NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Collaborating Centre for Women's and Children's Health (UK). Urinary Incontinence: The Management of Urinary Incontinence in Women. London: RCOG Press; 2006 Oct. (NICE Clinical Guidelines, No. 40.)

  • This publication is provided for historical reference only and the information may be out of date.

This publication is provided for historical reference only and the information may be out of date.

Cover of Urinary Incontinence

Urinary Incontinence: The Management of Urinary Incontinence in Women.

Show details

4Conservative management

In this chapter we refer to those therapies used for UI that do not involve surgery. These include lifestyle interventions, physical, behavioural, drug and complementary therapies, and non-therapeutic interventions (products that collect or contain leakage). The preventive use of physical and behavioural therapies and of lifestyle interventions is also considered.

4.1. Lifestyle interventions

Studies considered for the lifestyle interventions question

Evidence described in this section is derived from studies that investigated the effects of modifying the specified lifestyle factors on UI- or OAB-related outcomes. Where no such interventional studies were identified, other study designs investigating how these lifestyle factors may affect the prevalence, or incidence, of UI or OAB were considered. Several observational studies have considered the possible association between lifestyle factors and UI, many of which include both men and women.

4.1.1. Bowel habit

No studies were identified that addressed the effects of modifying bowel habit on UI in women. Three observational studies in women considered whether bowel habit is a risk factor for UI.

One observational study compared the history of bowel function in women with uterovaginal prolapse (n = 23, ten of whom had ‘minor’ stress UI symptoms), women with stress UI (n = 23) and a control group (n = 27). Straining at stool as a young adult was reported by significantly more women with prolapse or stress UI than the control group (61% versus 30% versus 4%), as was bowel frequency of less than twice a week as young adults in women with prolapse compared with control (48% versus 8%).187 [EL = 2+]

Another cohort study reported that bowel urgency was associated with risk of OAB at 1 year (n = 12 570).14 [EL = 2+]

A cross-sectional study considered the effects of constipation and straining at stool on lower urinary tract symptoms. Stress UI, urgency and hesitancy were associated with both constipation and straining at stool. Sensation of incomplete emptying, post-void dribble and straining were associated with straining at stool (n = 487).188 [EL = 3]

A further cross-sectional survey reported that constipation was associated with a risk of stress and urge UI, although constipation was not defined (n = 6006).189 [EL = 3]

4.1.2. Dietary factors

No studies were identified that addressed the effects of modifying dietary factors, including alcohol consumption. A cohort study investigated the association between the intake of certain foods, energy, minerals and vitamins and the 1 year incidence of stress UI or OAB in women aged 40 years or above. The data indicate that certain quantities of some foods may be associated with reduced risk of new-onset OAB (chicken, vegetables, bread, protein, vitamin D and potassium) or new-onset stress UI (bread), and some quantities may be associated with an increased risk of new-onset OAB or stress UI (carbonated drinks) or stress UI (high fat, cholesterol, vitamin B12 and zinc intake) n = 6424.190–192 [EL = 2+]

A cross-sectional study did not find an association between alcohol use and urgency (n = 1059; 50% women).193 [EL = 3]

4.1.3. Caffeine

One RCT evaluated the effects of reducing caffeine intake to a maximum of 100 mg/day in addition to bladder training compared with bladder training alone in men and women with OAB with or without UI (n = 74). At 1 month, significantly greater reductions in urgency episodes and frequency were seen in the caffeine reduction group, with no significant differences between groups in reductions in urge UI episodes.194 [EL = 1+]

Four observational studies investigated the relationship between caffeine consumption and UI or OAB.195–198 Two of these studies evaluated the effects of caffeine on urodynamic parameters in women, as follows:

  • A case–control study reported that the risk of DO was significantly higher with high versus minimal caffeine intake (OR 2.4, 95% CI 1.1 to 6.5). The risk with moderate versus minimal caffeine intake was not statistically significant (OR 1.5, 95% CI 0.1 to 7.2, n = 259).195 [EL = 2+]
  • A significant increase in detrusor pressure rise on bladder filling was seen after 200 mg caffeine intake in women with DO. No significant changes were identified in other urodynamic parameters in either group (women with DO or asymptomatic women, n = 30).196 [EL = 3]

The other two studies reported the effects of modifying caffeine intake on subjective or objective outcomes:197,198

  • During the initial 2–4 week self-monitoring phase of a behaviour management programme in women,199 daily intake of caffeine, urine loss, daytime leakage episodes, and frequency fell and daily fluid intake increased. None of the changes in outcomes was significantly associated with reduced caffeine intake (n = 34).197 [EL = 3]
  • In a series of older people with psychiatric conditions who underwent a 13 week programme of alternating caffeine intake or abstinence, day and night leakage episodes were higher during periods of caffeine intake (n = 14; eight women).198 [EL = 3]

Four cross-sectional studies investigated the association between caffeine intake and UI or OAB.193,200–202 The findings in the individual studies were as follows:

  • no association between coffee intake and urgency (n = 1059; 50% women)193
  • increased risk of UI with tea intake (n = 6876)200
  • nocturia was more common in women who drank tea in the evening (no numerical data presented; n = 3669)201
  • the risks for difficulty in emptying the bladder and having a weak urinary stream were higher in women who drank coffee (n = 297).202 [EL = 3]

4.1.4. Fluid intake

In one RCT in women with UI (type unspecified), no significant changes in leakage episodes were reported after modifying daily fluid intake for 5 weeks. Adherence to fluid intake protocols was reported to be poor (n = 32).203 [EL = 1−]

A further crossover RCT considered the effects of fluid manipulation over a 3 week period in women with stress UI or idiopathic DO. Fluid manipulation consisted of caffeine restriction for 1 week followed by increased or decreased fluid intake in association with continued caffeine restriction for 2 weeks. Caffeine restriction alone did not lead to statistically significant reductions in any outcome (leakage or urgency episodes, frequency, 24 hour pad test). Increasing fluid intake led to a significant increase in urgency episodes in women with DO, with no significant effects on other outcomes. After reducing fluid intake, significant reductions in leakage and urgency episodes, and in frequency, were seen (n = 84; 69 analysed).204 [EL = 1−]

Based on women with stress UI or DO enrolled in a study of behaviour management,205 weak correlation between fluid intake and diurnal and nocturnal frequency and leakage episodes was reported over a 1 week period (n = 126).206 [EL = 3]

4.1.5. Smoking

No studies were identified that addressed the effects of smoking cessation on UI in women. One cohort study found significantly increased incidence of stress UI or OAB in current smokers compared with never smokers at 1 year (n = 6424).190 [EL = 2+] A case–control study reported significantly higher prevalence of smoking among women with UI than women who were continent (OR 4.2, 95% CI 2.2 to 8.2). In women with UI, the prevalence of urge UI was significantly higher among smokers than non-smokers (n = 160).207 [EL = 2−]

Six cross-sectional surveys considered the relationship between smoking and UI in women. Three surveys reported no association (n = 297, n = 486, n = 6037; 56% women).202,208,209 The others reported a positive association between current smoking and UI (one study, n = 6876)200 or nocturia (one study, n = 3669),201 and between former smoking and UI (two studies, n = 6876; 1059 [50% women]).193,200 [EL = 3]

4.1.6. Weight

One RCT evaluated the effects of a 3 month weight reduction programme in overweight women with UI. Reductions in weight and leakage episodes and improvements in QOL were significantly greater in the group assigned weight reduction compared with no intervention. Beyond the randomisation phase, women were offered continued intervention for 6 months, after which symptoms were still improved compared with baseline (n = 48; 40 analysed).210 [EL = 1−]

Three case series reported the effects of surgically induced weight loss on UI in morbidly obese women:

  • In 12 women with stress, urge or mixed UI who had mean weight loss of 33% following gastric bypass surgery, nine were subjectively cured of UI with no significant change in frequency at mean follow-up of 14 months.211 [EL = 3]
  • In women who had lost 50% or more of their excess weight following bariatric surgery, the prevalence of stress UI fell from 61% to 12% after stabilisation of weight loss (2–5 years) (n = 138).212 [EL = 3]
  • In men and women who lost 46% of their excess bodyweight following laparoscopic adjustable gastric banding, 64% reported that their stress UI was better. Symptoms were unchanged in the remainder (n = 195; 83% women).213

A fourth case series evaluated the effects of a 3 month weight reduction programme on leakage episodes in ten women with UI (six urge, three mixed, one stress). Seven women reported at least 50% reduction in leakage episodes (all those who lost at least 5% of their weight, and one-quarter of those who lost less than 5% of their weight).214 [EL = 3]

Two cohort studies190,215 and five cross-sectional studies188,200,201,208,209 investigated the relationship between BMI and UI or OAB. One cohort study found significantly increased 1 year incidence of stress UI or OAB in women with BMI more than 30, compared with a BMI of 20–25 (n = 6424).190 [EL = 2+] The second cohort study, which investigated the prevalence of stress or urge UI in participants of high- or low-impact exercise, reported that BMI was associated with risk for regular stress and urge UI but gave no specific detail (n = 104).215 [EL = 2+]

The cross-sectional studies found that the prevalence or risk of UI or OAB was higher with increased BMI, specifically:

  • increased risk of UI with BMI greater than 25 (n = 6876)200
  • women with regular UI had the highest mean BMI (n = 486)208
  • significantly higher prevalence of UI in women with BMI greater than 29 (n = 6037; 56% women)209
  • increased risk of two or more nocturia episodes versus one episode for women with BMI of 30 or more versus less than 20 (n = 3669)201
  • increased risk of UI or urgency with increasing BMI, although BMI thresholds were not defined (n = 487)188
  • increased risk of urge UI and a trend towards increased risk of urgency in women within the highest versus lowest BMI quartiles (n = 297)202
  • increased risk of both stress and urge UI in women within the highest BMI quartile (n = 6006).189 [EL = 3]

4.1.7. Physical exercise

No controlled studies were identified that addressed the effects of physical exercise on UI in women. A cohort study investigated the prevalence of stress or urge UI in past US Olympians who had participated in long-term high-impact exercise (gymnastics or track and field) in the past compared with low-impact exercise (swimming). No significant difference in the prevalence of stress or urge UI was identified between high- or low-impact exercise groups (n = 104).215 [EL = 2+] Another cohort study evaluated the effects of physical activity before, during and after first childbirth; the analysis suggested that prepregnancy high-impact activity may be associated with risk of UI (n = 665).216 [EL = 2+]

Three cross-sectional studies investigated the prevalence of UI in women who exercise compared with those who do not. The three studies found that the overall prevalence of UI was not significantly different between groups (total n = 1677).217–219 [EL = 3] A further cross-sectional study reported that urgency was less likely in women who exercise at least weekly (n = 6006).189 [EL = 3]

Evidence statements for lifestyle interventions

There is a lack of high-quality prospective controlled trials evaluating the effects of modifying lifestyle factors in women with UI or OAB. [EL = 4]

Observational studies suggest that an increased caffeine intake may be associated with OAB and UI. [EL = 3] There is some evidence that caffeine reduction leads to less urgency and frequency when used in addition to bladder training. [EL = 1+]

Only RCTs of poor quality exist for modifying fluid intake, which were inconclusive. [EL = 1−] There is evidence of an association between obesity and UI or OAB, and in obese women weight reduction of at least 5% is associated with relief of UI symptoms. [EL = 3]

Constipation (bowel frequency of less than twice a week), and increased straining at stool in early adult life, may be associated with an increased tendency to prolapse and UI but no evidence on the effect of modifying bowel habit on continence was identified. [EL = 2+] Some dietary factors may increase the risk of developing UI or OAB although there is no evidence in relation to the effects of modifying these factors. [EL = 2+] Most observational studies suggest that smoking is associated with an increased risk of UI and OAB although there is no evidence relating to smoking cessation in the management of these symptoms. [EL = 3] There are conflicting data in relation to the association between physical exercise and UI prevalence. [EL = 3]

From evidence to recommendations

Where there is consistent evidence that a lifestyle factor appears to increase the risk of UI or OAB, or where there is consistent evidence of benefit from modifying a lifestyle factor, the GDG has recommended interventions in relation to these (caffeine, weight). Both excessive and inadequate fluid intake may lead to lower urinary tract symptoms; this should be considered on an individual basis. The overall lack of consistency in findings indicates a need for further research in this area.

Recommendations for lifestyle interventions

A trial of caffeine reduction is recommended for the treatment of women with OAB. [D]

Consider advising modification of high or low fluid intake for the treatment of women with UI or OAB. [D (GPP)]

Women with UI or OAB who have a body mass index greater than 30 should be advised to lose weight. [D]

Research recommendation for lifestyle interventions

There is a need for prospective interventional studies in all areas of lifestyle interventions to evaluate the effects of modifying these factors on UI and OAB.

4.2. Physical therapies

A variety of physical therapies are used in the management of UI in women. Pelvic floor muscle training (PFMT) involves recruiting pelvic floor muscles for muscle strengthening and skill training. Contraction of pelvic floor muscles causes inward lift of the muscles, with resultant increase in urethral closure pressure, stabilisation and resistance to downward movement.220 Biofeedback can promote awareness of the physiological action of pelvic floor muscles by visual, tactile or auditory means, for example by manometry or electromyography (EMG).221,222 Weighted vaginal cones are cone-shaped appliances of various weights that can be used to facilitate strengthening of pelvic floor muscles. Passive and active contraction of the pelvic floor muscles prevent the cones from slipping out of the vagina.222 Electrical stimulation involves the application of electrical current, usually via vaginal electrodes, to stimulate the pelvic floor muscles via their nerve supply, or to aim to normalise reflex activity.221,223

Studies considered for the physical therapies question

Evidence described in this section is derived from RCTs. Two systematic reviews published on the Cochrane library (PFMT, and weighted vaginal cones) in women with UI collate much of the RCT data identified in the systematic searches.224,225 Owing to overlap of studies within Cochrane reviews, and because they include abstracts that have not subsequently been published, the studies were considered individually alongside all other relevant primary data.

Overall, the studies of physical therapies are heterogeneous in terms of the treatment programmes used, duration of treatment and/or follow-up, the populations and number of individuals enrolled and the outcomes measured. Several studies considered more than one intervention.

4.2.1. Pelvic floor muscle training

A wide range of different PFMT programmes was used across the RCTs, varying in duration, in number and type of contractions and repetitions. Daily PFMT was used in most. Further detail of the PFMT programmes used is provided in the relevant subsections.

PFMT versus no treatment or sham PFMT

Six RCTs compared PFMT with no treatment in a total of 422 women (211 having either treatment). Four enrolled women with stress UI,226–229 and one included women with stress or mixed UI (8.9%).230 The remaining study included women with stress, mixed or urge UI (urge UI in 16%), in which those with urge UI were treated with bladder training and those with mixed UI with bladder training plus PFMT; this study is considered in the behavioural therapies section (4.3).231,232 Some results for the 60% of women from this study who had stress UI were reported separately and are considered here.232 Other than one quasi-RCT,231,232 the studies were considered to be of good quality. [EL = 1+]

Women were advised to undertake daily PFMT in five studies. The number of contractions instructed varied, other than one study that only evaluated the ‘knack’ over a 1 week period.227

The lowest number of contractions across the other studies was 8–12 contractions three times a day (plus an exercise class every week) and the highest was 20 contractions four times a day, increasing to 200 per day.226,228,230,231 In a comparison with duloxetine, PFMT involved a target of 200 contractions per week (over 4 days), and the ‘knack’. Half the studies provided other support, including an audiotape plus group training once a week226 or a leaflet in addition to initial instruction.230,231 The woman’s ability to contract the pelvic floor muscle (PFM) was checked by vaginal palpation in five studies,226–229,231,232 one of which excluded women unable to contract the PFM.227

Duration of treatment was 3 months in most studies, although this varied from 1 week to 6 months. Most studies considered cure rates and changes in leakage episodes at the end of treatment, which showed significantly greater improvements with PFMT compared with no treatment, in the five studies that recommended daily PFMT. The results for PFMT groups versus no treatment were:

  • subjective cure rates of 16% and 56% versus 3% (two studies)226,230
  • success (combined subjective cure and improvement) in 85% versus 0% (one study)232
  • objective cure rates (1 hour pad test or negative stress test) of 44% and 65% versus 0% and 7% (two studies)226,232
  • reductions in leakage episodes of 54% or 72% versus 6% and an increase of 10% (three studies; no numerical data in one).226,230,232

Three studies also considered short pad test results, each of which showed significantly less leakage in women undergoing PFMT.226–228 One RCT used a 24 hour pad test where no difference was identified.226 Both those studies that did226–228,231,232 and those that did not230 assess PFM contraction prior to treatment showed efficacy of active treatment groups over control. [EL = 3]

Five year follow-up has been reported of women from a 3 month study during which the control group was offered PFMT.233 At 5 years in women with stress, mixed or urge UI (those with stress or mixed undergoing PFMT), 69% reported improvement or dryness compared with pretreatment. However, a non-significant increase in leakage episodes was seen in women with stress UI (n = 110).233 [EL = 3]

In a comparison of PFMT with sham PFMT (and with duloxetine 80 mg with or without PFMT), sham PFMT involved contracting hip abductor muscles. No significant differences were reported between PFMT and sham PFMT groups in any outcome (leakage episodes, global improvement, QOL) after 12 weeks’ treatment. It is unique to this trial that PFMT apparently gave no significant benefit over sham treatment.229 [EL = 1+]

Adverse effects

The majority of studies comparing PFMT with no treatment did not consider adverse effects. One RCT reported isolated adverse effects (pain, and an ‘uncomfortable feeling’ during exercise).231 Another RCT reported that no adverse effects occurred.226 Adverse effects occurring in the PFMT and no treatment arms of the duloxetine study were pooled, and therefore a distinction between these interventions is not possible.229

Different pelvic floor muscle training regimens

Several RCTs compared different PFMT regimens, or different methods of delivering PFMT.

Intensive versus standard regimens

Four RCTs compared ‘intensive’ with ‘standard’ PFMT. The PFMT programmes and the populations evaluated were as follows:

  • an exercise class every week in addition to standard PFMT (individual instruction, clinic biofeedback, 8–12 contractions three times a day, contraction checked by palpation) versus standard PFMT alone; 6 months’ treatment, women with stress UI (n = 52)234,235 [EL = 1+]
  • a program of PFMT with bladder training for women with frequency and urgency (PFMT consisting of individualised instruction, target 80–100 contractions/day) versus standard postnatal care (which could include information on pelvic floor exercises); in women with stress, urge (15%) or mixed UI (31%) 3 months postpartum, assessed after 1 year’s treatment,236 and at 6 years (n = 747)237 [EL = 1++]
  • PFMT with or without vaginal cones versus standard PFMT (antenatal and postnatal instruction) in women with UI 3 months postpartum, assessed after 1 year’s treatment and at 24–44 months postpartum; awareness of contraction checked by perineometry (n = 145)238 [EL = 1−]
  • one-to-one instruction while in hospital postnatally, with an option to attend two postnatal pelvic floor exercise classes, versus standard care (verbal promotion of exercises, plus explanatory leaflet); effects assessed at 6 months postpartum (n = 190)239 [EL = 1+]

The rate of subjective cure or improvement was significantly higher in the intensive group in the study that considered this outcome (96% versus 66%; cure rates alone 9% versus 0%).234,235 Two of three studies in postpartum women found significantly lower UI prevalence in the intensive groups (60% versus 69% and 50% versus 76%),236,238 while the third reported no significant difference (60% versus 46%).239 The findings of pad tests were inconsistent (two studies).234,235,238

Longer term follow-up is available from two studies. At 6 years, no significant differences were found between ‘intensive’ and ‘standard’ groups in UI prevalence, severity or leakage episodes in the 69% of women followed up.236,237 [EL = 1++] Five year follow-up of the intensive PFMT arm (23 women) of one RCT234,235 has also been published,240 and a 15 year follow-up of both treatment arms.241 Data at 15 years show no differences in urinary outcomes or satisfaction between groups in the 91% of women followed up.241

Group versus individual training

Two RCTs of 3 months’ duration compared groups with individual PFMT in women with stress, mixed or urge UI (n = 530, n = 44).242 [EL = 1+] 243 [EL = 1−] Group sizes were eight to ten242 or four to twelve.243 In the smaller of the two RCTs (n = 44), women also underwent bladder training. Neither study found significant differences between the two methods in the outcomes evaluated (leakage, UI severity, self-reported change in symptoms, pad test, QOL or frequency).

None of the RCTs comparing different PFMT methods considered adverse effects.

PFMT and drug treatment

One RCT compared PFMT with intravaginal oestrogen in women with stress UI (3 months’ treatment with follow-up at 12 months) but no between-group analyses were reported.228 [EL = 1+] Three RCTs evaluated combined PFMT and drug therapy (estriol, tolterodine and duloxetine).229,244,245

The PFMT regimen varied across these studies: five contractions per hour,228 75 contractions per day,245 15 minutes per day,244 and a target of 200 contractions per week over a 4 day period.229 Pelvic floor muscle contraction was checked by vaginal palpation in two studies.228,229

Oral estriol 1 mg plus PFMT was compared with PFMT alone in postmenopausal women with stress UI (n = 73; 66 analysed). A higher cure rate was reported with estriol plus PFMT compared with PFMT alone, at 2 years (78% versus 68%); no other outcomes were reported.244 [EL = 1−]

The effects of adding PFMT to tolterodine 2 mg b.d. compared with tolterodine alone was considered in men and women with frequency, urgency and urge UI (n = 480; 75% women). After 6 months’ treatment, no significant differences were seen between tolterodine plus PFMT versus tolterodine alone in changes in any outcome. The reductions in symptoms with combined therapy versus tolterodine alone were: urge UI episodes 64% versus 70%, frequency 23% versus 27%, and urgency 79% versus 83%. Overall 82% versus 86% considered themselves improved. Adverse effects reported with tolterodine (with or without PFMT) were dry mouth, headache, constipation, nausea, dry eyes and dizziness.245 [EL = 1++]

Duloxetine 80 mg daily (with or without PFMT) was compared with PFMT and with no active treatment (sham PFMT and placebo drug) in women with stress UI (n = 201). Significantly greater reductions in leakage episodes were reported with duloxetine (with or without PFMT) compared with PFMT alone after 3 months’ treatment. Global improvement and I-QOL scores indicated greater improvement with duloxetine plus PFMT compared with no active treatment. Discontinuation and adverse effect rates (nausea, dizziness, dry mouth, constipation, insomnia, somnolence, asthenia) were significantly higher in duloxetine-treated groups compared with PFMT or no active treatment combined.229 [EL = 1+]

4.2.2. Vaginal cones

Ten RCTs evaluated the use of weighted vaginal cones in women with UI compared with PFMT or electrical stimulation, or in combination with PFMT.226,238,246–253 Other than one study that enrolled women with stress, mixed or urge UI (postnatally),238 all studies included women with stress UI. Between 37 and 145 women were evaluated in each study.

The protocol for cone use differed across studies. Seven used a range of weights, increasing according to ability to retain the cone (20–70 g,226,247,252,253 20–100 g,238,250 or 50–100 g246). One study used a fixed weight of 150 g,251 while two studies did not specify weights used.248,249 In five studies, women were instructed to hold the cones in place twice or three times a day for 10–15 minutes.238,246–248,250 Once daily use for between 5 and 25 minutes was advised in another five studies.226,249,251–253

In studies involving PFMT, daily PFMT was undertaken. The number of contractions ranged from 24 (eight contractions three times a day) to 100 where specified.226,238,246–248 PFMT was individually tailored in three studies,247,249,252 and included the ‘knack’ in two.247,252 Seven studies stated that ability to contract PFM was checked at baseline.226,238,246–248,252,253

Cones versus no active treatment

Cones were compared with no treatment within one RCT (total n = 107). Significantly greater improvement in leakage and social activity indices were seen with cones after 6 months’ treatment. No significant differences were seen between groups in other outcomes (subjective or objective cure, leakage episodes).226 [EL = 1+]

Cones versus PFMT

Cones were compared with PFMT in four studies of 3–6 months’ duration.226,246,247,249 Two of these studies had other treatment arms (electrical stimulation and an untreated control,226 biofeedback249). A further study, also described under the intensive versus standard PFMT section, compared ‘intensive’ PFMT (PFMT/cones/PFMT plus cones) with standard postnatal care but only analysed results for women who completed the study. Results were presented for the intensive group as a whole, except for the proportion cured, where there was no significant difference between cone and PFMT groups at 1 year postpartum.238 [EL = 1−]

Two studies of 3 months’ duration reported no significant differences between cones and PFMT in improvements in outcomes evaluated: leakage episodes, subjective improvement, subjective or objective cure rates (n = 60);247 [EL = 1+] or leakage episodes, PFM strength or QOL (KHQ) (n = 101).249 [EL = 1−]

The third study found significantly greater improvement with PFMT versus cones in the short pad test, leakage episodes, leakage index and PFM strength, and a significant difference in subjective and objective cure rates (56% versus 7% and 44% versus 15%). No significant differences were seen in 24 hour pad test results after 6 months’ treatment (n = 107).226 [EL = 1+] The fourth study, which only analysed results for those who completed 4 months’ treatment, found a significantly greater reduction in leakage on the stress pad test with cones versus PFMT. No significant differences were found in other outcomes (PFM strength and subjective assessment) (n = 37).246 [EL = 1−]

Cones versus electrical stimulation

Cones were compared with electrical stimulation in three studies, none of which reported significant differences between groups in any outcome, whether assessed at 1 or 6 months. Women in one study also undertook PFMT.250 Outcomes evaluated across two studies were: short and 24 hour pad tests, subjective cure, leakage episodes, leakage and social activity indexes, and pelvic floor muscle strength.226,250 [EL = 1+] The remaining study only considered urethral pressure and pad test results, and gave inadequate data and detail of methods for evaluation (n = 20).253 [EL = 1−]

Cones in combination with PFMT

Cones and PFMT in combination were compared with PFMT alone in a 3 month study. Limited results were given (urodynamics only available for 59%), with no between-group analysis for subjective assessments; however, a similar proportion of women in both groups reported cure or improvement (n = 46).252 [EL = 1+]

Cones and PFMT in combination were compared with electrical stimulation in two studies of 6 weeks’ duration. One reported no differences between groups in leakage episodes or frequency, while no between-group analyses were reported for other outcomes (n = 40).248 [EL = 1−] The other study found no significant differences between groups in any outcome (urethral or vaginal pressures, PFM contraction and subjective assessment) (n = 120).251 [EL = 1+]

Adverse effects

One study reported adverse effects, which were four reports in 27 cone users (one abdominal pain, one bleeding, two vaginitis) and two reports in the electrical stimulation group (tenderness and bleeding; discomfort).226 No adverse effects were reported in the PFMT or untreated control groups. Motivation problems were very common in the cone and electrical stimulation groups in one study (52% and 32%).226

The withdrawal rate from cone therapy was 47% in a study that noted this, compared with none with PFMT.247 Differences in withdrawal rates between groups were not apparent in other studies.

4.2.3. Biofeedback

Most data regarding biofeedback relate its use in conjunction with PFMT, rather than as an isolated intervention. A variety of biofeedback methods were used across these studies, differing in the probes used (vaginal probes with EMG electrodes, pressure-sensitive intravaginal devices), in the feedback provided (visual and/or auditory) and the setting in which biofeedback was undertaken (home or clinic). Additionally, a few studies used electrodes/rectal catheters to monitor muscle activity or abdominal pressure.

Only one study compared biofeedback alone with PFMT (and with cones, n = 101). No significant differences were found between the three groups in improvements in outcomes evaluated at 3 months: leakage episodes, PFM strength and QOL (KHQ).249 [EL = 1−]

Eleven RCTs compared biofeedback-assisted PFMT with PFMT alone, in women with stress UI (eight studies),254–263 stress or mixed UI (one study),230 stress or urge UI (one study)264 or OAB (one study).265 Treatment duration ranged from 4 weeks to 6 months, with the number of women per study ranging from 22 to 103; most included fewer than 50 women. Of the 11 studies, four were considered to be of poor quality [EL = 1−]255,258,261–263 and seven of good quality.230,254,256,257,259,260,264,265 [EL = 1+]

The majority of the studies found no significant differences between biofeedback-assisted PFMT groups and PFMT alone in the outcomes measured (subjective or objective cure, QOL [BFLUTS] or social activity index scores).230,256,258–265 Cure rates in the seven studies that reported this outcome (though variously defined) ranged from 16% to 69% (median 30%) with PFMT and from 15% to 73% (median 50%) with biofeedback-assisted PFMT.230,254,256,258,261,264,265 Significant additional benefit, in terms of leakage episodes, was reported in one study that alternated biofeedback with electrical stimulation,262,263 and in PFM parameters in two studies.259,260,262,263

A further two RCTs evaluated different methods of biofeedback. One compared biofeedback by palpation with EMG in women with stress UI. There were no significant differences between the biofeedback methods in urinary outcomes after 8 weeks’ treatment (n = 50).266 [EL = 1+] The second study compared the use of a vaginal plus abdominal probe with a vaginal probe only, for 4 weeks. Greater improvement in QOL was seen using a vaginal probe only, with no differences reported in other outcomes (leakage, pelvic floor muscle strength or endurance) (n = 38).267 [EL = 1+]

Two RCTs considered adverse effects. None were reported in one,261 and another noted that two women (13%) found the vaginal probe uncomfortable, and that 17% from PFMT or biofeedback-assisted groups reported pain while training.259,260

4.2.4. Electrical stimulation

A range of various electrical stimulation methods and protocols were used across the RCTs, although the protocol used was poorly reported in some studies. Various types of current were used (interferential therapy, faradic stimulation, alternating pulse currents), with various current intensities. The setting (home or clinic), duration (15 to 30 minutes) and frequency (two to three times per week) of individual treatments also varied. Electrical stimulation parameters were generally tailored to the woman’s tolerance.

Electrical stimulation versus sham stimulation

Eight RCTs compared electrical stimulation with sham stimulation. Treatment duration ranged from 4 to 15 weeks, with the number of women recruited in each study ranging from 24 to 121.248,268–274 Four studies included women with stress UI,248,268,270,271 two included women with stress, mixed or urge UI,272,273 and two included women or men and women (57% women) with urge or urge-predominant UI.269,274 One study was of poor quality,248 [EL = 1−] and the others were of good quality.268–273 [EL = 1+]

The outcomes reported across these studies were leakage episodes, UI prevalence, pad tests, subjective cure or improvement, PFM strength, urodynamic parameters and QOL (SF-36, IIQ, UDI). The findings across these studies were inconsistent, with significant benefit with electrical stimulation versus sham stimulation reported for some but not all outcomes, and not across all studies. Not all studies reported between-group comparisons.

A further RCT compared electrical stimulation with ‘lower urinary tract exercises’ (PFMT and bladder training), and with both interventions combined, in women with DO (n = 68). At 9–11 weeks, no significant differences were found between groups in any outcome (detrusor activity index, leakage episodes, PFM strength).275 [EL = 1−]

PFMT versus electrical stimulation

Eight RCTs compared PFMT with electrical stimulation in women (six in women with stress UI,226,228,248,276–278 one in those with stress, mixed or urge UI,279 and one in those with OAB with urge UI265).

The RCTs involving women with stress UI recruited between 18 and 51 patients. Duration of treatment ranged from 6 weeks to 12 months. The PFMT group also used vaginal cones in one study.248 The quality of two studies was poor,248,278 [EL = 1−] while four were of good quality.226,228,276,277 [EL = 1+] None of the studies reported significant differences between groups in subjective or objective cure rates. The subjective cure rates ranged from 10% to 56% with PFMT, and from 4% to 12% with electrical stimulation, and objective cure rates from 10% to 54% versus 4% to 40%. Several other outcomes were reported in one study, where improvements in PFM parameters, short pad test results, leakage and social activity indexes were significantly greater in the PFMT group compared with electrical stimulation.226 No significant differences were reported between groups in leakage episodes or leakage frequency (three studies),226,248,277 or in 48 hour pad test results (one study).226 One of the studies also compared propantheline with electrical stimulation, which reported no significant differences between groups in subjective or objective cure or improvement.277

The RCT involving women with stress or mixed or urge UI (66% mixed) found no significant differences between PFMT and electrical stimulation groups in any outcome (subjective assessment, 48 hour pad test results, improvements in PFM strength or DO prevalence) after 8 weeks’ treatment (n = 35).279 [EL = 1+]

The RCT in women with OAB and urge UI reported significantly greater improvements in PFM parameters and QOL (KHQ) with PFMT compared with electrical stimulation, but no significant differences in self-reported cure or improvement after 12 weeks’ treatment (n = 103).265 [EL = 1+]

Electrical stimulation in combination with PFMT

Four RCTs evaluated electrical stimulation in combination with PFMT versus PFMT alone in women with stress UI (stress or urge in one RCT).278,280–282 The duration of treatment ranged from 1 to six months, with the number of women enrolled ranging from 14 to 57. One RCT also compared the combination with electrical stimulation alone.278 Three studies were of poor quality,278,280,282 [EL = 1−] and one of good quality.281 [EL = 1+]

Across these RCTs, electrical stimulation did not confer additional benefit to PFMT alone in the outcomes measured (self-reported cure or improvement, pad test, PFM parameters). No significant differences in self-reported cure or improvement were seen with electrical stimulation plus PFMT compared with electrical stimulation alone.

Adverse effects

Of all studies that considered the effectiveness of electrical stimulation, five considered adverse effects. None were reported in one study.275 Across the others, adverse effects or complications noted were: vaginal irritation (12–22%), pain (6–9%), and cases of faecal incontinence, discomfort, and tenderness and bleeding.226,268,269,277 One study reported difficulty in maintaining motivation in 32% of the electrical stimulation group.226

4.2.5. Transcutaneous electrical nerve stimulation and posterior tibial nerve stimulation

Transcutaneous electrical nerve stimulation (TENS) is a form of electrical nerve stimulation in which the electrodes are placed over the dermatomes of S2 to S4 for long periods of time, daily. The Stoller afferent nerve stimulator (SANS) involves inserting a fine gauge needle into the posterior tibial nerve just above the ankle. The tibial nerve then carries an electrical stimulation in an afferent direction to the sacral spine.

TENS

One 12 week crossover RCT compared TENS with oxybutynin in men and women (70% women) with idiopathic DO (n = 43). Significant improvements in functional capacity and frequency were reported with both treatments. No significant changes in SF-36 parameters were seen. Adverse effects reported were dry mouth, blurred vision and dry skin; these all had a lower incidence in the TENS group.283 [EL= 1+]

Two case series reported findings of TENS over the short term (1–3 weeks) in men and women with OAB (total n = 103, with 84 analysed). These generally indicated improvements in frequency, nocturia and urgency,284,285 and in urge UI.285 [EL = 3]

Posterior tibial nerve stimulation

Posterior tibial nerve stimulation was considered in one RCT,286 and in five case series of patients with OAB.287–293 Except for one study that included only women, others also included men (predominantly women). Patient numbers ranged from 15 to 90 (total 282).

Posterior nerve stimulation was delivered via a SANS device. The stimulation protocol involved a fixed pulse width and frequency, with amplitude tailored to individual patients. The treatment regimen varied from 30 minutes three or four times a week for 3 or 4 weeks, to 30–60 minutes once weekly for 8–12 weeks. End of treatment results were reported except for one study that had follow-up to 3 years (mean 21 months).292

The RCT evaluated the effects of oxybutynin 5 mg daily in addition to PTNS compared with PTNS alone in men and women (n = 43; 88% women). Improvements in frequency, urgency and urge UI episodes were noted with both treatments, with no significant difference between groups in ‘response rates’.286 [EL = 3]

The smallest case series noted that two-thirds of patients achieved cure or partial response (n = 15).290 Across the others, significant improvements in all outcomes were seen (leakage episodes, frequency and voided volumes).287–289,291–293 Quality of life (I-QOL and SF-36) was improved (two studies).287–289,293 In 11 patients from one case series,293 in whom treatment was maintained for a mean of 13 months, treatment was withheld for 6 weeks and then re-introduced. This study showed that symptoms deteriorated when treatment was withdrawn and that the majority of patients reported some improvement in symptoms 4 weeks after re-introducing treatment.294

There were case reports of adverse effects across the studies. Those related to site of needle insertion were haematoma, minor bleeding, temporary painful/numb feeling, throbbing and transient local tenderness. Systemic effects noted were diarrhoea, cramps, headache, lower back pain, moderate right foot pain and stomach discomfort (no numerical data). Adverse effects reported in the group treated with PTNS and oxybutynin were dry mouth and blurred vision.286

4.2.6. Magnetic therapy

Magnetic therapy aims to stimulate the pelvic floor muscles and/or sacral roots by placing them within an electromagnetic field.

Two RCTs compared magnetic stimulation therapy with sham stimulation delivered via a portable device for the treatment of UI for 8 weeks. In the first study, in women with stress, mixed or urge UI, significantly more women in the magnetic therapy group reported improvement in symptoms. No significant differences in improvements in other outcomes were seen (pad weight, PFM contraction, leakage episodes or nocturia). Two reports of a pulsating sensation in users of magnetic stimulation were noted.295 [EL = 1+] The second study, in women with urge-predominant mixed UI, reported a significantly higher ‘success’ rate with magnetic stimulation; between-group comparisons for other outcomes were not reported (frequency, nocturia). No women experienced adverse effects (n = 39).296 [EL = 1−]

Two case series considered the effects of 6 or 8 weeks of magnetic therapy using a special chair (two 20 minute sessions a week). Women in one had stress UI (a minority having urge-predominant mixed UI), and urge or mixed UI in the other. Results for 74 patients have been reported, which showed significant improvement in all outcomes considered (leakage episodes, pad test results, frequency and satisfaction). Adverse effects were not considered in one study, while none were reported in the other.297,298 [EL = 3]

4.2.7. Economic evidence for physical therapies

There is a lack of good-quality evidence about the clinical and cost effectiveness of conservative therapies for UI. In the absence of evidence of a difference in efficacy between treatment options, cost minimisation analysis may be used to determine the most cost effective. Cost minimisation was undertaken for the physical therapies PFMT, cones, biofeedback and electrical stimulation. Estimates of the costs of the conservative therapies and an explanation of how these estimates were derived are given in Appendix E.

Additionally, because the other conservative treatment option for stress UI is duloxetine (see Section 4.4.4), a decision tree model was developed to compare the cost effectiveness of PFMT and duloxetine, as a first-line treatment for women with moderate to severe stress UI (assumed to be 14 or more leakage episodes per week). Treatment effects and costs were based on a 52 week time frame. This is described in detail in Appendix F. Under baseline assumptions, PFMT ‘dominates’ duloxetine. This means that it is both more effective and less costly. The sensitivity analyses undertaken (and detailed in Appendix F) did not change this conclusion.

Evidence statements for physical therapies

Daily PFMT is an effective treatment for stress or mixed UI compared with no treatment over the short term. Other than occasional cases of pain or discomfort, no other adverse effects were noted. [EL = 1+] Women’s pelvic floor contraction was assessed at baseline in the majority of studies. Studies that did or did not assess pelvic floor muscle contraction prior to treatment both showed efficacy of active treatment compared with control. [EL = 3]

In studies of up to 1 year, higher intensity PFMT regimens confer greater subjective cure or improvement than lower intensity regimens. Over the longer term, differences between these groups are not sustained. [EL = 1+] There is a lack of evidence for optimum training regimens for PFMT. [EL = 4]

There is no additional benefit from the use of PFMT in women undergoing treatment with tolterodine for OAB. [EL = 1++]

In women with stress UI, vaginal cones are more effective than no treatment over the short term. There is no evidence of a difference in effectiveness between cones and PFMT. Compared with PFMT, cones are associated with more adherence problems. [EL = 1+] One study suggested that the training time for using vaginal cones is one-third of that for PFMT, which would make vaginal cones cheaper than PFMT. However, it is not clear what the appropriate training regimen should be for women using vaginal cones. Vaginal cones are not suitable for all women. Cones are inappropriate for use in some circumstances, such as when there is a moderate to severe prolapse, too narrow or too capacious a vagina causing difficulty with insertion or misplacement of the cone, untreated atrophic vaginitis, vaginal infection, or during menstruation or pregnancy. [EL = 4]

Evidence does not indicate additional benefit from biofeedback with PFMT in comparison with PFMT alone in treating UI. [EL = 1+] Biofeedback with PFMT is more costly than PFMT alone and therefore is not cost effective given a lack of additional benefit.

There is lack of consistency in the electrical stimulation protocols employed in available studies. There is limited evidence for the benefit of electrical stimulation versus sham electrical stimulation in the treatment of urge UI. [EL = 1+] There is no evidence of additional benefit of electrical stimulation in combination with PFMT compared with PFMT alone. [EL = 1−]

Case series of up to 3 weeks’ duration show improvements in frequency and nocturia with TENS, although the available data do not allow conclusions to be drawn as to the efficacy of TENS in women with UI. [EL = 3]

Data on posterior tibial nerve stimulation are mainly derived from case series of men and women with OAB, which show improvement in leakage episodes, frequency, voided volume and QOL with treatment for up to 3 months. Symptoms recur on treatment withdrawal. Adverse effects reported relate to needle insertion site (pain, tenderness, haematoma). Combining oxybutynin treatment with posterior tibial nerve stimulation did not lead to additional benefit. Overall the available data are inadequate to define the place of posterior tibial nerve stimulation for UI or OAB. [EL = 3]

There are limited data on the use of magnetic therapy for UI, and its role in the treatment of women with UI is unclear. [EL = 3]

An economic model constructed for the purposes of this guideline suggested that PFMT is more cost effective than duloxetine alone, as first-line treatment for stress UI. This result was generally not affected by making plausible changes to model parameters in favour of duloxetine. While the model was based on the best available clinical evidence, there is a lack of long-term effectiveness data for either treatment.

From evidence to recommendations

While there is no evidence of effectiveness for either biofeedback or electrical stimulation, the GDG considered that the information and support generated by biofeedback may assist motivation for some women, and that electrical stimulation may be of value for those who are unable to initiate a pelvic floor muscle contraction. [EL = 4]

In recommending the use of PFMT, the GDG considered that guidance should be given on the number of pelvic floor contractions to be undertaken within such a programme. Without clear evidence on optimal training regimens, the minimum number of daily pelvic floor muscle exercises advised across the studies was adopted by the GDG as the minimum number of contractions that women should be aiming for, that is 24 (eight contractions three times a day). Most studies evaluate 3 months’ treatment and, in the view of the GDG, this is an appropriate period of time to recommend PFMT before assessing its effectiveness.

Recommendations for physical therapies

A trial of supervised pelvic floor muscle training of at least 3 months’ duration should be offered as first-line treatment to women with stress or mixed UI. [A]

Pelvic floor muscle training programmes should comprise at least eight contractions performed three times per day. [A]

If pelvic floor muscle training is beneficial, an exercise programme should be maintained. [D (GPP)]

Perineometry or pelvic floor electromyography as biofeedback should not be used as a routine part of pelvic floor muscle training. [A]

Electrical stimulation should not routinely be used in the treatment of women with OAB. [D]

Electrical stimulation should not routinely be used in combination with pelvic floor muscle training. [A]

Electrical stimulation and/or biofeedback should be considered in women who cannot actively contract pelvic floor muscles in order to aid motivation and adherence to therapy. [D (GPP)]

Research recommendations for physical therapies

Studies investigating different pelvic floor muscle training regimens are required to establish the optimum method of delivering and undertaking this intervention.

The role of clinical pelvic floor assessment prior to PFMT should be investigated to determine whether it enhances the therapeutic effect of the intervention (Section 3.3).

Research into the optimal electrical stimulation parameters is required, to inform future clinical practice. Studies investigating the role of electrical stimulation in women who cannot contract the pelvic floor muscle are required.

There is a need for a robust evaluation of transcutaneous electrical nerve stimulation and posterior tibial nerve stimulation for the treatment of UI.

4.3. Behavioural therapies

Behavioural therapy involves an individual learning new patterns of response or re-establishing previously learnt behaviour to fit in with what is considered usual. Women with OAB (wet or dry) usually void more frequently than usual due to urgency. Women with stress UI also often void more frequently in the belief that they will pre-empt an involuntary urine loss associated with any increase in intra-abdominal pressure.

Various toileting programmes have been used. Bladder training (also described as bladder retraining, bladder drill, bladder re-education or bladder discipline) actively involves the individual, in attempting to increase the interval between the desire to void and the actual void.299 This may occur by mandatory schedules in which the individual may not use the toilet between set times for voiding, or a self-scheduled regimen where the patient gradually increases their inter-voiding times, and may use the toilet between times if the urge becomes unbearable.300

Studies considered for the behavioural therapies question

Evidence described in this section is derived from RCTs. Four systematic reviews of various toileting regimens have been published on the Cochrane library.299,301–304 The RCTs within these systematic reviews were considered individually. Further RCTs identified are also included here. Several studies included both men and women, none of which reported data separately by gender.

4.3.1. Bladder training

Bladder training versus control

Two RCTs compared bladder training with control in women (n = 60, n = 123).205,305 In the first study, supervised bladder training as inpatients towards a target voiding interval of 4 hours was compared with unsupervised training at home, in women with frequency, urgency and urge UI (two-thirds of whom also had stress UI). At 6 months follow-up (duration of intervention unclear), more women in the supervised group were continent or symptom-free.305 [EL = 1+] The second RCT in women with UI (type unspecified) compared bladder training (target voiding interval of 2.5–3 hours) with an untreated control group. Urine loss, leakage episodes and QOL (IIQ) were improved in the bladder training group after 6 weeks’ treatment.205 [EL = 1+] Neither study considered adverse effects.

Bladder training versus drug treatment

Two RCTs compared bladder training with drug treatment in women with urge or mixed UI (one oxybutynin,306 one a combination of flavoxate and imipramine307).

In women with urge UI, similar self-reported cure rates (about 73%) were seen with a 6 week bladder training programme (target voiding interval of 3–4 hours) and with oxybutynin (n = 81). Relapse occurred in 4% of the bladder training group, and in 44% of the oxybutynin group, at 6 months. About half of the oxybutynin group required dose reduction owing to adverse effects. No between-group comparisons were made for other outcomes.306 [EL = 1+]

No details of the bladder training programme were provided for the comparison with flavoxate plus imipramine (n = 50). Significantly more women were subjectively or objectively cured after 4 weeks’ bladder training than with drug therapy.307 [EL = 1+]

Bladder training in combination with drug treatment

Three double-blind (DB) RCTs evaluated the addition of antimuscarinic drug treatment to bladder training (one oxybutynin,308 one terodiline309 [no longer available in the UK], and one imipramine310). Bladder training aimed to reduce frequency by delaying voiding for as long as possible in two studies,308,309 and aimed at a 4 hourly voiding target in one.310 A further RCT compared tolterodine plus bladder training with tolterodine alone.311 All studies included men and women; two included elderly people,308,309 and two included a broader age group.310,311

In individuals with symptoms of urinary frequency, urgency and urge UI, a significant reduction in daytime frequency was seen with oxybutynin plus bladder training compared with placebo plus bladder training, after 6 weeks’ treatment (n = 60; 93% women). No significant differences were reported in other outcomes (daytime leakage episodes, nocturia, nocturnal enuresis, self-reported benefit, adverse effects).308 [EL = 1+]

No significant differences in frequency, leakage episodes or self-reported improvement were seen between terodiline or placebo in addition to bladder training in individuals with urinary frequency and urge UI, after 6 weeks’ treatment (n = 37; 88% women). Two adverse effects were noted with terodiline (one oesophagitis, one dry mouth).309 [EL = 1+]

In individuals with incontinence and ‘unstable bladders’, no significant differences were seen between imipramine plus bladder training and bladder training alone, in cure or urodynamic parameters with follow-up to 11 months. Dry mouth and constipation were reported with imipramine, with no adverse effects reported with bladder training (n = 33).310 [EL = 1−]

Bladder training in addition to tolterodine was compared with tolterodine alone in men and women (n = 501; 75% women) with urinary frequency, urgency, with or without urge UI (61% with). The aim of bladder training was five to six voids per day while maintaining the same fluid intake. Combined treatment resulted in reduced frequency and increase in volume voided versus tolterodine alone, with no differences between groups in leakage or urgency episodes, patient’s perception of change or adverse effects after 6 months’ treatment.311 [EL = 1++]

Bladder training versus PFMT

Two RCTs compared bladder training with biofeedback-assisted daily PFMT in women.312,313 One did not report the type of UI or report between-group comparisons for bladder training, PFMT or no treatment (n = 50).312 [EL = 1−]

The other RCT compared bladder training, biofeedback-assisted PFMT and the interventions in combination in women with stress, urge or mixed UI who had palpable pelvic floor contraction on vaginal examination.313 After 3 months’ treatment, a significantly greater reduction in leakage episodes was seen with combination treatment compared with monotherapy; this was not sustained after a further 3 months follow-up. No other significant differences were reported between groups (n = 204).313 [EL = 1+]

4.3.2. Multicomponent behavioural therapy

Eight RCTs evaluated the use of a multicomponent behavioural programme that included bladder training and PFMT.199,231,314–324 The bladder training methods used were urge strategies in three studies,314–319 bladder training in three studies (one also included fluid management),199,231,320 education in one study,321,322 and one study allocated PFMT or prompted voiding depending on the cognitive status of individuals.323,324 In four of the RCTs, biofeedback-assisted PFMT was used.199,314,315,317–319 PFMT involved daily exercises in six studies, and the programme was not described in two.320–322 Six studies checked the ability of individuals to contract the pelvic floor muscle at baseline.231,314–320,323,324 Seven RCTs enrolled women only, while one enrolled men and women.323,324

Compared with no active treatment or usual care

In four of the RCTs, the comparison group was no active treatment or usual care. Owing to the variety of behavioural methods used, each study is described individually. One RCT in women aged 55 years or above with stress, urge or mixed UI compared a 6 month sequential programme of behavioural management (self-monitoring including fluid management, bladder training, EMG-assisted PFMT) with no active treatment. The behavioural management group achieved significantly greater improvement in leakage episodes, 24 hour pad test, QOL (IIQ) and subjective severity assessment compared with control. Only 21% were followed up to 2 years, in whom improvements seemed to be maintained. No significant differences between groups were reported in voiding frequency or interval (n = 218).199 [EL= 1++]

Another RCT compared 3 months of behavioural therapy with untreated control in women with any type of incontinence. The behavioural strategies used were PFMT for stress UI, bladder training for urge UI and bladder training followed by PFMT for mixed UI. Overall, the results showed significant reduction in leakage episodes with behavioural therapy, and a higher proportion reporting improvement versus control (n = 110).231 [EL = 1+] Results for the 60% of women with stress UI were reported separately, which also showed significant reduction in leakage episodes.232

The third RCT, in women with any type of UI, reported significant reductions in leakage episodes with 6 weeks’ behavioural therapy (bladder training and PFMT) compared with no active treatment. No significant differences were reported in frequency (n = 152).320 [EL = 1+]

The fourth RCT compared a 10 week behavioural therapy programme (education, PFMT), with usual care in women with stress or urge UI. Significantly greater reductions in leakage episodes were reported with behavioural therapy, with no differences between groups in QOL (n = 145).321,322 [EL = 1+]

Compared with other active interventions

Two RCTs in women with urge or mixed UI compared a sequential 8 week programme of behavioural training (PFMT with anorectal biofeedback, urge strategies, repeat PFMT if needed, review and reinforcement) with oxybutynin, or self-help.314–318 Significant reductions in leakage episodes and in nocturia were seen with behavioural training versus oxybutynin, and oxybutynin versus placebo tablets, and significantly more reported satisfaction and improvement with behavioural training versus oxybutynin. Dry mouth and inability to void were significantly more frequent with oxybutynin than control groups (n = 197).314,315,317 [EL = 1+] In the study evaluating the same behavioural training programme with a group receiving biofeedback (vaginal palpation), and a self-help group receiving written instructions of the programme only, no significant differences were reported in satisfaction, improvement, QOL (IIQ, SF-36) or bladder capacity (n = 222).318 [EL = 1++]

After the initial 8 week period of one of these RCTs,314 women initially treated with behavioural therapy or oxybutynin who were not cured or not completely satisfied were offered the other treatment option in addition to the initial therapy for a further 8 weeks. Significant additional improvement in leakage episodes was seen in those who took oxybutynin in addition to continued behavioural treatment (n = 8) and in those who continued with oxybutynin and, in addition, underwent behavioural treatment (n = 27).325 [EL = 2+]

A further RCT in women with predominant stress UI also compared an 8 week sequential programme of behavioural training (PFMT with anorectal biofeedback, the ‘knack’, managing urgency, repeat PFMT if needed, review and reinforcement), with or without the addition of electrical stimulation, with a self-help group who received written instructions of the programme (n = 200). Leakage episodes were significantly reduced in both behavioural training groups versus self-help, and significantly more of the behavioural training (electrical stimulation) group reported ‘much better’ improvement (versus anorectal feedback) or satisfaction with treatment (versus self-help). No differences were reported in QOL (IIQ, SF-36), or bladder capacity. Vaginal irritation was reported in 6% of the electrical stimulation group.319 [EL = 1++]

One RCT assigned homebound adults with UI (type unspecified) to different treatment strategies depending on cognitive status: biofeedback-assisted PFMT versus control for the cognitively intact (n = 93; 91% women), or prompted voiding versus control for the cognitively impaired (n = 19; 68% women). Although reported as one trial, this was effectively two separate 8 week trials.323,324 PFMT was significantly more effective than control in reducing leakage episodes (the only outcome reported).323 [EL = 1+]

4.3.3. Prompted voiding

Prompted voiding and timed voiding are toileting programmes used in people who are not capable of independent toileting, such as the cognitively impaired. Prompted voiding teaches people to initiate their own toileting through requests for help and positive reinforcement from carers.303 It has been used in institutionalised patients with cognitive and mobility problems. They are asked regularly if they wish to void and only assisted to the toilet when there is a positive response.326

Five RCTs compared 1 hourly prompted voiding (three studies326–329) or 2 hourly prompted voiding (two studies324,330) with usual care, or ‘wet checks’ only. Four studies were conducted in cognitively impaired elderly nursing home residents,326–330 and one in homebound adults.324 One study enrolled women only,327 while the remainder enrolled both genders although predominantly women.

The findings of the RCTs of prompted voiding were as follows:

  • significant benefit with ‘functional incidental training’, which included daytime 2 hour prompted voiding, in terms of leakage and urine toileting ratio compared with usual care in a 32 week study (n = 190; 84% women)330 [EL = 1+]
  • significant reduction in wet episodes versus usual care after 13 weeks’ intervention, which appeared to be sustained after a further 22 weeks follow-up; no differences were reported between groups in improvement or self-initiated requests (n = 143; all women)327 [EL = 1+]
  • reductions in leakage and an increase in requests for toileting assistance during the 3 week intervention period versus wet checks only (n = 21; 71% women)328 [EL = 1+]
  • reductions in leakage and an increase in toileting into a receptacle over an intervention period of 10–20 days versus usual care (n = 126; 75% women)326,329 [EL = 1+]
  • No differences versus usual care in any outcome (leakage episodes, % wet, or self-initiated toileting) (n = 19; 68% women).324 [EL = 1+]

None of the studies evaluating prompted voiding considered adverse effects.

A placebo-controlled RCT evaluated the effects of oxybutynin in non-responders to prompted voiding (n = 75; 78% women).331 Significant improvement in leakage episodes was reported with oxybutynin after 20 days’ treatment (40% versus 18% had one or fewer episodes per day). No other outcomes were significantly different (change in leakage episodes, continent voids, volume voided).331 [EL = 1+]

4.3.4. Timed voiding

Timed voiding (scheduled, routine or regular toileting) is a passive toileting assistance programme that is initiated and maintained by a caregiver, for example for patients who cannot participate in independent toileting. Toileting is fixed by time or event, on a regular schedule or on a schedule to match the patient’s voiding pattern. The aim is to avoid incontinence episodes rather than restore bladder function.302

Three RCTs evaluated timed voiding in cognitively impaired elderly men and women (predominantly women) who were nursing home residents (two studies),332,333 or had caregiver support at home (one study).334 The comparator was no active treatment.

One 6 month RCT reported significant reduction in leakage episodes in the intervention group (scheduled toileting according to voiding pattern, mostly about 2 hours, and advice on fluid intake and environment; n = 118, 69% women).334 [EL = 1+] A cluster RCT of 36 weeks’ duration reported limited results indicating greater reductions in leakage episodes, with scheduled toileting (toileting within 30 minutes prior to an individual’s mean voiding time) but no differences between groups in volume voided (n = 113; 82% women).332 [EL = 1−]

The third RCT compared timed voiding (2 hourly) in combination with antimuscarinic drugs for urge UI, or PFMT for stress UI, with no active intervention for 8 weeks (n = 278; 83% women). Significant improvements in night-time leakage episodes were reported in the active intervention group, but not in daytime leakage or pad test findings.333 [EL = 1+]

None of the studies evaluating timed voiding considered adverse effects.

Evidence statements for behavioural therapies

Bladder training is more effective than no treatment in women with urge or mixed UI, at 6 months follow-up. In women with urge UI, bladder training had a similar subjective cure rate to oxybutynin after a 6 week programme but adverse effects and relapse rates were lower with bladder training. The combination of oxybutynin or tolterodine and bladder training programmes results in greater reduction in frequency of micturition but has not been shown to lead to further improvements in incontinence. Combination treatment of bladder training together with PFMT may confer a greater short-term benefit to women with stress, urge or mixed UI, but in the long term combination and monotherapies are equally effective. [EL = 1+]

A wide range of behavioural therapies have been used within multicomponent treatment regimens in women with stress, mixed or urge UI. All appear to show improvements in leakage episodes over comparators (no active treatment, drug therapy, written instructions, usual care) within a 6 week to 6 month time frame. [EL = 1+] No direct comparisons of single-component behavioural therapy with multicomponent behavioural therapies were identified.

Prompted voiding and timed voiding strategies lead to reduced leakage episodes in cognitively impaired men and women. [EL = 1+]

From evidence to recommendations

Bladder training is less costly than most antimuscarinic drug treatment and is not associated with adverse effects. [EL = 4]

Recommendations for behavioural therapies

Bladder training lasting for a minimum of 6 weeks should be offered as first-line treatment to women with urge or mixed UI. [A]

If women do not achieve satisfactory benefit from bladder training programmes, the combination of an antimuscarinic agent with bladder training should be considered if frequency is a troublesome symptom. [A]

In women with UI who also have cognitive impairment, prompted and timed voiding toileting programmes are recommended as strategies for reducing leakage episodes. [A]

Research recommendation for behavioural therapies

A direct comparison of single-component and multicomponent behavioural therapy is required.

4.4. Drug therapies

4.4.1. Antimuscarinic drugs

Drugs with antimuscarinic action are used to treat OAB. They block muscarinic receptors in the bladder, which reduces the ability of bladder muscle to contract and affects bladder sensation. The drugs differ in their selectivity for various muscarinic receptors, and some drugs have additional actions such as direct smooth muscle effects.

Several systematic reviews of antimuscarinic drugs for the treatment of UI and/or OAB were identified.335–340 These reviews included studies conducted only in men, and in patients with neurogenic bladders, both of which are outside the scope of this guideline. The fully published RCTs included in these reviews were thus considered individually, together with other relevant RCTs. In addition, because of the relatively short duration of most RCTs, case series (so-called ‘extension studies’) were also considered as these provide longer term data.

The use of antimuscarinic drugs with bladder training was considered in the behavioural section (4.3.1). Studies evaluating imipramine, oxybutynin and tolterodine in this context were identified, which showed that the combination of antimuscarinic drugs and bladder training programmes resulted in greater reduction in frequency but did not lead to further improvements in incontinence.308–311

In the following section, placebo-controlled studies of antimuscarinic drugs are considered first, followed by comparisons of the drugs.

Placebo-controlled trials of antimuscarinic drugs

Darifenacin

Two DB RCTs compared darifenacin extended release (ER) with placebo in men and women (n = 398, n = 561; ~85% women) with urge UI, frequency and urgency. In both studies, significantly greater improvement in leakage episodes, frequency, urgency episodes and severity were seen with darifenacin 7.5–15 mg compared with placebo after 12 weeks’ treatment. Reductions in leakage episodes of 62–73% were reported with darifenacin compared with 49–56% with placebo, the reductions in frequency episodes were in the range 15–19% versus 8–10%, and for urgency 28–29% versus 11–13%. No significant differences were seen between darifenacin and placebo groups in changes in nocturnal awakening due to OAB. Adverse effects occurring more frequently with darifenacin included constipation (14–21% versus ~7%), dry mouth (19–31% versus 9%) and headache (4–7% versus 2–5%).341,342 One of the studies also had a darifenacin 3.75 mg treatment arm, for which no formal comparisons were made against placebo.341 [EL = 1+]

A further two placebo-controlled RCTs of 2 and 12 weeks’ duration were designed to evaluate the effects of darifenacin ER 30 mg on the outcome of warning time in men and women with urgency (n = 72; 71% women) or urge UI (n = 439).343,344 One study was of poor quality343 [EL = 1−] while the other was of good quality.344 [EL = 1+] Neither study reported significant differences between groups in this outcome, nor in urgency episodes. Urge UI episodes were significantly reduced with darifenacin compared with placebo in the larger study.343

Flavoxate

Two DB placebo-controlled crossover RCTs evaluated 2 weeks’ treatment with flavoxate for idiopathic DO.345,346 The first RCT in men and women (n = 41; only 25 analysed; 48% women) found no significant differences between flavoxate 200 mg t.d.s. and placebo in any urodynamic parameters. Complete results were not given for frequency, the only other outcome.345 [EL = 1−] The second RCT, in women only, found no significant differences between flavoxate 200 mg q.d.s. and placebo in frequency (median per three days 25 versus 23), nocturia (medians 3 versus 0), or leakage episodes (medians 1 versus 0) after treatment (n = 20). The most common adverse effects reported across all treatment groups were dry mouth (5–7%), and nausea or heartburn (2–7%).346 [EL = 1+]

A DB randomised study compared two different daily doses of flavoxate (600 or 1200 mg), given for 4 weeks to women with sensory and/or motor urge syndrome or incontinence (n = 27). Symptoms were scored on a scale of 0 to 2; no results were provided for individual symptoms although it was reported that total scores fell from baseline in both groups. Of the urodynamic variables evaluated, greater benefit was seen with the 1200 mg dose in volume at first desire to void and in bladder volume at capacity. Nausea was reported by about 22% of the women.347 [EL = 1−]

A further RCT compared a combination of flavoxate and imipramine with bladder training. Significantly more women were subjectively or objectively cured after 4 weeks’ bladder training than with drug therapy (n = 50).307 [EL = 1+]

Imipramine and other tricyclic antidepressants

No placebo-controlled RCTs evaluating the use of imipramine for UI were identified. A DB placebo-controlled crossover RCT involving 3 week treatment periods evaluated doxepin (50–75 mg at night) in women with DO and frequency, urgency or urge UI, who had failed to respond to other drugs, mainly antimuscarinics (n = 19). Significantly greater reduction in night leakage episodes and frequency were seen with doxepin compared with placebo, and a greater increase in maximum cystometric capacity. No significant differences were reported between groups in day leakage episodes, frequency or the 1 hour pad test. More doxepin-treated women reported adverse effects than those treated with placebo (68% versus 16%).348 [EL = 1+]

Oxybutynin

Four RCTs of 8–12 weeks’ duration compared various formulations and/or doses of oxybutynin (oral in three, transdermal in one) and tolterodine, in studies that also included placebo arms. The studies generally showed greater benefit in efficacy outcomes with oxybutynin and tolterodine compared with placebo, although with varying statistical significance. Reductions in frequency of 15–21% were seen with oxybutynin and tolterodine compared with 10–11% with placebo, reductions in leakage episodes were 46–77% versus 19–46%, and subjective improvement rates were 38–73% versus 22–53%.349–352

Another placebo-controlled crossover RCT evaluated immediate release (IR) oxybutynin 2.5–5 mg t.d.s. in cognitively impaired elderly nursing home residents who had not responded to 2 hourly prompted voiding (n = 75; 78% women). Significant improvement in leakage episodes was reported with IR oxybutynin after 20 days’ treatment (40% versus 18% had one or fewer episodes per day). No other outcomes were significantly different (change in leakage episodes, continent voids).331 [EL = 1+]

Transdermal oxybutynin was evaluated in a DB placebo-controlled RCT in women with urge UI and frequency, 66% of who had mixed UI (n = 520). Three doses were assessed: 1.3, 2.6 and 3.9 mg. Significantly greater improvement in outcomes was seen with oxybutynin 3.9 mg compared with placebo after 12 weeks’ treatment (leakage episodes, frequency and IIQ scores) but not with other dosages. Of the 22% previously treated with antimuscarinics, ‘similar trends’ in results were reported.353 [EL = 1+] A further 12 weeks’ open use of transdermal oxybutynin by 411 patients generally showed sustained improvement in leakage episodes, frequency, and QOL (IIQ) with all doses. Application-site reactions and dry mouth were common with oxybutynin.353 [EL = 3]

One RCT evaluated 12 days’ treatment with intravesical oxybutynin (20 mg) compared with placebo in women with frequency and DO (n = 56; 43 analysed). Nine women withdrew from treatment, three from the active group owing to the need for daily catheterisation and six from the placebo group owing to lack of improvement; all were excluded from the analyses. In the remaining women who were assessed 2 weeks after treatment, improvements in frequency and bladder capacity were seen in both groups, with no between-group analyses reported.354 [EL = 1−]

Propantheline

No placebo-controlled studies were identified for propantheline.

Propiverine

A DB placebo-controlled RCT was conducted to evaluate the cardiac effects of propiverine in men and women with frequency, urgency, and mixed or urge UI during 4 weeks’ treatment (n = 107; 98 analysed; 79% women). Urinary outcomes were reported although the groups were not balanced at baseline for these outcomes. Additionally, no between-group comparisons were made. Of the cardiac monitoring undertaken, the only difference noted between groups was a greater increase in minimum heart rate on a 24 hour electrocardiogram with propiverine than with placebo.355 [EL = 1−]

A randomised study evaluated four dosages of propiverine (15, 30, 45 and 60 mg daily) given for 3 weeks to men and women with urge UI (43%) or urgency (57%) (n = 185; 98% women). Improvements from baseline were seen in all dosage groups in frequency and in urodynamic parameters. Blurred vision was the most common adverse effect (8–26%), followed by dry mouth (6–27%). A dose–response effect was evident for adverse effects. Subjective efficacy and tolerability showed optimum effects with the 30 mg dose.356 [EL = 1+]

Solifenacin

One DB dose-finding RCT evaluated solifenacin daily doses of 2.5, 5, 10 and 20 mg compared with tolterodine 2 mg b.d. and placebo in men and women with idiopathic DO (n = 225; ~60% women). After 4 weeks’ treatment, reduction in frequency was significantly greater with 5, 10 and 20 mg solifenacin compared with placebo (18–23% versus 9%). No significant differences were seen between solifenacin and placebo groups in leakage and urgency episodes. QOL (CONTILIFE) was significantly improved in all solifenacin groups compared with placebo. Adverse effects seen with solifenacin (dry mouth, constipation, blurred vision) showed a dose–response relationship.357 [EL = 1+]

Two DB RCTs of 12 weeks’ duration compared solifenacin 5 mg o.d. and 10 mg o.d. with placebo in men and women with OAB (the proportions of women were 75% and 82%).358,359 Overall, 57% had UI in one study (47% urge UI),358 and 93% in the other (63% urge UI).359 One study also had a tolterodine 2 mg b.d. treatment arm.359 About 33% in both studies had had prior drug treatment for OAB; treatment response in these patients was not considered separately. In one of the studies, reductions in frequency, leakage and urgency episodes were significantly greater with solifenacin 5 and 10 mg compared with placebo. Reduction in nocturia was significantly greater with solifenacin 10 mg versus placebo (n = 857).358 In the second study, significantly greater reductions in frequency were seen with both solifenacin doses and with tolterodine compared with placebo. Reductions in urgency and leakage episodes were significantly greater with solifenacin 5 and 10 mg compared with placebo (n = 1033).359

Across the studies, the improvements in each outcome with solifenacin, tolterodine and placebo were: frequency 17–22% versus 15% versus 8–13%; leakage episodes 47–61% versus 59% versus 28–29%; urgency episodes 51–55% versus 38% versus 33%; nocturia 25–39% solifenacin versus 16% placebo. Adverse effects reported in both studies were dry mouth, constipation and blurred vision, which occurred in more solifenacin-treated patients than with placebo.358,359 [EL = 1+]

A total of 1633 (91%) of the patients from the two 12 week RCTs took solifenacin 5 or 10 mg for up to 1 year. The results indicate continued benefit. Dry mouth was the most common adverse effect, reported by 21%.360 [EL = 3]

Results for QOL (KHQ) during both RCTs358,359 and the case series360 have been published separately. Pooled results from the RCTs show significantly greater improvements in nine of ten domains (except personal relationships) with solifenacin 5 mg or 10 mg compared with placebo. The longer term study suggested sustained improvement.361 [EL = 3]

Tolterodine

Three DB RCTs compared 4 weeks’ treatment with tolterodine 1 mg and 2 mg b.d. with placebo in men and women with frequency, urgency and/or urge UI (72–75%). The majority of patients were women (65–79% of n = 670).362–364 One enrolled patients aged at least 65 years.362 In one study, significantly greater reductions in frequency, urgency and leakage episodes were seen with tolterodine 2 mg b.d., although groups were not balanced at baseline in frequency or leakage. Dry mouth was significantly more common in tolterodine-treated patients. Response to treatment in patients who had previously received drug treatment for OAB was not considered separately.362 [EL = 1−]

In the second study, significantly greater reductions in leakage episodes were seen with both tolterodine doses compared with placebo (41% versus 17%), with no significant differences between tolterodine and placebo in improvements in frequency (13% versus 10% reporting reductions). Of the 75% of patients who had poor efficacy response to prior drug treatment, 37–51% across the groups in this study had a ‘good response’. Dry mouth was significantly more common in the tolterodine 2 mg b.d. group compared with placebo (34% versus 6%).364 [EL = 1+] The third RCT only reported urodynamic outcomes, with no bladder diary data or patients’ perception of change. Increases in volume at first contraction and maximum cystometric capacity were significantly greater with tolterodine 2 mg b.d. compared with placebo.363 [EL = 1+]

Following completion of these three studies, patients were offered continued treatment with tolterodine 2 mg b.d. for a further 12 months.365 Overall, 62% continued treatment for this duration, with 23% reducing the dose to 1 mg b.d. Reasons stated for withdrawal were adverse events (15%), loss to follow-up or withdrew consent (17%). The results for bladder diary variables indicate sustained benefit in those who continued treatment. Dry mouth was the most common adverse effect.365 [EL = 3]

One DB placebo-controlled RCT evaluated 12 weeks’ treatment with tolterodine 1 mg and 2 mg b.d. in men and women with DO (n = 316; 75% women). Significantly greater reductions in frequency were seen with both tolterodine groups compared with placebo (~20% versus 12%). More patients treated with tolterodine 2 mg b.d. reported improvement compared with other groups. No other significant differences were seen (leakage episodes, cure or adverse effects). Overall, 46% had had prior drug treatment for OAB; results in these patients were not considered separately.366 [EL = 1+]

Four RCTs of 8–12 weeks’ duration that compared oxybutynin (oral in three, transdermal in one) with tolterodine also had placebo arms. The studies generally showed greater benefit in efficacy outcomes with oxybutynin and tolterodine compared with placebo, although with varying statistical significance. Reductions in frequency of 15–21% were seen with oxybutynin and tolterodine, compared with 10–11% with placebo; reductions in leakage episodes ranged from 46% to 77% versus 19–46%, and subjective improvement rates ranged from 38% to 73% versus 22–53% (see later).349–352

Extended release tolterodine

Two formulations of tolterodine (2 mg b.d. and 4 mg ER o.d.) were compared in a 12 week placebo-controlled DB RCT in men and women with frequency and urge UI (n = 1529; 81% women). Two-thirds of the population were aged at least 65 years. Significantly greater improvements in outcomes were seen in both tolterodine groups compared with placebo (reductions in frequency of 15% versus 17% versus 11%; leakage episodes 46% versus 53% versus 30%). QOL was assessed using KHQ and SF-36 questionnaires. In both tolterodine groups, improvements in six or seven of the ten KHQ domains were greater than with placebo. No differences were seen between groups in the SF-36. Dry mouth was reported by significantly fewer patients treated with ER than standard tolterodine (23% versus 30%, compared with 8% placebo).367–373 [EL = 1+] No differences in efficacy or tolerability were seen in older (65 years or above) compared with younger patients.369

Following completion of this study, patients were offered continued treatment with ER tolterodine 4 mg for 12 months.374 Overall, 71% continued treatment (n = 1077; 82% women). Reasons stated for withdrawal included adverse events (10%) and loss to follow-up or withdrawal of consent (8%). The results for bladder diary variables indicated sustained benefit for those who continued treatment. Dry mouth was the most common adverse effect.374 [EL = 3]

A further DB placebo-controlled RCT compared ER tolterodine 4 mg o.d. with placebo in women with urge-predominant mixed UI, frequency and urgency over 8 weeks (n = 854). Significantly greater improvements were seen with tolterodine compared with placebo in frequency, urge leakage episodes, urgency, subjective improvement and QOL (KHQ). In women with predominant stress UI (25%), improvements in leakage episodes were not significantly different between groups. Significantly more tolterodine-treated women reported dry mouth.375 EL = 1++]

Trospium

Results from two placebo-controlled RCTs376,377 that evaluated 3 weeks’ treatment with trospium 20 mg b.d., in men and women with DO were pooled in a meta-analysis (n = 508; 69% women). The main outcomes were urodynamic parameters, with greater improvement reported with trospium in maximum cystometric capacity and volume at first contraction. More trospium-treated patients reported subjective cure or marked improvement of symptoms.378 [EL = 1+]

Three further placebo-controlled RCTs evaluated trospium.379–381 Two evaluated trospium 20 mg b.d. in men and women with symptoms of OAB (total n = 1170; 78% women). Improvements in efficacy outcomes were significantly greater with trospium after 12 weeks’ treatment (frequency, urgency, urge UI episodes, QOL [IIQ]). Dry mouth was reported by more trospium-treated patients (21% versus 6%).380,381 [EL = 1+] A smaller RCT of 4 weeks’ duration compared trospium 15 mg t.d.s. with placebo in men and women with urge UI. No differences in adverse effects or in maximum cystometric capacity were found between groups, although groups were not balanced at baseline for cystometric capacity (n = 46; ~92% women).379 [EL = 1−]

Comparisons of antimuscarinic drugs

Immediate release oxybutynin versus flavoxate

A DB crossover RCT compared flavoxate 400 mg t.d.s. with IR oxybutynin 5 mg t.d.s. in women with urgency (n = 50; only 41 analysed). No significant differences were found between groups in urodynamic findings or subjective cure or improvement after 4 weeks’ treatment. Bladder diary outcomes were assessed using a scoring system of 0 to 2, with no between-group analyses reported. Significantly more oxybutynin-treated women reported adverse effects (90% versus 27%).382 [EL = 1−]

Immediate release oxybutynin versus propantheline

A single-blind crossover RCT compared propantheline 15 mg t.d.s. with IR oxybutynin 5 mg t.d.s. in women with OAB (n = 23). Significantly greater increases in cystometric capacity were seen with oxybutynin compared with propantheline after 4 weeks’ treatment (36% versus 17%), with no differences between treatments in frequency, subjective improvement or adverse effects.383 [EL = 1+]

Immediate release oxybutynin versus propiverine

One DB placebo-controlled RCT compared 4 weeks’ treatment with IR oxybutynin 5 mg b.d. and propiverine 15 mg t.d.s. in men and women with urgency or urge UI (n = 366; 310 analysed; 93% women). Physician’s assessment of improvement, and incidence of adverse effects, was significantly higher with both drugs than with placebo. No other significant differences were identified between the three groups (frequency, urgency or cystometric capacity). Results for only 85% were analysed, with no explanation for withdrawals.384 [EL = 1−]

Oxybutynin versus tolterodine

Oxybutynin and tolterodine were compared in ten RCTs, six of which were DB comparisons of standard (immediate release) formulations of both drugs (oxybutynin 5 mg b.d. or t.d.s. with tolterodine 2 mg b.d.).349,350,385–388 The four other comparisons were: transdermal oxybutynin versus ER tolterodine;351 IR oxybutynin versus ER tolterodine;352 ER oxybutynin versus IR tolterodine;389,390 and ER formulations of both drugs.391–393

Immediate release oxybutynin versus tolterodine

Four DB studies compared the effectiveness of oxybutynin and tolterodine, in men and women (predominantly women; 67–77%) with OAB. Duration of treatment ranged from 8 to 12 weeks.349,350,385,386 Across the studies, between 27% and 60% of the study populations had previously received drug treatment for OAB or urge UI. The response to treatment in these groups was not considered separately in the study reports.

Two of the four RCTs compared tolterodine 2 mg b.d. with oxybutynin 5 mg b.d. (starting at a lower dose of 2.5 mg b.d.). The primary outcome of dry mouth occurred in significantly more oxybutynin-treated patients than with tolterodine (61% versus 37%). No significant differences were seen between groups in efficacy outcomes (n = 378).386 [EL = 1+] A study in Asian patients reported similar findings (n = 228).385 [EL = 1+]

Two placebo-controlled RCTs of identical design compared tolterodine 2 mg b.d. with oxybutynin 5 mg t.d.s., although one only reported efficacy data for completers (53% of those randomised). Both found no significant differences between treatments in frequency or in leakage episodes, while significantly more people treated with oxybutynin reported dry mouth. Improvements in efficacy outcomes were significantly greater with tolterodine and oxybutynin compared with placebo (n = 293),349 [EL = 1+] (n = 277; 147 analysed).350 [EL = 1−] Following completion of these two studies (and two other RCTs), patients were offered continued open treatment with tolterodine 2 mg b.d. for a further 9 months.394 Overall 70% continued treatment, with 13% reducing the dose to 1 mg b.d. Reasons stated for withdrawal were adverse events (9%), lack of efficacy (6%), loss to follow-up or withdrew consent (10%). Bladder diary variables indicated sustained benefit in those who continued treatment. Dry mouth was the most common adverse effect (28%).394 [EL = 3]

Two open RCTs with specific objectives provide further comparative data for tolterodine and oxybutynin in women with OAB. Duration of treatment was 10–12 weeks; one was a crossover study.387,388 The main objective of one study was to assess whether urodynamic grade predicts response to treatment; this also reported no differences between tolterodine 2 mg b.d. and oxybutynin 5 mg t.d.s. in frequency or cystometric capacity (n = 128; 107 analysed).387 [EL = 1−] The other RCT evaluated the impact of dry mouth with tolterodine 2 mg b.d. and oxybutynin 5 mg b.d. using a ‘Xerostomia Questionnaire.’ The authors reported no significant differences between groups in this outcome, but no numerical results were presented.388 [EL = 1+]

Transdermal oxybutynin versus extended release tolterodine

Transdermal oxybutynin 3.9 mg was compared with ER tolterodine 4 mg o.d. in a placebo-controlled DB RCT in men and women with frequency and urge UI (n = 361; 93% women). No significant differences were seen between active treatments in any outcome (frequency, leakage episodes, QOL, adverse effects) after 12 weeks’ treatment. Improvements in frequency were significantly greater with tolterodine compared with placebo, and significantly greater benefit was seen in leakage episodes and subjective cure or improvement with both active drugs compared with placebo.351 [EL = 1+]

Immediate release oxybutynin versus extended release tolterodine

Extended release tolterodine 4 mg o.d. was compared with IR oxybutynin 3 mg t.d.s. in a placebo-controlled DB RCT in men and women with urgency, frequency, and urge UI (n = 605; 70% women). No differences in efficacy measures were seen between tolterodine or oxybutynin, but fewer people treated with tolterodine experienced dry mouth after 12 weeks’ treatment. Improvements in leakage episodes and frequency were significantly greater with active treatment compared with placebo. No significant differences were seen between the three groups in the proportion of people reporting improvement. Overall, 24% had received prior drug treatment for OAB; the response in these patients was not considered separately.352 [EL = 1+] Treatment was continued in 31% of patients for up to 12 months; the results available indicate continued efficacy.395 [EL = 3]

Extended release oxybutynin versus immediate release tolterodine

One DB RCT compared IR tolterodine 2 mg b.d. with ER oxybutynin 10 mg o.d. in men and women with urge UI, of whom 40% had previously been treated with antimuscarinic drugs (n = 378; 83% women). Results were presented for study completers only (88%), with statistical analysis quoted for all patients randomised, which indicated consistent effects. Significantly greater reductions in leakage episodes (urge and total) and in frequency were seen with ER oxybutynin compared with IR tolterodine after 12 weeks’ treatment. No significant differences were identified between groups in adverse effects reported.389,390 [EL = 1+]

Extended release oxybutynin versus extended release tolterodine

One DB RCT compared ER formulations of both tolterodine and oxybutynin in women with OAB, 47% of whom had previously received antimuscarinic treatment. No significant differences were found between ER tolterodine 4 mg o.d. and ER oxybutynin 10 mg o.d. in changes in leakage episodes (urge or total), while reductions in frequency were reported to be significantly greater with ER oxybutynin (mean reductions of 28% versus 25%) with 12 weeks’ treatment. Dry mouth occurred in significantly more women in the oxybutynin group (30% versus 22%) (n = 790).391–393 [EL = 1+]

Immediate release oxybutynin versus trospium

One DB RCT compared IR oxybutynin 5 mg b.d. with trospium 20 mg b.d. in men and women with urge syndrome (n = 357; 86% women). After 1 year, improvements in frequency, urgency and cystometric capacity were seen in both groups, with between-group analysis only reported for cystometric capacity. The incidence of dry mouth and gastrointestinal effects was significantly higher with oxybutynin (50% versus 33% and 51% versus 39%, respectively).396 [EL = 1+]

Solifenacin versus extended release tolterodine

In men and women (n = 1200; 87% women) with OAB, solifenacin 5 to 10 mg (dose increased on patient request in 48%) was compared with ER tolterodine 4 mg daily in a 3 month study. The proportion of patients with UI was not reported. The aim of the study was to demonstrate non-inferiority of solifenacin to ER tolterodine in frequency of micturition, which was proven. The two drugs were not significantly different in terms of improvements in nocturia, but improvements in leakage and urgency episodes were significantly greater with solifenacin, as was the reduction in pad usage. The patients’ perception of their bladder condition was also reportedly improved to a greater extent with solifenacin than with tolterodine. The adverse effects listed occurred with both drugs: dry mouth (30% solifenacin versus 24% tolterodine), constipation (6% versus 3%) and blurred vision (1% versus 2%). Discontinuation rates were 6% versus 7%.397 [EL = 1+]

Different formulations of the same drug compared

Transdermal versus immediate release oral oxybutynin

Transdermal oxybutynin (2.6–5.2 mg) and IR oral oxybutynin (5 mg b.d. to 7.5 mg t.d.s.), were compared in a 6 week RCT in men and women (92% women) with DO, all of whom had urge UI and who were currently responding to oral oxybutynin (n = 76). No significant differences were identified in efficacy (leakage episodes, cure) between treatment groups. Significantly more patients reported dry mouth with oral oxybutynin (39% versus 82%).398 [EL = 1+]

Extended versus immediate release oxybutynin

Four DB RCTs compared ER and IR oral oxybutynin formulations in men and women with urge UI,399–402 (with frequency in two studies399,401). In three RCTs, the population had previously responded to oxybutynin or to other antimuscarinic treatment.400–402 Treatment duration ranged from 4 to 6 weeks. Three of the studies allowed dose titration of oxybutynin; the daily doses taken were within 5–30 mg (ER) or 5–20 mg (IR formulation).399,400,402 One study compared 10 mg given as a single ER daily dose or a twice daily dose of IR oxybutynin.401 Each study included men and women (n range 105 to 226), with the majority in each study being female (68–92%).

In both 6 week studies, only adverse effects were reported for all patients; efficacy outcomes were only reported for those who completed treatment,400 or who completed at least 2 weeks’ treatment and followed the protocol.399 [EL = 1−] The other two studies were of higher quality.401,402 [EL = 1+] Two of the studies reported no significant differences in efficacy or adverse effects between ER and IR oxybutynin.399,401 One study reported that the incidence of dry mouth was significantly lower with ER oxybutynin (68% versus 87%).400 One found no significant differences between ER and IR oxybutynin in incidence of dry mouth, although the cumulative rate of the first report of dry mouth, at a given dose, was significantly lower with ER oxybutynin.402

A case series evaluated treatment with ER oxybutynin 5–30 mg for 1 year in men and women with urge or mixed UI (n = 1067; 85% women). Overall, 46% continued treatment to 1 year; the main reasons for withdrawal were adverse effects (24%) and lack of efficacy (10%). Significant improvement was seen in impact on sleep and in the effects of leakage on lifestyle at 1 year (measured using the IIQ questionnaire).403 [EL = 3]

Extended versus immediate release tolterodine

In one placebo-controlled DB RCT, the reduction in leakage episodes was significantly greater with ER tolterodine 4 mg o.d. compared with IR tolterodine 2 mg b.d. (median reductions 71% versus 60% at 12 weeks) in men and women with frequency and urge UI (n = 1529; 81% women). Dry mouth was reported by significantly fewer patients treated with ER than IR tolterodine (23% versus 30%, compared with 8% placebo).367–373 [EL = 1+]

Economic evidence

From the health economics literature review, a total of seven articles were included as being relevant to the question of which is the most cost effective drug treatment for OAB. The design and the results of all included studies are presented in the evidence tables. All the studies included economic models, and the efficacy data used to populate these models was derived from a number of sources – trial data, literature review and expert opinion. All but two of the studies made some comparison between one or more of the formulations for oxybutynin and tolterodine. Two of the studies were based exclusively on a UK setting.404,405 One other study also considered the UK context, in addition to France and Austria.406

In the studies that considered oxybutynin, the general conclusion was that ER oxybutynin was cost effective. Two studies reported that it dominated IR tolterodine, being at least as cheap and more efficacious.404,407 In addition, a further two studies reported that ER oxybutynin dominated IR oxybutynin in addition to ER tolterodine.406,408 One UK study noted that IR oxybutynin was cheaper than ER oxybutynin and all formulations of tolterodine.405 It reported that IR oxybutynin and ER formulations of oxybutynin and tolterodine might all be considered cost effective contingent on the willingness of the NHS to pay for an additional incontinent-free week. However, it noted that IR tolterodine did not appear cost effective as it was more expensive but no more efficacious than ER formulations.

One Canadian study reported that tolterodine appeared cost effective in a population who had discontinued initial oxybutynin therapy.409 The comparison was against no further treatment and they argued that the incremental cost effectiveness ratio (ICER) fell within cost effectiveness thresholds for willingness to pay. Finally, a Swedish study compared the cost effectiveness of tolterodine against no treatment.410 They reported that the incremental cost per quality-adjusted life year (QALY) with tolterodine fell within the threshold generally accepted as cost effective.

Where treatments are of equivalent efficacy, the cheapest treatment will generally be the most cost effective. Since we did not find consistent evidence of greater efficacy of one antimuscarinic over another, a cost minimisation analysis has been adopted in this guideline. One possible criticism of such a cost minimisation approach is that it does not sufficiently take into account the differences between drugs and formulations in terms of their adverse effects and tolerability profile. While recognising a certain validity in this criticism, we believe that the cost minimisation approach was justified because of a lack of head-to-head trials, the difficulties of comparing across studies using different study designs, and evidence from actual practice showing low persistence with all antimuscarinic therapy.403,412 Furthermore, good prescribing practice requires that patients be reviewed when starting on new pharmacological treatment and therefore the costs of switching from a poorly tolerated drug are, at least partly, subsumed within this review process. The costs (see Appendix E) were based on a typical dose, as determined by the GDG, taken for 12 months and using prices published in BNF 50. Based on this, non-proprietary oxybutynin is the most cost effective.

Evidence statements for antimuscarinic drugs

Treatment with darifenacin, oxybutynin, solifenacin, tolterodine and trospium in women with OAB is associated with improvements in frequency, leakage episodes and quality of life. [EL = 1+] There is no evidence of a clinically important difference in efficacy between antimuscarinic drugs. Based on the cost minimisation analysis undertaken, non-proprietary immediate release oxybutynin is the most cost effective antimuscarinic drug.

Propiverine may be associated with an improvement in frequency. [EL = 1+] There is limited evidence that doxepin reduces night-time leakage episodes and nocturia. [EL = 1+] There is no evidence of efficacy for the use of flavoxate, propantheline or imipramine for the treatment of UI or OAB. [EL = 4]

Antimuscarinic adverse effects are common with all antimuscarinic drugs, and dry mouth is more likely with oral IR oxybutynin than tolterodine, trospium, ER or transdermal oxybutynin, but skin reactions are very common with the latter. [EL = 1+] In view of the high incidence of adverse effects and the time to maximum benefit the GDG believes that early treatment review is good practice. [EL = 4]

Recommendations for antimuscarinic drugs

Immediate release non-proprietary oxybutynin should be offered to women with OAB or mixed UI as first-line antimuscarinic drug treatment, if bladder training has been ineffective. If immediate release oxybutynin is not well tolerated, darifenacin, solifenacin, tolterodine, trospium or an extended release or transdermal formulation of oxybutynin should be considered as alternatives. Women should be counselled about the adverse effects of antimuscarinic drugs. [A]

An early treatment review should be undertaken following any change in antimuscarinic drug therapy. [D (GPP)]

Propiverine should be considered as an option to treat frequency of urination in women with OAB, but is not recommended for the treatment of UI. [A]

Flavoxate, propantheline and imipramine should not be used for the treatment of UI or OAB in women. [A]

Research recommendation for antimuscarinic drugs

There is a need for a comparison of the clinical effectiveness and cost effectiveness of drug therapy compared with other conservative therapy as first-line treatment for women with OAB or mixed UI.

4.4.2. Desmopressin

Desmopressin (also known as DDAVP) is a synthetic analogue of vasopressin or antidiuretic hormone, which acts by inhibiting diuresis while avoiding vasopressive effects. Used at night, it decreases nocturnal urine production.

Nocturia

Three DB placebo-controlled RCTs evaluated the use of desmopressin for nocturia. One evaluated 3 weeks’ treatment in women (n = 144),413 who were then offered continued desmopressin treatment for up to 1 year.414 The two other RCTs were smaller crossover studies that evaluated 2 weeks’ treatment (n = 17, n = 25),415,416 one of which enrolled men and women.415 An oral formulation of desmopressin was used in two studies (dose ranging from 100 to 400 μg),413–415 and an intranasal formulation in the third (dose 20 μg).416

In the 3 week RCT in women, significantly greater improvements were seen with desmopressin in all nocturia-related outcomes (nocturia episodes, volume of nocturnal voids, duration of sleep to first nocturnal void, diuresis, ratio of day/night to 24 hour urine volume), and a reduction in the ‘bothersome factor’ of nocturia (assessed using BFLUTS).413 [EL = 1+] Follow-up of women who took desmopressin for up to 1 year indicated sustained benefit in the outcomes measured (50% or greater reduction in nocturia, duration of sleep to first nocturnal void, ‘bothersome factor’).414 [EL = 3]

In the crossover study in men and women, nocturia episodes and nocturnal diuresis were significantly lower with desmopressin compared with placebo, with no significant change in either group in 24 hour diuresis.415 [EL = 1+] Nocturnal frequency and urine output were significantly reduced with desmopressin (from baseline and compared with placebo) in the crossover study in women only, with no significant change in diurnal outcomes.416 [EL = 1+]

Urinary incontinence

A placebo-controlled ‘pilot’ RCT evaluated the use of desmopressin in the treatment of women with daytime UI. Four sequences of desmopressin 40 μg (seven doses) or placebo (three doses) were evaluated, with treatment administered intranasally when required. At both 4 and 24 hours after dose administration, the number of periods with no leakage was greater with desmopressin compared with placebo, and the volume voided or leaked during a UI episode lower with active treatment, although the confidence intervals for all mean values overlapped, indicating that differences were not statistically significant.417 [EL = 1+]

Adverse effects

Adverse effects reported across the short-term studies included headache, nausea, hyponatraemia, abdominal pain, frequency, dry mouth, dizziness, fatigue, peripheral oedema and earache.415–417 In the longer term study (up to 1 year), the most frequent adverse effects related to treatment in women were: hyponatraemia 12% (none required treatment, none with sodium below 125 mmol/l); headache 7%; frequency, peripheral oedema and UTI (each 3%); and nausea and dizziness (each 2%).414

Evidence statements for desmopressin

The use of desmopressin significantly reduces nocturia. There is insufficient evidence that desmopressin reduces incontinence in adult women. A reduction in serum sodium is very common (more than 10%). [EL = 1+]

Symptomatic hyponatraemia due to therapy with desmopressin may be more common in elderly women, and is more likely to occur soon after treatment initiation. Pretreatment and early post-treatment (72 hours) serum sodium monitoring is recommended. Where there are new symptoms or a change in medication, further measurement of serum sodium is recommended. [EL = 4]

Recommendations for desmopressin

The use of desmopressin may be considered specifically to reduce nocturia in women with UI or OAB who find it a troublesome symptom. [A]

However, the use of desmopressin for nocturia in women with idiopathic UI is outside the UK marketing authorisation for the product. Informed consent to treatment should be obtained and documented.

4.4.3. Diuretics

Only one RCT involving women was identified. This was a small DB placebo-controlled crossover study that evaluated bumetanide 1 mg for the treatment of nocturia in men and women (n = 33; 28 completed; 13 women). Treatment was given 4–6 hours before bedtime for 2 weeks. Weekly nocturia episodes were significantly fewer after bumetanide treatment compared with placebo (10 versus 14).418 [EL = 1−] No studies evaluating furosemide in women were identified.

Evidence statement for diuretics

There is insufficient evidence to support the use of diuretics for the treatment of nocturia in women with UI. [EL = 1−]

4.4.4. Duloxetine

Duloxetine is a serotonin and noradrenaline reuptake inhibitor that acts chiefly in the sacral spinal cord. It is thought that the resultant increase in pudendal nerve activity increases urethral sphincter contraction and closure pressure. It is licensed for use in moderate to severe stress UI.

A systematic review has considered the effectiveness of serotonin and noradrenaline reuptake inhibitors (duloxetine) for the treatment of stress UI.419 Because some of the studies included in the review were only published as abstracts, and because some relevant studies were not included in the review, all relevant studies are considered individually.

Six DB placebo-controlled RCTs evaluated the effectiveness of duloxetine for the treatment of predominant stress UI in women, five of which were considered to be of good quality.420–424 [EL = 1+] A further RCT compared duloxetine (with or without PFMT) with PFMT alone or no active treatment (placebo drug and sham PFMT) in women with stress UI, 11% of whom had had prior continence surgery (n = 201).229

Of the six placebo-controlled RCTs, one was a 12 week dose-ranging study, comparing 20, 40 and 80 mg daily doses of duloxetine in women with at least four leakage episodes per week (n = 553).420 Three other 12 week studies, identical in design, evaluated duloxetine 80 mg in women with at least seven (mean about 17) leakage episodes per week (total n = 1635).421–423 The fifth study considered the impact of duloxetine 80 mg on QOL after 9 months’ treatment (n = 451).424 [EL = 1+] One study evaluated duloxetine use in women awaiting surgery for stress UI (n = 92).425 [EL = 1−] Across the six studies, between 8% and 18% had had prior continence surgery. Up to 35% of women across four studies performed PFMT.420–423

The outcomes evaluated across the 12 week studies were leakage episodes, voiding interval, QOL (I-QOL) and patient global impression of improvement (PGI-I). The findings for duloxetine 80 mg per day (40 mg b.d.) compared with placebo were as follows:

  • leakage episodes: significantly greater median reductions with duloxetine in each study (range of median reductions 50–64% versus 28–41% placebo)420–423
  • voiding interval: significantly greater increases with duloxetine in each study (range of mean increases 15–24 minutes versus 4–9 minutes)420–423
  • QOL: significantly greater improvement (I-QOL) with duloxetine in three studies;420–422 no significant difference between groups in one423
  • PGI-I: significantly more women reported improvement with duloxetine in three studies;420–422 no significant difference between groups in one.423 [EL = 1+]

The lower daily duloxetine dose of 20 mg b.d. (40 mg daily) was associated with significantly greater reductions in leakage episodes and in frequency compared with placebo, but not in QOL or PGI-I.420 [EL = 1+]

The RCT that focused on QOL as an outcome after 9 months’ treatment found no significant differences between duloxetine 80 mg and placebo in increases in I-QOL scores or in PGI-I at 3 or 9 months. About one-quarter of the women had dropped out of the study at the endpoint.424 [EL = 1+]

In the study of women awaiting surgery, significantly greater improvements in leakage episodes, I-QOL and PGI-I scores were seen with duloxetine compared with placebo. A ‘willingness to consider surgery’ questionnaire indicated that a greater proportion of women treated with duloxetine versus placebo would change their mind about having surgery, although seven women (one duloxetine, six placebo) were excluded from this analysis.425 [EL = 1−]

In the comparison of duloxetine 80 mg (with or without PFMT) with PFMT alone or no active treatment (sham PFMT and placebo drug), PFMT involved a target of 200 contractions a week (over 4 days), and the ‘knack’. Sham PFMT involved contracting hip abductor muscles. Significantly greater reductions in leakage episodes were reported with duloxetine (with or without PFMT) compared with PFMT alone after 12 weeks’ treatment. Global improvement and I-QOL scores indicated greater improvement in the duloxetine plus PFMT group compared with no active treatment.229 No significant differences were reported between the PFMT and placebo groups in any outcome (leakage episodes, global improvement, I-QOL scores) after 12 weeks’ treatment.229 [EL = 1+]

Adverse effects

Across all studies, significantly more women in all duloxetine dosage groups discontinued treatment owing to adverse effects compared with placebo (range 15–33% versus 0–6%). Nausea was significantly more common with all daily dosages of duloxetine (range 13–46% versus 2–13% placebo), and accounted for a significant proportion of withdrawals compared with placebo in one study.422 Other adverse effects that occurred significantly more commonly with duloxetine in two or more studies were dry mouth, constipation, fatigue, insomnia, dizziness, increased sweating, vomiting and somnolence.229,421–423,425 Adverse effects occurring in the PFMT and no active treatment arms of the duloxetine study were pooled, and thus a distinction between duloxetine and PFMT or no active treatment is not possible.229

Cost effectiveness of duloxetine

Duloxetine is the only drug therapy currently available for stress UI. As part of this guideline, we considered the cost effectiveness of duloxetine in order to inform recommendations about the sequencing of conservative therapies, something that could potentially have a large impact on clinical practice.

One published article considered the cost effectiveness of duloxetine.426 This described a state transition (Markov) model to evaluate the cost effectiveness of duloxetine alone or in combination with PFMT against ‘standard’ treatment (PFMT and surgery), either as a first-line treatment or as second line to PFMT for stress UI, over 2 years. The Markov approach was adopted so as to capture the effect of waiting times on access to services, with a concomitant effect on deferred costs and benefits. The results of the model suggested that, using a willingness to pay threshold of £30,000 per QALY, duloxetine was a cost effective treatment for stress UI. Using baseline assumptions, the authors reported that the ICER of duloxetine alone when used first-line was £8,730 per QALY and £5,854 per QALY when used first-line in combination with PFMT.426 When used second-line to PFMT, duloxetine dominates standard treatment.

The health economics model of the use of duloxetine as a first-line treatment for stress UI in women used in this study factored in delays for access to services.426 Clinical trials would normally try to eliminate differential timings of treatment, to ensure a like-for-like comparison. Similarly, it could be argued that economic evaluation should be neutral with respect to treatment timing so that the results are not contingent on a particular service configuration. Indeed, it seems that service delivery should be configured to produce cost effective health care rather than be a driver of what is deemed cost effective. Therefore, two additional health economics models, which did not consider access times to health services, were developed for this guideline. The first sought to compare the cost effectiveness of PFMT versus duloxetine as a first-line treatment for women with moderate to severe stress UI, which is assumed to be 14 or more leakage episodes per week. In this model, treatment effects and costs were based on a 52 week time frame. A second model, based on a 2 year follow-up, assessed the cost effectiveness of duloxetine versus surgery as a second-line treatment for women with moderate to severe stress UI in whom first-line treatment with PFMT had been unsuccessful. These models are described in detail in Appendix F.

Under baseline assumptions, PFMT ‘dominates’ duloxetine as a first-line treatment. This means that it is both more effective and less costly. The sensitivity analyses undertaken (and detailed in Appendix F) did not change this conclusion. The second-line treatment model suggests that surgery is more cost effective than duloxetine. Surgery is more costly but the ICER, in the baseline analysis, falls below the £20,000 per QALY threshold used by NICE as a willingness to pay benchmark for cost effectiveness.

Evidence statements for duloxetine

Short-term studies (up to 12 weeks) suggest that the use of duloxetine is associated with a reduction in leakage episodes, an increased voiding interval and improved quality of life in women with stress UI or mixed UI where stress-related leakage is the predominant symptom. Between-group differences are clinically small. Adverse effects, particularly nausea, and discontinuation rates are very common (more than 10%). There is a lack of long-term safety data. The combination of duloxetine and PFMT is more effective than no treatment. It remains unclear whether the combination is better than either treatment alone. [EL = 1+]

An economic model constructed for the purposes of this guideline suggested that PFMT is more cost effective than duloxetine alone as first-line treatment for stress UI. This result was generally not affected by making plausible changes to model parameters in favour of duloxetine. While the model was based on the best available clinical evidence, there is a lack of long-term effectiveness data for either treatment. A second model suggested that surgery was cost effective relative to duloxetine as a second-line treatment to PFMT. However, duloxetine was the lower cost treatment option and therefore its use does not necessarily impose opportunity costs on the NHS relative to surgery.

Recommendation for duloxetine

Duloxetine is not recommended as a first-line treatment for women with predominant stress UI. Duloxetine should not routinely be used as a second-line treatment for women with stress UI, although it may be offered as second-line therapy if women prefer pharmacological to surgical treatment or are not suitable for surgical treatment. If duloxetine is prescribed, women should be counselled about its adverse effects. [A]

4.4.5. Oestrogens

Oestrogens help to maintain health of the tissues that are essential for normal pressure transmission in the urethra. This includes the sphincter muscles, urothelium and vascular tissues, as well as the urethral secretions that may help to create a ‘seal’. Oestrogen replacement has been promoted as a solution to UI in postmenopausal women, although its chief mode of action is unclear.

Four systematic reviews of oestrogens for the treatment of UI and/or OAB were identified, which were completed at different times and which considered different questions.427–430 Therefore, studies included in these reviews were considered individually, together with other relevant RCTs.

Ten RCTs evaluated the use of oestrogens for the treatment of postmenopausal women with stress UI228,431–434 or other types of UI or OAB.435–439 Four further RCTs that primarily evaluated the effects of oestrogen on symptoms of vaginal atrophy reported some data for urological symptoms.440–443 Additionally, secondary analyses of data from three RCTs that were designed to evaluate the benefits and risks of hormone replacement therapy (HRT) provide data on UI and/or OAB.444–449

The comparator group was placebo in all except two studies. One of these two studies is also considered in the physical therapies section and involved a comparison of PFMT with oestrogen, electrical stimulation and no treatment.228 The second study compared two oestrogen preparations.436

Intravaginal oestrogens

Five RCTs evaluated the use of intravaginal oestrogens for UI and/or other urological symptoms. Two RCTs enrolled women with stress UI (n = 100). One reported significantly greater subjective improvement of incontinence with intravaginal estriol compared with placebo at 6 months (68% versus 16%, n = 88).431 [EL = 1+] No comparisons were made between the conjugated oestrogen cream and control groups in the 3 month RCT; cure rates were 12% versus 0.228 [EL = 1+]

One RCT compared two intravaginal oestrogen preparations (the estradiol vaginal ring with estriol pessaries) in women with urgency, frequency, stress or urge UI. No significant differences were seen between groups in any outcome after 6 months’ treatment (subjective improvement; responder or cure rates for urgency, frequency, nocturia, urge or stress UI). Responder rates across all outcomes ranged from 51% to 61%, and cure rates from 27% to 44% (n = 251).436 [EL = 1+]

Two placebo-controlled RCTs, in which an intravaginal oestrogen preparation was used to treat urogenital symptoms, reported the following:439,443 [EL = 1+]

  • The prevalence of UI and frequency/nocturia fell to a greater extent with an intravaginal estradiol tablet than with placebo at 1 year (UI prevalence 18% versus 10%; frequency/nocturia 38% versus 10%) but no between-group analyses were reported. At baseline, 28% of women had UI and 43% frequency or nocturia (n = 1612).439
  • Significantly more women who had urological symptoms (41–53%) reported improvement in symptoms (frequency, dysuria, urge or stress UI) with intravaginal 17β-estradiol versus placebo after 3 months’ treatment (63% versus 32%, n = 164).443

Systemic oestrogens

Systemic oestrogens for UI or OAB

Three RCTs evaluated oral oestrogens for the treatment of stress UI for 3 or 6 months.432–434 The oestrogens evaluated were conjugated equine oestrogen (CEE) with medroxyprogesterone acetate (MPA), given for 10 days a cycle;432 estradiol;433 and estrone.434 No significant differences were seen between oestrogen and placebo groups in any outcome across the studies (leakage episodes, pad tests, frequency, QOL, perception of improvement, objective cure).432–434 [EL = 1+]

Two RCTs evaluated systemic oestrogens for the treatment of stress or urge UI.435,438 One reported no differences between a subcutaneous estradiol or placebo implant in subjective outcomes (self-reported cure, leakage episodes, frequency) after 6 months’ treatment (n = 40).438 [EL = 1+] In the second RCT, improvements in leakage episodes, frequency and urgency were seen after 3 months’ treatment with oral estriol and placebo, but no between-group differences were reported (n = 56).435 [EL = 1−]

A further RCT evaluated CEE+MPA in female nursing home residents who were incontinent. No significant differences were found between CEE+MPA and placebo groups in any outcome (leakage, bladder capacity), although only data from 21 of the 32 women randomised, who completed 6 months’ treatment, were analysed.437 [EL = 1−]

Systemic oestrogens for urogenital symptoms

Three RCTs that primarily evaluated the effects of systemic oestrogen on symptoms of vaginal atrophy reported some continence data.440–442 One of these studies compared oral estradiol and estriol with placebo in women with stress or mixed UI. At 4 months, a higher cure rate was reported for women on active treatment compared with placebo, although no baseline data were given (n = 29).440 [EL = 1−] A second RCT comparing oral estriol with placebo in women with stress, urge or mixed UI was of unclear duration (3 or 6 months) and only reported that symptoms were alleviated in the majority of women with urge or mixed UI (n = 34).441 [EL = 1−] No significant changes in frequency were reported with oral estriol or placebo in a 10 week RCT investigating the effects of oestrogen on vaginal flora, cytology and urogenital symptoms (n = 35).442 [EL = 1+]

Studies evaluating HRT for other indications

Data from two RCTs that were designed to evaluate the benefits and risks of HRT have been analysed with respect to continence outcomes. In the ‘HERS’ RCT,445 which compared CEE+MPA with placebo, 55% of women had UI (stress, mixed or urge) at baseline (n = 1525).444 After 4 years’ treatment, significantly fewer women in the HRT group reported improvement and significantly more reported worsening of UI symptoms, compared with the placebo group. Leakage episodes were increased in the HRT group compared with placebo.444 In women who did not have UI at baseline, the risk of reporting any type of UI at study end was also significantly higher in the HRT group.450 [EL = 1++]

In women enrolled in the Women’s Health Initiative (WHI) RCTs (CEE+MPA versus placebo447 or CEE versus placebo448), 85% had continence data at baseline and at 1 year (n = 23 296). In women who were continent at the beginning of the study (35%), the relative risk of incident UI of any type at 1 year was significantly higher in the CEE+MPA and CEE groups compared with placebo. When the relative risk of each type of UI was considered separately, all results remained significant except for the risk of urge UI in the CEE+MPA versus placebo study. The relative risk of worsening prevalent UI (leakage quantity and episodes, limitations of daily activities, bother factor) was also significantly higher with HRT compared with placebo.446 [EL = 1++]

A further placebo-controlled RCT, which evaluated both CEE and raloxifene for the prevention of osteoporosis in postmenopausal women, reported a significantly higher incidence of UI with CEE compared with other treatment groups at 3 years. Fewer women treated with oestrogen reported improvement of pre-existing UI (n = 619).449 [EL = 1+]

Adverse effects

Adverse effects reported across the studies of intravaginal oestrogens (mostly uncommon) included vaginal irritation or discomfort, burning and itching,431,436 breast pain,436 and vaginal spotting or discharge.439 One study reported that no systemic adverse effects occurred.431

Adverse effects reported with systemic oestrogens included breast tenderness,437,438 vaginal spotting437 and increased vaginal discharge.439 No adverse effects were reported in two RCTs.435,440 Adverse effects were not considered in four RCTs.432,433,441,442 In the HERS study, the risk of venous thromboembolism was significantly higher with CEE+MPA compared with placebo.445 In the WHI studies, the risk of stroke was significantly higher with CEE+MPA (mean follow-up 5.2 years) and with CEE (mean follow-up 6.8 years), compared with placebo.447,448 The risk of coronary heart disease, venous thromboembolism and invasive breast cancer was also significantly higher with CEE+MPA compared with placebo.447

Evidence statements for oestrogens

Short-term studies (up to 6 months) of intravaginal oestrogens suggest some improvement in symptoms of incontinence and frequency in postmenopausal women who have urogenital symptoms secondary to vaginal atrophy. There is a lack of evidence to support the use of intravaginal oestrogens for the treatment of UI. [EL = 1+]

Systemic oestrogen does not confer any benefit in women with UI and there is evidence that it may increase the likelihood of developing incontinence in postmenopausal women. Systemic oestrogens are associated with increased risk of systemic adverse effects such as thromboembolism. [EL = 1+]

Recommendations for oestrogens

Systemic hormone replacement therapy is not recommended for the treatment of UI. [A]

Intravaginal oestrogens are recommended for the treatment of OAB symptoms in post-menopausal women with vaginal atrophy. [A]

4.5. Non-therapeutic interventions

This section covers the use of products that collect or contain leakage (e.g. absorbent products, urinals and toileting aids, catheters) and products used to prevent leakage (e.g. devices that support the bladder neck, intra- or extraurethral devices).

Studies considered for the non-therapeutic interventions section

Little primary research evidence was identified that addressed the guideline questions. No studies were identified that evaluated the effects of containment on maintenance of independent living, rates of institutionalism, or return to work, and only one study considered QOL (an evaluation of pessaries for UI, described below). Published consensus statements and narrative reviews that discussed issues relevant to the circumstances of use of containment products, or catheterisation, were used as a basis for the recommendations.32,451–453 [EL = 4]

4.5.1. Absorbent products, urinals and toileting aids

One RCT compared a conservative management strategy with the use of absorbent products (pads and pants) for 6 months (n = 90). Conservative management involved providing estriol (depending on oestrogen status), PFMT (six training sessions), bladder training (for urge or mixed UI), electrical stimulation, and pads and pants. Significantly greater reductions in UI severity and impact, and leakage episodes, were seen with conservative management compared with the control group. No significant differences were seen between groups in frequency. The pads and pants arm showed no change in incontinence impact at 6 months.454 [EL = 1+]

Evidence statement for absorbent products, urinals and toileting aids

Pads and pants are ineffective in the treatment of UI. [EL = 1+] There is no evidence to support the use of hand-held urinals and toileting aids in the treatment of UI. There is a variety of such products available. There is a lack of evidence for their use in management. However, they are used by women in maintaining social continence. [EL = 4]

From evidence to recommendation

The GDG recognises that some women may not wish to pursue active interventions for UI and that absorbent products, urinals and toileting aids are an alternative management option in such circumstances. However, the GDG felt that women must be fully aware of all possible treatment options before adopting this course of action. In addition, the GDG agreed that the use of these products should be considered for women awaiting definitive treatment.

Recommendation for absorbent products, urinals and toileting aids

Absorbent products, hand-held urinals and toileting aids should not be considered as a treatment for UI. They should be used only as:

  • a coping strategy pending definitive treatment
  • an adjunct to other ongoing therapy
  • long-term management of UI only after treatment options have been explored. [D (GPP)]

4.5.2. Catheters

The care of patients with long-term urinary catheters is covered within the NICE clinical guideline Infection Control: Prevention of Healthcare-Associated Infection in Primary and Community Care.22 Within that guideline, the recommendation that intermittent catheterisation is preferred to indwelling catheterisation was informed by a systematic review of risk factors for UTI, in adults with spinal cord dysfunction. The GDG’s view is that the evidence relating to adults with spinal cord dysfunction is relevant to the use of catheters in women with idiopathic UI.455 The systematic review found eight cohort studies that considered risk of infection according to type of catheter used by men and/or women with spinal cord injuries. In seven of eight studies, patients using intermittent catheters, had fewer infections or lower prevalence of bacteriuria than those who used indwelling catheters (follow-up ranging from about 1 to 2.5 years). However, none of the primary studies adjusted for baseline differences between groups (total n = 1153). [EL = 2+]

Evidence statement for catheters

Intermittent catheterisation is associated with reduced risk of UTI compared with indwelling catheterisation. [EL = 2+]

From evidence to recommendations

In the absence of evidence on long-term catheterisation in women, but alongside the systematic review described, the GDG considered that suprapubic catheterisation is preferable to indwelling urethral catheterisation owing to reduced risk of urethral and other complications (symptomatic UTI, and ‘bypassing’). Suprapubic catheterisation is not without risk, particularly at initial insertion, although the benefits and risks of this approach have not been fully established. Long-term medical management of suprapubic catheterisation may be problematic if healthcare providers lack knowledge and expertise in this area, and if the homebound patient lacks rapid access to medical care if a problem arises.

The population for whom catheterisation is recommended was also determined by GDG consensus.

The GDG feels that the use of an indwelling catheter in a woman with OAB may be associated with an increase in detrusor activity and therefore an increased tendency to ‘bypassing’ (urine leakage around the catheter). Assuming they empty their bladder completely, as most patients with idiopathic DO will, a catheter is unlikely to achieve continence.

Recommendations for catheters

Bladder catheterisation (intermittent or indwelling urethral or suprapubic) should be considered for women in whom persistent urinary retention is causing incontinence, symptomatic infections or renal dysfunction, and in whom this cannot otherwise be corrected. Healthcare professionals should be aware, and explain to women, that the use of indwelling catheters in urge UI may not result in continence. [D (GPP)]

Intermittent urethral catheters

Intermittent urethral catheterisation should be used for women with urinary retention who can be taught to self-catheterise or who have a carer who can perform the technique. [C]

Indwelling urethral catheters

Careful consideration should be given to the impact of long-term indwelling urethral catheterisation. The practicalities, benefits and risks should be discussed with the patient or, if appropriate, her carer. Indications for the use of long-term indwelling urethral catheters for women with UI include:

  • chronic urinary retention in women who are unable to manage intermittent self-catheterisation
  • skin wounds, pressure ulcers or irritations that are being contaminated by urine
  • distress or disruption caused by bed and clothing changes
  • where a woman expresses a preference for this form of management. [D (GPP)]

Indwelling suprapubic catheters

Indwelling suprapubic catheters should be considered as an alternative to long-term urethral catheters. Healthcare professionals should be aware, and explain to women, that they may be associated with lower rates of symptomatic UTI, ‘bypassing’, and urethral complications than indwelling urethral catheters. [D (GPP)]

4.5.3. Products to prevent leakage

Studies reporting the use of the following products were identified:

  • intravaginal devices: Continence Guard (also known as the Conveen Contiguard),456–462 Contrelle® continence tampon,462 Contiform®,463 bladder neck support prosthesis (Introl)464
  • meatal devices: FemAssist®,465,466 CapSure® shield,467 Miniguard468–470 (also called ‘continence control pad’/Impress Softpatch471)
  • intraurethral devices: a urethral plug472,473 (the ViVa Plug, also called Alive474), FemSoft® insert,475 Reliance® 476,477 (one of the studies was a controlled trial versus NEAT).478

Two of these products are known to be available or could be obtained by women in the UK (Contrelle® Activgard [formerly known as Conveen Contiguard, or Continence Guard], and FemSoft®). Neither product can be provided on NHS prescription. The Rocket® incontinence device is also available, but no studies were found regarding its use. Many of the other products listed are known to have been withdrawn from the UK market for commercial reasons (Contiform®, FemAssist®, CapSure®,479 Reliance®).

While there is no evidence to support the use of menstrual tampons in the management of UI, GDG members are aware that menstrual tampons are used by many women to support the bladder neck and to prevent leakage. Manufactures of these products do not recommend this usage and state that they should be used only during menstruation.

Evidence relating to the use of the available products

A case series of women with stress or mixed UI who used FemSoft® (mean follow-up 15 months), reported that leakage episodes were fewer and more pad tests were negative with the device in place, compared with without the device. Symptomatic UTI was very common (47%), and the incidence of insertion trauma, haematuria, spotting, cystoscopic evidence of bladder or urethral irritation or trauma and device migration were common (n = 150).480 [EL = 3]

Four case series456–459,461 (patient numbers ranging from 15 to 38) and one crossover RCT462 (n = 94; 62 completed and analysed) found that the majority of users of the Continence Guard reported cure or improvement (57–89%) after 3–5 weeks’ treatment (up to 1 year in one study). Adverse effects reported were expulsion of the device (8–47%),456,457,462 voiding difficulties (11–14%),456,457,462 UTI (11%)458 and vaginal irritation (23%).462 Three studies found no vaginal irritation or erosion on gynaecological examination.456–459 No adverse effects were reported in two studies.459,461

Evidence statement for products to prevent leakage

There is limited evidence of efficacy for Contrelle® Activgard (formerly known as the Continence Guard or Conveen Contiguard) and FemSoft® in the management of UI. Adverse effects, in particular UTI, are very common. [EL = 3]

From evidence to recommendation

In the GDG’s view, some women find these products beneficial for occasional use in certain circumstances as a preventive strategy.

Recommendation for products to prevent leakage

Intravaginal and intraurethral devices are not recommended for the routine management of UI in women. Women should not be advised to consider such devices other than for occasional use when necessary to prevent leakage, for example during physical exercise. [D (GPP)]

4.5.4. Pessaries

A small case series reported that QOL improved (IIQ) in women with stress or mixed UI and POP who used a ring pessary with diaphragm for 1 year (six of 38 women enrolled).481 [EL = 3]

Evidence statement for pessaries

The limited evidence available does not support the use of ring pessaries for the treatment of UI in women whether or not there is prolapse present. [EL = 3]

4.6. Complementary therapies

Women who do not find conventional treatments acceptable often explore the use of complementary therapies for UI, and as adjuncts to conventional treatments.

Studies considered for the complementary therapies section

Most of the articles identified reported the use of acupuncture or hypnotherapy for UI, or were narrative reviews regarding the use of complementary therapies for UI. Use of traditional Chinese medicines was also mentioned in a narrative review, but no further references to their use for UI were found. 482

4.6.1. Acupuncture

Three RCTs483–486 and three case series487–490 evaluated the use of acupuncture for UI or OAB in women. Across these studies, the acupuncture points and duration of stimulation used varied. Duration of treatment ranged from 2 to 4 weeks. All were considered to be of poor quality because of lack of information or for only analysing results for women who completed treatment.

One RCT assessed the effects of daily acupuncture at acupoints Sp-6 and St-36 on nocturnal frequency in elderly people on long-stay hospital wards. The median reduction in frequency in the acupuncture group after 2 weeks’ treatment (20 minutes per day) was −2.0 (95% CI −1.0 to −3.0). No significant change was seen in the placebo group, who received mock TENS. Two publications of this RCT were identified; one stated that 15 of the 20 studied were women, another stated that 17 were women.483,484 [EL = 1−]

The acupuncture treatment given in the second RCT, to women with stress UI, depended on what the deficiency was considered to be. A total of 30 sessions were given every other day. Significantly more women treated with acupuncture than placebo were improved (assessed clinically and urodynamically) after treatment.485 [EL = 1−]

An RCT in women with OAB with urge UI reported significant improvements in frequency, urgency and QOL (UDI and IIQ) after 4 weeks’ acupuncture treatment compared with placebo acupuncture (designed for relaxation). Changes in leakage episodes were not significantly different between groups. Adverse effects reported were bruising or bleeding from acupuncture sites (23%) and minor discomfort on needle placement (25%) (n = 85; 74 analysed).486 [EL = 1−]

Case series

Three case series evaluated acupuncture for UI or OAB in a total of 87 patients (84 women).487–489,491 The symptoms being treated were frequency, urgency and dysuria,487 ‘lower urinary tract symptoms’489 and urge or mixed UI.491 Treatment consisted of a single session, or 6 or 12 weeks’ regular treatment. The acupuncture points used also varied across studies.

Symptomatic improvement was reported in 53–60% of patients (assessed at 3 or 8 months).487–489,491 No adverse effects were reported. Longer term follow-up (about 5 years) of 21 patients show that symptoms recur, and that repeated treatment may be necessary.488 [EL = 3]

4.6.2. Hypnosis

The studies identified in relation to hypnotherapy in women with UI consisted of case series and case reports.

In the largest case series of women with incontinence due to DO, they underwent 12 sessions of hypnotherapy over 1 month, which involved symptom removal by direct suggestion and ‘ego strengthening’. At the end of the 12 sessions, the majority of women were subjectively cured or improved, with the remainder unchanged (n = 50). Objective cure or improvement (on cystometry) was seen in the majority at 3 months (n = 44).492 Limited follow-up data at 2 years for 30 of the women have been reported. Of the women who were subjectively or objectively cured at 3 months, fewer than half remained cured (n = 30).493 [EL = 3]

In another publication, four cases (three women) of hypnotherapy for DO were reported. Hypnotherapy involved three 1 hour sessions, including anxiety control methods, ego strengthening, training in self-hypnosis, age progression, explanation of stable bladder function and ‘hand-on-abdomen technique’. Two of the three women reported remission of symptoms at 6 months.494 [EL = 3] A report of two women with UI who were ‘successfully’ treated with hypnotic techniques and waking counselling was also identified.495 [EL = 3]

4.6.3. Herbal medicines

One report described the use of a tablet preparation containing crataeva (Crataeva nurvala, a herb used in traditional Hindu science of medicine) and equisetum (horsetail) to treat women with symptoms of urge and/or stress UI for 12 weeks (n = 8). Quality of life (UDI) showed significant positive change to perceptions of frequency, leakage related to urgency or activity and difficulty emptying the bladder. All parameters of the IIQ questionnaire except physical recreation and household chores improved significantly.496 [EL = 3]

Evidence statements for complementary therapies

Poor-quality evidence shows that acupuncture may reduce nocturia and both stress and urge incontinence in the short term (up to 4 weeks) but it is unclear whether any particular acupuncture treatment is more effective than others. [EL = 3]

There is limited evidence that hypnotherapy for women with UI secondary to detrusor overactivity offers some benefit over the short term (up to 6 months). About half of women relapsed over a 2 year period. [EL = 3] There is a lack of evidence on herbal medicines for UI or OAB.

The GDG recognises that, despite the limited and poor-quality evidence available, some women may wish to explore complementary therapies for their incontinence [EL = 4].

Recommendation for complementary therapies

Complementary therapies are not recommended for the treatment of UI or OAB. [D]

4.7. Preventive use of conservative therapies

Studies considered for this section

Evidence described in this section is derived from RCTs. Two systematic reviews of the use of physical therapies for prevention of UI were identified.497,498 The RCTs within these systematic reviews were considered individually if they addressed effectiveness. Any further RCTs identified are also included here.

No studies were identified that evaluated the use of lifestyle interventions for prevention of UI or OAB.

4.7.1. Behavioural therapy

One RCT evaluated a multicomponent behavioural modification programme comprising initial education, PFMT and bladder training in older women who had no UI (39%) or minimal UI (defined as one to five wet days in the previous year) (n = 480 randomised; 359 analysed). At 1 year follow-up, significantly more women maintained or improved their continence status compared with an untreated control group. Significantly greater improvements in frequency and voiding interval were also seen in the behavioural modification group versus control. Adverse effects were not considered.499,500 [EL = 1−]

4.7.2. Physical therapies

Preventive use during pregnancy

Four RCTs compared more structured PFMT with usual care during pregnancy (from weeks 18 or 20) in women in their first pregnancy.501–504 One study enrolled women with increased bladder neck mobility,501 which has been shown to be predictive of postnatal stress UI.505 Women with UI were excluded from one study,501 whereas in two studies between 25% and 32% had UI at baseline.503,504 Between 72 and 1169 women were enrolled in these studies (total n = 1810); the proportion completing follow-up or responding to questionnaire follow-up was noted to be low in three studies (64–86%).

The PFMT programmes involved daily exercises with between 42 and 72 contractions.501–504 The individual’s ability to contract the pelvic floor muscle was checked at baseline. The comparison group was ‘usual care’ in each study, which comprised: ‘usual’ information from midwife or GP;503 routine antenatal care (likely to have received verbal advice on pelvic floor exercises);501 and routine care, no systematic PFMT programme.502,504 Two of the studies reported that women in the control group also undertook PFMT regularly (20%502 and 51%501).

In the first study, significantly fewer women who had been randomised to a 12 week PFMT programme reported UI at the end of treatment, and at 3 months postpartum, compared with those receiving usual care. Greater benefit was seen with PFMT in the number of leakage episodes and in pelvic floor muscle strength. No adverse effects were reported.503 [EL = 1++]

In a study of women with increased bladder neck mobility, significantly fewer reported stress UI or had a positive 1 hour pad test at 3 months postpartum, after a structured PFMT programme, compared with those receiving usual care. No significant differences were seen between groups in changes in bladder neck mobility, pelvic floor muscle strength or in QOL (KHQ). Women in the PFMT group had significantly higher scores in the general health domain of SF-36 compared with the usual care group.501 [EL = 1+]

No numerical data were reported in the third RCT, although the authors noted that no significant differences were seen between groups in UI or pelvic floor muscle strength at 12 months postpartum (n = 46). Adverse effects were not considered.502 [EL = 1−]

The largest study considered risk of urinary outcomes during the antenatal period and up to 6 months postpartum. At antenatal week 36, there was a trend towards reduced risk of any type of UI and of fewer leakage episodes in the PFMT group, although no difference between groups was statistically significant. At 6 months postpartum, differences between groups were less, again with none being significant (n = 1169).504 [EL = 1+]

Preventive use after pregnancy

Four controlled trials evaluated PFMT for the prevention of UI in postpartum women. UI was reported by 17–32% of women across all studies at baseline.506–509 The intervention was started 24 or 48 hours after delivery in primi- or multiparous women in two RCTs.506,507 A more structured 4 or 8 week PFMT programme was compared with usual care in both studies. At 3 months postpartum, the following results were seen:

  • significantly lower prevalence of UI in the PFMT group following the 8 week treatment programme (n = 676); this difference was not sustained at 1 year (n = 569)510 [EL = 1+]
  • no significant differences in UI prevalence between groups following the 4 week treatment programme (n = 1609; 89% of those randomised).507 [EL = 1−]

A further two studies (one RCT,509 one cohort508) recruited women 8 weeks postpartum. The RCT compared a 6 week programme of PFMT with biofeedback and electrical stimulation, with usual care in primiparous women (n = 107). Stress UI prevalence was not significantly different between groups at 10 months postpartum. However, the prevalence of stress UI differed between groups at baseline (31% PFMT versus 16% control), which was not accounted for in the analysis at 10 months.509 [EL = 1−]

The cohort study reported a significantly lower stress UI prevalence in women (41% of whom had UI at baseline) who had undergone a structured 8 week PFMT programme compared with usual care, both at the end of the intervention (n = 198)508 and at 1 year postpartum (n = 162).511 No significant differences were found between groups in leakage index or social activity index. [EL = 2+]

Evidence statement for preventive use of physical therapies

There is evidence that PFMT used during a first pregnancy reduces the prevalence of UI at 3 months following delivery. [EL = 1+] The effects in the longer term are inconsistent and the impact of subsequent pregnancies unknown. [EL = 4]

Recommendation for preventive use of physical therapies

Pelvic floor muscle training should be offered to women in their first pregnancy as a preventive strategy for UI. [A]

Research recommendation for preventive use of conservative therapies

Further studies need to be undertaken to evaluate the role and effectiveness of physical and behavioural therapies and lifestyle modifications in the prevention of UI in women. Long-term outcomes in particular should be evaluated.

4.8. Optimal sequence and timescales for conservative therapies

The GDG’s view was that conservative management should be pursued prior to surgical procedures in the treatment of UI. The factors affecting the GDG’s decisions regarding first-line conservative therapies are summarised below.

For stress UI, the GDG considered the cost effectiveness of PFMT and duloxetine as first-line treatment. The conclusion was that PFMT dominated and it was therefore recommended as the first-line intervention for stress UI.

For OAB (with or without urge UI), cost minimisation was used to determine whether bladder training or antimuscarinic drug treatment should be offered as first-line treatment. Bladder training was recommended as the first-line intervention as it is less costly and not associated with adverse effects.

In the GDG’s view, the management of mixed UI depends upon which symptom predominates (i.e. stress or urge UI).

In the absence of evidence for optimal duration of treatment, the GDG considered that a 3 month trial period of PFMT is sufficient to determine whether treatment is effective and tolerated. A shorter trial period of 6 weeks would usually be appropriate for bladder training.

Copyright © 2006, National Collaborating Centre for Women’s and Children’s Health.

No part of this publication may be reproduced, stored or transmitted in any form or by any means, without the prior written permission of the publisher or, in the case of reprographic reproduction, in accordance with the terms of licences issued by the Copyright Licensing Agency in the UK [www.cla.co.uk]. Enquiries concerning reproduction outside the terms stated here should be sent to the publisher at the UK address printed on this page.

The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant laws and regulations and therefore for general use.

Bookshelf ID: NBK57207

Views

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...