cure or improvement	59/60	55/60	RR 5.36 (95% CI 0.61 to 47.36)
	I‐QoL score^a	96.0 (15.2), 60	89.0 (11.2), 60	Favours ES plus surgery MD 7.00 (95% CI 2.22 to 11.78)
	ICIQ‐SF score^b	2.0 (2.5), 60	5.0 (3.1), 60	Favours ES plus surgery MD −3.00 (95% CI −4.01 to −1.99)
Secondary outcomes
Min 2015	Incontinence episodes per 24 hours	^c4.3 (1), 60	^c3 (1.3), 60	Favours ES plus surgery MD −0.70 (95% CI −1.19 to −0.21)
	Micturitions per 24 hours	^c7.67 (1.67), 60	^c8 (1.3), 60	MD −0.33 (95% CI −0.87 to 0.21)
	Pad test (g/h)	1.8 (1.21), 60	3.0 (1.08), 60	Favours ES plus surgery MD −1.20 (95% CI −1.61 to −0.79)
	Adverse effects: Urgency Dysuria Medial thigh pain	^d9/60 3/60 3/60 3/60	^d9/60 4/60 2/60 3/60	RR 1.00 (95% CI 0.37 to 2.72)

Study	Outcome	Surface ES: mean (SD), N or n/N	Intravaginal ES: mean (SD), N or n/N	Result
Primary outcomes
Correia 2014	King's Health Questionnaire Incontinence impact score^a	6.66 (13.80), 15	4.44 (11.73), 15	MD 2.22 (95% CI −6.95, 11.39)
Secondary outcomes
Correia 2014	1 hour pad test (g)	3.31 (12.10), 15	0.41 (0.78), 15	MD 2.90 (95% CI −3.24, 9.04)
Tertiary outcomes
Correia 2014	Pelvic floor muscle strength measured by PERFECT^b	2.53 (0.83), 15	2.66 (0.81), 15	MD −0.13 (95% CI −0.72, 0.46)

Study	Outcome	Low intensity ES plus PFMT: mean (SD), N or n/N, unless otherwise stated	Maximal intensity ES plus PFMT: mean (SD), N or n/N, unless otherwise stated	Result
Primary outcomes
Knight 1998	Subjective cure or improvement	9/25	16/24	Favours maximal intensity RR 0.28 (95% CI 0.09 to 0.91)
Secondary outcomes
Knight 1998	Subjective assessment of symptoms (1‐5 scale, higher score = better)	Median (range), N: 3.3 (1 to 5), 19	Median (range), N: 3.9 (3 to 5), 20	Not estimable
Knight 1998	Pad test (g)	Median (range), N: 2.9 (0.0‐50.9), 19)	Median (range), N: 1.5 (0.0 to 28.1), 20	Not estimable
Tertiary outcomes
Knight 1998	Objective cure	6/19	11/20	RR 0.38 (95% CI 0.10 to 1.40)
Knight 1998	Objective cure or improvement	10/25	16/24	RR 0.33 (95% CI 0.10 to 1.07)

Study	Outcome	Low frequency ES: mean (SD), N or n/N	Medium frequency ES: mean (SD), N or n/N	Result
Primary outcomes
Alves 2011	SUI‐related discomfort (10 cm VAS)	0.5 (0.4), 10	0.6 (0.7), 10	MD −0.10 (95% CI −0.60 to 0.40)
Secondary outcomes
Alves 2011	1 hour pad test (g)	1.2 (NR), 10	1 (NR), 10	Not estimable
Tertiary outcomes
Alves 2011	Objective cure (Laycock and Green criteria)	6 months: 4/10 12 months: 10/10	6 months: 9/10 12 months: 10/10	6 months: RR 0.07 (95% CI 0.01 to 0.84) 10 months: not estimable
Alves 2011	Perineal pressure (mmHg)	9.82 (2.87), 10	8.59 (5.47), 10	MD 1.23 (95% CI −2.60, 5.06)

Study	Outcome	ES (faradism) plus PFMT: mean (SD), N or n/N	ES (interferential) plus PFMT: mean (SD), N or n/N	Result
Primary outcomes
Wilson 1987	Subjective cure or improvement	10/15	9/15	RR 1.33 (95% CI 0.30 to 5.91)
Secondary outcomes
Wilson 1987	Micturitions per 24 hours	7.8 (2.0), 15	8.0 (1.8), 15	MD (95% CI −0.20 −1.56 to 1.16)
Wilson 1987	Pads per 24 hours	1.3 (1.4), 15	1.6 (2.3), 15	MD −0.30 (95% CI −1.66 to 1.06)
Tertiary outcomes
Wilson 1987	Maximum urethral closure pressure (cmH₂O)	58.0 (15.4), 14	56.0 (16.7), 12	MD 2.00 (95% CI −10.42 to 14.42)

Study	Outcome	Conventional ES: mean (SD), N or n/N	Dynamic ES: mean (SD), N or n/N	Result
Primary outcomes
Huebner 2011	Change in perception of bother of UI symptoms (VAS)	−2.2 (3.2), 33	−2.9 (2.9), 28	MD 0.70 (95% CI −0.83 to 2.23)
Huebner 2011	Change in King's Health Questionnaire score	−20.7 (5.3), 33	−24.8 (5.3), 28	Favours dynamic ES MD 4.10 (95% CI 1.43 to 6.77)
Tertiary outcomes
Huebner 2011	Oxford score^a	1.9 (0.9), 33	1.8 (0.7), 27	MD 0.10 (95% CI −0.30 to 0.50)

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Subjective cure	2	101	Risk Ratio (M‐H, Fixed, 95% CI)	2.31 [1.06, 5.02]
2 Subjective cure or improvement	5	347	Risk Ratio (M‐H, Fixed, 95% CI)	1.73 [1.41, 2.11]
3 Quality of life (higher score = worse quality of life)	4	250	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.72 [‐0.99, ‐0.45]
4 Pad test (g)	3	110	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.71 [‐1.11, ‐0.31]
5 Adverse effects	3	103	Risk Ratio (M‐H, Fixed, 95% CI)	5.96 [0.30, 118.70]

PERMALINK

Electrical stimulation with non‐implanted devices for stress urinary incontinence in women

Fiona Stewart

Bary Berghmans

Kari Bø

Cathryn MA Glazener

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

Plain language summary

Summary of findings

Background

Description of the condition

Description of the intervention

How the intervention might work

Why it is important to do this review

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

1. Description of electrical stimulation interventions.

Types of outcome measures

Primary outcomes

Secondary outcomes

Tertiary outcomes

Search methods for identification of studies

Electronic searches

Economic data searches

Searching other resources

Data collection and analysis

Selection of studies

Data extraction and management

Assessment of risk of bias in included studies

Measures of treatment effect

Unit of analysis issues

Dealing with missing data

Assessment of heterogeneity

Assessment of reporting biases

Data synthesis

Subgroup analysis and investigation of heterogeneity

Sensitivity analysis

'Summary of findings' table

Results

Description of studies

Results of the search

1.

Included studies

Design

Sample size

Setting

Participants

Interventions

2. Description of sham electrical stimulation interventions.

Follow‐up

Outcomes

Excluded studies

Ongoing studies

Risk of bias in included studies

2.

3.

Allocation

Random sequence generation

Allocation concealment

Blinding

Blinding of participants and personnel

Blinding of outcome assessment

Incomplete outcome data

Selective reporting

Other potential sources of bias

Effects of interventions

Summary of findings for the main comparison. Electrical stimulation versus no active treatment.

Summary of findings 2. Electrical stimulation versus sham treatment.

Summary of findings 3. Electrical stimulation versus pelvic floor muscle training.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Subjective cure	3	158	Risk Ratio (M‐H, Random, 95% CI)	2.21 [0.38, 12.73]
2 Subjective cure or improvement	5	236	Risk Ratio (M‐H, Random, 95% CI)	2.03 [1.02, 4.07]
3 Number of incontinence episodes per 24 h	3	181	Mean Difference (IV, Fixed, 95% CI)	‐1.34 [‐2.02, ‐0.66]
4 Number of micturitions per 24 h	3	163	Mean Difference (IV, Fixed, 95% CI)	‐0.46 [‐1.38, 0.46]
5 Number of pads per week	2	97	Mean Difference (IV, Fixed, 95% CI)	‐0.78 [‐1.23, ‐0.33]
6 Pad test (g)	2	137	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.89 [‐1.27, ‐0.52]
7 Adverse effects	4	233	Risk Ratio (M‐H, Fixed, 95% CI)	2.01 [0.52, 7.67]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Subjective cure	4	143	Risk Ratio (M‐H, Random, 95% CI)	0.51 [0.16, 1.63]
2 Subjective cure or improvement	7	244	Risk Ratio (M‐H, Fixed, 95% CI)	0.85 [0.70, 1.03]
3 Adverse effects	3	121	Risk Ratio (M‐H, Fixed, 95% CI)	5.0 [0.25, 99.16]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Subjective cure	3	157	Risk Ratio (M‐H, Fixed, 95% CI)	1.04 [0.70, 1.54]
2 Subjective cure or improvement	5	331	Risk Ratio (M‐H, Fixed, 95% CI)	1.09 [0.97, 1.21]
3 Quality of life (I‐QoL)	2	96	Mean Difference (IV, Fixed, 95% CI)	1.59 [‐3.72, 6.90]
4 Number of incontinence episodes per 24 h	2	96	Mean Difference (IV, Fixed, 95% CI)	0.10 [‐0.13, 0.33]
5 Pad test (g)	4	239	Std. Mean Difference (IV, Fixed, 95% CI)	0.06 [‐0.20, 0.31]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Subjective cure	2	123	Risk Ratio (M‐H, Fixed, 95% CI)	1.45 [0.96, 2.20]
2 Subjective cure or improvement	2	123	Risk Ratio (M‐H, Fixed, 95% CI)	1.53 [1.08, 2.18]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Subjective cure	3	99	Risk Ratio (M‐H, Fixed, 95% CI)	0.76 [0.38, 1.52]
2 Subjective cure or improvement	6	308	Risk Ratio (M‐H, Fixed, 95% CI)	1.10 [0.95, 1.28]
3 Quality of life (higher score = worse quality of life)	4	193	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.35 [‐0.64, ‐0.05]
4 Subjective assessment (VAS)	3	150	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.57 [‐0.90, ‐0.24]
5 Women requesting surgery at end of follow‐up	2	82	Risk Ratio (M‐H, Fixed, 95% CI)	0.91 [0.59, 1.41]
6 Number of incontinence episodes per 24 h	4	275	Mean Difference (IV, Fixed, 95% CI)	‐0.33 [‐0.59, ‐0.06]
7 Number of micturitions per 24 h	2	66	Mean Difference (IV, Fixed, 95% CI)	‐0.13 [‐1.46, 1.20]
8 Pad test (g)	4	346	Std. Mean Difference (IV, Random, 95% CI)	‐0.20 [‐0.61, 0.21]

Methods	Study design: RCT Multicentre or single‐centre: not reported Setting: not reported Period: October 1992 to January 1994 Sample size: not reported Follow‐up:: unclear
Participants	N: 47 randomised and analysed Age: 24‐82 years Sex: women Inclusion criteria: genuine stress urinary incontinence or detrusor instability Exclusion criteria: not reported
Interventions	For detrusor overactivity women only (DI) A: probanthine B: electrical stimulation (ES) 2nd RCT in patients with GSUI C: PFMT D: ES Participant numbers unclear
Outcomes	Cure ‐ defined as cessation of incontinence Improvement ‐ defined as reduction in frequency of voids per 24 h by ≥ 50% or ≤ 10 voids per 24 h, or decrease number of pads per 24 h by ≥ 50%
Notes	No usable data
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not reported
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding of participants and personnel (performance bias)   All outcomes	Unclear risk	Not reported
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Not reported
Incomplete outcome data (attrition bias)   All outcomes	Unclear risk	Not reported
Selective reporting (reporting bias)	Unclear risk	Insufficient information
Other bias	Unclear risk	Insufficient information

Methods	Study design: prospective randomised study
Participants	Women with mild genuine stress incontinence (USI) N: 102 randomised and analysed Mean (SD) age: not reported Inclusion criteria: women with SUI. Diagnosis based on history, urodynamics, pad test, pelvic floor grading, perineometry. Exclusion criteria: not reported
Interventions	A (n = 50) ‐ conservative treatment. Instructed to perform 20 maximal PFM contractions 3 times a day, use of cones twice a day, and one 30 min instructor‐led session per week doing a series of pelvic exercises B (n = 52) ‐ electrical stimulation. 12 × 30 min sessions over 6 weeks. 20 min maximal functional electrical stimulation and 10 min EMG/pressure biofeedback
Outcomes	Women reporting continence Pad tests (g) Urethral pressure profile (cmH₂O)
Notes	—
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	"[P]rospective randomised study"
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding of participants and personnel (performance bias)   All outcomes	Unclear risk	Not possible
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Not reported
Incomplete outcome data (attrition bias)   All outcomes	High risk	Group A ‐ 12/50 and Group B ‐ 6/52 lost to follow‐up. No explanation.
Selective reporting (reporting bias)	Unclear risk	Insufficient information to determine if planned outcomes are reported in full
Other bias	Low risk	Nothing to indicate other risk of bias

Methods	Study design: randomised Multicentre or single‐centre: single Setting: Italy Period: not reported Details of sample size calculation: not reported Follow‐up: assume same as length of treatment, i.e. approx. 2.5 weeks for group A, 4 weeks group B
Participants	N: 20 randomised Mean (SD) age: A ‐ 49.5 (14.5) years; B ‐ 41.5 (7.4) years Inclusion criteria: women with GSI Exclusion criteria: detrusor instability, inversion of perineal command, absent contraction of the pubococcygeal muscle, neurological disease, unwillingness to collaborate
Interventions	A (n = 10) ‐ functional electrical stimulation. Lying in dorsal position, 12 × 30 min sessions on consecutive days, excluding Saturdays and Sundays. Pulse width 4 seconds, 8 seconds recovery phase, 15 min at frequency 20 Hz, 15 at 50 Hz, intensity according to the participant's sensations B (n = 10) ‐ vaginal cones. Femcon set of 5 cones, same size and volume, weight 20‐70 g. To select treatment cone, static and dynamic tests were carried out. Static test: standing patient required to hold cone in place for 1 min without any voluntary contractions. Dynamic test: series of exercises requiring voluntary contraction of PFM, e.g. up and down a step‐ladder, skipping With treatment cone women were taught a series of exercises starting from positions facilitating the holding of the cone in place to the upright position, then invited to practise with the heaviest cone they could hold in the vagina for 1 min, aiming to reach 20‐25 min, then begin with next heaviest cone. Once familiar with method, instructed to perform exercises at home 25‐35 min per day for 4 weeks.
Outcomes	Subjective assessment of SUI (10 point VAS) Pad test (g) (not specified if 1‐hour) Change in maximum urethral pressure (cmH2O)
Notes	—
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	"[R]andomly divided into two equal groups"
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding of participants and personnel (performance bias)   All outcomes	Unclear risk	Impossible to blind participants. No blinding reported
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Not reported
Incomplete outcome data (attrition bias)   All outcomes	Unclear risk	Group B 2/10 refused to undergo follow‐up urethral pressure profile. Unclear if they were included in other analysis
Selective reporting (reporting bias)	Low risk	Outcomes seem to be reported in full
Other bias	High risk	Groups not comparable at baselines in terms of age

Methods	Study design: quasi‐RCT Multicentre or single‐centre: not reported Setting: UK Period: not reported Details of sample size calculation: not reported Follow‐up: 3 months' treatment
Participants	N: 58 randomised Mean (range) age: 50 (25‐72) Inclusion criteria: urodynamic diagnosis of uncomplicated GSI Exclusion criteria: not reported
Interventions	A (n = 20) ‐ conservative treatment. Pelvic floor muscle exercise programme. "Home exercise programme of thrice daily 'fast' and 'held' contractions in sitting and standing was given and patients were encouraged to perform these throughout the 3 months." B (n = 20) ‐ electrical stimulation plus PFME. Interferential therapy, 20 min 10‐40 Hz using vaginal electrodes C (n = 18) ‐ sham ES plus PFME. Placebo interferential therapy, 20 min A + B + C ‐ 5 appointments in month 1, 12 in month 2 (thrice weekly), 3 in month 3. Month 1 is identical for all three groups (i.e. ES and sham ES introduced in month 2)
Outcomes	Perceived severity of leakage (VAS) Micturitions per 48h 48hr pad tests (g) Quality of life Perceived effect on life (VAS)
Notes	—
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	"[A]ssigned consecutively to one of 3 treatment groups"
Allocation concealment (selection bias)	High risk	"[A]ssigned consecutively to one of 3 treatment groups"
Blinding of participants and personnel (performance bias)   All outcomes	Unclear risk	Not possible to blind all participants, other blinding not reported
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Not reported
Incomplete outcome data (attrition bias)   All outcomes	Unclear risk	A: 12/20 withdrew B: 9/20 withdrew C: 9/20 withdrew No explanation for withdrawals. Only participants completing the trial are included in analysis
Selective reporting (reporting bias)	Unclear risk	Outcomes seem to be reported in full
Other bias	High risk	Groups not balanced at baseline in terms of objectively measured leakage

Methods	Study design: RCT Single centre Setting: Department of a Regional Hospital in Perth, Western, Australia Period: not reported Sample size: 50 participants in each group would be sufficient to give 0.8 power at the 0.05 alpha level for for two‐sided alternative. Calculation of sample size was performed using the PASS statistical software (NCSS, Kaysville, Utah, USA). Follow‐up: 4 weeks
Participants	N: 24 randomised and analysed Sex: women Mean age (SD): A (n =12) ‐ 52.1 (17.5) years; B (n = 12) ‐ 55.1 (15.1) years Inclusion criteria: women aged 20 years or older with stress or urgency urinary incontinence Exclusion criteria: altered mental state, urinary incontinence caused by problems other than stress or urge, transient incontinence, or severe disability requiring full assistance with all acts of daily living
Interventions	A (n = 12) ‐ pelvic floor muscle exercises (PFME). 12 sessions (3 per week for 4 weeks): 10 sets of 5 contractions with 30 second rest between each set. Then repeated after an hour. B (n = 12) ‐ interferential therapy (ITT) plus PFME. 12 sessions (3 per week for 4 weeks) of 50 pelvic floor contractions followed by ITT with Nemectrodyne 5 stimulator then another 50 contractions. 2 anterior flat electrodes placed over obturator foramen 1.5 cm‐2 cm lateral to symphasis, 2 posterior electrodes placed medial to ischial tuberosities either side of anus. ITT was at highest tolerable frequency between 0‐100 Hz for 15 min (session 1), then 30 min for sessions 2‐12.
Outcomes	Pad test (g) Frequency (number of micturitions per day) Nocturia (number of nocturia episodes per night) Improvement in stop/start test, defined as change from unable to stop to being able to slow, or change from able to slow to able to stop
Notes	No usable data. Not stratified by stress/urgency incontinence
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	"Participants were randomly allocated as soon as they gave written consent, using the sealed envelope method".
Allocation concealment (selection bias)	Low risk	Sealed envelopes
Blinding of participants and personnel (performance bias)   All outcomes	Unclear risk	Not possible to blind participants, other blinding not reported
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Assessor was blinded
Incomplete outcome data (attrition bias)   All outcomes	Low risk	No withdrawals reported. Outcomes seem to be reported for all participants who were randomised.
Selective reporting (reporting bias)	Low risk	Outcomes seem to be reported in full
Other bias	Low risk	Nothing to indicate any other source of bias