Dumoulin C, Adewuyi T, Booth J, Bradley C, Burgio K, Hagen S, Hunter K, Imamura M, Morin M, Morkved S, Thakar R, Wallace S, Williams K

Dumoulin C, Adewuyi T, Booth J, Bradley C, Burgio K, Hagen S, Hunter K, Imamura M, Morin M, Morkved S, Thakar R, Wallace S, Williams K. Adult conservative management. In: Abrams P, Cardozo L, Wagg A, Wein A, ed. Incontinence: 6th International Consultation on Incontinence, Tokyo, September 2016. Bristol, UK: International Continence Society (ICS) and International Consultation on Urological Diseases (ICUD), 2017, pp.1443-1628.

URL link to article: https://www.ics.org/education/icspublications/icibooks/6thicibook

Date deposited:

31/10/2017

Embargo release date:

01 November 2018

Newcastle University ePrints - eprint.ncl.ac.uk

Committee 12

ADULT CONSERVATIVE MANAGEMENT

Chair C. DUMOULIN (CANADA)

Members T. ADEWUYI (UK) J. BOOTH (UK) C. BRADLEY (USA) K. BURGIO (USA) S. HAGEN (UK) K. HUNTER (CANADA) M. IMAMURA (UK) M. MORIN (CANADA) S. MORKVED (NORWAY) R. THAKAR (UK) S. WALLACE (UK) K. WILLIAMS (UK)

The present chapter would not have been possible without the work of the previous editors, Katherine Moore [2013], Jean hay Smith [2008] and don Wilson [2000, 2004]. They produced the remarkable framework for the present review. In addition, we would like to acknowledge the indispensable contribution of Pauline Campbell to the reviewing, analysis and reporting of POP section. Finally, this chapter would not have been possible without the untiring dedication of Yvonne Ruella.

1 CONTENTS

3.2.2 Are VC As Effective as Other Treatments? 60 INTRODUCTION 6 3.2.3 Are VC Combined with PFMT Better than PFMT Alone? ...... 60 URINARY INCONTINENCE IN WOMEN 7 3.3. Other Lower Urinary Tract Symptoms 1. Lifestyle Modification Interventions ...... 7 (LUTS)...... 61 1.1. Prevention ...... 8 3.4. Factors Affecting Outcome ...... 61 1.2. Treatment ...... 8 4. Electrical Stimulation (EStim) ...... 61 1.2.1 Weight Loss by Obese or Overweight 4.1. Prevention ...... 62 Women ...... 8 4.2. Treatment ...... 68 1.2.2 Physical Activity ...... 10 4.2.1 Is EStim Better than No Active Treatment 1.2.3 Strenuous Physical Activity ...... 10 (Placebo, Sham, Control or No Treatment) for Treatment of UI? ...... 68 1.2.4 Smoking ...... 10 4.2.2 Is One Type of EStim Better than Another 1.2.5 Dietary Modification (Elimination Diet) .... 11 in the Treatment of UI? ...... 68 1.2.6 Constipation ...... 14 4.2.3 Is EStim Better than Other Treatments for 1.3. Other LUTS ...... 14 UI? ...... 70 1.4. Factors Affecting Outcome ...... 14 4.2.4 Does the Addition of EStim to Other Treatments Add Any Benefit in the 2. Pelvic Floor Muscle Training (PFMT) ...... 14 Treatment of UI? ...... 70 2.1. Prevention and Treatment (Pregnant and 4.3. Other LUTS ...... 72 Postnatal Women Only) ...... 14 4.4. Factors Affecting Outcome ...... 72 2.1.1 Is PFMT Effective in the Prevention of UI in Childbearing Women? ...... 14 5. Posterior Tibial Nerve Stimulation (PTNS) ...... 73 2.1.2 Is PFMT Effective in the Treatment of UI in Childbearing Women? ...... 15 5.1. Prevention ...... 77 2.1.3 Is PFMT Effective in the Mixed Prevention 5.2. Treatment ...... 77 and Treatment of UI in Childbearing 5.2.1 Is PTNS Better than No Active Treatment Women? ...... 16 (Placebo, Sham, Control or No Treatment) 2.2. Prevention (Other Women) ...... 22 for Treatment of UI? ...... 77 2.3. Treatment (Other Women) ...... 22 5.2.2 Is One Type of PTNS Better than Another in the Treatment of UI? ...... 77 2.3.1 Is PFMT Better than No Treatment, a Placebo, or a Control Group Treatment? 22 5.2.3 Is PTNS Better than Other Treatments for Treatment of UI? ...... 77 2.3.2 Is One Type of PFMT Programme Better than Another?...... 23 5.2.4 Does the Addition of PTNS to Other Treatments Add Any Benefit in the 2.3.3 Is PFMT Better than Other Treatments? .. 43 Treatment of UI? ...... 80 2.3.4 Does the Addition of PFMT to Other 5.3. Other LUTS ...... 82 Treatments Add Benefit? ...... 48 5.4. Factors Affecting Outcome ...... 82 3. Weighted Vaginal Cones (VC) ...... 52 6. Magnetic Stimulation (MStim)...... 82 3.1. Prevention ...... 52 6.1. Prevention ...... 83 3.2. Treatment ...... 52 6.2. Treatment ...... 83 3.2.1 Are VC Better than No Treatment, Placebo or Control Treatments? ...... 60 6.2.1 Is MStim Better Than no Active Treatment (Placebo, Control or no Treatment)?...... 83

2 6.2.2 Is One Approach to MStim Better than 9.3.1 Lifestyle Modification Intervention ...... 97 Another? ...... 85 9.3.2 PFMT in Antenatal and Postnatal women 97 6.2.3 Is MStim Better than Other Treatments? . 85 9.3.3 PFMT in Women (Others) ...... 97 6.2.4 Does the Addition of MStim to Other Treatments Add Any Benefit in the 9.3.4 Vaginal Cones ...... 98 Treatment of UI? ...... 85 9.3.5 EStim/PTNS/MStim ...... 98 6.3. Other LUTS ...... 85 9.3.6 Scheduled Voiding Regimen ...... 98 6.4. Factors Affecting Outcome ...... 85 7. Scheduled Voiding Regimens ...... 85 PELVIC ORGAN PROLAPSE (POP) 98 7.1. Prevention ...... 87 1. Lifestyle Modification Intervention ...... 98 7.2. Treatment ...... 87 1.1. Prevention ...... 98 7.2.1 What Is the Most Appropriate Bladder 1.1.1 Quality of Data ...... 98 Training (BT) Protocol? ...... 87 1.1.1.1 Association Between POP and 7.2.2 Is BT Better than No Treatment, Placebo or Occupation and Physical Activity...... 98 Control Treatments? ...... 89 1.1.1.2 Association Between POP and Body 7.2.3 Is BT Better than Other Treatments? ...... 89 Weight ...... 99 7.2.4 Can Any Other Treatment Be Added to BT 1.1.1.3 Association Between POP and Smoking to Add Benefit? ...... 90 ...... 99 7.2.5 Does the Addition of BT to Other Treatment 1.1.1.4 Association Between POP and Bowel Add Benefit? ...... 90 Function ...... 105 7.2.6 Is Timed Voiding Effective at Treating UI? 1.1.1.5 Association Between POP and Nutrition 91 ...... 105 7.3. Other LUTS ...... 91 1.1.2 Results ...... 105 7.4. Factors Affecting Outcome ...... 92 1.1.2.1 Association Between POP and Occupation and Physical Activity...... 105 8. Complementary and Alternative Medicines ...... 92 1.1.2.2 Association Between POP and Body Weight ...... 106 8.1. Prevention ...... 93 1.1.2.3 Association Between POP and Smoking 8.2. Treatment ...... 93 ...... 106 8.2.1 Acupuncture for Treatment of SUI ...... 93 1.1.2.4 Association Between POP and Bowel 8.2.2 Acupuncture for Treatment of OAB, UUI Function ...... 106 and Mixed UI ...... 93 1.1.2.5 Association Between POP and Nutrition 9. Future Research Directions ...... 95 ...... 107 9.1. Summary ...... 95 1.2. Treatment ...... 107 9.2. Recommendations for practice ...... 95 2. Pelvic Floor Muscle Training ...... 108 9.2.1 Lifestyle Intervention ...... 95 2.1. Prevention ...... 108 9.2.2 PFMT (Principal Recommendation) ...... 95 2.2. Treatment ...... 109 9.2.3 Cones ...... 96 2.2.1 PFMT Alone ...... 109 9.2.4 EStim ...... 96 2.2.2 PFMT and Surgery ...... 121 9.2.5 PTNS ...... 96 2.2.3 PFMT and Pessary ...... 124 9.2.6 Magnetic Stimulation ...... 96 3. Pessaries ...... 126 9.2.7 Bladder Training ...... 97 3.1. Prevention of POP with Pessaries ...... 127 9.2.8 Complementary and Alternative Medicine 3.2. Treatment of POP with Pessaries ...... 127 97 3.2.1 Pessary Alone ...... 127 9.3. Future Research Direction ...... 97 3 3.2.2 Pessary Versus no Treatment ...... 127 2.5. Pre-Operative and/or Post-Operative TURP PFMT ...... 158 3.2.3 Pessary and PFMT ...... 127 2.6. Post-Operative TURP PFMT for Incontinent 3.2.4 Pessary Versus Surgery ...... 127 Men ...... 158 3.2.5 Comparison of One Pessary to Another 128 2.7. Factors Affecting Outcome ...... 158 4. Recommendations ...... 136 2.8. PFMT for Other LUTS ...... 159 4.1. Recommendations for Practice ...... 136 2.8.1 PFMT for Post-Micturition Dribble (PMD) 4.1.1 Lifestyle Modification...... 136 159 4.1.2 Pelvic Floor Muscle Training (PFMT) ..... 136 3. Electrical Stimulation (EStim) ...... 159 4.1.3 Pessaries ...... 136 3.1. Prevention of UI ...... 159 4.2. Future Research Directions ...... 136 3.2. Treatment of UI ...... 159 4.2.1 Lifestyle Interventions ...... 136 3.2.1 Is EStim Better than No Treatment, Placebo or Control Treatments? ...... 159 4.2.2 Pelvic Floor Muscle Training (PFMT) ..... 136 3.2.2 Is One Approach to EStim Better Than 4.2.3 Pessarie for Prolapse ...... 137 Another? ...... 159 3.2.3 Is EStim Better Than Other Treatments?159 URINARY INCONTINENCE IN MEN 137 3.2.4 Does the Addition of EStim to Other 1. Lifestyle ...... 137 Treatments Add Benefit? ...... 159 1.1. Weight Loss by Obese or Overweight Men 3.3. Other LUTS ...... 159 ...... 137 3.4. Factors Affecting Outcome ...... 159 1.2. Smoking ...... 138 4. Magnetic Stimulation (MStim)...... 159 1.3. Dietary Modification in Men ...... 138 4.1. Prevention of UI ...... 159 2. Pelvic Floor Muscle Training (PFMT) ..... 154 4.2. Treatment of UI ...... 159 2.1. Pre-Operative RP PFMT ...... 154 4.2.1 Is MStim Better Than No Treatment, 2.2. Pre-Operative and/or Post-Operative RP Placebo or Control Treatment? ...... 159 PFMT, Post RP Continence Status Not 4.2.2 Is One Approach to MStim Better Than Established Prior to Intervention ...... 154 Another? ...... 159 2.2.1 Pre-Operative PFMT Instruction with Post- 4.2.3 Is MStim Better Than Other Treatments? Operative Home PFMT Versus Control .. 154 159 2.2.2 Pre-Operative PFMT Instruction Followed 4.3. Other LUTS ...... 159 by supervised Post-Operative PFMT Versus Post-Operative PFMT ...... 155 4.4. Factors Affecting Outcome ...... 159 2.2.3 Post-Operative PFMT Immediately Before 5. Penile Vibratory Stimulation (PVS) ...... 160 or After Catheter Removal (No Pre- 5.1. Prevention of UI ...... 160 Operative Instruction) ...... 155 5.2. Treatment of UI ...... 160 2.3. Post-Operative RP PFMT for Incontinent Men ...... 156 5.3. Other LUTS ...... 160 2.3.1 PFMT with Digital Rectal Feedback (DRE) 6. Scheduled Voiding Regimens ...... 160 After Radical Prostatectomy ...... 156 6.1. Prevention of UI ...... 160 2.3.2 PFMT with BF After Radical Prostatectomy 157 6.2. Treatment of UI ...... 160 2.3.3 PFMT Plus or minus BF with EStim or 6.2.1 Bladder Training ...... 160 MStim after Radical Prostatectomy ...... 157 6.2.2 Timed Voiding ...... 161 2.3.4 PFMT Compared to or Plus Other 6.3. Other LUTS ...... 161 Interventions After Radical Prostatectomy 157 6.4. Factors Affecting Outcome ...... 161 2.4. Pre-Operative TURP PFMT ...... 158

4 7. Complementary and Alternative Medicines ...... 161 7.1. Prevention of UI ...... 161 7.2. Treatment UI ...... 161 7.2.1 What Is the Most Effective Acupuncture Protocol? ...... 161 7.2.2 Acupuncture Versus No Treatment, Sham Acupuncture or Any Other Treatment ... 161 7.3. Other LUTS ...... 161 7.4. Factors Affecting Outcome ...... 161 8. Summary ...... 161 8.1. Recommendations for Practice ...... 161 8.2. Future Research Directions ...... 162 8.2.1 Lifestyle Interventions ...... 162 8.2.2 Pelvic Floor Muscle Training (PFMT) ..... 162 8.2.3 Electrical Stimulation (EStim) and Magnetic Stimulation (MStim) ...... 162 8.2.4 Scheduled Voiding Regimens ...... 163

URINARY INCONTINENCE IN MEN AND WOMEN 163 1. Posterior Tibial Nerve Stimulation (PTNS) ...... 163 1.1. Prevention of UI ...... 163 1.2. Treatment of UI ...... 163 1.2.1 Is PTNS Better Than No Treatment, Placebo or Control Treatments for UI? .. 163 1.2.2 Is PTNS Better Than Other Treatments for UI? ...... 164 1.2.3 Does the Addition of PTNS to Other UI Treatment Add Any Benefit? ...... 165 1.2.4 What is the Best PTNS Protocol for UI in Adults?...... 166 1.3. What is the Effect of PTNS on LUTS Other Than UI? ...... 166 1.4. Factors Affecting Outcomes ...... 170

REFERENCES 172

5 ADULT CONSERVATIVE MANAGEMENT

C. DUMOULIN (CANADA) T. ADEWUYI (UK), J. BOOTH (UK), C. BRADLEY (USA), K. BURGIO (USA), S. HAGEN (UK), K. HUNTER (CANADA), M. IMAMURA (UK), M. MORIN (CANADA), S. MORKVED (NORWAY), R. THAKAR (UK), S. WALLACE (UK), K. WILLIAMS (UK)

The updated literature search covered publications INTRODUCTION from July 1st, 2011 to September 9th, 2015 according to the following search strategy: Conservative management of urinary incontinence (UI) or pelvic organ prolapse (POP) is defined as any  Cochrane Incontinence systematic reviews intervention not involving surgical or pharmacological approaches (1). These comprise lifestyle modifica- Relevant Cochrane systematic reviews were identi- tion, pelvic floor muscle training (PFMT) (either alone fied, for each section of this Chapter, by two of the or with the addition of biofeedback), vaginal cones, authors (CD, SW) assessing the full list of reviews electrical stimulation, magnetic stimulation, posterior (Cochrane Incontinence (3) or on request). The leads tibial nerve stimulation, scheduled voiding regimens, for each section were given lists of these relevant re- complementary/alternative medicines and supportive views. Additional searches of the Cochrane Inconti- rings/pessaries. In this chapter, we cover studies that nence Specialised Register were provided to each provide comparative data between a conservative section author to bring the searches up to date or, if management approach and the absence of treatment there was a gap in review coverage, a full search of the Register was conducted (search date: September or placebo, between two conservative management th approaches or between a conservative management 9 , 2015). approach and medications or surgery.  Additional searches for this ICI Chapter Conservative management approaches are consid- Electronic searches ered relatively low cost and non-invasive, with minimal adverse effects that are typically guided by a We identified relevant trials from the Cochrane Incon- healthcare professional and depend on user partici- tinence Specialised Register. For more details of the pation. It is generally accepted that conservative in- search methods used to build the Register please see terventions are part of initial management at the pri- the Group's module in the Cochrane Library (4). The mary care level for individuals suffering from either UI Register contains trials identified from the Cochrane or POP. Conservative intervention is also indicated Central Register of Controlled Trials (CENTRAL), for those for whom other treatments are inappropri- MEDLINE, MEDLINE In-Process, ClinicalTrials.gov, ate, for example, those unwilling to undergo or unfit WHO ICTRP, UK Clinical Research Network Portfolio for surgery and women who plan future pregnancies. and hand searching of journals and conference pro- Other indications include individuals awaiting surgery ceedings. The search covered: or who wish to delay surgery and those whose symp-  For sections covered by existing Cochrane re- toms are not severe enough for surgical intervention. views – from the date of last search for the pub- The research base on which to judge the effective- lished version of the review to September 9th, ness of conservative interventions is growing alt- 2015. hough some aspects, such as lifestyle modification interventions and alternative and complementary  For sections not covered by existing Cochrane reviews – complete search of all contents of the medicines in particular, require considerably more th well-designed trials. In this chapter we add to the Register on September 9 , 2015. th 2013 5 Edition ICI chapter on conservative manage-  Searching other sources ment (2), integrate the evidence and upgrade the recommendations from the 5th ICI. We also add a new We searched the reference lists of relevant articles section on posterior tibial nerve stimulation with its for other relevant trials. levels of evidence and recommendation. We did not impose any language or other limitations on any of the searches described above.

6 In continuity with the last edition of the ICI chapter on conservative management and to ensure consistency URINARY INCONTINENCE IN throughout the chapter, the primary outcomes to be WOMEN reported were patient-reported outcomes: cure and cure and improvement, UI-specific or POP-specific quality of life or symptoms-validated questionnaire 1. LIFESTYLE MODIFICATION and/or number of leakages (as per the diary). Cure and improvement were based on patients' self-report INTERVENTIONS when available; if not, cure and improvement based on quantifiable measures such as a diary and pad Lifestyle modification Interventions for women tests were reported. A number of lifestyle factors may play a role in either Secondary outcomes were clinician-reported out- the development or resolution of UI. Lifestyle modifi- comes (pad test, POPQ score, etc.). These were cho- cation interventions are defined as application of in- sen in relation to each subsection’s specificity and in terventions in management of lifestyle-related health continuity with our previous chapter. problems (1). They are low cost, non-invasive altera- tions in lifestyle such as weight loss; dietary changes; Each subsection concludes with the level of evi- fluid intake modification; reduction in caffeinated, car- dence, grade of recommendations and factors affect- bonated and alcoholic drinks; avoidance of constipa- th ing outcome. Separate chapters in this 6 Edition ad- tion; stopping smoking; and physical activity. dress incontinence in the frail older person, children, pessaries and products for continence, those with As reported in the 5th ICI (2013) evidence for the effi- neurological conditions and those suffering from fae- cacy of lifestyle modification interventions for urinary cal incontinence. incontinence is lacking, there are few lifestyle modification interventions that have received scrutiny in Where summary statistics are presented, the raw randomised controlled trials, although weight loss data from which these are derived can be found in the and dietary factors have been examined and pro- tables within each subsection, in trial reports and sys- spective studies and observational studies have been tematic reviews cited in the chapter. The Chapter is undertaken for diet, smoking and constipation. For intended as a stand-alone document and adds to the more details, refer to ICI 5th Edition (2). comprehensive reports from previous ICI editions. This review section updates the evidence on the use of lifestyle modification interventions to improve incontinence and related symptoms in females. It revealed 10 potentially eligible new trials. Table 1 illustrates the number of studies included in the 5th ICI edition as well as the new studies identified in the current update. Characteristics of each new trial are presented in subsequent tables for each lifestyle modification intervention for which there is new data.

Table 1 Studies included in the updated review 2016

Lifestyle modification ICI 2013 ICI 2016 Total Interventions Weight loss 3 RCT 2 RCT 5 RCT 4 prospective cohort 1 cohort study 13 cohort studies 7 cross sectional cohort 1 meta-analysis 1 case control study 1 retrospective cohort 1 prospective longitudinal 1 meta- analysis 1 case control study study 1 prospective longitudinal study Physical activity 2 prospective cohort 1 prospective cohort study 3 prospective cohort studies 2 observational cohort 1 case control study 2 observational cohort study 1 case control study Smoking 2 case control studies 1 RCT (pilot, secondary 1 RCT pilot analysis) 2 case control studies Dietary factors 3 observational studies 1 case control study 3 observational studies 1 prospective cohort study 1 case control study 1 cross sectional cohort 1 prospective cohort study study 1 cross sectional cohort study

7 Questions addressed are: were randomised to an intensive weight loss programme or a diabetes support and education group.  Are lifestyle modification interventions (e.g. Self-report of incontinence, nocturia and daytime weight loss, increased/decreased physical activ- voiding frequency were recorded at baseline and 1 ity/ smoking cessation/ dietary or fluid change) year. Each kilogramme of weight loss reduced the effective in the prevention of UI? odds of developing UI at one year by 3% (OR 0.97,  Are lifestyle modification interventions (e.g. 95% CI (0.95-0.99); p=0.1). There was uncertainty weight loss, increased/decreased physical activ- over the risk of selection bias for several key param- ity/ smoking cessation/ dietary or fluid change) eters, including: random sequence generation and al- better than no treatment, placebo or control in location concealment. Performance bias was unclear the treatment of UI? as the blinding of participants and staff was not undertaken. The study ensured the blinding of outcome  Is one lifestyle modification intervention (e.g. assessment but did not provide complete outcome weight loss, increased/decreased physical activ- data and provided only selective reporting indicating ity/ smoking cessation/ dietary or fluid change) possible reporting bias. better than another? A single study not identified in our previous searches In 2015 a Cochrane review was published (5) which from 2008 was discovered (8). This prospective lon- included all available trials of lifestyle modification in- gitudinal study examined the effect of weight loss of- terventions for urinary incontinence. These trials are fered as part of a weight reduction programme which included in this review, along with new trial data since included low calorie diet and exercise, 64 obese in- the Cochrane search was undertaken (October 27th, continent women were included. Weight loss of ≥5% 2014) as well as other cohort evidence where no trial of baseline weight was associated with significant re- data are available. duction in pad test loss (median difference, 19 g; 95% confidence interval, 13-28 g; p < 0.001). They also 1.1. Prevention report a statistically significant improvement in quality Are lifestyle modification interventions (e.g. of life measures. They suggest that weight reduction weight loss, increased/decreased physical activ- of 5% of initial body weight can improve UI severity ity/ smoking cessation/ dietary or fluid change) and its effects on quality of life in obese women. effective in the prevention of UI? There was uncertainty over the risk of bias in all parameters for this study due to the poor reporting style. In the last ICI chapter (5th ICI, 2013), there were no trials on prevention of UI. No new trials have been Further analysis on the effect of weight loss on published. changes in health-related quality of life are reported by Pinto et al. 2012 (9). In a study conducted for the Based on the clear lack of evidence in the literature Programme to Reduce Incontinence by Diet and Ex- of RCTs or observational studies of lifestyle modifica- ercise (PRIDE), a longitudinal cohort analysis was tion interventions to prevent UI, no evidence based undertaken of 338 obese and overweight women with recommendations can be made. UI and health related quality of life assessed at baseline, 6 and 18 months. They concluded that weight Prevention should be an area for future research in- loss and increased physical activity but not reduction vestment and comprise robust economic evaluation in frequency of UI were strongly associated with im- to determine the benefits of prevention strategies in provements in QoL. specific age groups. More recently in 2014 a systematic review and meta- analysis of non-surgical weight loss interventions in 1.2. Treatment overweight women was published by Vissers et al. 1.2.1 Weight Loss by Obese or Overweight (10) drawing on the studies from the ICI 2013 chapter Women and the more recent Cochrane review (2015); this concluded that whilst few (n=6) studies were availa- Urinary incontinence and obesity are both common ble, non-surgical weight loss should be considered problems for women. It is recognised that obese standard practice for overweight women with UI. This women have higher intra-abdominal pressures than review did not identify any additional studies for inclu- non-obese women, and it has been suggested that sion here. It is important to note that this review this chronically elevated pressure may predispose to sought to determine the effect of non-surgical weight incontinence by weakening pelvic floor support struc- loss interventions on urinary incontinence in over- tures (6). weight women, but, in fact, included studies of both obese and overweight women. One new trial was identified which examined weight loss in overweight/obese women (7) (Table 2). The study was conducted to determine the effect of an intensive weight loss programme over 4 years on a subset of female participants (n=2739), previously outlined in the Look AHEAD trial(7). All participants

8 Table 2 Summary of data on weight loss

Author, Study Notes (side effects, loss of follow up, Comparator N Age Outcomes/results Follow up year population risk of bias) Phelan Intensive weight-loss vs a 2994 Women in a Group 1 Decrease of at least 2 incontinence 12 months Random sequence generation: unclear 2012(1) structured education diabetes trial 57.8 episodes a week assessed by a Allocation concealment – unclear programme on weight Group 2 validated self report questionnaire. Blinding of participants – high risk loss 58.1 Mean weight loss at 1 year Group 1 : 7.7 Blinding of outcome assessment – low risk group 2 0.7; p value<0.0001 Incomplete outcome data – unclear risk Selective reporting – unclear risk Breyer Intensive weight loss 1910 Men in a 59.9 Decrease in incontinence episodes or 12 months Random sequence generation : unclear 2014(2) versus a structured diabetes trial cure assessed by a validated self-report Allocation concealment – unclear education programme on questionnaire Blinding of participants – high risk weight loss Blinding of outcome assessment – low risk Incomplete outcome data – unclear risk Selective reporting – unclear risk Auwad Commercial programme 64 Obese women 52.2 Decrease in pad test loss 18 months Blinding of outcome assessment – unclear 2008(3) of diet and exercise with risk versus no control group urodynamically Incomplete outcome data – unclear risk proved urinary Selective reporting – unclear risk incontinence Pinto Behavioural weight loss 338 Overweight and 53 +- 11 Health related QoL 18 months Secondary analysis of data: 2012(4) intervention versus a obese women years Random sequence generation : low risk structured education with UI (SD) Allocation concealment – low risk programme on weight Blinding of participants – high risk loss Blinding of outcome assessment – low risk Incomplete outcome data – low risk Selective reporting – unclear risk 1. Phelan S, Kanaya AM, Subak LL, Hogan PE, Espeland MA, Wing RR, et al. Weight Loss Prevents Urinary Incontinence in Women With Type 2 Diabetes: Results From the Look AHEAD Trial. Journal of Urology. 2012;187(3):939-44. 2. Breyer BN, Phelan S, Hogan PE, Rosen RC, Kitabchi AE, Wing RR, et al. Intensive Lifestyle Intervention Reduces Urinary Incontinence in Overweight/Obese Men with Type 2 Diabetes: Results from the Look AHEAD Trial. Journal of Urology. 2014;192(1):144-9. 3. Auwad W, Steggles P, Bombieri L, Waterfield M, Wilkin T, Freeman R. Moderate weight loss in obese women with urinary incontinence: a prospective longitudinal study. Int Urogynecol J Pelvic Floor Dysfunct. 2008;19(9):1251-9. 4. Pinto AM, Subak LL, Nakagawa S, Vittinghoff E, Wing RR, Kusek JW, et al. The effect of weight loss on changes in health-related quality of life among overweight and obese women with urinary incontinence. Quality of Life Research. 2012;21(10):1685-94.

9 Summary 1.2.3 Strenuous Physical Activity Evidence from 1 RCT, 2 cohort studies and one meta- The previous ICI identified no research evidence analysis support lifestyle modification interventions which suggested that strenuous exercise causes UI. promoting weight loss as a tool to reduce urinary in- Minimal uncontrolled data did suggest that women continence in women who are overweight or obese. engaged in occupations which involved heavy lifting Weight loss of 5% of initial body weight has an impact may be predisposed to genital prolapse or UI (12). A on the reduction of urinary incontinence symptoms case control study conducted by Nygaard et al. (13) and the odds of developing UI at one year can be re- to determine lifetime physical activity and stress induced by 3% for every kilogram lost by overweight continence enrolled 1528 women between the ages and obese women. (Level of Evidence: 1) of 39-65. Each participant underwent a Pelvic Organ Prolapse Quantification examination, 213 cases were Recommendations identified and matched 1:1 to controls. Physical activity was measured using the Lifetime Physical Activity Weight loss as a non-surgical intervention, should be Questionnaire in which women recall their past phys- recommended to obese and overweight women with ical activity. Lifetime strenuous activity was not asso- UI. (Grade of Recommendation: A) ciated with SUI (OR, 1.11; 95% CI, 0.99-1.25) and increased lifetime activity did not affect the odds of de- 1.2.2 Physical Activity veloping SUI (OR, 1.11; 0.99-1.25; p=0.06) (Table 3). No new RCT data have been published on either the benefits of moderate physical activity or the adverse Summary effect of strenuous activity on UI in women. One rele- Evidence for the association between strenuous vant prospective cohort study (11) and one case con- physical activity and incontinence need to be repli- trol study have been reported since the last review cated using larger populations and more robust de- Table 3. sign before recommendations can be made. No fur- The prospective cohort study (11) determined that ther data was identified. (Level of Evidence: 3) low physical activity may be a causal factor in the de- Recommendations: velopment of overactive bladder. This study developed a hypothetical causative model for OAB involv- In the absence of robust randomised controlled trials ing secondary analysis from the Leicestershire UK (Grade of Recommendation: C) MRC Incontinence study and included 3411 women free from OAB and 277 incident cases aged 40+. The 1.2.4 Smoking study used graphical chain modelling to EStimate the associations between variables and identify the likely In 2014, a pilot study investigated the effect of smok- causal pathways. Low physical activity was a direct ing cessation on overactive bladder symptoms (14). risk factor linked to the development of over active The authors explored the hypothesis that smoking bladder (OAB) with incontinence (RR 2.47;95% CI abstinence would be associated with improvements 1.82, 3.36). They identified the need for further rein OAB symptoms. The sample was randomised to search to demonstrate a causal link between lifestyle one of three arms: 1) very low nicotine cigarettes; 2) and OAB. This study was difficult to combine as it very low nicotine cigarettes and a 21g nicotine patch; used sophisticated statistical modelling to determine 3) patch alone. All participants had at least one OAB a causative model for use in future research. This re- symptom including urinary frequency, urgency, noc- search indicated that poor lifestyle factors causally turia or urgency urinary incontinence, smoked 10-40 linked to diabetes and obesity and suggest that this cigarettes a day for a year and were in good physical may contribute to the onset of OAB. The authors iden- and mental health. The ICIQ-OAB was used at base- tify that low physical activity could be an important line and 12 weeks, 6 weeks after completion of the modifiable causal factor for OAB operating via path- trial intervention. Outcomes included self-report of ways involving obesity or diabetes. smoking cessation and OAB symptoms and score. 96 (47%) of the 202 participants met the inclusion criteria Summary for this secondary analysis and 57 (59%) (37 women and 20 men) completed the study and were included Non RCT evidence is building which suggests that in the analysis. Data was not presented separately for moderate exercise decreases the incidence of UI alt- women and men. Those who were abstinent from hough the level of evidence remains low (Level of smoking at 12 weeks were more likely to have a re- Evidence: 3). duction in urinary frequency (p=0.042) and the ICIQ- OAB score showed no change for those who were Recommendations: abstinent or those who resumed smoking (4.6+0.5 vs In the absence of robust randomised controlled trials 4.3+0.3, respectively, p=ns). The authors conclude (Grade of Recommendation: C) that smoking cessation and its effect on OAB would need to undergo more rigorous evaluation in further trials with larger sample size to determine the effect of smoking abstinence on OAB symptoms (Table 4).

10 incomplete outcome data and selective reporting was Summary also high. Data suggest that smoking increases the risk of more Hirayama and Lees study comprised a sample of 683 severe UI. This small new study indicates that urinary men and 298 women (results for men are reported frequency may be improved by smoking abstinence. separately) aged 40-75 who completed a food fre- (Level of Evidence: 3) quency questionnaire and the Consultation on Incon- Recommendations tinence short Form (ICI-SF). The data showed a slight increase in the risk of UI at the highest level of caf- There remains a need for further prospective studies feine consumption (similar to the US data), but this of smoking cessation for urinary symptoms (Grade of was not significant after adjusting for confounding Recommendation: C) factors with OR: 95% CI, 1.12 (0.57-2.22) for women. The risk of bias reporting was largely unclear in this 1.2.5 Dietary Modification (Elimination Diet) study with blinding of participants and personnel and blinding of outcome assessment, incomplete out- Dietary factors can be divided into three groups: Diet; Fluid Intake and Caffeine. There have been no new come data and selective reporting all of high risk. The published RCTs evaluating the effect of diet or fluid sample of women did not show an association between caffeine and UI, they suggest the need for fur- intake on urinary incontinence, but one small new RCT (15), three new observational studies (1 case ther larger samples to explore any association further. control study, 1 prospective cohort study, 1 cross sectional cohort study) examining the effect of caffeine on urinary incontinence, one in men presented in sec- Summary tion IV (16), one in women (17) and one in both men There have been no new data on dietary content and women (18) (Table 5). modification and UI since 2013, the existing evidence A small double blind randomised crossover study by suggests that dietary content may play a role in uri- Wells et al. of 14 patients with OAB was conducted in nary incontinence (Level of Evidence: 3). which women were allocated to one of two groups. Minimal evidence exists on the role of macronutrient Group A were allocated to a 14 day caffeinated drink intake and reduction of UI. Fluid intake may play a period followed by a 14 day de-caffeinated period. minor role in the pathogenesis of UI (Level of Evi- Groups B were allocated to a 14 day de-caffeinated dence: 2). period followed by a 14 day caffeinated period. Both were preceded by a 14 day run-in period. The primary Caffeine consumption is likely to play a role in exac- outcome was reported episodes of urgency, fre- erbating UI and related symptoms such as urgency quency and volume per void recorded in a 3 day di- and frequency. Small clinical trials suggest that de- ary. Secondary outcomes included OAB symptom se- creasing caffeine intake improves continence. (Level verity and health related quality of life (QOL). Of the of Evidence: 2), new epidemiological evidence from eleven women who completed the study, a significant a large cross sectional study supports this conclusion reduction of urgency (p<0.1) and frequency (p<0.5) however it does not change the Level of Evidence: were reported. Risk of bias was minimised using a Level 2. double blind design, but conclusions must be cau- Recommendations tious due to the very small sample size (15). Recommendations for fluid intake and dietary intake Three studies examined the association between caf- have not changed (Grade of Recommendation: B). feine and UI, Gleason et al. (2013) (17) examined the Further RCTs to assess the role of diet in UI are war- association in women in the US and Hyrayama et al. ranted. (2012)(18) in men and women in Japan and Davis et al. (2012)(16) in a male sample in the US. A reduction in caffeine intake is recommended for those with incontinence and related symptoms In the cross sectional female US survey study, UI sta- (Grade of Recommendation: B) Larger RCTs to as- tus was collected using the Incontinence Severity In- sess the effect of caffeine and other dietary factors dex and food diaries were used to collect data on caf- are feasible and important. feine intake in 4309 women aged 20 or over. This study found an association between caffeine consumption in the highest quartile (>204mg day) was associated with any UI (prevalence odds ratio 1.47, 95% CI 1.07, 2.01), but not moderate or severe. Caf- feine consumption at the very highest level was found to be associated with UI in a very large sample of US women. All parameters in terms of risk of bias reporting were unclear or high risk. The blinding of participants and personnel and blinding of outcome assessment were at high risk of bias and the likelihood of

11 Table 3 Summary of data on physical activity

Author, Comparator Follow N Study population Age Assessment Outcomes year groups up McGrother Development of a 3411 Women in a UK 40+ Food frequency questionnaire no Low physical activity linked to the onset of 2012(1) Causative model epidemiological study of and OAB recorded in a postal OAB. Older women reporting less activity for OAB UI free from OAB at questionaire than their peers had more than double the baseline risk of developing OAB. Nygaard Case control study 1538 Women seeking 39- POP assessment, severity no SUI odds increased with overall lifetime 2015 (2) gynaecological primary 65 index score (indicating SUI), activity (OR 1.20 per 70 additional metabolic care Lifetime physical activity equivalent of task-h/wk; 95% CI, 1.02-1.41. questionnaire 1. McGrother CW, Donaldson MMK, Thompson J, Wagg A, Tincello DG, Manktelow BN. Etiology of overactive bladder: A diet and lifestyle model for diabetes and obesity in older women. Neurourology and Urodynamics. 2012;31(4):487-95. 2. Nygaard IE, Shaw JM, Bardsley T, Egger MJ. Lifetime physical activity and female stress urinary incontinence. American Journal of Obstetrics and Gynecology. 2015;213(1).

Table 4 Summary of data on smoking cessation

Study Follow Notes (side effects, loss of follow up, Author, year Comparator N Age Outcomes/results population up risk of bias) Wyman et al. Low nicotine cigarette 202 Subset of adult 18-70 ICIQ-OAB 12 weeks Subset of participants: 2014 (1) alone vs low nicotine participants in a Self-reported smoking Random sequence generation: uncertain cigarette combined with smoking status and cessation Allocation concealment – uncertain 21mg nicotine patch vs cessation trial Blinding of participants – uncertain patch alone Blinding of outcome assessment – uncertain Incomplete outcome data – low risk Selective reporting – unclear risk 1. Wyman J, Allen A, Hertsgaard L, Overson E, Allen S, Hatsukami D. Effect of Smoking Cessation on Overactive Bladder Symptoms in Adults: A Pilot Study. Neurourology and Urodynamics. 2014;33(6):866-7

Table 5 Summary of data on caffeine consumption

Author, year Comparator N Study population age Outcomes/results Follow Notes (limitations) up Davis 2013 (1) Association 3960 Publically available data 20+ Incontinence Severity no Cross sectional design means that no between caffeine from the National Health Index (ISI) causation can be determined. The ISI needs and UI in US men and Nutrition additional validation for use in men. 25% of

12 Author, year Comparator N Study population age Outcomes/results Follow Notes (limitations) up Examination Surveys Structured dietary the sample was excluded due to missing UI 2005-2006 and 2007- recall and dietary data. 2008 in men Gleason 2013 (2) Association 4309 Publically available data 20+ Incontinence Severity no The cross sectional design means that no between caffeine from the National Health Index (ISI) causation can be determined. Analysis and UI in US and Nutrition Structured dietary depended on self report. Questions for UI women Examination Surveys recall type were not validated and data was not 2005-2006 and 2007- analysed according to UI type or severity. 2008 in women Hirayama 2012(3) Association 981 Community dwelling 40-75 ICIQ-SF 18 Cross sectional study: No cause-effect between caffeine Japanese men and months relationship could be established due to the use and UI in women study design. Classification of urine loss was Japanese adults based on self report

Wells 2014 (4) Group A were 14 Women patient with 18+ Reported episodes of no Small double blind randomised crossover allocated to a 14 community OAB urgency, frequency study day caffeinated dwelling and volume per voids drink period women recorded in a 3 day followed by a 14 diary day de-caffeinated period. Groups B were allocated to a 14 day de- caffeinated period followed by a 14 day caffeinated period. 1. Davis NJ, Vaughan CP, Johnson TM, Goode PS, Burgio KL, Redden DT, et al. Caffeine Intake and its Association with Urinary Incontinence in United States Men: Results from National and Nutrition Examination Surveys 2005-2006 and 2007-2008. Journal of Urology. 2013;189(6):2170-4. 2. Gleason JL, Richter HE, Redden DT, Goode PS, Burgio KL, Markland AD. Caffeine and urinary incontinence in US women. International Urogynecology Journal. 2013;24(2):295-302. 3. Hirayama F, Lee AH. Is caffeine intake associated with urinary incontinence in Japanese adults? J Prev Med Public Health. 2012;45(3):204-8. 4. Wells MJ, Jamieson K, Markham TC, Green SM, Fader MJ. The effect of caffeinated versus decaffeinated drinks on overactive bladder: a double-blind, randomized, crossover study. Journal of wound, ostomy, and continence nursing : official publication of The Wound, Ostomy and Continence Nurses Society. 2014;41(4):371-8.

13 1.2.6 Constipation  Is one type of PFMT programme better than another in the treatment of UI? No new trials on constipation were found. Therefore, the evidence remains the same (small observational  Is PFMT better than other treatments in the treat- studies) suggesting that chronic straining may be a ment of UI? risk factor for the development of UI (Level of Evi- dence: 3).  Does the addition of PFMT to other treatments add any benefit in the treatment of UI? Recommendation  What factors might affect the outcome of PFMT Further research is needed to define the role of strain- in the treatment of UI? ing in the pathogenesis of UI.  What is the effect of PFMT on other lower urinary tract symptoms (LUTS)? 1.3. Other LUTS 2.1. Prevention and Treatment (Pregnant No data were found. and Postnatal Women Only) 1.4. Factors Affecting Outcome This subsection specifically considers PFMT for the No data were found. prevention and treatment of UI in pregnant and postnatal women (called childbearing women). As the physiological changes of childbearing can affect PFM 2. PELVIC FLOOR MUSCLE function, it is possible that the effect of PFMT might TRAINING (PFMT) differ in this group compared to non-childbearing women; therefore, this group is treated separately. Pelvic floor muscle training (PFMT) is defined as exercise to improve PFM strength, endurance, power, Since the last ICI (2013), 11 randomised controlled relaxation or a combination of these parameters (1). trials aiming at prevention and/or treatment of UI were PFMT remains a key factor in the treatment of UI. Be- identified and reviewed for this subsection (29) (30) cause pelvic floor muscle (PFM) integrity appears to (31) (32) (33) (34) (35) (36) (37) (38) (39) (Table 6). play an important role in the continence mechanism An updated Cochrane review was published in 2012, (see report from Chapter 4: Pathophysiology), there analysing data from 22 trials involving 8485 women is a biological rationale to support the use of PFMT in (40). Trials in this section have been grouped in 3 ar- preventing and treating SUI in women (19-21). The eas: 1) trials of PFMT for prevention of UI (performed role of PFMT in the treatment of UUI came later, when in women without UI symptoms when randomised); it was recognized that PFM contraction can also be 2) trials of PFMT for treatment of UI; and 3) trials of used to occlude the urethra to prevent leakage during PFMT for the prevention and treatment of UI (partici- detrusor contraction, as well as inhibit and suppress pants with and without UI symptoms enrolled). Trials detrusor contraction (22, 23). More details about the were further separated into those beginning during biological rationale regarding PFMT for SUI and UUI pregnancy (antenatal) or postnatal. can be found in the last edition of this chapter (2). The primary outcome of interest was self-reported UI PFMT is an intervention that involves the understand- (cure, improvement, number of leakage episodes). ing of PFM activation and the pursuit of a repeated Other outcomes of interest included adherence PFM exercise programme over time. Because effec- measures. tiveness depends on the participant's adherence dur- 2.1.1 Is PFMT Effective in the Prevention of ing the intervention and afterwards (in the mainte- UI in Childbearing Women? nance phase), a better understanding of adherence mechanisms and how they can be promoted is of ma- This section addresses the question of PFMT effec- jor importance. The 2011 International Continence tiveness for primary and secondary prevention of UI Society State of the Science Seminar on Adherence in childbearing women. Clinically it can be difficult to produced 4 papers and a Consensus Statement re- effectively screen trial participants to ensure that a viewing present literature and making recommenda- disease process is altogether absent (for primary pre- tions to increase PFMT adherence that could be use- vention studies) or present, although asymptomatic ful, both in the clinical and research settings UI (24- (for secondary prevention). Trials investigating pre- 28) vention of UI usually enrol people purely on the basis of the absence of symptoms. Thus, the trials in this This section presents evidence for the use of PFMT section likely represent a combination of primary and in the prevention and treatment of UI in women. secondary prevention effects. Questions addressed are: Since last chapter edition three new prevention trials  Is PFMT effective in the prevention of UI? were found (30, 36, 38) (Table 6) adding to the six  Is PFMT better than no treatment, placebo or previously existing trials. control treatments in the treatment of UI? Two studies (30) (38) recruited nulliparous or primiparous women during pregnancy, and one recruited

14 "pregnant women" (36). Primigravidae were recruited Research recommendation at weeks 6-9 (30), 10-14 (38) and 14-20 (36). In all trials, PFMT began during pregnancy while controls Additional trials with long-term follow-up (more than received usual antenatal care, which may have in- 12 months postnatal) are needed to determine long- cluded advice on PFMT, from their maternity caregiv- term benefits of antenatal PFMT. ers (30) (36) (38) There were some variations in the The only study including multiparous women is a PFMT parameters (intensity and supervision) (Table quasi-randomized trial. Thus, large and good-quality 6). RCTs are needed to investigate the effect of antepar- Quality of data tum PFMT on preventing postpartum UI in multiparous women. Two were RCTs (30, 38), while one was a quasi-randomized trial (36). Allocation concealment appeared 2.1.2 Is PFMT Effective in the Treatment of UI adequate only in the trial by Pelaez (2014) (38). The in Childbearing Women? outcome assessors were not blinded in the trials by Kocaoz (2013) (36) and Pelaez (2014)(38), while Since last chapter edition two trials assessing the Barakat (2011)(30) gives no information about blind- treatment effect of postnatal PFMT were found (29) ing of evaluators. Dropout rates were 16% (30), 10- (34) (Table 6) adding to the four previously existing 14% (38) and 22% (36), and quite similar in both trials. study groups. Outcomes were measured at various Kim (2012) recruited a mix of primiparous and multip- times: outcome in late pregnancy was measured only arous incontinent women less than six weeks (34) af- after delivery (30), 36 weeks of pregnancy (38); 28 ter delivery, while Åhlund (2013) recruited only pri- and 32 weeks of pregnancy then at three months miparous women 10-16 weeks after delivery (29). postpartum (36). None of the trials applied intention- The control group in the trial by Åhlund received to treat analysis. standard care, which included ante and postpartum Results advice on PFMT (1), whereas the control group/unsupervised training group in Kim’s study were instructed  Late pregnancy (34 weeks or later): Pelaez (38) in PFMT and followed the same PFMT programme as found a statistically significantly lower frequency the supervised training group (34). The PFMT inter- of UI (p<0.001) in the intervention group com- ventions varied (Table 6). pared to the control group between 36 and 40 weeks of pregnancy, and concluded that PFMT Quality of data was effective in primary prevention of UI in pri- Both studies were RCTs, but random allocation con- miparous pregnant women, while Barakat (30) cealment was inadequate. In one trial (34) evaluators reported no difference in frequency of urine loss were blinded to group allocation. Åhlund (2013) and between the exercise (supervised moderate Kim (2012) reported an 8% and 1% loss to follow up, physical exercise including PFMT) and control respectively. Similar dropout rates were found in both groups in late pregnancy. study groups in both trials. Outcomes were meas-  Mid postpartum (three to six month postpartum): ured, before, 3 months after delivery (34) and 6 Kocaoz (36) reported a statistically significant months after delivery (29). Kim did not report analysis difference between the intervention and control by intention-to-treat, while information is lacking in the groups in terms of development of stress urinary trial by Åhlund. th nd incontinence at the 28 and 32 weeks of preg- Results nancy and the 12th postpartum week (p<0.05). In the trial by Åhlund the results showed significantly Summary improved continence in both groups, however there was no between group comparison. Kim reported sig- Two studies documented the effectiveness of PFMT nificant difference in UI symptoms in favour of the su- on primary prevention of UI during pregnancy (36, pervised PFMT group. 38). One study included only nulliparous women and reported UI in late pregnancy (38). The other study Summary also included multiparous women and reported UI in late pregnancy and 12 weeks postpartum (36). (Level Data from one trial showed significantly better treat- of Evidence: 1) ment effect of supervised PFMT compared to unsupervised PFMT. The other trial did not compare dif- Recommendations ference in treatment effect between groups. The ad- Offer continent, pregnant women a supervised (in- dition of the new trials does not change the level of cluding regular health professional contact) and in- evidence. (Level of Evidence: 1) tensive strengthening PFMT programme to prevent Recommendations antepartum and postpartum UI (Grade of Recom- mendation: A) PFMT should be offered as first line conservative therapy to women with persistent UI symptoms three

15 months after delivery. (Grade of Recommendation: Results A) i) Antepartum PFMT versus usual care An ‘intensive’ PFMT programme (in terms of supervision and exercise content) is likely to increase the Ko (2011), Miquelutti (2013) and Stafne (2012) treatment effect. (Grade of Recommendation: B) showed that women who were randomised to antepartum PFMT had significantly less risk of UI in late Research recommendation pregnancy compared to a group receiving usual care in late pregnancy. There is a need for at least one large, pragmatic, well- conducted and explicitly reported trial with long term Three of the new five new mixed prevention and treat- follow-up (five plus years) of postpartum PFMT that ment trials assessing the effect of antenatal PFMT investigates the long term effect of ‘intensive’ PFMT. (35) (37) (39), reported a significant effect of PFMT during pregnancy, and 3 months after delivery (35). 2.1.3 Is PFMT Effective in the Mixed Preven- Bø (2011) found no difference in late pregnancy and tion and Treatment of UI in Childbear- 3 months postpartum, between a group following an ing Women? aerobic fitness class including PFMT and a control group. While Fritel (2015) showed no difference in Since last chapter edition six mixed prevention and prevalence or severity of UI between a group receiv- treatment trials were found adding to the ten previ- ing written instructions and a group receiving an ad- ously existing trials. Five trials assessed the effect of ditional eight PFMT sessions with a midwife or phys- antenatal PFMT (31) (35) (37) (39) (32), and one the iotherapist in late pregnancy, 2 and 12 months post- effect of postnatal PFMT (33) (Table 6). partum. In five trials, nulliparous, primiparous or multiparous Some adherence outcomes were included for four women were randomised to either supervised ante- antepartum (31) (35) (32) (39) trials; >80% of the partum PFMT or usual antepartum care (31) (32) (35) PFMT women attended every group session, and at (37) (39). One study randomised nulliparous women 36 gestational weeks, 87% reported PFMT practice to supervised postpartum PFMT or usual postpartum at least 75% of the time (35); 40% of the exercise care (33). Ko (2011) and Miguelutti (2013) recruited group attended >80% of the weekly exercise classes nulliparous women and Stafne (2012) multiparous (31); 67% of the women in the intervention group and women between 16-24 weeks’ gestation. Bø’s study 40% in the control group performed PFMT three times (2011) differed in that the primary aim was to assess per week or more (39); 54% in the intervention and the effect of regular exercise on weight gain in preg- 63% in the control group performed postpartum nancy. Primiparous, sedentary women within the first PFMT (32). 24 weeks of pregnancy were recruited. Fritel (2015) included nulliparous women between 20-28 weeks of ii) Postpartum PFMT versus usual care or no PFMT gestation. The only new mixed prevention and treatment trial as- i) Antepartum PFMT versus usual care sessing the effect of postnatal PFMT (33), found no differences in the prevalence of UI between a super- The PFMT interventions varied (Table 6). vised PFMT group and a control group receiving instructions in correct PFM contractions and written in- Quality of data formation. Adherence was 96% in the intervention  Antepartum PFMT versus usual care: The four group (33). RCTs had adequate random allocation generation and concealment and thus had a low risk of bias (31) (32) (37) (39). In two of the trials a blinded method of collecting patient-reported incontinence symptom data was used (32) (39). Sample sizes varied between 105 and 855 women (31) (39), and the proportions of lost to follow up were 20%, 2%, 33% and 11% respectively (31) (37) (32) (39). Losses to follow up in intervention and control groups were quite similar in all trials. Intention-to-treat analysis was performed by all, except for Bø (2011).  Postpartum PFMT versus usual care or no PFMT: Random allocation generation and concealment was adequate and a blinded method of collecting patient-reported incontinence symptom was used in the trial by Hilde (2013) (33). Lost to follow up was 9%, in the control 14% and in the exercise group 3%. Analysis was by intention-to-treat.

16 Table 6 Summary of PFMT data on prevention and treatment (pregnant and postnatal women)

Author, Comparator N Study population Modality details or parameters Outcomes/results Follow Notes (side effects, year up loss of follow up…) Ahlund Control group (n=49): N=98 Primiparous women Three visits with midwife during the Significantly (p<0.05) no Loss to follow up: 16% 2013(1) instructions and control of with SUI, 10-16 intervention period. A 6 months home improved continence CG: 7/49 correct PFM contractions weeks after delivery PFM program including three fast score (0-20) in both IG: 9/49 and written information contractions, three times 8-12 slow groups at 6 months about PFMT. Multi centre velocity (6 seconds), close to maximum postpartum. Sweden contractions Difference between Intervention group (n=49): groups was not reported. instructions and control of correct PFM contractions, written information about PFMT and follow up by midwife Bo 2011 Control (n=53): standard N= Nulliparous, 12 -16 weeks of aerobic exercise classes No significant difference oss to follow up: 20% (2) care 105 sedentary women twice per week during pregnancy, between groups: no CG: 11/53 within the first 24 including intensive PFMT (in a group) led IG: 10/52 Intervention (n=52): 12-16 weeks of pregnancy by aerobic instructor. Additional home Self reported UI at 36-38 wk’ aerobic fitness class With and without UI exercises 10 max contractions (each held weeks of pregnancy: including PFMT at inclusion. for six seconds) and to the last 4 were 3- Control: 7/53 4 fast contractions added x 3, per day. Intervention: 9/52 Single centre Correct VPFMC was not checked at Norway enrolment. Self reported UI at 3 months postpartum: Control: 6/53 Intervention: 5/52 Barakat Control group (n=40): N= 80 Sedentary women Exercises to strengthen the PFM included No significant difference no Loss to follow up: 16% 2011 (3) standard care within 6-9 weeks of in a general exercise program with 45 in frequency of urine loss CG: 6/40 pregnancy. minutes sessions x 3 per week, during (CIQ-SF Incontinence IG: 7/40 Intervention group (n=40): Without UI at the last trimester. No further details about classification) in late Physical conditioning inclusion. the PFMT program. pregnancy. program 45 minutes 3 (Questionnaire answered times per week from Single centre after delivery). inclusion to week 38-39. Spain PFMT included during the last trimester.

17 Author, Comparator N Study population Modality details or parameters Outcomes/results Follow Notes (side effects, year up loss of follow up…) Fritel 2015 Control group (n=142): N=282 Nulliparous women The PFMT group received eight individual No significant difference Yes Loss to follow up: 33% (4) standard care including between 20-28 training sessions supervised by a midwife between groups in UI 12 CG: 45/142 written instructions about weeks of gestation. or a physiotherapist, once per week severity (ICIQ-UI SF months IG: 47/140 PFMT With and without UI between the sixth and eight month of score) and prevalence at at inclusion. pregnancy. Each session lasted 20-30 2 and 12 months Intervention group minutes and included evaluation of pelvic postpartum: (n=140): Individual Multicentre floor muscle contraction. Sessions UI severity at 2 months supervised PFMT France consisted of standing contractions (5 postpartum (mean, SD): minutes), lying contractions (10 minutes) Control: 2.3 (±3.4) and learning how to start a pelvic floor Intervention 1.7 (±2.9) muscle contraction just before exerting intraabdominal pressure (the Knack). UI severity at 12 months Women were encouraged to perform postpartum: daily PFMT. No specific instructions on Control: 2.1 (±3.3) the number or intensity of the Intervention: 1.9 (±3.7) contractions were given. Hilde 2013 Control group (n=88): N=175 Singleton PFMT in groups supervised by No significant differences no Loss to follow up: 9% (5) receiving instructions in primiparous women physiotherapist once per week in 16 in the percentage of CG: 3/88 correct PFM contractions Who delivered weeks (starting 6-8 weeks after delivery), women with UI at six IG: 12/87 and written information vaginally after more and daily home training with three sets of months postpartum than 32 weeks of 8 to 12 contractions close to maximum between a supervised Intervention group (n=87): gestation PFMT group and a supervised PFMT control group receiving Single centre instructions in correct Norway PFM contractions and written information. UI 6 months postpartum: Control: 39% Intervention: 35% Kim 2012 Control (n=10): N=18, Mix of primiparous PFMT in various positions (20 repetitions Significant difference no Loss to follow up: 10% (6) unsupervised PRMT and multiparous with 10 second holding), abdominal (p=0.001) in change in CG: 9/10 incontinent women strengthening exercises and trunk values for UI symptoms IG: 9/10 Intervention (n=10): less than six weeks stabilisation using a therapeutic ball. ((BFLUTS) between Supervised PFMT after delivery. All with Twenty three one-hour sessions of groups, in favour of the UI at inclusion. training with physiotherapist, three times IG per week in an 8 week period. Additional (mean, SD): Single centre daily home training program Control: -18 (± 5.5) Korea Intervention: -27 (± 6.2)

18 Author, Comparator N Study population Modality details or parameters Outcomes/results Follow Notes (side effects, year up loss of follow up…) Ko Control: Routine N=300 Nulliparous women The PFMT group met individually with a Self reported UI at 36 wk’ Yes Loss to follow up: 0% 2011 (7) antenatal care. recruited at 16-24 physical therapist for instruction and pregnancy p<0.01: 6 weeks of pregnancy. assessment of correct contraction. The Control: 76/150 (51%) months Intervention: Individual With and without UI PFMT included three sets of eight Intervention: 52/150 PFMT with at inclusion. contractions repeated twice daily. (34%) physiotherapist once per Additional group therapy occurred weekly Self reported UI at 3days week between 20-36 Single centre, in 45 minute sessions over a 12 week postpartum p=0.06: weeks pregnancy with Taiwan period. Control:62/150 (41%) additional home exercises Intervention:46/150 (30%) 3 sets of 8 contractions Self reported UI at 6 (each held for 6 seconds) weeks postpartum repeated twice daily. p=0.06: Instructed to contract the Control: 53/150 (35%) PFM when coughing or Intervention: 38/150 sneezing. (25%) Self reported UI at 6 months postpartum p=0.04: Control: 42/150 (27%) Intervention: 25/150 (16%) Kocaoz Quasi randomized trial N=136 Continent, pregnant Correct voluntary PFM contraction was A statistically significant yes Loss to follow up: 25% 2013 (8) Control group (n=68): ? women at week 14- checked prior to training difference was found 12th CG: 18/68 20 of pregnancy. Daily PFMT, three sets of 10 between control and week IG: 16/68 Intervention group (n=68): Without UI at contractions. Maximal contraction of the intervention group in SUI postpar Instructions in PFMT and inclusion. PFM and hold for 10 seconds, and quick development at 28th and tum information. contractions 32nd weeks of gestation Single centre and the 12th week Turkey postpartum.

19 Author, Comparator N Study population Modality details or parameters Outcomes/results Follow Notes (side effects, year up loss of follow up…) Miquelutti Control group (n=100): N=205 Nulliparous low risk On the days of prenatal visits; consisted The risk of UI was no Loss to follow up: 2% 2013 (9) standard care women at 18 weeks of physical exercises, educational significantly lower in the CG: 2/102 of gestation. activities and instructions on exercises to intervention group at 30 IG: 6/103 Intervention group (n=97): With and without UI be performed at home. The PFMT should weeks of pregnancy: Additional physical and at inclusion. be performed daily at home and included CG 62% educational activities rapid (30 times) and sustained maximal IG 43% including information Multi centre contractions (20 times holding for 10 RR 0.69 (95% CI 0.51- about PFMT. Brazil seconds) in sitting and standing positions. 0.93) and 36 weeks of pregnancy: CG 68% IG 41% RR 0.60 (95% CI 0.45- 0.81) Pelaez Control group (n=96) N=169 Healthy, primiparous Correct voluntary PFM contraction was At 36 weeks of pregnancy no Loss to follow up: 16% 2014 (10) pregnant women in checked prior to training there was statistically CG: 7/96 Intervention group (n=73) gestational week 10- The intervention consisted of 70-78 group significant (p=0.001) IG: 10/73 14. sessions; 10 minutes of PFMT three difference in frequency of Without UI at times per week in 22 weeks UI and in ICIQ-UI SF inclusion. score in favour of the intervention group. Single centre CG 2.7 (SD4.1) Spain IG 0.2 (SD 1.2) Stafne Control group (n=426): N=855 Multiparous women The PFMT group met individually with a Self-reported UI at 34-38 no Loss to follow up: 11% 2012 (11) standard care between 16-24 physical therapist for instruction and weeks of pregnancy: CG: 61/426 weeks gestation. assessment of correct contraction, and Any UI (p=004): IG: 32/429 Intervention group With and without UI PFMT protocol was similar to that CG 192/365 (53%) (n=429): 12 weeks of at inclusion. described previously by Mørkved (1997, IG 166/397 (42%) PFMT 2003), including three sets of eight Multi centre contractions repeated twice daily. UI once per week or more Norway Additional group therapy occurred weekly (p=0.004): in 45 minute sessions over a 12 week CG 68/365 (19%) period. IG 44/397 (11%) 1. Ahlund S, Nordgren B, Wilander EL, Wiklund I, Friden C. Is home-based pelvic floor muscle training effective in treatment of urinary incontinence after birth in primiparous women? A randomized controlled trial. Acta Obstetricia et Gynecologica Scandinavica. 2013;92(8):909-15. 2. Bo K, Haakstad LA. Is pelvic floor muscle training effective when taught in a general fitness class in pregnancy? A randomised controlled trial. Physiotherapy. 2011;97(3):190-5.

20 3. Barakat R, Pelaez M, Montejo R, Luaces M, Zakynthinaki M. Exercise during pregnancy improves maternal health perception: A randomized controlled trial. Amer- ican journal of obstetrics and gynecology. 2011;204(5):402. 4. Fritel X, de TR, Bader G, Savary D, Gueye A, Deffieux X, et al. Preventing Urinary Incontinence With Supervised Prenatal Pelvic Floor Exercises: A Randomized Controlled Trial. Obstetrics & Gynecology. 2015;126(2):370-7. 5. Hilde G, Staer-Jensen J, Siafarikas F, Ellstrom EM, Bo K. Postpartum pelvic floor muscle training and urinary incontinence: a randomized controlled trial [Erratum appears in: Obstet Gynecol. 2014 Sep;124(3):639]. Obstetrics & Gynecology. 2013;122(6):1231-8. 6. Kim EY, Kim SY, Oh DW. Pelvic floor muscle exercises utilizing trunk stabilization for treating postpartum urinary incontinence: randomized controlled pilot trial of supervised versus unsupervised training. Clinical Rehabilitation. 2012;26(2):132-41. 7. Ko PC, Liang CC, Chang SD, Lee JT, Chao AS, Cheng PJ. A randomized controlled trial of antenatal pelvic floor exercises to prevent and treat urinary incontinence. Int Urogynecol J. 2011;22(1):17-22. 8. Kocaoz S, Eroglu K, Sivaslioglu AA. Role of pelvic floor muscle exercises in the prevention of stress urinary incontinence during pregnancy and the postpartum period. Gynecologic & Obstetric Investigation. 2013;75(1):34-40. 9. Miquelutti MA, Cecatti JG, Makuch MY. Evaluation of a birth preparation program on lumbopelvic pain, urinary incontinence, anxiety and exercise: A randomized controlled trial. BMC Pregnancy and Childbirth. 2013;13(154). 10. Pelaez M, Gonzalez-Cerron S, Montejo R, Barakat R. Pelvic floor muscle training included in a pregnancy exercise program is effective in primary prevention of urinary incontinence: a randomized controlled trial. Neurourology & Urodynamics. 2014;33(1):67-71. 11. Stafne S, Salvesen K, Romundstad P, Torjusen I, Morkved S. Does regular exercise including pelvic floor muscle training prevent urinary and anal incontinence during pregnancy? A randomised controlled trial. BJOG: an International Journal of Obstetrics & Gynaecology. 2012;119(10):1270-80.

21 In this trial, randomisation was conducted prior to eli- Summary gibility assessment, leading to a relatively high rate of The effect of antepartum PFMT or postpartum PFMT, non-completion (97/238 in the treatment group and in groups of women where some did and some did 65/242 in the control group; 162/480 overall). Analy- not have prior UI symptoms, varied by study with sis was not ITT. The method of randomisation was some showing a benefit on UI prevalence whereas not described but the randomization block size was others did not. The characteristics of the new trials, set at 16 to minimize the chance of prior recruiter all methodologically robust, that demonstrated re- knowledge of subject assignment. The assessors duced UI prevalence in late pregnancy (35) (37) (39) were not blinded. and six months postpartum (35), were high adherence to a supervised PFM strength training program Summary and home exercises. (Level of Evidence: 2) There is preliminary evidence that PFMT may help However, a similar training protocol in postpartum prevent UI in older women. (Level of Evidence: 2) women by Hilde (2013) did not show any difference between groups. Even if the combined results from However, more definitive trials are needed to clarify the published trials still are not conclusive, the benefit the effects of this multi-component program or other of PFMT in the majority of the studies and the lack of approaches to using PFMT as a prevention strategy adverse effects of PFMT should be taken into ac- in older women. count. Recommendations Recommendations Without robust randomised controlled trials on the Health providers should carefully consider the preventive effect of PFMT on UI the Grade of Rec- cost/benefit of population-based approaches to ommendation is Grade C (New). health professional taught antepartum or postpartum With limited new data on the effects of preventive PFMT, that is, health professional instruction to all PFMT on UI in older women, this association should pregnant or postpartum women regardless of their be investigated further. current or prior continence status (Grade of Recommendation antepartum PFMT: A) 2.3. Treatment (Other Women) (Grade of Recommendation postpartum PFMT: B) 2.3.1 Is PFMT Better than No Treatment, a Placebo, or a Control Group Treat- Where a population approach is used, the ‘best’ evi- ment? dence to date suggests the following: (a) an intervention comprising of a daily home PFMT and weekly This updated literature review identified 19 new trials physiotherapist-led exercise classes for 12 weeks, that compared PFMT to no treatment, sham treat- starting at 16-24 weeks’ gestation for pregnant ment, or control treatment. Six trials were excluded women, and (b) an individually taught strengthening from this summary because they selected women PFMT programme that incorporates adherence strat- with “OAB,” “lower urinary tract symptoms,” or “pelvic egies for postpartum women who have had a forceps floor dysfunction,” thereby including women without delivery or a vaginal delivery of a large baby (4000g urinary incontinence (42-47). Some included conti- or more). (Grade of Recommendation: C) nent women, while others did not provide sufficient information to determine the continence status of par- 2.2. Prevention (Other Women) ticipants. One additional trial was excluded due to insufficient detail about the intervention needed to dis- Although there are multiple trials of PFMT for preven- cern whether PFMT was part of the treatment (48), tion of UI in peri-partum women (see section II.2.1) and one was not considered because the sample and in men undergoing prostatectomy (see section consisted of residents in long-term care facilities (49). IV.2), there is little research on prevention of UI in The 11 additional trials included in the analysis were non-childbearing women. A single randomised trial diverse and included studies from 10 countries (See (41) was found that investigated the preventive ef- Table 7). Samples included young, nulliparous volley- fects of a multi-component behavioural modification ball athletes, older women, peri-and post-menopau- program, including PFMT, Bladder Training (BT), and sal women, obese women, and gynaecological can- other behavioural skills, compared to no intervention cer survivors. Most samples were drawn from com- in 359 older women who were essentially continent munity-dwelling populations. (0-5 days of incontinent episodes in previous year). The intervention was a 2-hour class followed 2-4 PFMT Details weeks later with an individualized session to test Most trials involved 12 weeks (3 months) of interven- PFMT technique and reinforce adherence. After 12 tion, but some were of 6 weeks (50) or 8 weeks dura- months, continence status was the same or better in tion (51). Most involved supervised PFMT accompa- 56% of the prevention group compared to 41% of the nied by a regimen of home-based exercises by a control (p=0.01). physiotherapist or nurse. One study tested a new smart phone application (52) with no face-to-face

22 contact. In several trials, PFMT was delivered on an were reported in two (54, 56), no differences in a third individual basis (50, 53-57), while others were con- (55), the within and between group differences did not ducted in a group class (51, 58). Some trials were re- reach statistical significance (57). stricted to PFMT with or without functional PFM contraction to prevent SUI episodes (stress strategies, Summary the Knack) (50, 57), while in others, PFMT was embedded in a multi-component programme with other PFMT is effective as a stand-alone therapy, as part of behavioural or exercise components (Table 7). One multi-component therapies embedding PFMT with trial included PFMT in a broader programme of gen- concomitant behavioural strategies, lifestyle eral physical exercise (58). changes, and as part of more general physical exercise programmes to improve physical function in Risk of Bias older women. Some of these 11 trials may have been limited by Results expand the evidence base to include PFMT small sample size. Only 3 included more than 100 implemented by mobile technology, with a potentially women, 3 included 50-100 women, and 5 had fewer broader reach, cost savings, and impact on rural than 50 participants. Two trials reported intent-to- health. treat analysis (52, 55). Three reported the method of randomisation (54, 55, 58); five reported concealment Benefits are shown across age cohorts and UI type, (52, 55, 58-60); and two reported using blinded out- in various cultural contexts, using several different come assessors (55, 58). Two studies were reported training regimens, and assessed by multiple outcome as abstracts making it difficult to assess some as- measures. (Level of Evidence: 1) pects of the methods and data quality (52, 55). Recommendations Results Supervised PFMT should be offered as a first-line A variety of outcome measures were used, including conservative therapy for women of all ages with uri- bladder diaries, validated questionnaires, global pa- nary incontinence (Grade of Recommendation: A). tient ratings, and pad test. Most trials used multiple measurements to evaluate outcomes. 2.3.2 Is One Type of PFMT Programme Better than Another? Bladder diaries were used in 7 studies to assess change in incontinent episode frequency (52, 54-59). A number of factors can influence the outcome of a In all studies, reductions in incontinence episode fre- PFMT programme such as the way it is taught and/or quency were significantly greater in the treatment supervised, the parameters of the actual exercises, group. and adherence to the training regimen. What, therefore, is the most effective PFMT programme? The UDI or UDI-6 was used in 4 trials to report change in urinary symptoms (54, 55, 57, 60). Signifi- This updated search revealed 24 new potentially eli- cant differences were found in the two larger trials gible RCTs to address this question. Of them, one (54, 55). The ICIQ-UI was used as an outcome meas- was excluded due to insufficient information on the ure in two trials (51, 52), both demonstrating greater incontinence outcome and inter-group comparison benefit for PFMT compared to control. (61) and three because they did not report any incontinence outcomes but focused on PFM outcomes (62- In 4 trials, the IIQ-7 was used to assess change in 64). Table 8 shows the studies included in the previ- impact of UI. Two found significant between group dif- ous edition as well as those identified in this update. ferences (54, 59); one found significant change in the Characteristics of each new RCT are presented in Ta- treatment group, but not in the control group; and the ble 9. third did not observe improvement on this condition- specific quality of life measure (60). The PGI-I was the primary outcome in one study that found 80% of women in the treatment group and 40% in the control group were “much better” or “very much better” (60). Similarly, in another trial, 55.7% of women in the treatment group were “better” compared to 5% in the control group (52).

Urine loss, assessed by 1-hour pad test (with provo- cation), was the primary outcome for one trial (50). Reductions in urine loss were greater in both treatment groups (individualized PFMT and group administered PFMT) compared to the control group. In four other studies that used a pad test as a secondary outcome, significant reduction and group differences

23 Table 7 Summary of data on PFMT vs no treatment, a placebo or a control group treatment

Study Comparator N Study Population Intervention Outcomes/Results Notes Abdulaziz No therapy N=56 Obese, older, peri- 3 months of supervised PFMT taught Fewer UI episodes on self-report Small N may have affected 2012 (1) Tx: 29 menopausal Saudi at physio clinic; 10 repetitions of 8 (not BD) and better score on VAS results CG: 27 women with SUI (40- maximal contractions each, progressed (but p>0.05 for group diff) 50 years) to 12; 36 sessions Tx group: significant change and 90% cured CG: no change and 19% cured Asklund Delayed treatment N=123 Women with weekly 3 months of Smart phone app; PFMT Web-based questionnaires and 2- ITT analysis 2015 Tx: 62 SUI (>18 years; mean exercises at different levels (6 basic day leakage diary; ICIQ-UI-SF, Allocation concealment with Abstract CG: 61 45 years) and 6 advanced); graphic support and PGII, IEF, ICIQ-LUTSqol, use of sequentially numbered (2) functions for statistics and reminders. incontinence aids, patient envelopes No face-to-face contact satisfaction (Tx group only). Tx group showed greater improvements on all outcomes (p<0.001). Celiker Wait-list N=130 Women with SUI/MUI 12-weeks of PFMT, 30 min sessions, UDI-6, IIQ-7, 3-day bladder diary, Not ITT; randomization 2015 (3) Tx: 65 Excluded weak supervised by physiotherapist, X3 per stop test, pad test. computer generated; CG: 65 muscles week initially, taught by perineal Tx group showed significant allocation concealment palpation, individually prescribed, improvement on all outcomes and Tx included home-based exercise. Included advice greater improvements than CG behavioral/lifestyle advice. on bladder hygiene (p<0.0001). Dumoulin Education N=48 Postmenopausal 12 weeks of Individualized PFMT, 7-day BD, 24-hour pad test, UDI. Randomization computer 2011 (4) program on Tx: 24 women with weekly 30-min PT sessions; daily Post-treatment, Tx group had generated Abstract osteoporosis (3 CG: 24 osteoporosis and SUI, home exercise; BF, urge control, BT, greater improvements on IEF Allocation concealment hours) and follow- UUI, or MUI dietary advice and constipation advice. (p=0.04) and UDI (p=0.04), but not ITT analysis up phone calls (55+ years) pad test. Blinded assessors At 1-year follow up, there were No adverse events significant differences on IEF Tx included (p=0.04), UDI (p=0.03) and pad behavioral/lifestyle advice. test (p=0.01). Ferreira Summary of N=32 Nulliparous, female 3 months of PFMR; “educational 7-day BD and pad test (15 min Randomization by lottery 2014 (5) PFMR education Tx: 16 volleyball athletes with action,” awareness of PFM, and during practice). (participant drawing paper program via CG: 16 SUI (13-30 years) informational pamphlet; weekly visits; Tx group showed greater from a box) pamphlet home exercise - 30 sustained reductions on IEF (p<0.001) and Small N contractions and 4 quick contractions pad test (P<0.001). daily; daily bladder diary for awareness

24 Study Comparator N Study Population Intervention Outcomes/Results Notes Karger No Tx N=50 Elderly women with Two months; eight 45-min group ICIQ-UI and self-esteem. Tx included other types of 2015 (6) Tx: 25 SUI (60-74 years) training classes taught by a nurse; 8- Group differences observed on exercise. CG: 25 12 contractions in each of 3 positions, ICIQ-UI (p=0.001), and on 6-8 second contractions with 3-4 fast individual items of frequency, contractions added; home exercise: 8- amount of leakage, and QoL. 12 high intensity contractions, X3 daily. Strength, awareness, and relaxation training for other muscle groups. Kim 2011 General health N=127 Elderly Japanese 3 months, multi-dimensional exercise Primary outcome was cure based Not ITT (7) class monthly for 3 Tx: 63 women with SUI, UUI, Tx; twice weekly group classes, PFM on 7-day BD and ratings on 5-point Computer generated months CG :64 or MUI and general strength exercise. scale (in BD); ICIQ randomization (70+ years) PFMT: 10 fast contractions (3 sec), 10 Tx group showed greater change Randomization procedure sustained contractions (8-10 sec); in urine leak score (p=0.007) and blinded lying, sitting, and standing positions. cure. Blinded assessors One-hour class X1/month during 7- Tx group 44.1% cure rate post-Tx month follow-up. Home exercise: 2-3 and 39.3% at 7-month follow up vs sets as learned in class, at least 1.6% and 1.6% in CG (p<0.001). X3/week for 30 min/day Leong Advice and N=56 Older Chinese women 12-week PT program, 30-min individual 7-day BD, IIQ-SF-7, 11-point VAS Computer generated 2015 (8) pamphlet on Tx: 27 with SUI, UUI, or MUI training sessions, 8 sessions taught for perception of improvement and randomization and management of UI CG: 28 (55+ years) with manual palpation and verbal satisfaction. allocation concealment feedback; progressive exercise Tx group showed greater Tx accompanied by urge program starting with 10 slow improvement in IEF (p<0.001) and suppression and bladder submaximal contractions (5 sec) IIQ-7 (p=0.001) and greater training increasing to 25 per session; urge perception of improvement suppression + bladder training (BT). (p=0.004) and satisfaction (p=0.001). Reductions significant in Tx group only. McLean No treatment N=40 Women with 12-week individualized PFMT program 3-day BD, pad test, IIQ-7, UDI-6. Randomization by 2013 (9) Tx: 20 predominant SUI with weekly 30-min sessions; PFMT Tx group showed greater changes automated computer CG: 20 (18+ years) taught by manual palpation; home on IEF (intragroup p=0.007) and algorithm. exercise: 3 sets of 12 PFM IIQ-7 (intragroup p=0.0003); Primary purpose of the contractions daily Neither group had significant study was to examine change on pad test or UDI-6 urethral morphology and mobility. Measures of UI were secondary outcomes.

25 Study Comparator N Study Population Intervention Outcomes/Results Notes Pereira No treatment N=49 Women with SUI only 6 weeks PFMT in group or individual Primary outcome: pad test; Randomization was by 2011 (10) Group: (18+ years) sessions; two 1-hour sessions weekly secondary King’s questionnaire participants drawing 17 (12 sessions); taught by vaginal and satisfaction. envelope from a box; Individ: palpation in both Tx groups; 3 s and 5- Both Tx groups showed greater assessments not blinded; 17 CG:15 10 s contractions; repetitions and improvement on pad test than CG not ITT duration progressed; average of 100 (p<0.0001). Improvements shown Purpose of the study was to contractions per session. in both Tx groups but not CG. examine group vs individual. No adverse events. Small N Rutledge Usual care N=40 Gyn cancer survivors 12 weeks; single 15-min training PGI-I and UDI-6 Randomization generated 2014 (11) (no handout) Tx: 20 with UI session; taught using levator ani PGII: 80% in Tx group vs 40% in by random number table. CG: 20 (37-79 years) palpation; 10 contractions 5 s duration; CG were “much better” or “very Allocation concealment handout on behavioral management much better” (p=0.02). Not ITT analysis tips: fluid intake, reduction of caffeine UDI-6: Groups not significantly Treatment included and bladder irritants; constipation different: Tx 70% vs CG 50% behavioral/lifestyle advice. management; Home exercise: 3 sets of reported lack of bother (p=0.62). exercises daily for 12 weeks. 1. Abdulaziz K, Hasan T. Role of pelvic floor muscle therapy in obese perimenopausal females with stress incontinence: A randomized control trial. The Internet Journal of Gynecology and Obstetrics. 2012;16(2):34-42. 2. Asklund I, editor Treatment of stress urinary incontinence via a smartphone application. Report from an ongoing randomised controlled study. Medicine 20 Confer- ence; 2014: JMIR Publications Inc., Toronto, Canada. 3. Tosun OC, Mutlu EK, Ergenoglu A, Yeniel A, Tosun G, Malkoc M, et al. Does pelvic floor muscle training abolish symptoms of urinary incontinence? A randomized controlled trial. Clinical rehabilitation. 2015;29(6):525-37. 4. Dumoulin C, Sran M, Lieblich P, Wilson P. Physiotherapy significantly reduces leakage in postmenopausal women with osteoporosis and urinary incontinence: result of a parallel randomised controlled trial. Neurology & Urodynamics. 2011;30(6):985. 5. Ferreira S, Ferreira M, Carvalhais A, Santos PC, Rocha P, Brochado G. Reeducation of pelvic floor muscles in volleyball athletes. Revista da Associação Médica Brasileira. 2014;60(5):428-33. 6. Jahromi MK, Talebizadeh M, Mirzaei M. The Effect of Pelvic Muscle Exercises on Urinary Incontinency and Self-Esteem of Elderly Females With Stress Urinary Incontinency, 2013. Global journal of health science. 2015;7(2):71. 7. Kim H, Yoshida H, Suzuki T. The effects of multidimensional exercise treatment on community-dwelling elderly Japanese women with stress, urge, and mixed urinary incontinence: a randomized controlled trial. International journal of nursing studies. 2011;48(10):1165-72. 8. Leong B, Mok N. Effectiveness of a new standardised Urinary Continence Physiotherapy Programme for community-dwelling older women in Hong Kong. Hong Kong Medical Journal 2015;21(1):30-7.

26 9. McLean L, Varette K, Gentilcore‐Saulnier E, Harvey MA, Baker K, Sauerbrei E. Pelvic floor muscle training in women with stress urinary incontinence causes hypertrophy of the urethral sphincters and reduces bladder neck mobility during coughing. Neurourology and urodynamics. 2013;32(8):1096-102. 10. Pereira VS, Correia GN, Driusso P. Individual and group pelvic floor muscle training versus no treatment in female stress urinary incontinence: a randomized controlled pilot study. European Journal of Obstetrics & Gynecology and Reproductive Biology. 2011;159(2):465-71. 11. Rutledge TL, Rogers R, Lee S-J, Muller CY. A pilot randomized control trial to evaluate pelvic floor muscle training for urinary incontinence among gynecologic cancer survivors. Gynecologic oncology. 2014;132(1):154-8.

Table 8 Studies comparing different PFMT programs included in the previous review (5th ICI) and current update (6th ICI)

Studies included in the previous New studies identified in this update Total review (5th ICI) (6th ICI) 1. Supervision of training: amount of contact with health professionals 6 4 10

2. Supervision of training: individual versus group supervision 6 2 8

3. Exercise program: direct versus indirect exercises 6 2 8

4. Exercise program: generic versus individualized exercises 1 0 1

5. Exercise program: submaximal versus near maximal contractions 1 0 1

6. Exercise program: daily versus three times per week 1 0 1

7. Exercise program: addition of upright exercise position 1 0 1

8. Exercise program: addition of strength training to motor learning 1 1 2

9. Exercise program: addition of abdominal or hip muscle exercises 1 3 4

10. Exercise program: addition of intravaginal resistance device 3 2 5

11. Exercise program: addition of adherence strategy 2 0 2

12. Exercise program: addition of biofeedback 9 6 15

27 Table 9 Summary of data on different PFMT programs comparisons

Author, year Comparator N Study population Modality details or parameters Outcomes/Results Follow-up Notes

Cruz et al. Supervised PFMT 79 Primigravidas with UI Supervised group – 5-6 biweekly Self-reported cure: 3 weeks Large dropout 2014 (1) - (43) at 21-26 weeks of sessions supervised by a Non- sign. difference btw groups in rate: Abstract vs pregnancy physiotherapist cure rate (p=0.052) 23/43 Unsupervised (36) Supervised 16/20 (74%) supervised Unsupervised group: same Unsupervised 10/21 (48%) PFMT treatment protocol ICIQ-SF : 15/36 Supervised group pre-Tx8.1±3.7 to unsupervised post-Tx1.2±2.5 PFMT Unsupervised PFMT group pre- Tx7.7±5.0 to post-Tx 4.7±5.6 Sign. difference btw treatments (p=0.016) Delgado et al. PFMT (24) 52 Women with SUI or PFMT: As below without device Reported cure (based on the Q11 of 16 weeks of Dropouts (at 6 2013 (2) vs MUI (stress PFMT +resistance; 5 quick and 5 the ICIQFluts) treatment, month) PFMT + resistance predominant) slow (sustained), high-intensity PFMT 0/13 outcomes PFMT 9/24 Abstract was device (28) contractions daily. Advised to hold PFMT+resistance 1/15 assessed at PFMT + included in the contractions as long as possible, Non-sign. difference btw groups post-Tx and at resistance ICI 5th edition relaxing their PFM for an (p=0.429) 6 month 15/28 equivalent time before repeating Improvement (post-Tx) follow-up the process. Intravaginal PFMT 10/19 resistance: instructions to use the PFMT+resistance 11/21 Pelvic-Toner Device concurrently Non-sign. difference btw groups whilst exercising. Two clinic visits and one phone call

28 Author, year Comparator N Study population Modality details or parameters Outcomes/Results Follow-up Notes

Donahoe- PFMT (5) 11 11 women with SUI PFMT: 10 repetitions of 10s PFM ICIQ-IU-SF 12 weeks of Dropouts: Filmore et al. vs (urodynamic contraction/5-10s rest, 10 Delta between treatments -5.0±4.7 treatment, PFMT (1/5) 2011 (3) PFMT + hip diagnosis) repetitions of 1-2s contraction 10s (p=0.35) outcomes PFMT + hip muscle training (6) contraction, PFM and transverse UDI-6 assessed muscle abdominis contraction Delta between treatments 10.25±3.5 post-Tx training (4/6) PFMT + hip muscle training: (p=0.043) in favor of combined PFMT training while performing treatment Results should resisted hip contraction with a ball Incontinence severity index be interpreted /elastic band, 10 repetitions of 5s Delta between treatments -5.25±2.48 with caution (PFM contraction+TrA+squeezing (p=0.139) (n=11) a ball) and 10 repetitions of 5s (PFM contraction+TrA+hip external rotation resisted with an elastic band) Both groups were instructed to perform the exercises twice a day, 7 days/week for 12 weeks. Both also had 3 sessions supervised with a physiotherapist Ferreira et al. Supervised PFMT 38 Women with mild to Both groups received a leaflet and Self-reported cure/improvement: 6 month of Dropout 2012(4) (20) moderate SUI attended a 1h educational session The rate of women who thought they treatment, Supervised (pilot RCT) vs (urodynamic including verification of adequate were cured, almost cured, or feeling outcomes PFMT 3/20 Unsupervised diagnosis) PFM contraction better was 100% in supervised PFMT assessed Unsupervised (home) PFMT (18) Supervised PFMT: and 64.7% in home PFMT (p=.018) post-Tx PFMT 1/18 Same as home program + weekly 7-day bladder diary: 45 min session Supervised PFMT pre-Tx 11.3±5.1 to Unsupervised (Home) PFMT: post-Tx 4.5±5.1 (p<.05) 8-10 PFM contraction, 10s Unsupervised PFMT pre-Tx 11.3±5.7 to contraction followed by four quick post-Tx 6.4±7.5 (p<0.05) contractions/5-10s rest, three Similar reductions on pad test were times a day, in different positions registered in both groups (p=0.125) for 6 months with monthly Pad test: motivational phone call Supervised PFMT pre-Tx 3.6±1.9 to post-Tx 1.90±1.5 (p<0.05) Unsupervised PFMT pre-Tx 3.8±3.0 to post-Tx 2.0±2.0 (p<0.05) Similar reductions on pad test were registered in both groups (p=0.530)

29 Author, year Comparator N Study population Modality details or parameters Outcomes/Results Follow-up Notes

Fitz et al. 2015 Supervised (25) 63 Women with Supervised: 24 sessions, twice Bladder diary: 3 months of 13 dropouts (5)- abstract vs predominance of SUI weekly. 3 sets of 30 contractions Post-Tx score in supervised 0.6±0.9 treatment, (on is not Unsupervised (25) and more than 2 g of followed by 3 fast-twitch and unsupervised group 0.9±1.4 outcome mentioned in leakage (pad test) contractions Non-sign. difference btw groups assessed at which group) (p=0.944) post-Tx Unsupervised: Same exercise Pad test protocol, appointment once a Supervised 3.8±6.6 and unsupervised month for exercise adjustment 16.0±28.2 Sign. difference btw groups (p=0.026) QoL Non sign. difference btw groups (0.661) Galea et al. PFMT (11) 23 Women with SUI or PFMT: correct PFM contraction King’s health questionnaire 10 weeks of PFMT 0/11 2013 (6) vs UUI was verified through vaginal Non-sign. difference btw groups treatment, PFMT+BF PFMT+BF (clinic) palpation at the first visit 7-day bladder diary outcomes 1/12 (12) (over 60 years old) PFMT pre-Tx 2(5) to post-Tx 2(3) assessed at 3 PFMT+BF: Ultrasound imaging PFMT+BF pre-Tx 8(8) to post-Tx 3(5) month post-Tx was used as a visual BF. The Non-sign. difference btw groups visual feedback was progressively (p=0.868) reduced throughout sessions Pad test PFMT pre-Tx 2.1(11.3) to post-Tx 0.2(4.2) PFMT+BF pre-Tx 6(54) to post-Tx 0(4.8) Non-sign. difference btw groups (p=0.428)

30 Author, year Comparator N Study population Modality details or parameters Outcomes/Results Follow-up Notes

Hirakawa et al. PFMT (23) 46 Women with SUI PFMT: 5 individually supervised ICIQ-UI-SF: 12 weeks of Dropouts: 2013 (7) vs (more than 1 sessions over 12 weeks of PFMT pre-Tx 12.0±3.5 to post-Tx treatment, PFMT 3/23 PFMT +BF (home episode per week) treatment. PFM contraction taught 8.3±3.5 (p<0.001) outcomes vs and clinic) (23) with palpation of the perineal PFMT+BF: pre-Tx 11.2±3.9 to post-Tx assessed PFMT +BF body.10 repetitions of 5s 7.8±3.3 (p=0.002) post-Tx 4/23 contraction/10s rest, 10 repetitions Non-sign. difference btw groups of 2s contraction/4s rest, Knack, King’s health questionnaire twice daily Non-sign. difference btw groups Bladder diary PFMT +EMG biofeedback: same PFMT pre-Tx 1.9±1.8 to post-Tx as above but with EMG BF at 1.2±1.4 (p=0.028) home and at the clinic PFMT+BF: pre-Tx 1.2±0.7 to post-Tx 0.8±1.2 (non-sign.) Non-sign. difference btw groups Pad test: PFMT pre-Tx 11.7±18.9 to post-Tx 7.7±15.4 (non-sign.) PFMT+BF: pre-Tx 21.3±38.2 to post-Tx 9.9±15.1 (non-sign.) Non-sign. difference btw groups

31 Author, year Comparator N Study population Modality details or parameters Outcomes/Results Follow-up Notes

Jordre et al. Resisted hip 30 30 SUI women (with Resisted hip rotation: Subjective improvement 6 weeks of Dropout 2014 (8) rotation training at least 2 SUI -10 hip external/internal rotation Hip training 69.6±21.4 post-Tx treatment, Resisted hip (14) episodes per month) during breathing PFMT 52.8±32.2 post-Tx outcomes rotation 2/14 vs -10 repetitions each of hip external Non-sign. difference btw groups assessed vs PFMT (16) rotation (resistance band) and (p=0.24) post-Tx PFMT1/16 internal rotation (with a ball) UDI No PFM contraction Hip training: pre-Tx 142.3±69.8 to post- PFMT: Tx 79.8±56.6 (p=0.02) 20 5s contraction and 20 quick PFMT: pre-Tx 145.2±54.0 to post-Tx flicks 77.7±60.4 (p<0.01) Vaginal palpation was performed Non-sign. difference btw groups only when required (p=0.90) IIQ Hip training: pre-Tx 151.3±77.2 to post- Both programs were performed Tx 122.5±68.2 (p=0.03) twice daily. PFMT: pre-Tx 140.7±39.4 to post-Tx Weekly recheck were scheduled 119.8±59.6 (p<0.01) (in person or over the phone) Non-sign. difference btw groups (p=0.70) Bladder diary Hip training: pre-Tx 5.2±6.2 to post-Tx 0.1±0.1 (p<0.01) PFMT: pre-Tx 3.3±3.5 to post-Tx 1.1±3.4 (p<0.01) Women received hip training had sign. more reduction of UI episodes (p=0.03) Jungin et al. PFMT with bladder 80 Women with SUI or PFMT with bladder neck effective Improvement: 3 months of Dropouts: 2014 (9) neck effective MUI relearning: PFM contraction and PFMT with bladder neck effective treatment, PFMT with relearning (42) Knack taught with ultrasound 10(28%) had some improvement and outcomes bladder neck vs biofeedback 15 (42%) reported great improvement assessed effective PFMT (38) 3 sessions supervised over 3 PFMT 12 (46%) and 8 (31%) had some post-Tx relearning months of treatment and great improvement, respectively 6/42 PFMT: PFMT with EMG Non-sign. difference btw the two PFMT 7/38 biofeedback - 3 supervised groups (p=0.365) sessions. There were 14 women in the PFMT PFM contractions (80% of the group who switch group after trial maximum, 8s contraction/10 s rest, completion for 10 min)

32 Author, year Comparator N Study population Modality details or parameters Outcomes/Results Follow-up Notes

Kashanian et PFMT (50) 91 Women with SUI or PFMT: 6-8s contraction/6s rest Number of UI (self-reported) 12 weeks of Dropouts: al. 2011 (10) vs MUI Twice daily for 15 min The data were categorized in ranges of treatment, PFMT 4/50 PFMT + resistance different frequencies. outcomes PFMT + device (41) PFMT + resistance device: twice Sign. changes in both groups assessed 3 resistance daily for 15 min (p<0.0001) month post-Tx device 2/41 Non-sign. difference btw groups UDI PFMT pre-Tx 45.1±15.5 to post-Tx 71.8±11.2 (p<0.0001) PFMT + resistance pre-Tx 39.9±13.5 to post-Tx 69.9±10.1 (p<0.0001) Non-sign. difference btw groups (p=0.418) IIQ PFMT pre-Tx 44.7±23.0 to post-Tx 67.7±20.20 (p<0.0001) PFMT+resistance pre-Tx 37.2±12.10 to post-Tx 62.52±12.13 (p<0.0001) Non-sign. difference btw groups (p=0.162) Kim et al. 2012 Supervised PFMT 20 Women with Supervised PFMT and trunk Bristol Female Lower Urinary Tract 8 weeks of Dropout: (11) + trunk postpartum UI stabilization: 23 1-h sessions, 3x/ Symptom questionnaire – treatment, Supervised stabilization (10) week, for eight weeks; PFM Urinary symptoms outcomes 1/10 vs contractions in different positions (changes) assessed Unsupervised Unsupervised Supervised –27.22±6.20 post-Tx 1/10 PFMT + trunk Unsupervised: same protocol but Unsupervised–18.22±5.49 stabilization (10) only one demonstration session Quality of life Supervised–5.33±2.96 Unsupervised –1.78±3.93) Total score Supervised –32.56±8.17 Unsupervised –20.00±6.67) All outcomes favored the supervised group (P<0.05)

33 Author, year Comparator N Study population Modality details or parameters Outcomes/Results Follow-up Notes

Konstantinidou PFMT (25) 46 Women with SUI or PFMT monotherapy KHQ 3 months of Dropouts are et al. 2013 vs MUI (stress PFMT: pre-Tx 342.9±28.2 to post-Tx treatment, not reported (12)- abstract PFMT + transverse predominant) Transverse abdominis exercises 198.3±18.4 (p<0.0001) outcomes abdominis combined with PFMT PFMT+TrA : pre-Tx 327.1±38.4 to post- assessed exercises (21) (minimum of 7 Tx 193.1±24.7 (p=0.0003) post-Tx Results were episodes per week Limited information was provided Non-sign. difference btw the two also presented and a 3-4 score at about the training protocol groups (p=0.76) in two other the Oxford scale) Bladder diary abstracts PFMT: pre-Tx 4.36±0.6 to post-Tx (13, 14) 1.8±0.3, (p<0.0001) It should be PFMT+TrA : pre-Tx 3.4±0.4 to post-Tx underlined that 1.2±0.2 (p<0.0001) lower sample Non-sign. difference btw groups size were reported in the 2011 abstract. Lamb et al. Group PFMT (111) 174 Women with SUI Group PFMT: Three 1h session for Symptoms severity index 3 weeks of Dropout: 2009 (15) vs and/or UUI 3 weeks, Group changes from baseline treatment, Group PFMT Individual PFMT average of 10 women per class, 2.34±0.44 outcomes 13/111 (63) individual assessment with PFM Individual changes from baseline assessed 6 Individual examination if required 1.71±0.57 months after PFMT 3/63 Non-sign. difference btw groups randomisation Individual PFMT: Same as group (p=0.38) intervention but individually Self-rated assessment of treatment supervised benefit Group changes from baseline 5.37±0.29 Individual changes from baseline - 6.34±0.38 Higher perceived benefit for individual treatment (p=.0462) Incontinence related QoL Group changes from baseline - 14.40±1.75 Individual changes from baseline - 14.81±2.50 Non-sign. difference btw groups (p=.89)

34 Author, year Comparator N Study population Modality details or parameters Outcomes/Results Follow-up Notes

Liebergall- Paula method 245 Women with SUI Paula : 12 individual 45-min Frequency of UI 12 weeks of Dropout Wischnitzer et (119) session (low or high) treatment, Paula 55/119 al. 2013 (16) vs No sign. deterioration of UI in 6 months follow-up at 6- PFMT 47/126 PFMT (126) PFMT : 6 group sessions over 12 in both groups month post-Tx (This is a weeks Paula method: 25 (39.7%) The two follow-up of the reported a low frequency rate of groups 2009 study) UI episodes differed PFMT: 18 (22.8%) in the PFMT group according to reported a low frequency the amount of Sign. difference btw groups (p=0.03) supervision received (PFMT had group supervision and less sessions). Manonai 2013 PFMT 61 Women with SUI PFMT: verbal instruction about Incontinence related QoL 16 weeks of Dropouts: one (17) -abstract vs PFM contraction, 15 min of PFM PFMT pre-Tx 51.1±15.9 to post-Tx treatment, in each groups PFMT +BF (home) exercises, three times a day for 16 70.6±15.5 (sign.) outcomes weeks PFMT+BF: pre-Tx 53.9±18.3 to post-Tx assessed (sample size in 72.6±10.81 (sign.) post-Tx each group not PFMT+BF: same as above but Non-sign. difference btw groups specified) with pressure BF

Marques et al. PFMT (15) 35 Women with SUI Both groups received 20 Bladder diary (5 days) 10 weeks of Dropout not 2014 (18)- vs (urodynamic individually supervised sessions, Non-sign difference btw the two groups treatment, reported abstract PFMT + hip diagnosis) twice a week (p=0.172) outcomes muscle training assessed (19) protocol (20) PFMT: post-Tx PFM contraction verified by vaginal palpation in the first 8 sessions

PFMT: same PFMT program with the addition of hip muscle strengthening (with progressive increases in loads)

35 Author, year Comparator N Study population Modality details or parameters Outcomes/Results Follow-up Notes

Pereira et al. Group PFMT (17) 49 Women with SUI Group PFMT: Two 1h weekly 1-h pad test: 6 weeks of Dropout: 2011 (20) vs sessions for 6 weeks. Group: pre-Tx 1.88±2.85 to post-Tx treatment, Group PFMT Individual PFMT Approximately 100 contractions 0.46±0.45 (p=0.05) outcomes 2/17 (17) (3s contraction/6s rest and 5-10 s Individual: pre-Tx 4.22±5.21 to post-Tx assessed Individual vs contraction/10-20 rest). 0.45±0.90 (p=.0006) post-Tx PFMT 2/17 Control (15) Contraction taught individually at Non-sign. difference btw groups Control 0/15 baseline King health questionnaire Only the personal relationships and No adverse Individual PFMT emotion differed btw the two groups effects Same as above but individually and favored individual PFMT (p.043) reported supervised

Control (15) No treatment Ong et al. 2015 PFMT (19) 40 Women with SUI PFMT: 4 sessions individually Subjective cure (based on the Q6 of 16 weeks of Dropouts: (21) vs supervised (20-min), 3-5 sets of 10 the Australian pelvic floor treatment, PFMT 3/19 PFMT+ BF (home repetitions (3-10s contraction/ 3- questionnaire) outcomes PFMT+ Results also and clinic) (21) 10s rest). 3-5 sets of 10 repetitions PFMT 10/16 assessed at 4 biofeedback presented in (2s contraction/2s rest) PFMT+BF 12/21 weeks of 0/21 the following Non-sign. difference btw groups treatment and abstracts (22, PFMT + biofeedback: (p=.742) post-Tx 23) Same as above but with use of the Vibrance device at home and in Australian pelvic floor questionnaire the clinic (total score) PFMT post-Tx 7.8±5.1 PFMT+BF post-Tx 11.3±8.3 Non-sign. difference btw groups (p=0.157)

36 Author, year Comparator N Study population Modality details or parameters Outcomes/Results Follow-up Notes

Prudencio PFMT (51) 156 SUI women The 3 groups received 20 sessions Cure of SUI 3 months Dropout not 2014 Vs of 45 min (twice a week for 3 Non-sign. difference btw groups specified (24) abstract PFMT+VC (55) months). (p=0.267) Vs PFMT included isolated rapid and PFMT 39.2% (20/51continent) PFMT+BF (clinic) sustained contractions of the PFM PFMT+BF 44% (22/50continent) (50) and at the end functional exercises KHQ: with contraction of the same The 3 groups significantly improved muscle group by differentiating just from baseline (p<0.001) the use or not of associated PFMT preTx74±18 to postTx43±17 instruments. (p<0.001) PFMT+BF preTx70±21 to postTx50±15 (No further information was Sign. Difference btw groups (p<0.001). provided on treatment in the It is not specified which groups differ abstract) Shin et al. PFMT (30) 60 Women with UI (type PFMT: 20 min, twice a day, ICIQ-SF: 6 months of Dropouts : 2012 (25) vs not specified) 5x/week for 6 months PFMT pre-Tx 8.8±4.7 to post-Tx treatment, PFMT 6/30 PFMT+BF (home) 8.2±5.1 (non-sign.) outcomes vs Results also (30) PFMT +BF: same as above but PFMT+BF: pre-Tx 9.0±3.9 to post-Tx assessed at 6 PFMT+BF presented in with the pressure device 7.0±2.7 (p<0.05) weeks and 6 (home) 0/30 the following KHQ months abstracts PFMT pre-Tx 6.84±15.9 to post-Tx (26, 27) 59.2±19.2 (non-sign.) PFMT+BF: pre-Tx 69.5±14.2 to post-Tx 48.7±12.3 (p<0.001) Bladder diary PFMT pre-Tx 9.7±3.3 to post-Tx 8.0±3.2 (non-sign.) PFMT+BF: pre-Tx 8.6±4.0 to post-Tx 5.1±4.6 (p<0.001) Comparison btw groups not reported. Sign. reduction of leakage frequency, KHQ and ICIQ occurred in BF group but not in PFMT group 1. Cruz C, Riesco ML, Zanetti M. Supervised pelvic floor muscle training to treat urinary incontinence during pregnancy: A randomized controlled trial. Neurourol Urodyn. 2014;33(6(Abstract#403)):867-8. 2. Delgado D, White P, Trochez R, Drake MJ. A pilot randomised controlled trial of the pelvic toner devicein female stress urinary incontinence. Int Urogynecol J 2013;24:1739-45.

37 3. Donahoe-Fillmore B, Chomy W, Brahler CJ, Ingley A, Kennedy J, Osterfeld V. A comparison of two pelvic floor muscle training programs in females with stress urinary incontinence: A pilot study. The Journal of Applied Research. 2011;11(2):73-83. 4. Ferreira M, Clara P, Duarte JA, Rodrigues R. Exercise programmes for women with stress urinary incontinence. Primary Health Care. 2012;22(3):24-7. 5. Fitz FF, Stupp L, Costa TF, Sartori MG, Girao MJ, Castro RA. Supervised versus non-supervised pelvic floor muscle training for stress urinary incontinence: Randomized controlled trial. Int Urogyn J. 2015;26(Suppl1):145-6. 6. Galea M, Tisseverasinghe S, Sherburn M. A randomised controlled trial of transabdominal ultrasound biofeedback for pelvic floor muscle training in older women with urinary incontinence. Australian and New Zealand Continence Journal. 2013;19(2):38-44. 7. Hirakawa T, Suzuki S, Kato K, Gotoh M, Yoshikawa Y. Randomized controlled trial of pelvic floor muscle training with or without biofeedback for urinary incontinence. International Urogynecology Journal. 2013;24(8):1347-54. 8. Jordre B, Schweinle W. Comparing resisted hip rotation with pelvic floor muscle training in women with stress urinary incontinence: A pilot study. J Womens Health Phys Therap. 2014;38(2):81-9. 9. Junginger B, Metz M, Baessler K. Comparison of a bladder neck effective pelvic floor rehabilitation program and EMG-biofeedback augmented pelvic floor muscle training: a randomized controlled trial Neurourology and Urodynamics 2014;33(6):970-1. 10. Kashanian M, Ali SS, Nazemi M, Bahasadri S. Evaluation of the effect of pelvic floor muscle training (PFMT or Kegel exercise) and assisted pelvic floor muscle training (APFMT) by a resistance device (Kegelmaster device) on the urinary incontinence in women "comparison between them : a randomized trial". European Journal of Obstetrics & Gynecology and Reproductive Biology. 2011;159(1):218-23. 11. Kim EY, Kim SY, Oh DW. Pelvic floor muscle exercises utilizing trunk stabilization for treating postpartum urinary incontinence: randomized controlled pilot trial of supervised versus unsupervised training. Clinical Rehabilitation. 2012;26(2):132-41. 12. Konstantinidou E, Kalaitzi M, Mytilekas K, Mikos T, Ioannides E, Hatzichristou D. Is there a role for training of the transversus abdominis muscles in the physiotherapy schemes applied in the treatment of female urinary incontinence? . Proceedings of the 43rd Annual Meeting of the International Continence Society (ICS). 2013. 13. Konstantinidou E, Kalaitzi M, Mytilekas K, Ioannides E-I, Hatzichristou D, Apostolidis A. Does the type of physiotherapy affect the quality of life and clinical outcomes in female urinary incontinence? A comparative study of two physiotherapy schemes European Urology Supplements 2013;12(1):e733. 14. Konstantinidou E, Mikos T, Kalaitzi M, Oeconomou A, Mytilekas K, Papameletiou V. Real-time ultrasonographic evaluation of the levator ani and transversus abdominis muscles: is there a role in female urinary incontinence? . Neurourology and Urodynamics 2011;30(6):1111-2. 15. Lamb SE, Pepper J, Lall R, Jorstad-Stein EC, Clark MD, Hill L, et al. Group treatments for sensitive health care problems: a randomised controlled trial of group versus individual physiotherapy sessions for female urinary incontinence. BMC Womens Health. 2009;9:26. 16. Liebergall-Wischnitzer M, Paltiel O, Lavy Y, Shveiky D, Manor O, Hochner-Celnikier D. Long-term efficacy of Paula method as compared with pelvic floor muscle training for stress urinary incontinence in women. J Wound Ostomy Continence Nurs. 2013;40(1):90-6. 17. Manonai J, Kamthaworn S, Petsarb K, Wattanayingcharoenchai R. Development of a pelvic floor muscle strength evaluation device Neurourology and Urodynamics. 2013;32(6):657-8.

38 18. Marques S, Haddad J, Passaro A, Silveira S, Baracat E, Ferreira E. Effectiveness of the strengthening of pelvic floor muscles, adductors of hip, gluteus maximus and gluteus medius in the treatment of stress urinary incontinence: blind randomized clinical trial - partial results. Proceedings of the 44th Annual Meeting of the International Continence Society (ICS). 2014. 19. Marques S. Comparison of the Efficacy of Two Kinesiotherapy Protocols for Stress Urinary Incontinence in Women: Randomized Blind Clinical Trial. Ref ID: 61587 Trials registry number(s): NCT01948713. 2013. 20. Pereira VS, Correia GN, Driusso P. Individual and group pelvic floor muscle training versus no treatment in female stress urinary incontinence: a randomized controlled pilot study. European Journal of Obstetrics & Gynecology and Reproductive Biology. 2011;159(2):465-71. 21. Ong TA, Khong SY, Ng KL, Ting JRS, Kamal N, Yeoh WS, et al. Using the Vibrance Kegel Device With Pelvic Floor Muscle Exercise for Stress Urinary Incontinence: A Randomized Controlled Pilot Study. Urology. 2015;86(3):487-91. 22. Jesse T, Teng A, Azad R, Keng L, Su Y, Prevathe P, et al. A randomized control trial to compare the effectiveness of pelvic floor exercises with the vibrance kegel device compared to standard kegel pelvic floor exercises for the treatment of stress urinary incontinence in females International Journal of Urology 2012;19 (Suppl A1):214. 23. Ng K, Ting J, Ong T, Khong S, Razack A. Randomised controlled trial comparing standard pelvic floor muscle exercises versus vibrance kegel device enhanced pelvic floor muscle exercises in women with urinary stress incontinence. BJU International. 2013;111(Suppl 1):107. 24. Prudencio C, Barbosa A, Derobio AL, Anezio A, Vesentini G, Almeida AP, editors. Comparison of three physiotherapy methods for treatment of stress urinary incontinence: impact in quality of life and muscle function. Proceedings of the 44th Annual Meeting of the International Continence Society (ICS), 2014 Oct 20-24, 2014; 2014; Rio de Janeiro, Brazil. 25. Shin J, Sul C, Na Y, Song K, Lim J, Yun C, et al. Effectiveness of perinometer biofeedback pelvic floor muscle exerciser with ExTT-101tm in female stress urinary incontinence. Proceedings of the 42nd Annual Meeting of the International Continence (ICS) (Beijing, China). 2012. 26. Shin J, Song K, Na Y, Lim J, Sul C, Hwang E, et al. Effectiveness of Biofeedback Training with EXTT-101TM in Female Stress Urinary Incontinence. Urology. 2011;78(3):S399. 27. Na Y, Lim J, Song K, Sul C, Shin J, Oh T, et al. Effectiveness of biofeedback training with ExTT-101 (trademark) in female stress urinary incontinence: Portable self education system. International Urogynecology Journal. 2011;22(3):S1819.

39 Quality of data Subgroup 1.2: additional phone calls Of the 20 RCTs, 12 provided sufficient detail support- No new studies investigated the effect of adding su- ing adequate randomisation and allocation conceal- pervisory phone calls to PFMT programmes. ment methods (34, 50, 65-74). Of these, seven RCTs stated that the outcome assessors were blinded (34, Subgroup 1.3: individual supervision versus no su- 65-70). Pereira et al. (50) reported adequate random- pervision isation and concealment but the evaluators were not Four new RCTs were included in this comparison (34, blinded. Of the remaining eight RCTs, seven did not 65, 71, 75). The PFMT programmes evaluated dif- report sufficient information on allocation conceal- fered with regard to the amount of health-professional ment methods and had unblinded assessors (75-81). contact but were similar in type and quantity of PFM None of the included trials was large and half (10/20) exercises. included only 20 to 51 participants per group (65, 68, Self-reported cure: Women in the supervised group 69, 72, 74, 75, 77, 79-81). Eight had fewer than 20 were more likely to report cure in comparison with participants per comparison group (34, 50, 70, 71, 73, women following unsupervised PFMT (71). Although 76, 78, 82). Six of the eight were reported as a pilot non-statistically significant, Cruz et al. (65) also re- study (70-73, 76, 78). Only two RCTs had a larger ported a higher cure rate in the supervised group. sample size involving a total of 174 (66) and 245 women (67). Improvement: Two RCTs reported improvement in UI symptoms as assessed with validated questionnaires In four RCTs, a dropout rate of more than 46% in one (34, 65). Both showed that women receiving super- study arm was reported. For instance, Dohanhoe et vised treatment reduced their UI symptoms signifi- al. (78) reported 60% dropout in the PFMT combined cantly more than women in the unsupervised arm (34, with hip training and 20% in the PFMT group. 65). Delgado et al. (72) had 37% and 57% in the PFMT group and PFMT combined with resistance, Leakage episodes: There was no statistically signifi- respectively. Liebergall-Wischnitzer et al. (67) cant difference in the number of leakage episodes reported a 46% and 37% dropout rate in the paula when comparing supervised and unsupervised PFMT and PFMT group, respectively while Cruz et al. (65) in women (71, 75). had 42-53% dropout in the supervised and Pad tests: Two of the four RCTs measured UI severity unsupervised PFMT group. The results of these with pad testing (71, 75). Fitz et al. (13) showed that RCTs should therefore be interpreted with extreme supervised treatment resulted in a more significant caution (65, 67, 72, 78). reduction of leakage than unsupervised PFMT while Ongoing RCTs Ferreira et al. (71) reported a non-significant difference. Based on a research on protocol registry database, there is currently one ongoing large non-inferiority  Comparison 2. Supervision of training: individual RCT comparing group versus individual supervised versus group PFMT (83) (comparison 2). Involving 364 elderly Two new RCTs were included in this comparison; women with SUI or MUI, this study is expected to be both studies used the same PFMT program and dif- completed by March 2017. Moreover, the study of fered only according to the type of supervision (group Buen et al. (84) is also ongoing and will compare the vs individual). It should also be underlined that Pe- efficacy of PFMT and Pilates in preventing and reira et al. (50) conducted an individual assessment treating UI (N=80) (comparison 3). The study of of the PFM to teach active PFM contraction to all Navarro et al. (85) will evaluate the efficacy of women prior to class training while Lamb al. (66) in- hypopressive exercises in 78 women with various cluded this assessment only when required. pelvic floor disorders including UI (comparison 3). Expected in July 2017, the study of Radlinger et al. Self-reported cure: No data were reported. (51) will assess the benefit of adding motor learning to PFMT in 96 women suffering from SUI or MUI Improvement: Non-significant differences in the (Comparison 8). Rao et al. (86) undertook a study symptom severity index were found between group comparing the addition of abdominal muscle training and individual PFMT in the study of Lamb et al. (66) to PFMT in 300 women (Comparison 9). Three However, both Lamb et al. (66) and Pereira et al. (50) upcoming RCTs will assess the benefit of adding favoured individual treatment with regard to self-rated clinic-based BF to PFMT (87-89). treatment benefits and the impact of UI (personal relationships and emotion domains), respectively. The Results study of Pereira et al. (50) involving 17 women per group, may be insufficiently powered to detect signif-  Comparison 1. Supervision of training: amount of icant difference between groups. In the study of Lamb contact with health professionals? et al. (66), a large sample was included (n=174). Subgroup 1.1: additional group supervision However, the information was limited in regard to No further studies investigated the effect of adding an additional supervised group exercise session.

40 whether this sample was sufficient to conclude treat- group 3.5, non-significant differences were found be- ment equivalence (as per sample size calculation in tween hip rotator training and PFMT with regard to non-inferiority design). subjective UI improvement and UI symptoms and related impact (76). Leakage episodes: No data were reported. Leakage episodes: Jordre et al. (76), subgroup 3.5, Pad test: Only the study of Pereira et al. (50) used the reported that women who received hip training had pad test and reported a non-significant difference be- significantly fewer UI episodes than women in the tween the two groups. PFMT group.  Comparison 3. Exercise programme: direct ver- Pad test: Neither of the two new RCTs used pad-test- sus indirect exercises ing measurements. This comparison encompassed six subgroups evalu-  Comparison 4. Exercise programme: generic ating direct versus indirect training. In “direct” PFMT, versus individualized exercises women specifically performed voluntary contractions of the pelvic floor muscles while in the “indirect” No further evidence was available. PFMT, women focused on other muscle groups in or-  Comparison 5. Exercise programme: submaxi- der to facilitate or stimulate pelvic floor muscle con- mal versus near maximal contractions tractions. No further evidence was available. Subgroup 3.1: PFMT versus sham/imitation  Comparison 6. Exercise programme: daily ver- No new studies were found comparing PFMT to sham sus three times per week. or imitation PFMT treatments (e.g. crossing the an- kles and pulling the legs apart). No further evidence was available. Subgroup 3.2: PFMT versus the 'Paula method'  Comparison 7. Exercise programme: addition of upright exercise position. The previous ICI chapter on conservative management included the study of Liebergall-Wischnitzer et No further evidence was available. al. (90) comparing PFMT to the Paula method. The 6- month follow-up of this study was included in the cur-  Comparison 8. Exercise programme: addition of rent review (67). It should be emphasized that the two strength training to motor learning. groups did not receive the same amount of supervi- Junginger et al. (68) evaluated PFMT and motor re- sion by a health professional because the Paula learning taught with ultrasound imaging in compari- group was individually supervised and the PFMT son to regular PFMT using EMG biofeedback in group had group teaching. women with SUI or MUI. Both groups had the same Subgroup 3.3: PFMT versus the 'Sapsford' approach amount of contact with health professionals. Our search revealed no new RCTs evaluating incon- Self-reported cure: No data were reported. tinence outcomes for this subgroup. Improvement: There was no difference between the Subgroup 3.4: PFMT versus Pilates two groups in the number who reported ‘some’ or ‘great’ improvement. However, women in the PFMT No new RCT were found evaluating Pilates treatment group were more likely to switch to the motor relearn- in comparison to PFMT. ing group at RCT completion. Subgroup 3.5: PFMT versus hip rotator training Leakage episodes: No data reported. One RCT was included in this subgroup evaluating Pad tests: The pad-testing measure was not included the effectiveness of hip external and internal rotator (68). training compared to PFMT (76).  Comparison 9. Exercise programme: addition of Subgroup 3.6: PFMT versus hypopressive training abdominal or hip muscle exercises. Our search revealed no RCT including hypopressive This comparison encompassed two subgroups eval- training conducted in women with UI. uating the addition of either abdominal muscle (subgroup 9.1) or hip muscle (subgroup 9.2) training to Self-reported cure: No data were reported in the dif- PFMT. ferent sub-groups. Subgroup 9.1: PFMT vs PFMT+ abdominal muscle Improvement: Liebergall-Wischnitzer et al. (67) in exercises subgroup 3.2 showed that UI results were maintained at the 6-month follow-up in both the Paula method One new RCT evaluates the effectiveness of adding and the PFMT groups. Women who received the transverse abdominis exercises to PFMT (77). Lim- Paula method intervention were more likely to report ited information was available about the training pro- a lower rate of UI than women after PFMT. In sub- tocol in this abstract report.

41 Subgroup 9.2: PFMT vs PFMT+ hip muscle exercises comparison groups were similar according to the amount of supervision and the intensity of training; Two RCTs investigated the addition of hip muscle ex- they differed only with regard to the addition of bio- ercises to PFMT in women with SUI (78, 82). The feedback. same amount of health-professional supervision was given to each group (78, 82). Two RCTs investigated clinic-based BF using pressure perineometry (79) and ultrasound imaging (70). Self-reported cure: No trial reported this outcome. Prudencio et al. (79) included women with SUI while Improvement: There was no significant benefit of Galea et al. (70) also evaluated women with UUI. Du- adding abdominal muscle exercises to PFMT in the ration of treatment varied from 10 weeks (70) to 3 study of Konstantinidou et al. (77) (subgroup 9.1). months (79). Likewise, Donahoe-Fillmore et al. (78) showed a non- Four RCTs evaluated home-based biofeedback using significant difference between adding hip muscle ex- a Vibrance device (73), EMG BF (74) and pressure ercises to PFMT training as assessed with the King’s BF (80, 81). Of these, three were conducted in health question and the Incontinence Severity Index women with SUI (73, 74, 81) and one did not specify (sub-group 9.2). Results from the UDI-6 question- the type of UI of women included (80). In all RCTs, naire, favoured the combined treatment (78). These the same PFMT program was used in the comparison findings should be interpreted with caution given the groups and differed only with regards to the addition small sample size of Donahoe-Fillmore et al.’s study of BF. Treatment duration varied from 12 weeks (74), (78) (n=11). 16 weeks (73, 81) and 6 months (80). Leakage episodes: Non-significant differences in Self-reported cure: There was no statistically signifi- leakage episodes were found when adding ab- cant difference in terms of the number of women who dominal muscle (77) or hip muscle exercises (82). indicated they were ‘cured’ between PFMT alone or Pad and paper towel tests: No data were reported. combined with home-based (73) or clinic-based BF (79).  Comparison 10. Exercise programme: addition of intravaginal resistance device Improvement: Galea et al. (70) reported a non-significant benefit of adding clinic-based BF as assessed The study of Delgado et al. (2009) (91), presented as with the King’s health questionnaire. Using the same an abstract, was included in the previous ICI edition. questionnaire, Prudencio et al. (79) conversely re- The complete published manuscript has now been in- ported a significant difference between groups but did cluded in the current review (72) along with a new not reveal which of the three treatment arms differed. study evaluating the addition of an intravaginal re- With regard to home-based BF, three of the four sistance device to PFMT in women with SUI or MUI RCTs reported a non-significant difference between (69). Both RCTs were the same in all aspects except PFMT combined or not with BF. Shin et al. (44) did that one group used an intravaginal device designed not conduct any inter-group comparisons but re- to increase resistance to the PFM contraction. In both ported a significant reduction of incontinence only in RCTs, the resistance device was composed of a the group which received BF. spring-loaded device with two limbs (69, 72). Leakage episodes: Galea et al. (70) found a non-sig- Self-reported cure: There was no statistically signifi- nificant difference in leakage episodes in women re- cant difference between the groups in terms of the ceiving or not clinic-based BF. Likewise, non-signifi- number of women who indicated they were ‘cured’ cant benefits were found when adding home-based (72). BF to PFMT (74). Improvement: There was no statistically significant Pad and paper towel tests: Galea et al. (70) reported difference between the groups in terms of self-re- a non-significant benefit of adding clinic-based BF to ported improvement and reduction of symptoms as PFMT as assessed with pad-testing measurements. assessed with standardised questionnaires (69, 72). Also, Hirakawa et al. (74) found a non-significant difference between women receiving or not home- Leakage episodes: No data were reported for this based BF. outcome.

Pad and paper towel tests: No data reported.  Comparison 11. Exercise programme: addition of adherence strategy No new RCTs investigated the efficacy of adding adherence strategies to PFMT.  Comparison 12 Teaching programme: Addition of biofeedback (BF) Six new RCTs compared PFMT alone to PFMT assisted with BF (70, 73, 74, 79-81). In these RCTs, the

42 small to moderate differences. Finding the best ap- Summary proach to PFMT remains among the highest research Based on current evidence, PFMT with regular (e.g. priorities. Future studies should be sufficiently pow- weekly) supervision is better than PFMT with little or ered to detect clinically important differences. no supervision (Level of Evidence: 1). However, data were unclear as to whether supervision was 2.3.3 Is PFMT Better than Other Treatments? more effective in individual or group settings. Suffi- Trials were considered for inclusion in this section if ciently powered studies using appropriate design they compared PFMT with another stand-alone inter- should be undertaken to investigate this comparison. vention, e.g. vaginal cones, bladder training, drug Based on limited evidence (6 previous RCTs and 2 therapy. The 2013 ICI review concluded that PFMT is new RCTs), 'indirect' methods of PFMT (e.g. the better than EStim, BT, or vaginal cones for women 'Paula method' or 'Sapsford' approach) are not better with SUI and better than duloxetine because of its than direct PFMT, However, some data were con- side-effects. PFMT and surgery were both effective, founded by differences in the amount of contact time but PFMT was recommended as first-line therapy be- with health professionals or the small sample size, cause it is less invasive. For women with UUI or MUI, which made it difficult to detect clinically significant PFMT and BT were both deemed effective, with some differences (Level of Evidence 2: unchanged). evidence to suggest an advantage for PFMT. Evi- dence was sparse for a comparison between PFMT No robust recommendation can be made with regard and drug therapies to the type or specification of training (i.e. generic versus individualized exercises, submaximal versus Seven new trials were found that compared PFMT to near maximal contractions, daily versus three times another stand-alone treatment. per week, addition of upright exercise position). PFMT was compared to vaginal cones (92-94), sur- There remains insufficient evidence as to whether the gery (95), continence pessary, (96, 97), drug therapy combination of PFMT with other treatment modalities (98), and bladder training (98). One trial was a head- (i.e. motor learning, abdominal- or hip-muscle train- to-head comparison of PFMT to surgery (95). The ing, intra-vaginal resistance device) could increase its other trials had more than two arms. They included efficacy. PFMT alone and another treatment alone, but were also designed to evaluate combinations of treatments With regard to clinic-based BF, new evidence (2 stud- (See Tables 10 and 13). ies) reported no statistically significant differences between BF-assisted and non-BF groups for self-re- Trials addressed the following comparisons: ported cure, improvement, or frequency of leakage i) PFMT versus vaginal cones (VC): episodes (Level of Evidence: 1). One recent Cochrane review compared PFMT to VC Likewise, there is no statistically significant differ- (92). This review considered the eleven studies in- ences between home BF and non-BF groups for self- cluded in this consultation along with two other RCTs reported cure, improvement, frequency of leakage (93, 94). Moreover, our literature search revealed the episodes and pad-test measures in women with SUI study of Golmakani et al. (60). Details of these new (Level of Evidence: 2). trials are presented in Table 13. Recommendations ii) PFMT versus EStim: Clinicians should offer and provide the most intensive No new trials have been published. health professional-led PFMT programme possible within service constraints. (Grade of Recommenda- iii) PFMT versus Bladder training (BT): tion: A). The Kafri trial examined the effects of PFMT alone Although studies are limited, there does not appear to and BT alone in the context of its 4-arm design (98). be clear benefit for adding other modalities (i.e. motor All groups had significant improvement at 3 and 12 learning, abdominal- or hip-muscle training, intra-vag- months on all parameters, including UUI episodes. inal resistance device) to PFMT (Grade of Recom- But, no significant time X group interactions were mendation: B) found that would indicate a difference between PFMT and BT (Table 10). There is no clear benefit from adding clinic- (Grade of Recommendation: A) or home-based BF (Grade iv) PFMT versus drug therapy: of Recommendation: B) to a PFMT program. The trial by Kafri and colleagues examined the effects of PFMT alone compared to anticholinergic drug ther- Implications for research apy alone in the context of a 4-arm design (98). All Comparisons of PFMT approaches are, de facto, groups had significant improvement at 3 and 12 comparisons of two active treatments. It is therefore months on all parameters, including UUI episodes difficult to determine which approach is best unless but, no significant time X group interactions were (a) the differences in outcome are large or (b) the found that would indicate a difference between PFMT RCTs are designed with sufficient sample size to find and drug therapy (Table 10).

43 v) PFMT versus surgery: vi) PFMT vs continence pessary: The single trial of surgery compared PFMT to mid- Randomization method and allocation concealment urethral sling in women with stress predominant UI were reported. Analyses were performed by ITT. (95). It was a multi-site trial involving 23 medical centres and 83 physiotherapists (Table 10). Results and Summary vi) PFMT vs continence pessary: i) PFMT versus VCs: A single new trial compared PFMT to continence pes- As underlined in Herbison et al.’s meta-analysis (92) sary in women with stress predominant UI (96, 97). It comprising 13 RCTs, most studies had a small sam- was a multi-site trial conducted by the Pelvic Floor ple size and differed according to the PFMT regimen Disorders Network of the National Institute on Child (92). There was also limited overlap between out- Health and Human Development (NICHD) (Table 10). comes. No statistically significant differences between PFMT and VC were found in subjective im- Quality of data provement or cure (reported in six RCTs) (Risk ratio i) PFMT versus VC: (RR) for failure 0.97, 95% CI 0.75 to 1.24) (92). There were no significant differences in subjective cure (re- Randomization and adequate allocation concealment ported in five RCTs) (RR for failure 1.01, 95% CI 0.91 were reported in the three new RCTs (93, 94, 99). Of to 1.13) (92). Pooled data from these RCTs showed these, two indicated that outcome assessors were substantial heterogeneity. Likewise, inconsistency blinded (93, 99). In the third study, the evaluator was between RCTs was noted in relation to leakage epi- unblinded (94). Sample sizes were small ranging sodes whereas a non-significant difference between from 15 to 30 women with SUI per group (93, 94, 99). treatments was observed (mean difference 0.00, 95% Dropout rates were higher in the VC groups in two CI -0.20 to 0.20) (92). With regard to improvement in studies (93, 99); no loss to follow up was reported in pad test, non-significant differences were found in the the study of Pereira et al. (94). Substantial attrition meta-analysis based on six RCTs (RR for failure 1.00, occurred in the study of Harvey et al. (93) with drop- 95% CI 0.76 to 1.31). However, the new study of outs reaching 41% in the PFMT group and 72% in the Golmakani et al. (60) favoured PFMT when using the VC group. Follow-up beyond the post-treatment eval- 1-h pad test and the leakage index. With regard to uation was reported by Pereira et al. (94) in a subse- quality of life improvement, the three new RCTs found quent publication with a 12-month post-treatment as- a non-significant difference between PFMT and VC sessment (100). treatments (93, 94, 99). Three RCTs included in the ii) PFMT versus EStim: Not applicable. Herbison review as well as the study of Golmakani et al. (60) reported the inability for some women to use iii) PFMT versus BT: VCs and adverse effects such as pain, vaginitis, bleeding and a sense of unpleasantness or inconven- Risk of Bias - This trial was not a head-to-head com- ience. parison of PFMT and drug therapy, but rather a comparison of four treatment approaches including BT In the 14 RCTs (13 from the new Cochrane review, and combined therapy. Randomization method and 11 of which were discussed in the previous ICI chap- allocation concealment were described. Analysis was ter edition and 1 additional RCT) that compared conducted by ITT. It was not clear whether assessors PFMT with VC in women with SUI, limited evidence were blinded. suggests that VC appear to have similar effects or are not superior to PFMT (Level of Evidence: 1). There iv) PFMT versus drug therapy: were no statistically significant differences between This trial was not a head-to-head comparison of interventions in subjective improvement or cure (re- PFMT and drug therapy, but rather a comparison of ported in six RCTs), subjective cure (reported in five four treatment approaches including BT and com- RCTs), or leakage episodes per day (reported in four bined therapy. Randomization method and allocation RCTs), and no improvement in the pad test (reported concealment were described. Analysis was con- in six RCTs). The additional RCT favoured PFMT us- ducted by ITT. It was not clear whether assessors ing both the 1-h pad test and the leakage index. VC were blinded. treatment may be inappropriate in some cases due to the inability to use and potential side effects as re- v) PFMT versus surgery: ported in four RCTs (i.e. pain, vaginitis, bleeding and a sense of unpleasantness or inconvenience). Randomization method and allocation concealment were reported. Analysis included ITT and other ap- ii) PFMT versus EStim: proaches. It was not clear whether assessors were blinded. Fifteen women in the surgery group and 28 No new trials were found. Previous pooled data in the PFMT group did not start treatment after ran- demonstrated that self-reported cure and cure/im- domization. The ITT analyses did not include these provement were more likely in PFMT than in EStim individuals. Nineteen women in the surgery group groups (Level of Evidence: 1). and 28 in the PFMT group were lost to follow-up after starting treatment.

44 iii) PFMT versus BT:  Surgery is more effective than PFMT, but potential benefit should be weighed against potential There is evidence for an advantage of PFMT over BT adverse events. PFMT should be offered as first for women with SUI (Level of Evidence: 2). line therapy due to its being less invasive. There does not appear to be a significant difference (Grade of Recommendation: B New). between PFMT and BT in UUI and MUI women  PFMT and continence pessary are both effective (Level of Evidence: 2). in first line conservative therapy (Grade of Rec- iv) PFMT versus drug therapy: ommendation: B New). There is weak evidence that PFMT is more beneficial For women with SUI or MUI: than drug therapy, but not enough evidence to  VC do not appear to be better than PFMT in the change previous recommendations (Level of Evi- dence: 2). treatment of UI. PFMT should be recommended as first-line conservative therapy (Grade of Rec- v) PFMT versus surgery: ommendation: B). On the primary outcome, PGI-I at 12 months, per- VC with supervised training sessions by a trained ceived improvement was significantly greater in the health professional can be offered to women who can surgery group compared to the PFMT group: 90.8% and are prepared to use them (Grade of Recommen- of women in the surgery group reported being “much dation: B). better” or “very much better” compared to 64.6% in the PFMT group. Change in the UDI UI domain and VC may be inappropriate for some women due to inability to insert or retain the cone or because of side OAB domain were also significantly greater with surgery and a higher proportion of women in the surgery effects and discomfort. group had subjective and objective cure. For women with UUI or MUI: The one new trial of PFMT compared to surgery ap-  PFMT and BT are effective first-line conservative peared well-designed and adequately powered, therapy (Grade of Recommendation: B). providing evidence that mid-urethral sling may be more effective than PFMT for treatment of SUI in  PFMT is better than oxybutynin as first line ther- women (Level of Evidence: 2). apy (Grade of Recommendation: B). vi) PFMT vs continence pessary: For women with UUI The PFMT/behavioural group had better outcomes  PFMT and BT are effective first line conservative on the PFDI SUI subscale and greater patient satis- therapy (Grade of Recommendation: B). faction. The groups were not significantly different on the PGI-I, UI episode frequency, or outcome Research recommendation measures, including domains of the PFDI and PFIQ. Larger, good quality trials are needed to address There may be some benefit for PFMT over conti- each of the above comparisons if these are of interest nence pessary alone, the evidence is not strong to women. In planning comparisons researchers enough to recommend one treatment over the other should consider carefully the potential impact of dif- (Level of Evidence: 2; Grade of Recommendation: ferent levels of supervisory intensity between groups, D). particularly in comparisons of conservative therapies. Recommendations For women with SUI:  PFMT and VC are both effective as conservative therapy, although PFMT is better because inability of use and side effects are experienced with VC in some women (Grade of Recommenda- tion: B).  PFMT is better than EStim as first line conservative therapy (Grade of Recommendation: B).  PFMT is better than BT as first line conservative therapy (Grade of Recommendation: B).  PFMT and drug therapy are both effective as first line therapy, although PFMT is better because of side effects experienced with drug therapy (Grade of Recommendation: B).

45 Table 10 Summary of data on PFMT vs other treatments

Author, Study Comparator N Intervention Outcomes/Results Notes year Population

Kafri Drug therapy N=164 Women with PFMT based on National Number of voids and UUI episodes, Comparison was in the context of 2013 (1) (anticholinergic) DT: 42 UUI, No SUI Institute for Health and Clinical QOL-ruI, Urogyn VAS. 4-arm trial, including arm for BT BT: 41 (45-75 years) Excellence (NICE) All groups had significant improvement at combined Tx. PFMT: 40 recommendations. 3 and 12 months on all parameters Randomization method and Conbined: 41 Included behavioral guidance (p<0.001). allocation concealment described. 3 sets of 8-12 slow maximal No significant timeXgroup interactions ITT analysis contractions sustained for 6-8 (except favoring combination group) Not clear whether assessors were s in different positons. blinded Daily home-based exercise prescribed. Richter Intravaginal N=446 Women with 12 weeks of supervised PFMT Primary: PGI-I and SUI subscale of the Comparison was in context of 3- 2010 (2) pessary PFMT: 146 predominant and behavioral strategies PFDI at 3 months (FU 6 & 12 months) arm trial, including arm for Kenton Pessary: 149 SUI (PFM pre-contraction for SUI; Secondary: Satisfaction, PFDI, PFIQ combined Tx 2012 (3) Combined: (18+ years) urge suppression for PFMT/behavioral group had better Multi-site trial 151 concomitant UUI). outcomes on the PFDI SUI subscale Randomization method and Implemented in 4 visits at 2- (49% vs 33% reporting no bothersome allocation concealment reported week intervals. Home exercise: SUI symptoms, p=0.006) and greater ITT analysis written prescriptions X3 daily satisfaction (75% vs 63%, p=0.02). exercise between visits. Groups not significantly different on the PGI-I (49% vs 40% “much better” or “very much better,” (p=0.09), UI episodes on bladder diary, or other outcome measures including other domains of the PFDI and PFIQ.

46 Author, Study Comparator N Intervention Outcomes/Results Notes year Population

Labrie Mid-urethral sling N=460 Women with Supervised, somewhat Assessed at 12 months Multi-site trial (23) 2013 (4) PFMT: 230 predominant individualized program Primary: PGI-I, 90.8% in surgery group Randomization computerized on Surgery: 230 SUI conducted by 83 PTs were “much better” or “very much better” central server (35-80 years) according to Dutch guidelines. vs 64.4% in PFMT. Allocation not concealed Treatments given at 1-week or Secondary: UDI, IIQ, PGI-S Not clear whether assessors were 2-week intervals with an Both groups had significant improvement blinded. intended 9 sessions in 9-18 in UDI and IIQ domain scores. Change in 65 adverse events reported - all in weeks. Goal to build up to 8 to UDI greater for surgery than PFMT on UI surgery group. 12 maximal contractions X3 and OAB domains (P<0.001, .02) Analysis included modified ITT and per day. Use of touch, tapping, Subjective cure: (single question) 85.2% other approaches. massage, BF, or functional in surgery vs 53.4% in PFMT 15 in surgery group and 28 in electrical stimulation allowed to Objective cure: (cough test) 76.5% in PFMT group did not start treatment increase muscle awareness as surgery vs 58.8% in PFMT after randomization. needed. After 12 months, 99 women (49.0%) had 19 in surgery group and 28 in crossed over to surgery after a mean PFMT group were lost to follow-up 31.7 weeks. after starting Tx.

1. Kafri R, Deutscher D, Shames J, Golombp J, Melzer I. Randomized trial of a comparison of rehabilitation or drug therapy for urgency urinary incontinence: 1-year follow-up. Int Urogynecol J. 2013;24(7):1181-9. 2. Richter HE, Burgio KL, Brubaker L, Nygaard IE, Ye W, Weidner A, et al. Continence pessary compared with behavioral therapy or combined therapy for stress incontinence: a randomized controlled trial. Obstet Gynecol. 2010;115(3):609-17. 3. Kenton K, Barber M, Wang L, Hsu Y, Rahn D, Whitcomb E, et al. Pelvic floor symptoms improve similarly after pessary and behavioral treatment for stress incontinence. Female Pelvic Med Reconstr Surg. 2012;18(2):118-21. 4. Labrie J, Berghmans BL, Fischer K, Milani AL, van der Wijk I, Smalbraak DJ, et al. Surgery versus physiotherapy for stress urinary incontinence. N Engl J Med. 2013;369(12):1124-33.

47 2.3.4 Does the Addition of PFMT to Other There is some evidence that there may be benefit in Treatments Add Benefit? adding PFMT to BT (Level of Evidence: 2). Six trials addressed the following comparisons (Table There is no evidence of improved outcomes with peri- 11): operative PFMT for women undergoing surgery for apical prolapse and SUI, or for adding PFMT to EStim i) PFMT+VC vs VC: Both the Herbison et al. meta- in women with SUI (Level of Evidence: 2). analysis (92) and the ICI 5th edition (2) included two studies comparing combined PFMT∕ VC versus VC. Recommendation No new RCTs with the same comparators were iden- In women using VC, it does not appear to help to add tified. PFMT (Grade of Recommendation: C). ii) PFMT + EStim vs EStim: A single trial examined For the treatment of SUI, UUI, or MUI in women, con- the effects of adding PFMT to vaginal EStim in 48 sider a combination of PFMT and BT rather then BT women with SUI (101). No significant differences alone (Grade of Recommendation: C). were found between the groups. For treatment of UUI (tolterodine) but not for SUI (du- iii) PFMT+ BT vs BT: Two trials examined the effects loxetine), consider adding PFMT to drug therapy of adding PFMT to BT, one in 108 women with SUI, (Grade of Recommendation: B), UUI, or MUI (102), the other in 164 women with UUI (no SUI) (98). In the first, significantly more patients When treating women with SUI and vaginal atrophy, in the combined therapy group reported cure or im- consider combining PFMT and intravaginal oestrogen provement and greater improvements on several sec- over estrogen alone (Grade of Recommendation: C ondary outcomes. In the second trial, all groups had New). significant improvement, but there were no significant interactions indicating differential group effects. Note, many of these recommendations are based on single RCTs of variable quality. Larger, good-quality iv) PFMT + drug therapy vs drug therapy alone: A sin- RCTs are needed to address each of the above com- gle trial examined the effects of adding PFMT to in- parisons. Furthermore, these studies examine the ef- travaginal oestriol in 206 postmenopausal women fects of combining therapies as an initial approach. with SUI and vaginal atrophy (103). Combined ther- Less is known about the effects of combining thera- apy resulted in significantly greater improvement than pies in a stepped fashion when women do not oestriol only. achieve the desired outcomes with a single therapy (i.e. active PFMT than resisted PFMT with cones v) PFMT + continence pessary vs continence pes- such as in progressive muscle training overload). sary: A single trial examined combined pessary + PFMT to pessary alone in 446 women with stress predominant UI (97). The combined therapy group had significantly better outcomes on the PGI-I and the PFDI and reported greater satisfaction with treatment. vi) PFMT + surgery vs surgery: A single, multi-centre trial examined the effects of perioperative behavioural and pelvic floor muscle training (BPMT) in 374 women undergoing surgery to treat both apical prolapse and SUI (104). The group receiving BPMT resulted in no greater improvements in urinary symptoms compared to usual peri-operative care at 6 or 24-month follow up.

Summary The literature on the effect of adding PFMT to another stand-alone therapy remains relatively small. There is no evidence of benefit in adding PFMT to VC in women with SUI (Level of Evidence: 2). There may be some benefit for adding PFMT when using a continence pessary (Level of Evidence: 2). PFMT may add benefit in terms of reducing SUI over intravaginal estrogen alone when treating women with vaginal atrophy (Level of Evidence: 2).

48 Table 11 Summary of data on PFMT + another treatment vs the other treatment

Study Comparator N Study Population Intervention Outcomes/Results Notes Richter 2010 Continence N=446 Women with stress Pessary + behavioral Primary: PGI-I and SUI subscale of Multi-site trial conducted by the (1) pessary alone predominant UI training UDI at 3 months NICHD Pelvic Floor Disorders Behavioral training program Results - Combined therapy group Network with randomization to - supervised PFMT + had significantly better outcomes on 3 arms: pessary alone, behavioral strategies: PFMT the PGI-I (53.3% vs 39.6% reporting behavioral training alone, pre-contraction for SUI and “much better” or “very much better,” combined Tx urge suppression for p=.02) and the PFDI (44.0% vs Data quality – Randomization concomitant UUI. Treatment 32.9% reporting no bothersome SUI method and allocation implemented in 4 visits at 2- symptoms (p=.05). Combined group concealment were reported. week intervals and included also reported greater satisfaction Analysis conducted by ITT written prescriptions for with treatment (78.7% vs 63.1%, home exercise X3 daily p=.003). The groups were not between visits. significantly different on number of UI episodes on bladder diary. Group differences not sustained to the 12- month follow up. Barber 2014 (2) Surgery alone N=374 Women undergoing Surgery + perioperative Primary: UDI at 6 months Multi-center 2X2 factorial trial for prolapse surgery to treat both behavioral and pelvic floor Results – Perioperative BPMT not with 2 distinct randomizations: and SUI apical vaginal prolapse muscle training (BPMT) associated with greater First: to perioperative BPMT or and SUI PFMT program - Single improvements in urinary symptoms usual care session 2-4 weeks prior to at 6 months or 24-month follow-up Second: to one of 2 surgical surgery; 4 postoperative approaches for prolapse, each visits (2, 4-6, 8, & 12 weeks) with concomitant retropubic Individualized, supervised midurethral sling for SUI. program of progressive Data quality – Randomization home PFME, education on method and allocation healthy bladder and bowel concealment reported. habits, and behavioral strategies to reduce UI episodes (PFM pre- contraction for SUI, urge suppression for UUI) Home PFME prescribed for X3 daily with duration increasing to 10 sec. Capobianco 2012 (3) Intravaginal N=206 Postmenopausal Estrogen + PFMT with Outcome based on change in Data quality - Randomization estrogen alone women with SUI and EStim (per Castro, 25) patient rating of SUI as none, mild, method not described.

49 Study Comparator N Study Population Intervention Outcomes/Results Notes (estriol ovules, vaginal atrophy, no PFMT was for 6 months and moderate, or severe before and after Group allocation was 1m daily for 2 DO included EStim. Further 6 months of treatment. concealed. wks, then 2.wk details lacking, but per Results - Combined therapy resulted Analysis was by ITT for 6 months) reference to a previous in significantly greater improvement AEs were reported article, PFMT was than estriol only. 73.5% (61/83) of Not clear all women actually conducted in 45-minute treated and 9.7% (10/103) showed had SUI. group sessions X3 per week improvement in UI (p<.01) under the supervision of a physiotherapist. Kafri 2013 (4) BT alone N=164 Women with UUI, no Bladder training + PF rehab Number of voids and UUI episodes, Comparison was in context of SUI and behavioral. voids/24 hours, QoL-rUI, Urogyn 4-arm trial: BT alone, PFMT (ages 45-75) PFMT program based on the VAS, number of pads alone, drug therapy alone, and National Institute for Health Results - analysis was by repeated combined Tx. and Clinical Excellence measures ANOVA with all four Data quality –Randomization (NICE) recommendations. groups, including tests for main method and allocation Women practiced 3 sets of effects and group X time concealment described. 8-12 slow maximal interactions. However, statistical Analysis conducted by ITT. contractions sustained for 6- analysis did not include comparisons Not clear whether assessors 8 s in different positons. between individual groups. were blinded. Daily home-based exercise All groups had significant was prescribed, and urge improvement at 3 and 12 months on suppression using PFM all parameters, (p<.001) but there contraction was taught. were no significant interactions. UI episodes decreased by 1.7/day in BT and 4.0/day in combined therapy, but a test of this difference was not conducted. Only CPFR showed significant reduction in voids/24 hours. Kaya 2015 BT alone N=108 Women with SUI, UUI, BT + PFMT for 6 weeks Primary: global rating of Data quality –Randomization (5) MUI High-intensity PFMT improvement on a 4-point scale method and allocation program conducted in 4 (worse, unchanged, improved, concealment described. visits across 6 weeks by an cured). Secondary: UI severity, Analysis conducted by ITT. experienced physical symptom distress, QOL, UI Not clear whether assessors therapist. Home-based episodes, voids/day. were blinded. exercise regimen initiated Results – Significantly more patients with 5 sets of exercises per in the combined therapy group day and progressed to 30 reported cure or improvement (100% vs 82.7%, p=.001). Greater

50 Study Comparator N Study Population Intervention Outcomes/Results Notes sets per day (600 PFM improvements observed for contractions). combined therapy on secondary outcomes, including severity of UI (ISI; p=.001), UDI-6 (p=.001), IIQ-7 (p=.005), and incontinent episode frequency per bladder diary (p=.024). Furst 2014 EStim alone N=48 Women with SUI EStim + PFMT for 3 months Improvements in urinary symptoms, Data quality – Randomization (6) Vaginal probe; (Mean age 49.6) PFMT program – individually IEF per bladder diary, satisfaction method described; allocation two 30-min designed by PT, 2 30-min (perception of need or not to repeat concealed. sessions per sessions per week. or change Tx); assessments at 3, 12 Analysis not conducted by ITT week; PFMT and EStim conducted and 96 months. (only included patients who had frequencies of on alternate days. Results – Both groups showed not received any additional 4Hz and 50Hz; No home PFME program improvement in leakage episodes therapy) fixed intensity and voiding intervals. (20mA); 4 s No significant between group stimulation/rest differences. cycles 1. Richter HE, Burgio KL, Brubaker L, Nygaard IE, Ye W, Weidner A, et al. Continence pessary compared with behavioral therapy or combined therapy for stress incontinence: a randomized controlled trial. Obstet Gynecol. 2010;115(3):609-17. 2. Barber MD, Brubaker L, Burgio KL, Richter HE, Nygaard I, Weidner AC, et al. Comparison of 2 transvaginal surgical approaches and perioperative behavioral therapy for apical vaginal prolapse: the OPTIMAL randomized trial. JAMA. 2014;311(10):1023-34. 3. Capobianco G, Donolo E, Borghero G, Dessole F, Cherchi PL, Dessole S. Effects of intravaginal estriol and pelvic floor rehabilitation on urogenital aging in postmenopausal women. Archives of gynecology and obstetrics. 2012;285(2):397-403. 4. Kafri R, Deutscher D, Shames J, Golombp J, Melzer I. Randomized trial of a comparison of rehabilitation or drug therapy for urgency urinary incontinence: 1-year follow-up. Int Urogynecol J. 2013;24(7):1181-9. 5. Kaya S, Akbayrak T, Gursen C, Beksac S. Short-term effect of adding pelvic floor muscle training to bladder training for female urinary incontinence: a randomized controlled trial. International Urogynecology Journal. 2015;26(2):285-93. 6. Furst MC, Mendonca RR, Rodrigues AO, Matos LL, Pompeo AC, Bezerra CA. Long-term results of a clinical trial comparing isolated vaginal stimulation with combined treatment for women with stress incontinence. Einstein. 2014;12(2):168-74.

51  Are VC as effective as other treatments for the 3. WEIGHTED VAGINAL CONES (VC) treatment of UI?

Weighted vaginal cones (VC) were developed as a  Are VC combined with PFMT better than PFMT method for testing PFM function and to provide pro- alone for the treatment of UI? gressive muscular overload during PFM strengthening exercises (105). In theory, when a cone is inserted 3.1. Prevention into the vagina, the sensation of ‘losing the cone’ pro- No previous (prior to last Consultation) or new RCT vides strong sensory feedback that prompts the PFM investigating either the primary or secondary preven- to contract to prevent the cone from slipping out. tion effects of training with VC for women with UI were Women start in a standing position with a weighted found. cone held inside the vagina for at least one minute, incrementally adding time and increased cone weight 3.2. Treatment whilst standing or walking. The goal is to walk around A Cochrane review specifically addressing the effec- for 20 minutes without losing the cone; the gradual tiveness of VC in the treatment of UI was updated in increase in cone weight maintains muscle overload March 2013 (92) This meta-analysis and four new over the course of the exercise programme. studies form the basis of this subsection (79, 99, 107, There are various cone weights and sizes (Figure 1). 108). Table 12 illustrates the number of studies in- However, the effectiveness of the VC training method cluded in the 5th ICI and in the Cochrane review of is unclear. Because orientation of the vagina is not Herbison et al. (92) as well as the new studies iden- completely vertical, some women can retain the cone tified in the current update. Characteristics of each without actually contracting the pelvic floor. Radiology RCT not included in the previous ICI are presented in has also demonstrated that the cones can rest in a Table 13. transverse position (106). Depending on the axis of the vagina, women need to produce different force intensities to retain the cone. Thus, using VC as a measure of PFM function may not be a valid method. Finally, some women may find it impossible to insert the cones due to a narrowed vaginal opening or, conversely, to retain it due to an enlarged vaginal opening, prolapse, or an insufficient PFM contraction, one incapable of holding even the lightest cone. This section examines the evidence for VC in the prevention and treatment of UI in women. Questions addressed:  Are VC better than no treatment, placebo or control for the prevention of UI?  Are VC better than no treatment, placebo or control for the treatment of UI?

Figure 1 Weighted vaginal cones

52 Table 12 Studies of VC included in the previous review (5th ICI), in the Cochrane meta-analysis of Herbison et al. and current update (6th ICI)

Studies included in the previous Studies included in the Cochrane New studies identified in this update review (5th ICI) review of Herbison et al. (6th ICI) PFMT vs VCs 11 13 1 (section II.2.3.3i) PFMT/VCs vs VCs 2 2 0 (section II.2.3.4i) VCs vs no treatment, placebo or control 4 5 0 treatments VCs vs EStim 5 6 0 PFMT/VCs vs PFMT 2 2 3

53 Table 13 Summary of data on VC

Author, Study Modality details or Comparator N Outcomes/Results Follow-up Notes year population parameters

Golmakani PFMT (30) 60 Women with SUI Behavioral Intervention Detection SUI severity questionnaire 12 weeks of Dropouts: 2014 (1) vs (at least 3 including PFMT: 10 to 40 (mild/moderate/severe): treatment, VCs 5/30 VCs (30) episodes per repetitions daily, Sign. reduction in both groups (p≤0.04) treatment, PFMT 4/30 week) Knack and urge, suppression Non-sign. difference btw treatments outcomes technique, (p=0.52) assessed post- Bladder training Leakage index: Tx Side effects reported Sign. reduction in both groups for VCs: Candida VCs: PFMT pre-Tx 16.2±4.5 to post-Tx 3.2±0.15 vaginitis and difficulty 6 cones (20 to 70 g) (p<0.001) with the device Cone held passively (without VCs pre-Tx 16.6±4.8 to post-Tx 6.1±2.5 reported in 5 women PFM contraction) for 15 min (p<0.001) twice a day Greater changes in PFMT than VCs Active PFM contractions with (p=0.001) the cone inserted (30 x 5-s 7-day diary contractions) Sign. reduction in both groups Supervision included weekly PFMT pre-Tx 1.2±1.3 to post-Tx 0.5±0.5 phone call and in-person (p=0.001) assessment every 2 weeks VC pre-Tx 1.0±1.0 to post-Tx 0.8±0.7 (p=0.001) Non-sign. difference btw treatments (p=0.52) 1-h pad test Sign. reduction in both groups PFMT pre-Tx 35.8±6.8 to post-Tx 12.8±3.8 (p<0.001) VC pre-Tx 36.1±7.1 to post-Tx 19.5±5.2 (p<0.001) Greater change in PFMT than VC (p=0.008) Incontinence QoL and King’s Health questionnaires: Sign. improvement in both groups. Non-sign. difference btw treatments

54 Author, Study Modality details or Comparator N Outcomes/Results Follow-up Notes year population parameters Harvey PFMT (19) 44 Women with PFMT: 10 weekly sessions of Cure (based on pad test): 10 weeks, Dropout 2006 vs urodynamically biofeedback with trained nurse PFMT 2/19 outcomes PFMT: 12/29 (2) - VCs (25) proven SUI VCs: Hold for 15 min daily VCs 2/25 assessed at 6 VCs 18/25 abstract Women from both groups decreased their months post-Tx pad weight, but High dropout rate- No the difference btw groups was non-sign. potential explanations (p=.39) provided PFMT:6.0 g, 95% CI −2.3 to 14.3 VCs: 3.0 g 95% IC 0.3 to 5.7 UDI-6: All women had reduced scores, but the changes were neither clinically nor statistically significant I-QoL: Subjects in both arms showed an improvement on I-QoL, with no difference btw groups (p=0.86)

55 Author, Study Modality details or Comparator N Outcomes/Results Follow-up Notes year population parameters Pereira VCs (15) 45 Postmenopausal VCs: 12 sessions, 2x 40 1-h pad test: 6 weeks of Dropout 2012 vs women with SUI min/week, 5 cones 20-100 g, it VCs pre-Tx7.36±8.76 to FU 0.36±0.38 treatment, VCs: 0/15 (3) PFMT (15) (at least 1 is reported that the same (p<.001) outcomes PFMT:2/15 vs episode of exercises as the PFMT group PFMT pre-Tx3.70±4.35 to FU 0.19±0.27 assessed 6 Control: 2/15 Control (15) urinary leakage were achieved with the cone (p<.001) weeks after in the previous inserted Control pre-Tx3.87±5.56 to post Tx treatment No adverse effects month) PFMT: 100 3.65±4.94 (p=.19) reported in the three contractions/session, 6 weeks Non-significant difference btw the VC and groups with twice-weekly sessions of PFMT groups (p=.10) 40 min, 3s contraction/6s rest Significant difference btw the VCs and and some help for 5-10s/ 10- control group (p<.001) 20s rest King’s Health Questionnaire (quality of life Control: no treatment for 6 - incontinence impact) weeks, after which they VCs pre-Tx75.56±32.0 to FU 17.78±17.2 received physiotherapy (p<.001) PFMT pre-Tx55.82±39.32 to post-Tx 7.69±14.6 (p<.001) Control pre-Tx59.98±33.81 to post Tx 57.84±29.48 (p=.39) Non-significant difference btw the VC and PFMT groups (p=.32) Significant difference btw the VC and control groups (p<.001)

56 Author, Study Modality details or Comparator N Outcomes/Results Follow-up Notes year population parameters Porta Roda VC+PFMT (37) 70 SUI or stress VCs (vaginal sphere, pelvic Subjective cure/improvement 6 months of Dropouts: 2013 vs predominant gym) and PFMT twice/day, 5 VC+PFMT 24/30 treatment, PFMT +VCs: 2/37 (4) abstract PFMT (33) MUI in parous days/week PFMT 26/35 outcomes PFMT: 3/33 and poster women PFMT same as above RR 1.08 95% CI 0.83 to 1.40 (non-sign. assessed at Both treatment arms were difference btw groups) each visit as well supervised (5 sessions) ICIQ-UI-SF: as post-Tx Mild adverse events at Sign. improvement in both groups the beginning of (improvement occurred earlier in the treatment (i.e. second PFMT+VC group) treatment visit). 4 in VC+PFMT pre-Tx 8.6±2.6 to post-Tx PFMT + VCs 4.8±3.5 (p<0.01) 1 in PFMT PFMT pre-Tx 8.9±2.3 to post-Tx 5.7±3.5 No events were (p<0.01) reported in the PFMT+VCs showed higher improvement subsequent visits than PFMT at the 3-month visit (p<0.05) The nature of the (non-sign. for the 6-month visit) events is not specified 1-hour pad-test: Sign. improvement in the PFMT+VC group pre-Tx3.3±4.8 to post-Tx 1.9±3.7 (p<0.01) Non-sign. improvement in the PFMT group PFMT pre- Tx 1.5±3.0 to post-Tx 1.9±3.3 (non-sign.) Non-sign. difference btw treatments King’s Health Questionnaire (KHQ): Non-sign. difference from baseline and btw treatment PFMT+VC showed higher adherence to treatment than PFMT alone

57 Author, Study Modality details or Comparator N Outcomes/Results Follow-up Notes year population parameters Prudencio PFMT (51) 156 SUI women 3 groups received: 20 sessions Cure of SUI 3 months of Dropout not specified 2014 vs of 45 min (twice a week for 3 PFMT+VCs 30/55 treatment, (5) abstract PFMT+VCs (55) months) PFMT 20/51 outcomes vs PFMT included isolated rapid RR 1.39 95% CI .91 to 2.11 assessed post- PFMT+BF (50) and sustained contractions and KHQ: Tx functional exercises All groups significantly improved from (differentiating just the use or baseline (p<0.001) not of associated instruments) PFMT pre-Tx74±18 to post-Tx 43±17 (no further information was PFMT+VCs pre-Tx62±17 to post-Tx31±8 provided on VCs in the Difference btw groups (p<0.001). It is not abstract) specified which of the 3 groups differed Santos 2009 VCs (21) 45 Women with SUI VCs: Participants attended 2 Subjective cure or improvement: 4 months of No loss to follow-up (6) vs (urodynamic sessions of 45 min per week. VCs 13/21 vs EStim 14/24 treatment with EStim (24) diagnosis) Cones 20-100 g RR: 1.06 (0.66-1.71) outcomes EStim: 2x 20-min weekly Leakage episodes per day: assessed post- session for 4 months VCs 0.21 (0.24) vs EStim 0.33 (0.79) Tx supervised by a Mean difference -0.12 (-0.45 to 0.21) physiotherapist. Intravaginal Improvement on pad test: EStim, Frequency 50 Hz, VCs 10/21 vs EStim 12/24 Intensity10-100 mA and pulse RR 0.95 (0.52- 1.73) duration 1 ms

58 Author, Study Modality details or Comparator N Outcomes/Results Follow-up Notes year population parameters Stupp 2011 PFMT (22) 44 Women with SUI PFMT (3 sets of 8 contractions. Subjective cure (SUI during cough): 12 weeks of Dropouts (7) abstract or MUI 6s contraction: 8s rest and 3 Combined treatment group 12/21 treatment PFMT: 1/22 vs fast contractions) verbal PFMT alone 7/21 outcomes commands given by a RR 1.71 95% CI 0.84 to 3.48 assessed post- PFMT + proprioception PFMT+ physiotherapist KHQ: Tx and awareness training proprioception PFMT + proprioception and Sign. improvement in several domains of 1/22 and awareness awareness training, including 3 the KHQ in women in the combined training additional sessions with a treatment. Non- sign. improvement in the (including VC) physiotherapist PFMT group (22) a) session 1: education Statistical comparisons btw group are not (anatomy and function of the reported PFM), diaphragmatic breathing, visualization with a mirror b) session 2: proprioceptive technique with a VC c) session 3: PFMT contraction during increases in intra- abdominal pressure (Knack) 1. Golmakani N, Khadem N, Arabipoor A, Kerigh BF, Esmaily H. Behavioral Intervention Program versus Vaginal Cones on Stress Urinary Incontinence and Related Quality of Life: A Randomized Clinical Trial. Oman Med J. 2014;29(1):32-8. 2. Harvey MA, Johnston SL. A randomized, single-blind, prospective trial comparing pelvic floor physiotherapy with biofeedback versus weighted vaginal cones in the treatment of female genuine stress urinary incontinence: a pilot study. Int J Urogyn J. 2006;17(Suppl 2):S235-S6. 3. Pereira VS, de Melo MV, Correia GN, Driusso P. Vaginal cone for postmenopausal women with stress urinary incontinence: randomized, controlled trial. Climacteric. 2012;15(1):45-51. 4. Porta Roda O, Simo Gonzalez M, Reula Blasco MC, Diaz Lopez MA, Diaz Bellido P, Vara Paniagua J. Use of a vaginal spheres device in the conservative treatment of stress urinary incontinence: a randomized controlled trial Neurourol Urodyn. 2013;32(6):661-3. 5. Prudencio C, Barbosa A, Derobio AL, Anezio A, Vesentini G, Almeida AP, editors. Comparison of three physiotherapy methods for treatment of stress urinary incontinence: impact in quality of life and muscle function. Proceedings of the 44th Annual Meeting of the International Continence Society (ICS), 2014 Oct 20-24, 2014; 2014; Rio de Janeiro, Brazil. 6. Santos PF, Oliveira E, Zanetti MR, Arruda RM, Sartori MG, Girao MJ, et al. [Electrical stimulation of the pelvic floor versus vaginal cone therapy for the treatment of stress urinary incontinence]. Rev Bras Ginecol Obstet. 2009;31(9):447-52. 7. Stupp L, Yamamoto D, Fonseca T, Resende AM, Ploger C, Oliveira E. Proprioception and awareness training prior pelvic floor muscle exercises for treatment of urinary incontinence: randomized controlled trial. Int Urogyn J. 2011;22(Supp 1):S162-S4.

59 3.2.1 Are VC Better than No Treatment, Pla- This comparison is addressed in Section II.2.3.3. De- cebo or Control Treatments? tails of the VC and PFMT programs for each trial are presented in Table 13. The meta-analysis published by Herbison et al. (92) comprised 5 RCTs comparing VC to a control treat- ii) VC versus EStim ment. Except for Pereira et al. (94) (presented in Ta- Since the last ICI edition, a meta-analysis update was ble 13), these studies were part of the ICI previous published. (92) It comprised the same trials as in the edition. Our literature search identified no further ICI 5th edition as well as the study of Santos et al. RCT. (109) (presented in Table 13). Our literature search Quality of data revealed no further trials. In the study of Pereira et al. (94), adequate allocation Quality of data concealment and randomisation was reported. The i) VC versus PFMT: See Section II.2.3.3. evaluator was however not blinded to patient assig- nation. No attrition was reported in the VC group while ii) VC versus EStim 13% dropped out from the control group. No adverse effects were reported in this RCT. In the study of Santos et al. (109), participants were randomly assigned to treatments while blinding of as- Results sessors and allocation concealment were not clearly reported. No losses to follow-up were reported (109). As discussed in Herbison et al.’s meta-analysis (92), women in the VC groups were more likely to report Results they were cured than the controls (RR for failure 0.84, 95%CI 0.76 to 0.94). VCs were also better than con- i) VC versus PFMT: See Section II.2.3.3. trol treatment in the subjective reporting of cure or im- ii) VC versus EStim provement (RR for failure 0.72, 95%CI 0.52 to 0.99). Further, Pereira et al. (94) reported better inconti- As reported in the Herbison et al. meta-analysis (92), nence-related QOL and a significant reduction in the there was no statistically significant difference be- 1-h pad test for women in the VC group compared to tween VC and EStim in the pooled data of three trials the control treatment. with regard to self-reported cure (RR for failure 1.26 95% CI 0.85 to 1.87). Non-significant differences also Summary emerged from the pooled data of three RCTs with respect to improvement in pad test (RR 1.21 95%CI VCs with supervised training sessions by a trained 0.90 to 1.63) and leakage episodes (Mean difference health professional are better than control treatments -0.05 95% CI -0.27 to 0.17). Herbison et al. (92) re- for subjective reporting of cure or cure/improvement ported discomfort or side effects associated to both and the QOL impact on the treatment of SUI (Level EStim and VC (VC: abdominal pain, vaginitis, bleed- of Evidence: 1). ing, motivational problems and difficulties using the However, VC treatment may be inappropriate in VC. EStim: tenderness and bleeding, discomfort or some cases due to potential reported side effects motivational and other difficulties in using the EStim). (92). Summary Recommendations The meta-analysis of Herbison et al. (92) including For women with SUI, VCs with supervised training six RCTs revealed no significant difference between sessions by a trained health professional may be of- VCs and EStim in terms of self-reported cure, fered as a first-line conservative therapy to those who cure/improvement, improvement in pad test or the can and are prepared to use them (Grade of Recom- number of leakage episodes; both the VC and EStim mendation: B) groups reported adverse events. Trained health professional assessment is recom- VC and EStim seem equally effective in the treatment mended. VC may be inappropriate in some cases due of SUI and MUI. (Level of Evidence: 1). Side effects to inability to insert or retain the cone or because of and discomfort appear to limit their utility in clinical side effects and discomfort. (Grade of Recommen- practice. (Grade of Recommendation: D). dation: D). 3.2.3 Are VC Combined with PFMT Better 3.2.2 Are VC As Effective as Other Treat- than PFMT Alone? ments? Both Herbison et al.’s meta-analysis(92) and the ICI VC have been compared with PFMT and EStim, but 5th edition relied on two RCTs comparing combined not with other therapies such as drug treatment, BT PFMT/VC to PFMT alone. Our search revealed three or surgery. additional studies presented as published abstracts (79, 107, 108). Details of the PFMT/VC and PFMT are i) VC versus PFMT presented in Table 13.

60 Quality of data increasing understanding of the neuroanatomy and physiology of the central and peripheral nervous sys- Limited information on method and results were avail- tems. The mechanisms of action vary depending on able from these published abstracts (79, 107, 108). the cause(s) of UI and the structure(s) being targeted Although participants were randomly assigned, no in- e.g. PFM or detrusor, peripheral or central nervous formation on allocation concealment and blinding of system. In general, the aim of EStim for SUI appears assessors was provided. The sample sizes were rel- to be to increase proprioception and/or to improve the atively small with no more than 39 patients per treat- muscle function of an atrophied or weak PFM, while ment arm. Dropouts ranged from 0 (79) to 13% (108) for UUI the objective seems to be to inhibit detrusor in the PFMT group and 0% (79, 108) to 5% (107) in overactivity (DO) (1). the combined treatment. EStim is provided by clinic-based mains powered ma- Results chines or portable battery powered stimulators (Fig- Including two RCTs, Herbison’s meta-analysis re- ure 2) with a seemingly infinite combination of current ported no significant differences between PFMT/VC types, waveforms, frequencies, intensities, electrode and PFMT alone for either cure or cure/improvement. types and placements (Figure 3). Without a clear bio- Likewise, the results from the three additional RCT logical rationale, it is difficult to make choices about failed to show any significant difference between different ways of delivering EStim. Additional confu- treatments in terms of subjective cure (79, 108) and sion is created by the relatively rapid developments subjective cure/improvement (107). Porta Roda et al. in the area of EStim, and a wide variety of stimulation reported that improvement occurred earlier in the devices and protocols that have been developed combined group (after 3 months of treatment) but this even for the same condition. was not maintained at the end of the treatment (6 Finally, the nomenclature used to describe EStim re- months of treatment) (107). mains inconsistent. EStim has not only been described based on the type of current being used (e.g. Summary faradic, interferential), but also on the structures tar- Limited evidence suggests no benefit from adding geted (e.g. neuromuscular), the current intensity (e.g. VCs to PFMT for women with SUI (Level of Evi- dence: 2). Recommendations No recommendation is possible for combination intervention. Adequately powered studies are needed to confirm or refute the advantages of adding VCs to PFMT (No Recommendation).

3.3. Other Lower Urinary Tract Symptoms (LUTS) In the previous ICI, two RCTs reported data on the efficacy of VC on urgency and nocturia (Williams et al. 2006) and nocturia (Gameiro et al. 2010). Our search revealed no new RCTs on other lower urinary tract symptoms.

3.4. Factors Affecting Outcome None of the RCTs above addressed the effect of age Figure 2 Neuromuscular electrical stimulation equip- or other factors on the outcome of VC training. None- ment theless, in the 22 RCTs included here, on average, 22% of the women being treated with VCs (range 0 to 63%) withdrew from the study or dropped out. Alt- hough few RCTs examined the causes of attrition, among those that reported causal factors low compliance, motivational problems, unpleasantness, aes- thetic dislike, discomfort, and bleeding were impli- cated although no one reason predominated.

4. ELECTRICAL STIMULATION (ESTIM) Figure 3 Neuromuscular electrical stimulation elec- The theoretical basis of neuromuscular electrical trodes stimulation (EStim) interventions has emerged with

61 low-intensity, or maximal stimulation), and the pro- or pattern of incontinence: (1) studies with all or at posed mechanism of action (e.g. neuromodulation). least 50% of participants having SUI alone or a pre- In this section, EStim type and parameters are re- dominant symptom of SUI (as defined by trial investi- ported in line with International Continence Society gators), (2) studies with all or >50% of participants definitions (1). having UUI alone or a predominant symptom of UUI (as defined by trial investigators), (3) other studies of This section presents the evidence for the use of ES- participants with UI in which neither stress- or ur- tim in the prevention and treatment of UI in women. gency-UI represented a predominant symptom in the Questions addressed are: study population (‘UI all types’ hereafter), and (4)  Is EStim effective in the prevention of UI? studies of overactive bladder (OAB) or DO in which it was unclear whether all participants had UI.  Is EStim better than no active treatment (placebo, sham, control or no treatment) for treat- Single EStimates with 95% confidence intervals (CI) ment of UI? were derived for each study comparison using odds ratios (OR) for dichotomous variables or mean differ-  Is one type of EStim better than another in the ence (MD) for continuous variables. Summary ESti- treatment of UI? mates were calculated using random effects models if there was more than one study reporting the same  Is EStim better than other treatments in the treat- outcome (meta-analysis). ment of UI? ‘Risk of bias’ in the included studies was assessed for  Does the addition of EStim to other treatments allocation concealment (selection bias) and com- add any benefit in the treatment of UI? pleteness of outcome data (attrition bias), using rele-  What is the effect of EStim on other LUTS? vant items in a standard tool developed by the Cochrane Urinary Incontinence Group (111). Risk of  What factors might affect the outcome of EStim bias regarding blinding to the allocated intervention in the treatment of UI? was high in most included studies: blinding of participants and care providers is not always feasible (other Eligible interventions were non-invasive EStim with- than the use of sham EStim), and blinding of outcome out implanted electrodes. (Magnetic stimulation and assessors is equally difficult for self-reported out- posterior tibial nerve stimulation are described in Sec- comes such as cure, improvement and quality of life. tions II.5 and II.6). Other criteria for inclusion were (1) randomised or quasi-randomised (alternate alloca- Description of intervention in included studies of tion) trial design, (2) women with UI or other LUTS, EStim (3) no participants with incontinence due to neurolog- Nine new trials were identified for this update, making ical or cognitive impairment and (4) no pregnant or a total of 42 trials included in this section. All new tri- postpartum women (within 12 months of childbirth). als targeted women with SUI or predominant SUI. No Trial data reported in conference abstracts as well as new studies for predominant UUI, DO or OAB (incon- full-text papers were included. EStim compared with tinent or not) were found for this update. Findings for PFMT and vaginal cones are covered in previous sec- UUI, UI all types, DO and OAB are therefore un- tions (sections II.2.3.3 and II.3.2.2). This section fo- changed from 5th ICI. The number of included studies cuses on EStim compared with no active treatment or by dominant type or pattern of incontinence is sum- other conservative treatments. marised in Table 14. The primary outcomes were cure rates (the number In addition, there was one study, published in Portu- of women with no urinary incontinence episode at guese, which was a three-arm trial comparing EStim time of assessment) and improvement rates (the and PFMT (N = 24), PFMT (N = 25) and control (N = number of women improved, including cure). There 22) (112). This trial could not be incorporated here, was considerable variability in the way these out- as an English translation was not available. comes were measured. Women’s self-report was given priority but for studies in which it was not re- The EStim parameters and protocols in this section ported, the rate based on diaries was used as a are summarised in Table 15. Some approaches to proxy; where diary data were also not reported, the treatment are now less common, such as the use of rate based on pad tests or any other definitions cho- interferential current or external electrodes. There sen by the trialists was used (110). Data on health- was considerable variation in the intervention proto- related quality of life and adverse effects were also col. Although the biological rationale and purpose of extracted. Data at the end of the prescribed treatment EStim might be different depending on diagnosis, phase, or at the first outcome measurement, if later, there was no consistency in the EStim protocols used were used in the analysis. Any treatment effects for women with SUI, UUI, UI all types, or DO. shown are likely to reflect maximum effect of each intervention. Data from further follow-up were also rec- 4.1. Prevention orded. No published trials were found. Due to the small number of available studies per intervention, data were sub-grouped by dominant type

62 Table 14 Studies of EStim included in the previous review (5th ICI) and current update (6th ICI)

Studies included in the previous New studies identified in this Total review (5th ICI) update (6th ICI)

EStim vs No active treatment SUI or predominant SUI 9 4 13 UUI or predominant UUI 3 0 3 UI all types 2 0 2 DO/OAB (dry or wet) 3 0 3 One type of EStim vs another SUI or predominant SUI 2 4 6 UUI or predominant UUI 0 0 0 DO/OAB (dry or wet) 2 0 2 EStim vs other treatment SUI or predominant SUI 1 No study found 1 UUI or predominant UUI 1 1 DO/OAB (wet or dry) 4 4 EStim+PFMT vs PFMT SUI or predominant SUI 8 (no BF) 2 (no BF) 10 2 (with BF) 1 (with BF) 3 UUI or predominant UUI 0 0 0 UI all types 2 (no BF) 0 2 BF = biofeedback

Table 15 Summary of EStim protocols

Current Pulse Shape & Treatment Author, year Current Frequency (Hz) Duty Cycle Electrodes Target UI Intensity Duration Duration Alves 2011 (1) Biphasic Max tolerable Study arm 1: Study arm 1: Single ratio: 1:2 (4s on, Single vaginal electrode 20min session, 2x SUI intensity 100ms; medium frequency 8s off) a week: 6wks Study arm 2: 2000Hz; 700ms Study arm 2: low frequency 50Hz

63 Current Pulse Shape & Treatment Author, year Current Frequency (Hz) Duty Cycle Electrodes Target UI Intensity Duration Duration Correia 2014 (2) Biphasic Max tolerable Duration: Single freq: 50Hz Single ratio: 1:2 (4s on, Study arm 1: 4 electrodes 20min session, 2x SUI intensity 700µsec 8s off) + 2s rise, 2s fall (2 in the suprapubic a week, by region and 2 medial to physiotherapist: the ischial tuberosity); 6wks Study arm 2: Single vaginal electrode Huebner 2011 (3) NR Range 20- NR Single freq: 50Hz Study arm 1: Active Single vaginal electrode 15min session, 2x SUI 80mA contracting of PFM 8s. a day, home After reaching the treatment with 5 maximum contraction, clinic visits: 12wks electrical stimulation was added for 8s. Resting 15s; Study arm 2: Stimulation 8s. Resting 15s. Active contracting of PFM 8s. Resting 15s. Jeyaseelan 2003* NR NR NR Range not defined A longer duty cycle than NR NR SUI (4) is traditionally used Lopès 2014 (5) Rectangular NR Duration: 3 frequencies NR Vaginal electrode 20min session, 3x SUI biphasic 400µsec available: 50Hz for a day, home SUI, 20Hz for MUI, treatment: 6mths 12.5Hz for pure urgency Maher 2009* (6) NR NR NR NR NR Study arm 1: external 30min session, 4x SUI electrodes; Study arm 2: a week, home single vaginal electrode treatment: 8wks (outcome reported at 4wks)

64 Current Pulse Shape & Treatment Author, year Current Frequency (Hz) Duty Cycle Electrodes Target UI Intensity Duration Duration Patil 2010 (7) Interferential Max tolerable NR Freq range: 0- NR 4 electrodes (2 flat First session SUI intensity 100Hz electrodes placed 15min, other anteriorly over the sessions 30min, obturator foramen, and 2 3x a week, by electrodes placed physiotherapist: posteriorly medial to 4wks. ischial tuberosity on either side of the anus) Pereira 2012 (8) NR Max tolerable Duration: Single freq: 50Hz Single ratio: 1:2 (4s on, 4 electrodes (2 in the 20min session, 2x SUI intensity 700µsec 8s off) suprapubic region and 2 a week, by medial to the ischial physiotherapist: tuberosity) 6wks Terlikowski 2013 NR NR 200 to 250µsec Freq range: 10- Single ratio: 1:2 (15s on, Single vaginal electrode 20min session, 2x SUI (9) 40Hz 30s off) a day, home treatment: 8 weeks.. * abstract only. Footnotes: EStim = electrical stimulation, freq= current frequency, PFM = Pelvic floor muscle, PFMC= Pelvic floor muscle contraction, VPFMC= voluntary Pelvic floor muscle contraction 1. Alves PG, Nunes FR, Guirro EC. Comparison between two different neuromuscular electrical stimulation protocols for the treatment of female stress urinary incontinence: a randomized controlled trial. Revista Brasileira de Fisioterapia. 2011;15(5):393-8. 2. Correia GN, Pereira VS, Hirakawa HS, Driusso P. Effects of surface and intravaginal electrical stimulation in the treatment of women with stress urinary incontinence: randomized controlled trial. European Journal of Obstetrics and Gynecology and Reproductive Biology. 2014;173(1):113-8. 3. Huebner M, Riegel K, Hinninghofen H, Wallwiener D, Tunn R, Reisenauer C. Pelvic floor muscle training for stress urinary incontinence: a randomized, controlled trial comparing different conservative therapies. Physiother Res Int. 2011;16(3):133-40. 4. Jeyaseelan S, Oldham JA. Can the effects of pelvic floor muscle exercises be enhanced with a new pattern of electrical stimulation in women with stress incontinence (Abstract). Proceedings of the World Confederation for Physical Therapy (WCPT), 14th International Congress, 7-12 June, Barcelona. 2003. 5. Lopès P, Rimbault F, Scheffler M, Andre C, Cappelletti MC, Mares P. [Multicentric prospective randomized and controlled study assessing effectiveness of intravaginal electrostimulation at home compared to usual care in female patients with urinary incontinence and prior perineal reeducation]. [French]. Gynecologie, Obstetrique & Fertilite. 2014;42(11):779-86. 6. Maher RM, Crowe L, Caulfield B. Comparison of two methods of electrical muscle stimulation training of pelvic floor musculature in the treatment of stress urinary incontinence (Abstract). Journal of Women's Health Physical Therapy. 2009;33(1):24.

65 7. Patil SP, Nagrale AV, Ganvir SD. Additive effect of interferential therapy over pelvic floor exercises. International Journal of Therapy & Rehabilitation. 2010;17(11):596- 602. 8. Pereira VS, Bonioti L, Correia GN, Driusso P. [Effects of surface electrical stimulation in older women with stress urinary incontinence: a randomized controlled pilot study]. [Spanish]. Actas Urologicas Espanolas. 2012;36(8):491-6. 9. Terlikowski R, Dobrzycka B, Kinalski M, Kuryliszyn-Moskal A, Terlikowski SJ. Transvaginal electrical stimulation with surface-EMG biofeedback in managing stress urinary incontinence in women of premenopausal age: a double-blind, placebo-controlled, randomized clinical trial. International Urogynecology Journal. 2013;24(10):1631-8. Table 16 Summary of data on EStim vs no active treatment

N Duration Author, year Comparator Study population Outcome** randomised (months) SUI or predominantly SUI

Correia 2014 (1) ES (31) 48 SUI alone 1.5 Cure: NR Improvement: NR vs QoL via KHQ No treatment (17) Surface ES vs no treatment, N = 32, mean difference 54.45 lower, 95% CI 73.44 to 35.46 lower; Vaginal ES vs no treatment, N = 32, mean difference 56.67 points lower, 95% CI 75.30 to 38.04 lower Adverse effects: NR Lopès 2014 (2) ES (78) 164 SUI or SUI- 6 Cure: NR vs predominant MUI Self-reported improvement: 63/76 vs 58/85 No treatment (86) QoL via ICIQ: N = 161, mean difference in change from baseline 3.10 lower, 95% CI 4.39 to 1.81 lower Adverse effects: reported with no detail Pereira 2012 (3) ES (7) 14 SUI alone 1.5 Cure: NR Improvement: NR vs QoL via KHQ: N = 14, mean difference 42.95 lower, 95% CI 70.74 to No treatment (7) 15.16 lower Adverse effects: no events noted Terlikowski 2013 (4) ES + BF† (68) 102 USI alone 2 Self-reported cure: 29/64 vs. 2/29 vs Self-reported improvement: 41/64 vs 6/29 Sham ES + BF† (34) QoL via I-QoL: N = 93 at 8 weeks (end of treatment), mean difference 22.3 higher, 95% CI 15.52 to 29.08 higher; N = 93 at 16 weeks (8 8 weeks after treatment), mean difference 30.20 higher, 95% CI 22.18 to 38.22 higher Adverse effects: smarting and discomfort Note: For modality details or parameters, see Table 14.

66 NR = not reported; ICIQ = International Consultation of Incontinence Questionnaire (higher scores indicate greater impact on QoL, i.e. a higher score indicates a lower QoL); I-QoL = Incontinence Quality of Life Questionnaire (higher scores indicate higher QoL); KHQ = King’s Health Questionnaire (higher scores indicate greater impairment); **Source of cure and improvement outcome: women’s self-report was given priority but for studies in which it was not reported, quantification of outcomes based on diaries, pad tests or any other definitions chosen by trialists were used as a proxy. † Electromyography-assisted biofeedback 1. Correia GN, Pereira VS, Hirakawa HS, Driusso P. Effects of surface and intravaginal electrical stimulation in the treatment of women with stress urinary incontinence: randomized controlled trial. European Journal of Obstetrics and Gynecology and Reproductive Biology. 2014;173(1):113-8. 2. Lopès P, Rimbault F, Scheffler M, Andre C, Cappelletti MC, Mares P. [Multicentric prospective randomized and controlled study assessing effectiveness of intravaginal electrostimulation at home compared to usual care in female patients with urinary incontinence and prior perineal reeducation]. [French]. Gynecologie, Obstetrique & Fertilite. 2014;42(11):779-86. 3. Pereira VS, Bonioti L, Correia GN, Driusso P. [Effects of surface electrical stimulation in older women with stress urinary incontinence: a randomized controlled pilot study]. [Spanish]. Actas Urologicas Espanolas. 2012;36(8):491-6. 4. Terlikowski R, Dobrzycka B, Kinalski M, Kuryliszyn-Moskal A, Terlikowski SJ. Transvaginal electrical stimulation with surface-EMG biofeedback in managing stress urinary incontinence in women of premenopausal age: a double-blind, placebo-controlled, randomized clinical trial. International Urogynecology Journal. 2013;24(10):1631-8.

67 4.2. Treatment Incontinence Quality of life (I-QoL) questionnaire, and two used the King’s Health Questionnaire (KHQ) 4.2.1 Is EStim Better than No Active Treat- (113, 115). ment (Placebo, Sham, Control or No Treatment) for Treatment of UI? Adverse effects appeared uncommon. One new study reported bleeding in one and discomfort in Four new studies including 267 women with SUI or three of 64 participants using the active EStim device predominant SUI (113-116). Participants in one of and none of 29 participants in the sham EStim device these studies were women who had responded posi- (116). This is in line with two previous studies that re- tively to physiotherapy treatment for their UI (10-15 ported tenderness and bleeding (118), and vaginal ir- sessions), with EStim used to maintain the benefit of ritation, pain or infection (123) associated with the de- initial physiotherapy (114). One study (115) was a pi- vice. Another study reported that one participant each lot study for a newly identified study (113). in the treatment (N = 78) and control (usual care; N = 86) groups was lost to follow-up due to adverse ef- Characteristics of the new studies comparing EStim fects (no further detail was provided) (114). with no active treatment are presented in Table 16. No active treatment consisted of no treatment (113, 115), sham EStim (116) or usual care (114). One Summary study was a three-arm trial. Two of the arms using A total of 21 studies assessed the effect of EStim surface and intravaginal EStim (N = 31) were com- compared with no active treatment, including 13 in bined and compared with the ‘no treatment’ arm (N = women with SUI or predominant SUI, three in UUI or 17) (113). predominant UUI, two in all types of UI, and three in Quality of data DO/OAB (incontinent or no). Three studies reported adequate methods of alloca- Findings from update analysis using the additional data from newly identified studies were broadly simi- tion concealment, namely third party involvement th (113, 116) or opaque (and sealed) envelopes with lar to those from the 5 ICI. Included studies were third party involvement (115) in the allocation proce- generally assessed as having a high risk of bias. dure. One study did not describe methods used for EStim might be more effective than no treatment in allocation concealment (116). Trial results were re- improving symptoms and quality of life in women with ported for everyone who entered the trial in one study SUI and improving symptoms in women with UUI, alt- (115) but this was not done in the others. One study hough this may not result in cure (Level of Evidence: is funded by the manufacturer of the study stimulator 2). (114). Information on quality of life was sparsely reported Results particularly for UUI or DO, and the limited data that SUI or predominant SUI. When adding new studies were available were not consistent. Adverse effects to those previously reported, pooled data suggest appear uncommon but some women experienced that cure rates were, on average, higher for EStim discomfort with the treatment device. Scant data were compared with no active treatment but the difference available on long-term performance. was not statistically significant (N = 434, 22% vs 5%, Recommendations OR 2.43, 95% CI 0.89 to 6.60) (116-123). Improve- ment rates were statistically significantly higher for EStim might be better than no treatment to improve EStim compared with no active treatment (N = 613, symptoms and quality of life in SUI women (Grade of 53% vs 30%, OR 3.64, 95% CI 1.82 to 7.27) (114, Recommendation: B). 116-118, 120-124), although there was some evi- EStim may be considered for treatment to improve dence of statistical heterogeneity for the improvement symptoms for UUI (Grade of Recommendation: B). rate (I-squared = 45%). The result of this meta-analysis using additional data from newly identified stud- However, this recommendation should be viewed ies remains similar to the analysis performed in the with caution until the findings are supported or refuted 5th ICI, except that effect size was larger and confi- in further trials; it would be particularly useful if further dence interval for cure was wider, and effect size was trials used validated and reliable quality life measures smaller and confidence interval was narrower for im- as a primary outcome indicator particularly for UUI. provement. Quality of life was reported in seven studies, using di- 4.2.2 Is One Type of EStim Better than An- verse measures. Six studies, including 4 new studies, other in the Treatment of UI? found statistically significant differences favouring Four new studies including 145 women with SUI and EStim compared with no active treatment (113-116, predominant SUI (113, 126-128) were found. The 118, 119), and one previous study found no signifi- characteristics of the new studies comparing one type cant differences between the groups (125). All new of EStim with another are presented in Table 17. Dif- studies that reported this outcome favoured EStim: ferent variants of EStim were assessed, either alone one (114) used the International Consultation on In- (113, 126, 128) or as an adjunct to PFMT and bio- continence Questionnaire (ICIQ), one (116) used the feedback (127).

68 Table 17 Summary of data on different types of EStim comparisons

Author, N Study Duration Comparator Outcome** year randomised population (months)

SUI or predominantly SUI

Alves 2011 Medium 24 SUI alone 1.5 Cure based on pad test: 10/10 vs (1) frequency ES 10/10 vs Improvement: NR Low frequency QoL: NR ES (number Adverse effects: NR per group not reported) Correia Surface ES 31 SUI alone 1.5 Cure: NR 2014 (2) (15) Improvement: NR vs QoL via KHQ: N = 30, mean Vaginal ES difference 2.22 higher, 95% 6.95 (16) lower to 11.39 higher Adverse effects: NR Huebner Dynamic ES + 72 SUI or SUI- 3 Cure: NR 2011 (3) PFMT + BF† predominant Improvement: NR (36) MUI QoL via KHQ: N = 61, mean vs difference in change from baseline Conventional 4.10 lower, 95% CI 6.77 to 1.43 ES + PFMT + lower BF† (36) Adverse effects: allergic reaction to lubricant Maher External ES 18 SUI alone 2 (outcome Cure: NR 2009* (4) vs measured Improvement: NR (ongoing) Vaginal ES at 1 month) QoL: NR (number per Adverse effects: NR group not reported) Note: For modality details or parameters, see Table 14. NR = not reported; KHQ = King’s Health Questionnaire (higher scores indicate greater impairment); * abstract only; **Source of cure and improvement outcomes: women’s self-report was given priority but for studies in which it was not reported, quantification of outcomes based on diaries, pad tests or any other definitions chosen by trialists was used as a proxy. † Electromyography-assisted biofeedback 1. Alves PG, Nunes FR, Guirro EC. Comparison between two different neuromuscular electrical stimulation protocols for the treatment of female stress urinary incontinence: a randomized controlled trial. Revista Brasileira de Fisioterapia. 2011;15(5):393-8. 2. Correia GN, Pereira VS, Hirakawa HS, Driusso P. Effects of surface and intravaginal electrical stimulation in the treatment of women with stress urinary incontinence: randomized controlled trial. European Journal of Obstetrics and Gynecology and Reproductive Biology. 2014;173(1):113-8. 3. Huebner M, Riegel K, Hinninghofen H, Wallwiener D, Tunn R, Reisenauer C. Pelvic floor muscle training for stress urinary incontinence: a randomized, controlled trial comparing different conservative therapies. Physiother Res Int. 2011;16(3):133-40. 4. Maher RM, Crowe L, Caulfield B. Comparison of two methods of electrical muscle stimulation training of pelvic floor musculature in the treatment of stress urinary incontinence (Abstract). Journal of Women's Health Physical Therapy. 2009;33(1):24.

69 One study comparing external and vaginal EStim for on cure rates, quality of life and adverse effects women with SUI, available only as an abstract, did (Level of Evidence: 2). not report any of the specified outcomes and thus did not contribute to the analysis (128). Since there is lit- The other studies did not find clinically important dif- tle duplication of EStim interventions, it was not ferences between stimulation groups for the specified thought appropriate to combine study findings. outcomes; the studies were small and may have been underpowered. Further comparisons of EStim proto- Quality of data cols are needed. Allocation concealment was adequate in one study Recommendations which involved a third party in the allocation procedure (113). The other three studies did not mention For women with SUI maximal clinic-based EStim allocation concealment (126-128). Data were re- might be better than daily low-intensity home-based ported only for those who completed the trial in two EStim in improving symptoms (Grade of Recom- studies (126, 127) but it was unclear in the other two mendation: B). studies if trial results were reported for everyone who There is a need for studies to elucidate the purpose entered the trial (113, 128). and biological rationale for EStim in different diagnos- Results tic groups, so these can then be tested and compared in clinical trials. SUI or predominant SUI. One new small study comparing medium and low frequency EStim found no dif- 4.2.3 Is EStim Better than Other Treatments ference for cure rates (126). The results for improve- for UI? ment rates are as reported in two previous studies (129, 130). No new study was found that investigated whether EStim is better than other treatments for UI. The level While no information was available on quality of life of evidence and recommendations remains un- and adverse effects in the 5th ICI, two new studies re- changed from the previous review (5th ICI) formed on ported quality of life using the KHQ. The first study the basis of six studies (124, 131-135). comparing ‘dynamic’ EStim (contract pelvic floor muscles, and then add stimulation at maximal con- Summary traction) with ‘conventional’ EStim (perform stimulation, rest, and contract pelvic floor muscles) found a The 5th ICI included six studies, including one in statistically significant difference favouring ‘dynamic’ women with SUI, one in women with predominant ES, although study authors considered the difference UUI, and four in women with OAB/DO (wet or dry). to be small and not clinically important (127). The sec- Included studies were generally assessed as having ond study comparing surface and vaginal EStim a high risk of bias. With small numbers per compari- found no statistically significant difference between son group available, there is insufficient evidence to the groups (113). determine if EStim is better than vaginal oestrogens One new study in which EStim was performed as an in women with SUI, propantheline bromide in women adjunct to PFMT and biofeedback reported that one with UUI, or oxybutynin and tolterodine for DO (Level of 72 participants had an allergic reaction to biofeed- of Evidence: 2). back lubricant and withdrew from the study, although Recommendation it was unclear in which group these adverse effects occurred (127). Based on current limited evidence, EStim could be considered as an alternative to medical treatment. Summary Medical treatments (drugs) appear to be no more effective than EStim (Grade of Recommendation: B). A total of eight studies assessed the effect of one ap- These findings need to be investigated further with proach of EStim compared with another, with six in high quality trials, if it is a clinical question of interest women with SUI and predominant SUI and two in to women. women with DO. No study focusing on UUI or predominant UUI was identified. 4.2.4 Does the Addition of EStim to Other Findings from an updated analysis using the addi- Treatments Add Any Benefit in the tional data from newly identified studies were broadly Treatment of UI? similar to those from the 5th ICI. Included studies generally had a high risk of bias. There were eight small Three new studies including 52 women including trials comparing different EStim protocols; the clinical women with SUI and predominant SUI were found heterogeneity between studies meant it was not ap- (127, 136, 137). The characteristics of these new propriate to pool the data. studies are presented in Table 18. Two studies (136, 137) combined EStim with PFMT (EStim + PFMT vs Based on a single trial (130) for women with SUI, PFMT), while the other (127) combined EStim with maximal clinic-based stimulation may be more effec- PFMT and biofeedback (EStim + PFMT + BF vs tive than low-intensity home-based stimulation in im- PFMT + BF).. proving symptoms, although no data were available

70 Table 18 Summary of data on EStim + another treatment vs the other treatment

Study Duration Author Comparator N randomised Outcome** population (months) SUI or predominantly SUI

Jeyaseelan ES + PFMT 13? SUI alone 2 Cure: NR 2003* (1) (6?) Improvement: NR vs QoL via IIQ (% change from PFMT (7?) baseline): N = 6, median -27%, range -63 to 0, vs. N = 7, median 0%, range -67 to 200 QoL via UDI (% change from baseline): N = 6, median -32%, range -50 to -18, vs. N = 7, median 0%, range -43 to 180 Adverse effects: NR Patil 2010 (2) ES + PFMT 13 USI alone 1 Cure: NR (55) Improvement: NR vs QoL via IIQ: N = 102, mean PFMT (55) difference 12.53 lower, 95% CI 19.45 to 5.61 lower Adverse effects: NR Huebner 2011 ES + PFMT + 26 SUI or SUI- 3 Cure: NR (3) BF† (72) predominant Improvement: NR vs MUI QoL via KHQ PFMT + BF† Dynamic ES+PFMT+BF vs (36) PFMT+BF: N = 55, mean difference in change from baseline 4.60 lower, 95% CI 7.43 to 1.77 lower; Conventional ES+PFMT+BF vs PFMT+BF: N = 60, mean difference in change from baseline 0.50 lower, 95% CI 3.22 lower to 2.22 higher Adverse effects: allergic reaction to lubricant Note: For modality details or parameters, see Table 14. NR = not reported; IIQ = Incontinence Impact Questionnaire (lower scores indicate better QoL); KHQ = King’s Health Questionnaire (higher scores indicate greater impairment); UDI = Urogenital Distress Inventory (higher scores indicate greater discomfort); **Source of cure and improvement outcome: women’s self-report was given priority but for studies in which it was not reported, quantification of outcomes based on diaries, pad tests or any other definitions chosen by trialists were used as a proxy. † Electromyography-assisted biofeedback 1. Jeyaseelan S, Oldham JA. Can the effects of pelvic floor muscle exercises be enhanced with a new pattern of electrical stimulation in women with stress incontinence (Abstract). Proceedings of the World Confed- eration for Physical Therapy (WCPT), 14th International Congress, 7-12 June, Barcelona. 2003. 2. Patil SP, Nagrale AV, Ganvir SD. Additive effect of interferential therapy over pelvic floor exercises. Inter- national Journal of Therapy & Rehabilitation. 2010;17(11):596-602. 3. Huebner M, Riegel K, Hinninghofen H, Wallwiener D, Tunn R, Reisenauer C. Pelvic floor muscle training for stress urinary incontinence: a randomized, controlled trial comparing different conservative therapies. Physiother Res Int. 2011;16(3):133-40.

71 One study was a three-arm trial that included both clinically important. No significant difference was ‘dynamic’ EStim (stimulation added at maximum con- found for the group combining ‘conventional’ EStim traction of pelvic floor muscles) and ‘conventional’ with PFMT and biofeedback, compared with PFMT EStim (stimulation followed by rest and pelvic floor and biofeedback alone. muscle contraction) as two of the study arms (127). These arms were combined in the analysis. One In the same study (127), one of 72 participants had study is a pilot study, available only as abstract, with an allergic reaction to biofeedback lubricant and with- limited information (136). drew from the study, although it was unclear in which group this occurred. Quality of data Summary Allocation concealment was inadequate in one study which used opaque sealed envelopes but with no in- A total of 15 studies assessed the effect of EStim as dication of third party involvement in the allocation an adjunct to another treatment, compared with the procedure (137). The other two studies did not men- other treatment alone. All but two studies included tion allocation concealment (127, 136). In two stud- women with SUI or predominant SUI. No study focus- ies, data were reported only for those who completed ing on UUI or predominant UUI was identified. the trial (127, 137) and it was unclear in the other study if trial results were reported for everyone who Findings from update analysis using the additional entered the trial (136). data from newly identified studies were broadly similar to those from the 5th ICI. For comparisons of EStim Results with PFMT versus PFMT alone, there was no clear evidence of a difference between the groups in i) EStim + PFMT vs PFMT women with SUI or predominantly SUI in terms of SUI or predominant SUI. Two new studies assessing cure and improvement. Evidence for quality of life out- EStim as an adjunct to PFMT compared with PFMT comes was not consistent across studies (Level of alone, contributed no additional data in terms of cure, Evidence: 2). There was also no evidence to suggest improvement and adverse effects. The 5th ICI previ- that the addition of EStim to a BF-assisted PFMT was ous analysis for these outcomes therefore remains more effective than BF-assisted PFMT in women with unchanged. SUI (Level of Evidence: 2). A few women experienced adverse effects with EStim. There is no evi- Three studies reported data on quality of life with in- dence to draw any conclusion about the effect of add- consistent results. One previous study found no sta- ing EStim to PFMT for women with UUI. tistically significant differences between the group (138), whereas one new study using the Incontinence Recommendations Impact Questionnaire (IIQ) found a statistically signif- The addition of EStim to PFMT or BF-assisted PFMT icant difference favouring the group which combined programmes does not appear to add benefit (Grade EStim with PFMT (137). The results from another new of Recommendation: B); combinations of tech- study based on either the IIQ or the Urogenital Dis- niques need to be investigated further with high qual- tress Inventory were inconclusive, due to the pilot na- ity trials if this is a clinical/research question of inter- ture of the study and small sample size (N = 13) (136). est to women. ii) EStim + PFMT with biofeedback vs PFMT with biofeedback 4.3. Other LUTS SUI. One new study assessed EStim as an adjunct to In the 5th ICI, there were no trials that analysed the PFMT and biofeedback compared with PFMT and bi- effect of EStim in women with other LUTS alone.No ofeedback alone. This was a three-arm trial, as- new published trials were found. sessing a combination of either ‘dynamic’ or ‘conventional’ EStim with PFMT and biofeedback, compared 4.4. Factors Affecting Outcome with PFMT and biofeedback alone (127). None of the included trials addressed the effect of This study did not provide any data on cure and image, or any other factor, on outcomes of EStim. There provement rates (127). Thus, the 5th ICI analysis of was no clear indication from the included trials that improvement rates based on two studies (129, 130) EStim could not be tolerated by elderly. There is no remains unchanged, with no available information on reason, therefore, to either exclude older women from cure rates. studies of EStim, or not to offer EStim as part of a conservative management programme, except where While no information was available on quality of life recognised precaution with use of intra-vaginal elec- and adverse effects for this comparison in the 5th ICI, trodes in women with vaginal atrophy and contraindi- the new study (127) reported that quality of life scores cations such as a cardiac pacemaker are present. based on the King’s Health Questionnaire improved One newly identified study including elderly women statistically significantly for the group which combined (over 60 years old) highlighted the potential embar- ‘dynamic’ EStim with PFMT and biofeedback, com- rassment perceived by the elderly with regard to con- pared with PFMT and biofeedback alone, although ventional intra-vaginal stimulation and suggested that study authors did not consider the difference to be

72 surface EStim may be a more acceptable method for nerve root contains sensory fibres from the pelvic this population (115). floor and parasympathetic motor efferent fibres to the detrusor as well as the pelvic sphincters and the pel- Of note, in a prospective cohort of 3,198 women vic floor muscles. Afferent nerve stimulation can treated with home-managed EStim in Norway during therefore lead to activation of inhibitory sympathetic 1992-1994, there was no association between self- neurons and suppression of detrusor contraction reported improvement and age (139). In the same co- through a direct sacral route. Urodynamic studies hort, success rates as defined by clinicians were have shown that electrical stimulation of the posterior higher in younger individuals but this effect was not tibial nerve increases cystometric capacity and sup- significant after controlling for other factors. presses detrusor contraction (143-145). The full Aside from age, other factors may have the potential mechanism of action of treatment effect for PTNS is to mitigate treatment outcome. On the basis of trial not yet understood, however it is thought that the ob- reports to date, it appeared that there was considera- served effects may be related to a neuroplastic reor- ble variation in EStim protocols with no consistent ganisation of sacral spinal reflexes and regulation of pattern emerging. EStim protocols are also often cortical excitability (146, 147). poorly reported, lacking detail of stimulation parame- Percutaneous PTNS is performed as an outpatient ters, devices and methods of delivery. The wide procedure. It involves inserting a 34-gauge needle 3– range of protocols that have been tested may have 5 cm cephalad to the medial malleolus. The needle is affected the effect EStimates reported in this section. connected to a low-voltage stimulator device, and a ICS/IUGA has produced a physiotherapy interven- grounding pad is placed on the bottom of the foot just tions terminology paper in the last year, and authors below the smallest toe. Transcutaneous PTNS may are encouraged to refer to this paper when reporting be delivered either in clinic or self-administered at EStim parameters in a publication (1). It is not clear whether one diagnostic group may benefit more than another from EStim. It has been hy- pothesised that, in women with SUI who cannot vol- untarily contract the PFM to begin a PFMT programme, EStim might help initiate or substitute for a voluntary contraction. However, most studies focusing on the efficacy of EStim do use EStim to initiate or substitute for a voluntary PFM contraction (140). To date, there has been no trial addressing this hypothesis. As with all conservative therapy modalities, one of the key factors to the success or failure of EStim is treatment adherence. Some authors commented on adherence or reported adherence data. Of the new studies identified for this 6th ICI update, one study reported that adherence to EStim was satisfactory Figure 4 PTNS equipment (116) and another reported that the EStim regimen was adhered to by over >75% of participants (114). home. Self-adhesive electrodes are placed behind However, adherence measures were highly variable, the medial malleolus and 10cm proximal to this. The making comparison across studies difficult. positive lead is connected to the proximal electrode and the negative to the distal electrode and both are 5. POSTERIOR TIBIAL NERVE connected to a portable battery-powered stimulator. The intensity level of the stimulation current for per- STIMULATION (PTNS) cutaneous PTNS and transcutaneous PTNS is determined once correct positioning has been established Posterior tibial nerve stimulation (PTNS) is a form of by noting sensory and motor (hallux) reaction (Figure peripheral neurostimulation targeted towards symp- 4). tom relief of OAB and UUI (141). Indirect access to the sacral plexus is achieved by intermittent, electri- This section presents the evidence for the use of cal stimulation of the posterior tibial nerve, which lies PTNS in the prevention and treatment of UI in behind the medial malleolus. PTNS may be minimally women. Since this is a new section for the Conserva- invasive, involving insertion of a fine needle close to tive Management chapter, the same search strategy the nerve (Percutaneous PTNS), or non-invasive, us- as EStim section was used although no date re- ing skin surface electrodes applied to the medial mal- strictions have been applied to the literature search. leolar area (Transcutaneous PTNS)(141). PTNS aims to stimulate the sacral nerve plexus through the afferent fibres of the posterior tibial nerve, a mixed nerve containing L5-S3 fibres (142). The S3

73 Table 19 Summary of PTNS protocols

Pulse Shape & Frequency Treatment Author, year Current Current Intensity Duty Cycle Electrodes Target UI Duration (Hz) Duration Bellette 2009 (1) NR NR NR NR NR NR 2 months OAB

Finazzi-Agrò NR 0 – 10 mA; increased 200 20 Hz NR One surface electrode on 3 months UUI 2010 (2) until flexion of big toe microseconds the medial aspect of the or fanning of all toes calcaneous become noticeable Manriquez 2013* NR NR NR NR NR NR 3 months OAB (3)

Marques 2008* Biphasic VIF (variation of 200 10 Hz NR Two transcutaneous 1 month OAB (4) intensity and microseconds electrodes frequency) Peters 2010* (5) NR NR NR NR NR NR 3 months OAB

Preyer 2007* (6) NR NR NR NR NR NR 3 months UUI

Preyer 2015 (7) NR NR NR NR NR NR 3 months OAB

Sancaktar 2010 NR 0.5 – 10Ma; adjusted in 0.2 milliseconds 20 Hz NR NR 3 months OAB (8) accordance to patient’s tolerance Schreiner 2010 Continuou 10 – 50mA; according 200 milliseconds 10 Hz NR Negative electrode on 3 months UUI (9) s mode to sensitivity and hallux medial malleolus and the mobilisation positive electrode was 10cm proximal to this on the right leg along the nerve path Souto 2014 (10) NR NR 250 10 Hz NR Negative surface electrode 3 months OAB microseconds placed behind medial malleolus and positive electrode placed 10cm above it

74 Pulse Shape & Frequency Treatment Author, year Current Current Intensity Duty Cycle Electrodes Target UI Duration (Hz) Duration Vecchioli- NR NR NR NR NR NR 1.5 months OAB Scaldazza 2013 (11) * abstract only. NR = not reported 1. Bellette PO, Rodrigues-Palma PC, Hermann V, Riccetto C, Bigozzi M, Olivares JM. [Posterior tibial nerve stimulation in the management of overactive bladder: a prospective and controlled study]. [Spanish]. Actas Urologicas Espanolas. 2009;33(1):58-63. 2. Finazzi-Agrò E, Petta F, Sciobica F, Pasqualetti P, Musco S, Bove P. Percutaneous tibial nerve stimulation effects on detrusor overactivity incontinence are not due to a placebo effect: a randomized, double-blind, placebo controlled trial. Journal of Urology. 2010;184(5):2001-6. 3. Manriquez VI, Naser ME, Gomez M, Guzman R, Valdevenito R, Lecannelier J, et al. Transcutaneal tibial nerve stimulation versus long release oxibutinin in the treatment of patients with overactive bladder. A randomize control trial (Abstract number 013). International Urogynecology Journal and Pelvic Floor Dysfunction. 2013;24(Suppl 1):S14. 4. Marques A, Herrmann V, Ferreira N, Bellette P. Transcutaneous posterior tibial nerve stimulation in overactive bladder (Abstract number 471). Proceedings of the 38th Annual Meeting of the International Continence Society (ICS), 2008 Oct 20-24, Cairo, Egypt. 2008. 5. Peters K, Carrico D, Perez-Marrero R, Khan A, Wooldridge L, Davis G, et al. 12 week results from the SUMIT trial: percutaneous tibial nerve stimulation vs validated sham in those exposed to pharmacologic therapy (Abstract number 125). Neurourology and Urodynamics. 2010;29(6):988-9. 6. Preyer O, Gabriel B, Mailath-Pokorny M, Doerfler D, Laml T, Umek W, et al. Peripheral tibial neurostimulation (PTNS) versus tolterodine in the treatment of women with urge urinary incontinence and urge symptoms (Abstract number 246). International Urogynecology Journal and Pelvic Floor Dysfunction. 2007;18(Suppl 1):S139- S40. 7. Preyer O, Umek W, Laml T, Bjelic-Radisic V, Gabriel B, Mittlboeck M, et al. Percutaneous tibial nerve stimulation versus tolterodine for overactive bladder in women: a randomised controlled trial. European Journal of Obstetrics, Gynecology, & Reproductive Biology. 2015;191:51-6. 8. Sancaktar M, Ceyhan ST, Akyol I, Muhcu M, Alanbay I, Mutlu EC, et al. The outcome of adding peripheral neuromodulation (stoller afferent neuro-stimulation) to anti- muscarinic therapy in women with severe overactive bladder. Gynecological Endocrinology. 2010;26(10):729-32. 9. Schreiner L, dos Santos TG, Knorst MR, da Silva Filho IG. Randomized trial of transcutaneous tibial nerve stimulation to treat urge urinary incontinence in older women. International Urogynecology Journal. 2010;21(9):1065-70. 10. Souto SC, Reis LO, Palma T, Palma P, Denardi F. Prospective and randomized comparison of electrical stimulation of the posterior tibial nerve versus oxybutynin versus their combination for treatment of women with overactive bladder syndrome. World Journal of Urology. 2014;32(1):179-84. 11. Vecchioli-Scaldazza C, Morosetti C, Berouz A, Giannubilo W, Ferrara V. Solifenacin Succinate versus Percutaneous Tibial Nerve Stimulation in Women with Over- active Bladder Syndrome: Results of a Randomized Controlled Crossover Study. Gynecologic & Obstetric Investigation. 2013;75(4):230-4.

75 Table 20 Summary of data on PTNS vs no active treatment

N Duration Author, year Intervention Study population Outcome** randomised (months) UUI Finazzi-Agrò 2010 Percutaneous PTNS (18) vs Sham 35 UUI 3 Cure: NR (1) percutaneous PTNS (17) Improvement based on diary: 12/17 vs. 0/15 QoL via I-QoL (mean): N = 17, change from pre 69.6 to post 81.3, p = 0.025 vs. N = 15, change from pre 69.5 to post 70.6, p = 0.619 Adverse effects: no serious events DO/OAB Bellette 2009† (2) Transcutaneous PTNS (21) vs Sham 37 OAB 2 Not available† transcutaneous PTNS (16) Marques 2008* (3) Transcutaneous PTNS 43 OAB (some had SUI 1 Cure: NR vs Sham transcutaneous PTNS and UUI) Improvement: NR (number per group not reported) QoL: NR Adverse effects: NR Note: For modality details or parameters, see Table 19. NR = Not reported; I-QoL = Incontinence Quality of Life Questionnaire (higher scores indicate higher QoL); OAB-q = Overactive Bladder Questionnaire (higher scores indicating worse condition); * abstract only **Source of cure and improvement outcome: women’s self-report was given priority but for studies in which it was not reported, quantification of outcomes based on diaries, pad tests or any other definitions chosen by trialists were used as a proxy. † Publication in Spanish with English abstract. English translation of the main text was not available at the time of writing. 1. Finazzi-Agrò E, Petta F, Sciobica F, Pasqualetti P, Musco S, Bove P. Percutaneous tibial nerve stimulation effects on detrusor overactivity incontinence are not due to a placebo effect: a randomized, double-blind, placebo controlled trial. Journal of Urology. 2010;184(5):2001-6. 2. Bellette PO, Rodrigues-Palma PC, Hermann V, Riccetto C, Bigozzi M, Olivares JM. [Posterior tibial nerve stimulation in the management of overactive bladder: a prospective and controlled study]. [Spanish]. Actas Urologicas Espanolas. 2009;33(1):58-63. 3. Marques A, Herrmann V, Ferreira N, Bellette P. Transcutaneous posterior tibial nerve stimulation in overactive bladder (Abstract number 471). Proceedings of the 38th Annual Meeting of the International Continence Society (ICS), 2008 Oct 20-24, Cairo, Egypt. 2008.

76 Questions addressed are: Quality of data  Is PTNS effective in the prevention of UI? One study described using a computer generated randomisation list (149); other studies did not report  Is PTNS better than no active treatment (pla- randomisation techniques (148, 151). No study re- cebo, sham, control or no treatment) for treat- ported information on allocation concealment. Re- ment of UI? sults were reported for everyone who entered the trial  Is one type of PTNS better than another in the in one study (148), and another study reported 5% treatment of UI? attrition rate from PTNS group and 13% from the sham group and that the reasons for drop out from  Is PTNS better than other treatments in the treat- each group were unrelated to the treatment (149), ment of UI? however it was unclear whether the other study had missing data (151).  Does the addition of PTNS to other treatments add any benefit in the treatment of UI? Results  What is the effect of PTNS on other LUTS? (a) UUI. No included studies reported information on cure rates. One study (149) reported that improve-  What factors might affect the outcome of PTNS ment rates were significantly higher for PTNS com- in the treatment of UI? pared with no active treatment (reported p value <0.001). I-Qol scores found a significant difference Eligible interventions were PTNS/neurostimulation/ between the groups which was due to a lack of signif- neuromodulation, percutaneous or transcutaneous. icant change after the sham procedure and to the sig- Eligibility criteria for study participants and outcomes, nificant increase (improved quality of life) after PTNS as well as criteria used to assess ‘risk of bias’ in the (149). No serious adverse effects were reported in ei- included studies, were identical to those used in the ther group but patients in both groups reported occa- previous section on EStim (Section II.4). sional transient pain at the stimulation site (149). Description of intervention in included studies of (b) DO/OAB. There was no information on cure rates. PTNS A total of 11 studies were eligible and are summa- Summary rised in Table 19 (148-158). The three included studies were small (35-43 partici- 5.1. Prevention pants). All were generally assessed as having a high risk of bias. Data pooling was not possible. Data avail- No studies that investigated either primary or second- able from two studies on women with UUI or OAB ary prevention of UI or LUTS were identified. suggests PTNS may be more effective than no active 5.2. Treatment treatment in improving symptoms and quality of life, although no data were available on cure (Level of 5.2.1 Is PTNS Better than No Active Treat- Evidence: 2). The included studies reported no seri- ment (Placebo, Sham, Control or No ous adverse effects associated with either active or Treatment) for Treatment of UI? sham treatment (Level of Evidence: 2). No evidence was available for women with SUI or predominant Three studies compared PTNS with no active treat- SUI. ment (Table 20) (148, 149, 151). One study included women with UUI (149) and two included women with Recommendations OAB where some but not all participants had UI. No For women with UUI or OAB, PTNS may be more ef- study focusing on SUI or predominant SUI was found. fective than no active treatment in symptom control A placebo or a sham procedure was employed as a (Grade of Recommendation: C New). comparator in all three studies. In two studies, the More studies with larger sample sizes and consistent sham procedure was described as having electrodes and clear reporting of core outcomes would be bene- placed without turning on the electrical generator ficial in reaching a conclusion on the effectiveness of (148, 151). In another study (149), the electrical gen- PTNS over no active treatment. erator was turned on only for a few seconds to allow the patient to experience a mild sensation; prior to treatment, patients in both groups had been coun- 5.2.2 Is One Type of PTNS Better than An- selled that the subsequent absence of the sensation other in the Treatment of UI? was due to adaptation. The choice of a different nee- No study was found for this comparison. dle position in the sham group was to eliminate any presumed response that could arise from the acu- 5.2.3 Is PTNS Better than Other Treatments puncture effect from piercing the skin cephalad to the for Treatment of UI? medial malleolus (149). Five studies compared PTNS with drug treatment (Table 21).

77 Table 21 Summary of data on PTNS vs other treatments N Study Author, Duration Comparator random- popula- Source of outcome** year (months) ised tion UUI Preyer Percutaneous 31 UUI 3 Cure: NR 2007* (1) PTNS (16) Improvement: NR vs QoL via unspecified tool (higher scores Tolterodine (15) indicate improvement): mean 4.4 higher, 95% CI 1.7 to 7.1 higher vs mean 4.6 higher, 95% CI 2.1 to 7.0 higher, reported p value = 0.93 Adverse effects: unspecified events noted DO/OAB Manriquez Transcutaneous 56 OAB 3 Cure: NR 2013* (2) PTNS Improvement: NR vs QoL via OAB-q: no significant difference, Oxybutynin no data (number per group Adverse effects: NR not reported) Preyer Percutaneous 36 OAB 3 Cure: NR 2015 (3) PTNS (18) (wet or Improvement: NR vs dry) QoL via VAS, change from baseline, Tolterodine (18) median [range]: N = 18, 1.9 [0 to 8] vs N = 18, 2.7 [0 to 8.5], reported p-value = 0.07 Adverse effects: pain, dry mouth, dizziness Souto Transcutaneous 50 OAB 3 Cure: NR 2014 (4) PTNS (25) Improvement: NR vs QoL via ICIQ-SF, mean [range]: 7.2 [0 to Oxybutynin (25) 18] vs 9.8 [0 to 18] at 3 months (treatment end), 8.3 [0 to 20] vs 13.3 [8 to 20] at 6 months (i.e. 3 month after treatment) QoL via ICIQ-OAB, mean [range]: 5.9 [1 to 11] vs 4.6 [0 to 10] at 3 months, 6.1 [1 to 20] vs 9.2 [4 to 13] at 6 months Adverse effects: NR Vecchioli- Percutaneous 40 OAB 1.5 Cure: NR Scaldazza PTNS (20) (some Improvement: NR 2013† (5) vs had QoL via PGI-I (Wilcoxon test for paired Solifenacin UUI) sample, mean, SD): Succinate (SS) Group A (SSPTNS): post SS 2.9 (1.1), (20) post PTNS 2.1 (0.7); Group B (PTNSSS): post SS 3.1 (1.0), post PTNS 2.3 (0.7) QoL via OAB-q: significant improvement in all groups both with SS and PTNS Adverse effects: NR Note: For modality details or parameters, see Table 19. NR = not reported; ICIQ-SF = International Consultation on Incontinence-Short Form (higher scores indicate increased severity); ICIQ-OAB = International Consultation on Incontinence-OAB (higher scores indicate increased severity); OAB-q = Overactive Bladder Questionnaire (higher scores indicating worse condition); PGI- I = Patient Global Impression of Improvement Questionnaire (lower scores indicate greater improvement); VAS = Global response assessment on visual analogue scale (higher scores indicate greater impact on QoL); * abstract only **Source of outcome: women’s self-report was given priority but for studies in which it was not reported, quantification of outcomes based on diaries, pad tests or any other definitions chosen by trialists was used as a proxy.

78 † Crossover study with a three-month washout period for both groups. 1. Preyer O, Gabriel B, Mailath-Pokorny M, Doerfler D, Laml T, Umek W, et al. Peripheral tibial neurostimulation (PTNS) versus tolterodine in the treatment of women with urge urinary incontinence and urge symptoms (Abstract number 246). International Urogynecology Journal and Pelvic Floor Dysfunction. 2007;18(Suppl 1):S139-S40. 2. Manriquez VI, Naser ME, Gomez M, Guzman R, Valdevenito R, Lecannelier J, et al. Transcutaneal tibial nerve stimulation versus long release oxibutinin in the treatment of patients with overactive bladder. A randomize control trial (Abstract number 013). International Urogynecology Journal and Pelvic Floor Dys- function. 2013;24(Suppl 1):S14. 3. Preyer O, Umek W, Laml T, Bjelic-Radisic V, Gabriel B, Mittlboeck M, et al. Percutaneous tibial nerve stimulation versus tolterodine for overactive bladder in women: a randomised controlled trial. European Journal of Obstetrics, Gynecology, & Reproductive Biology. 2015;191:51-6. 4. Souto SC, Reis LO, Palma T, Palma P, Denardi F. Prospective and randomized comparison of electrical stimulation of the posterior tibial nerve versus oxybutynin versus their combination for treatment of women with overactive bladder syndrome. World Journal of Urology. 2014;32(1):179-84. 5. Vecchioli-Scaldazza C, Morosetti C, Berouz A, Giannubilo W, Ferrara V. Solifenacin Succinate versus Percutaneous Tibial Nerve Stimulation in Women with Overactive Bladder Syndrome: Results of a Ran- domized Controlled Crossover Study. Gynecologic & Obstetric Investigation. 2013;75(4):230-4.

PTNS was compared with solifenacin succinate in Results one (158), tolterodine in two (153, 154) and oxybutynin in two (150, 157). The target population had (a) UUI. There was no information on cure and im- OAB (with UI in some but not all of the participants) provement (153). Quality of life outcomes, measured except for one study (153) where it was UUI. No study using an unspecified tool, improved in both groups, focusing on SUI or predominant SUI was found. One although there was no statistically significant differ- study used a cross-over design (158). Another study ence between them (153). One study reported un- was a three-arm trial comparing PTNS versus oxy- specified adverse events in one of 16 participants butynin versus PTNS and oxybutynin; only the find- and six of 15 participants in the PTNS and Toltero- ings for PTNS vs oxybutynin are reported here (157). dine groups respectively; authors did not specify the nature of these adverse events (153). Quality of data (b) OAB. Data on cure or improvement were not re- One study described using a computer generated se- ported. Quality of life outcomes were reported in four quence using adaptive randomisation, where alloca- studies. Three of these found no significant difference tion was centralised using telephone (154). Permu- between the groups, based on the Overactive Blad- tated blocks were used in one study but it was unclear der Questionnaire in one study (150), OAB-q and the whether there was allocation concealment (150). Patient Global Impression of Improvement (PGI-I) in Other studies did not report such information on allo- another (158) and a visual analogue scale in the third cation concealment nor provided details on the ran- (154). One study using ICIQ-SF and ICIQ-OAB re- domisation process (153, 157, 158). In one study ported that post-treatment scores decreased (indicat- there was a dropout rate of 11% in both groups (154). ing symptom reduction) in both groups at 12 weeks of In another study (158), a two-arm crossover trial of treatment, although the score increased (indicating PTNS and Solifenacin where group A had Solifenacin symptoms worsened) significantly for the oxybutynin then PTNS after a washout period and group B had group in contrast to the PTNS group. The PTNS PTNS then Solifenacin, there was a dropout rate of group maintained their scores at 6 months follow-up 30% from the group A and 20% from group B; the (12 weeks after the cessation of treatment) (157). reasons were refusal of further treatment (10% vs One study reported that adverse events were ob- 5%), no indication for starting treatment (10% vs served in three of 18 (17%) participants in the PTNS 15%). A further 10% dropped out from group A as a group (mainly pain at the puncture site) and nine of result of adverse effects of treatment. Twenty-six per- 18 (50%) participants in the tolterodine group (mainly cent of participants were excluded from analysis in dry mouth and dizziness) (154). one study (157) because they failed to comply with 12 weeks of treatment and/or did not attend follow up at six months; the reasons for non-compliance or missing data were not reported. In one study (150), it was unclear whether results were reported for everyone who entered the trial, while in another (153) the authors did not account for all drop-outs.

79 are reported here (157). No study focusing on SUI or Summary predominant SUI was found. Five studies were included. These were small (36-56 Quality of data participants) and generally assessed as having a high risk of bias. There were limited and widely heteroge- All studies reported adequate methods of randomisa- neous data. Two studies (one for UUI, one for OAB) tion, including the use of a list of random numbers reported that quality of life improved over time for both (155), simple random number generator (156) and PTNS and drug treatment groups with no significant online randomisation (157). Allocation concealment difference between them, while two other studies (for was not mentioned in any of the study reports. In one OAB) found no significant difference in quality of life study, there was 10% missing data from the group between PTNS and drug treatment post-intervention who had tolterodine alone, while there were no miss- (Level of Evidence: 2). However, one study on ing data in the combination therapy group (155). In women with OAB found that the improvement in qual- another study, there was 5% missing data from the ity of life was longer lasting following the cessation of PTNS group only; the reported reason for the attrition treatment with PTNS than after tolterodine (Level of was due to health problems unrelated to therapy Evidence: 2). Data available from two studies on (156). In the third study, 24% dropped out of the oxy- women with UUI or OAB suggests that PTNS is as- butynin group while 16% were excluded from the sociated with a lower rate of adverse events than analysis in the combination therapy group. The rea- tolterodine and when they occurred, there were not sons patients failed to complete the trial were not re- as bothersome as those associated with tolterodine ported and it is unclear whether they were related to (Level of Evidence: 2). No evidence was available treatment (157). for women with SUI or predominant SUI. Results Recommendations (a) UUI. No information was available on cure rates. There is no significant difference between PTNS and One study reported that improvement rates were sta- tolterodine in terms of quality of life, however PTNS tistically significantly higher for the group combining may be considered as both may improve quality of life PTNS with PFMT and bladder training compared with (Grade of Recommendation: B New). PFMT and bladder training alone (156). None of the participants reported significant adverse effects PTNS may considered for women as it is associated (156). ICIQ-SF scores improved for both groups but with fewer and less bothersome adverse effects than the combination treatment group had a significantly those from drug treatment (Grade of Recommenda- greater improvement (156). Similar results were tion: B New). found for KHQ; the combination treatment group Recommendations showed a significantly greater improvement (156). More randomised controlled trials with large sample (b) OAB. No information was available on cure and sizes and clear and consistent reporting of core out- improvement. Two studies reported quality of life outcome data would be beneficial in reaching a firm con- comes. One study using the short form of Inconti- clusion on the effectiveness of PTNS over other treat- nence Impact Questionnaire (IIQ-7) reported that ments. post-treatment scores decreased (indicating less impact of UI on daily living) in both groups but this de- 5.2.4 Does the Addition of PTNS to Other crease in the group combining PTNS and drug was Treatments Add Any Benefit in the greater that the group using drug alone (155). An- Treatment of UI? other study reported that ICIQ-SF scores decreased (indicating symptom reduction) in both groups at 12 Three studies assessed the effect of PTNS as an ad- weeks of treatment, although the score increased (in- junct to another treatment, compared with the other dicating symptoms worsened) significantly for the treatment alone or with PTNS alone (Table 22). One group using drug alone compared with the group study included women with UUI (reported as OAB in- combining PTNS and drug at 6 month follow-up (12 continent) and combined PTNS with PFMT and blad- weeks after the cessation of treatment) (157). ICIQ- der training (PTNS plus PFMT plus bladder training OAB results from the same study found that the com- vs PFMT plus bladder training) (156). Two other stud- bination treatment group had a greater improvement ies included women with OAB (where some but not in OAB symptoms compared with the group using all participants had UI) and combined PTNS with a drug alone at 12 weeks of treatment; ICIQ-OAB drug. One study combined PTNS using the Stoller af- scores increased (worsening in symptoms) for the ferent neuro-stimulation (SANS) protocol with Tolter- group using drug alone in contrast to the other group odine (PTNS plus Tolterodine vs Tolterodine) (155). that maintained their scores at 6-month follow-up The other study was a three-arm trial of PTNS versus (157). No information was available on cure and im- Oxybutynin versus PTNS plus Oxybutynin; only the provement. findings for PTNS plus Oxybutynin versus Oxybutynin

80 Table 22 Summary of data on PTNS + another treatment vs the other treatment

Author, N Study Duration Comparator Outcome** year randomised population (months) UUI

Schreiner Transcutaneous 52 UUI 3 Cure: NR 2010 (1) PTNS + PFMT + Improvement based on diary: 19/25 vs 7/26 bladder training QoL via ICIQ-SF, change from baseline, (26) mean (SD): 7.2 (4.3) vs. 2.6 (3.3), reported vs p-value <0.001 PFMT + bladder QoL via KHQ: reported p-value <0.05 training (26) favouring the combination treatment group; no data Adverse effects: no events noted OAB

Sancaktar SANS 40 OAB 3 Cure: NR 2010 (2) (Percutaneous) + (some had Improvement: NR Tolterodine (20) UUI) QoL via IIQ-7, mean (SD): vs SANS + Tolterodine, change from pre 19.0 Tolterodine (20) (2.0) to post 9.0 (0.8), p <0.001; Tolterodine, change from pre 18.1 (2.5) to post 11.2 (2.7), p <0.001 Adverse effects: dry mouth, constipation, headache, local irritation of puncture site Souto TENS 50 OAB 3 Cure: NR 2014 (3) (Transcutaneous) Improvement: NR + Oxybutynin QoL via ICIQ-SF, mean [range]: 7.9 [0 to (25) 14] vs 9.8 [0 to 18] at 3 months (treatment vs end), 7.4 [0 to 14] vs 13.3 [8 to 20] at 6 Oxybutynin (25) months (i.e. 3 month after treatment) QoL via ICIQ-OAB, mean [range]: 2.9 [0 to 5] vs 4.6 [0 to 10] at 3 months, 3.0 [0 to 5] vs 9.2 [4 to 13] at 6 months Adverse effects: NR Note: For modality details or parameters, see Table 19. NR = not reported; ICIQ-SF = International Consultation on Incontinence-Short Form (higher scores indicate increased severity); ICIQ-OAB = International Consultation on Incontinence-OAB (higher scores indicate increased severity); IIQ-7 = short form of Incontinence Impact Questionnaire (lower scores indicate better QoL); KHQ = King’s Health Questionnaire (higher scores indicate greater impairment); SANS = Stoller Afferent Nerve Stimulation of the posterior tibial nerve; TENS = Transcutaneous Electrical Nerve Stimulation of the posterior tibial nerve; TTNS = Transcutaneous Tibial Nerve Stimulation; **Source of cure and improvement outcome: women’s self-report was given priority but for studies in which it was not reported, quantification of outcomes based on diaries, pad tests or any other definitions chosen by trialists was used as a proxy. 1. Schreiner L, dos Santos TG, Knorst MR, da Silva Filho IG. Randomized trial of transcutaneous tibial nerve stimulation to treat urge urinary incontinence in older women. International Urogynecology Journal. 2010;21(9):1065-70. 2. Sancaktar M, Ceyhan ST, Akyol I, Muhcu M, Alanbay I, Mutlu EC, et al. The outcome of adding peripheral neuromodulation (stoller afferent neuro-stimulation) to anti-muscarinic therapy in women with severe overactive bladder. Gynecological Endocrinology. 2010;26(10):729-32. 3. Souto SC, Reis LO, Palma T, Palma P, Denardi F. Prospective and randomized comparison of electrical stimulation of the posterior tibial nerve versus oxybutynin versus their combination for treatment of women with overactive bladder syndrome. World Journal of Urology. 2014;32(1):179-84.

81 Adverse events were reported by one study comparing PTNS plus tolterodine with tolterodine alone (155). In this study, severe dry mouth, severe constipation, headache and local irritation at puncture site were reported in both treatment groups, and skin irritation for the group using PTNS. Two of 18 participants (11%) experienced more than one adverse event in the tolterodine group compared with one of 20 participants (5%) in the group combining PTNS with tolterodine.

Summary Three small studies were included (40-52 participants). The included studies were assessed as having a high risk of bias. No data were available on cure Figure 5 MStim machine and improvement rates for this comparison. Data from one study suggests that the addition of PTNS to the PFM and sacral nerve roots without insertion of PFMT and bladder training was more effective in im- an anal or vaginal probe (161). For treatment, the in- proving symptoms and quality of life than PFMT and dividual is positioned in a chair. Within the seat is a bladder training alone in women with UUI (Level of magnetic field generator (therapy head) that is pow- Evidence: 2). ered and controlled by an external power unit (Figure Data from two studies suggest that adding PTNS to 5). A concentrated steep gradient magnetic field is di- drug treatment resulted in a greater improvement in rected vertically through the seat of the chair. When quality of life than the drug treatment alone in women seated, the individual’s perineum is centred in the with OAB, and this effect was sustained for a longer middle of the seat, which places the PFM and sphinc- term (6 months) for the treatment with PTNS than the ters directly on the primary axis of the pulsing mag- treatment without PTNS (Level of Evidence: 2). Ad- netic field (Figure 6). Because of their anatomical lo- verse events appear uncommon for either group in cation, it is thought that all tissues of the perineum the same study. No evidence was available for can be penetrated by the magnetic field. According to women with SUI or predominant SUI. Galloway (1999) no electricity, but only magnetic flux enters the body from the device. Goldberg (2000) has Recommendations suggested that, in contrast to electrical current, the PTNS may be considered for symptom control when conduction of magnetic energy is unaffected by tissue impedance, creating a theoretical advantage in its chosen in combination interventions by women with UUI or OAB (Grade of Recommendation: B New). This recommendation should be viewed with caution until these findings are supported or refuted in further trials.

5.3. Other LUTS No trials were identified that analysed the effect of PTNS in women with other LUTS alone, e.g. frequency of voiding, urgency and/or nocturia. 5.4. Factors Affecting Outcome The included studies did not address the effect of factors that could potentially affect the response to treatment with PTNS. A greater discussion of the factors that affect outcome is provided in the section on urinary incontinence in men and women (see section V.1.4).

6. MAGNETIC STIMULATION (MSTIM)

MStim has been developed for “non-invasive” stimulation of both central and peripheral nervous systems (159). MStim for the treatment of UI was reported for the first time in 1999 by Galloway (160). In contrast to EStim, extracorporeal magnetic innervation (more Figure 6 Pulsing magnetic field commonly called magnetic stimulation) stimulates

82 clinical application compared to EStim. Conventional Description of intervention magnetic stimulators deliver, at frequencies of 10 to th 50 Hz, repetitive pulses of current lasting less than In addition to eleven studies included in the 5 ICI, two new trials were identified for this update. The 100 μsec (161) and 275 s (160) in duration. Size number of included studies by dominant type or pat- and strength of the magnetic field is determined by tern of incontinence is summarised in Table 23. Table adjustments of this amplitude by the therapist (160). 24 illustrates the intervention characteristics of the Possible advantages of MStim are that it is performed two new trials. through clothing, needs no probes, skin preparation, or contact with the skin surface. On the other hand, 6.1. Prevention the need for repeated clinic based treatment sessions In the last ICI chapter, there were no trials on preven- is a potential disadvantage. In contrast to EStim, tion of UI or LUTS. No new published trials were MStim lacks portability, although a study by But in found. 2003 (162) reported the development of a portable small electromagnetic device (Pulsegen) for home 6.2. Treatment use that fit into the underwear and was designed for 6.2.1 Is MStim Better Than no Active Treat- continuous use for up to 8 weeks. ment (Placebo, Control or no Treat- The mechanism of action of MStim is not fully under- ment)? stood (163). Some authors have suggested that in SUI stimulation of the PFM causes external sphincter Two new studies including 162 women were identi- contraction (164), acts as a passive PFMT exercise fied (167, 168). In one study the majority of female (165), and increases maximal urethral closure pres- participants had SUI (refractory to first-line manage- sure (162). In UUI, MStim might suppress DO through ment) (167). The other study included all types of UI activation of pudendal nerve afferents blocking para- (168). sympathetic detrusor motor fibres at the spinal reflex Characteristics of the two new studies comparing arc, activation of inhibitory hypogastric sympathetic magnetic stimulation with no active treatment are pre- neurons, or a combination of both mechanisms (166). sented in Table 25. In both studies the no active treat- Stimulation of sympathetic fibres maintaining smooth ment consisted of sham treatment.One study as- muscle tone within the intrinsic urethral sphincter and sessed the efficacy of a magnetic stimulator devel- modulation of pudendal nerve afferent branches stim- oped for home use (168), whereas in the other study ulating an inhibitory spinal reflex at the S3 nerve root, treatment was provided in the outpatient clinic setting are also suggested to play a role in this mechanism (167). Since there was little duplication of MStim in- of action (166). terventions, or sample populations, in the eleven tri- In this section the evidence is considered for the use als, it was inappropriate to combine study findings. of MStim for the prevention and treatment of UI in Quality of data women. Questions addressed are: Allocation concealment was considered to be ade-  Is MStim effective in the prevention of UI? quate in one study that used an independent research assistant in the randomisation process (168).  Is MStim better than no active treatment (placebo, sham, control or no treatment) for the The other study did not describe methods used for treatment of UI? allocation concealment (167). In both studies, data were reported only for those participants who com-  Is one type of MStim better than another in the pleted the study (167, 168). In particular, in the study treatment of UI? by Wallis (2012), 19% (12/62) of participants from the intervention group and 15% (9/60) in the control  Is MStim better than other treatments in the treat- group withdrew from the study (168). ment of UI? Results  Does the addition of MStim to other treatments add any benefit in the treatment of UI? i) SUI. No additional data on cure and improvement were available for this update. The results for cure  What is the effect of MStim on other LUTS? and improvement rates are as reported in two previous studies (169, 170).  What factors might affect the outcome of MStim in the treatment of UI? Information on quality of life was provided in three studies, including one new study (167), but the results Eligible interventions were non-invasive magnetic were not consistent. Two previously reported studies stimulations. Eligibility criteria for study participants found no significant differences between the groups and outcomes, as well as criteria used to assess ‘risk (169, 170), whereas the new study reported better of bias’ in the included studies, were identical to those quality of life at 18 weeks after active treatment com- described in the previous sections on EStim and pared with sham treatment, based on the Urge-Uri- PTNS (Sections II.4 and II.5). nary Distress Inventory and the Overactive Bladder Questionnaire (167).

83 Table 23 Studies of MStim included in the previous review (5th ICI) and current update (6th ICI)

Studies included in the New studies identified in Total previous review (5th ICI) this update (6th ICI) MStim vs No active treatment SUI or predominant SUI 3 1 4 UUI or predominant UUI 1 0 1 MUI 1 0 1 UI all types 1 1 2 DO/OAB (wet or dry) 3 0 3 One type of MStim vs another SUI or predominant SUI 0 No study found 0 UUI or predominant UUI 0 0 UI all types 1 1 MStim vs other treatment No study found No study found 0 MStim+PFMT vs PFMT SUI or predominant SUI 1 No study found 1 UUI or predominant UUI 0 0

Table 24 Summary of data on MStim

Study Intervention description Tsai 2014 (1) A Magstim Rapid2 with a 70-mm figure-8 coil. Treatment administered over the bilateral third sacral roots, with the maximal tolerable intensity, which was typically 70% to 80% of the maximal stimulator output, for 20 minutes each day for 12 consecutive weekdays. The stimulation frequency and the stimulation-on time and - off time were fixed at 5Hz, 10 seconds, and 20 seconds, respectively. Wallis 2012 (2) A commercially available undergarment incorporating 15 static magnets of 800-1200 Gauss arranged anterior, posterior, and inferior to the pelvis. Women were asked to wear it for a minimum of 6 consecutive hours during the day and at least 6 hours overnight for 3 months. 1. Tsai PY, Wang CP, Hsieh CY, Tsai YA, Yeh SC, Chuang TY. Long-term sacral magnetic stimulation for refractory stress urinary incontinence. Archives of Physical Medicine & Rehabilitation. 2014;95(12):2231- 8. 2. Wallis MC, Davies EA, Thalib L, Griffiths S. Pelvic static magnetic stimulation to control urinary incontinence in older women: a randomized controlled trial. Clinical Medicine & Research. 2012;10(1):7-14 Two studies (one new) reported no adverse effects Summary (167, 169). A total of 11 studies assessed the effect of MStim ii) All types of UI. One new study that provided stimu- compared with no active treatment, including four fo- lation using an undergarment with embedded mag- cusing on SUI or predominant SUI, one on UUI, one nets reported that improvement rates were statisti- on MUI, two on all types of UI, and three on OAB/DO cally significantly higher for active compared with no (where only some but not all had UI). treatment, although the result did not hold in sensitivity analysis (168). No information was available re- Findings, using the additional data from newly identi- garding cure rates. fied studies were broadly similar to those in the 5th ICI. With respect to quality of life measures, no statistically significant difference was found based on the For women with SUI, MStim might be more effective Bristol Female Lower Urinary Tract Symptoms than sham in improving (not necessarily curing) (BFLUTS) questionnaire (168). symptoms (Level of Evidence: 2), In the same study, participants reported problems Data from two small trials of MStim (12 and 6 ses- with comfort and wearability of the garment, which sions) examining the effect on quality of life were con- were considered by study authors to have contributed flicting (Level of Evidence: 2). to a relatively high attrition rate of 20%. Commonly cited problems include the attraction to metal objects; For women with UUI, evidence from a small trial (171) embedded magnets sticking to each other and mak- suggests that active MStim might result in better qual- ing it difficult to put on; and also the garment being ity of life than sham (Level of Evidence: 2), although bulky and very warm to wear, especially in the sub- there is some uncertainty surrounding this, as data tropical climate of the study area (168). was limited and no statistical test was performed.

84 Active MStim was associated with higher cure rates 6.4. Factors Affecting Outcome than sham in a small trial (172) with women with MUI None of the included trials addressed the effect of (Level of Evidence: 2), and also higher cure and image, or any other factor, on outcome of MStim. In one provement rates in another small trial (162) with early prospective multi-centre study, factors predict- women with all types of UI (Level of Evidence: 2) but ing success of MStim were included (160). Treatment no such difference was observed for women with DO success was associated with no prior hysterectomy, (wet or dry) (Level of Evidence: 2). In general, ad- no prior anti-incontinence operations, UI symptoms verse effects appear uncommon. for fewer than 10 years, and no use of medications Recommendations known to cause UI. Brodak has suggested that detrusor response to MStim might be better in ‘thin’ indi- No recommendation is possible based on current viduals (presumably due to a shorter distance be- conflicting evidence (Grade of recommendation D). tween the stimulating coil and the sacral nerve roots) and at low bladder volumes (175). Overall, little is 6.2.2 Is One Approach to MStim Better than known about the factors affecting the outcome of Another? MStim. No new study comparing one approach of MStim with another was found for this update. The level of evidence and recommendations remains unchanged 7. SCHEDULED VOIDING REGIMENS from the 5th ICI. Scheduled voiding regimens Summary and Recommendation This section examines the evidence on use of sched- The 5th ICI included one study with women presenting uled voiding regimens in cognitively intact, non-insti- UI symptoms (173). There is insufficient evidence to tutionalized women with UUI, SUI, and MUI and pro- determine if one type of MStim is better than another vides recommendations for their use in clinical prac- (Level of Evidence: 2). No recommendation is pos- tice. A summary of the search strategy and inclu- sible (Grade of Recommendation: D). sion/exclusion criteria for selecting studies for review is provided (see section I). The chapters on the Frail 6.2.3 Is MStim Better than Other Treatments? Elderly and Neurogenic Incontinence provide detailed

th review of scheduled voiding regimens that are used There were no trials for this comparison in the 5 ICI in those with cognitive impairment, or UI secondary to chapter. No new published trials were found for this central nervous system or spinal cord disease (see update. chapter 11). 6.2.4 Does the Addition of MStim to Other Types of Scheduled Voiding Regimens Treatments Add Any Benefit in the Treatment of UI? Bladder training is a broad term often used to describe any type of a scheduled toileting intervention. No new study was found investigating the effect of This has created conceptual confusion in interpreting adding MStim to other treatments compared with the research reports where few details are provided other other treatments alone. The level of evidence and than the statement that bladder training was used. recommendations remains unchanged. The types of scheduled voiding regimens can be cat- egorised as: bladder training, timed voiding, habit Summary and Recommendation training, and prompted voiding (176). Although these regimens share a common feature of a toileting The previous 5th ICI included one study in women with schedule, they differ on the basis of adjustments to SUI (174). The addition of MStim to PFMT does not the voiding schedule, the active or passive involve- appear to be beneficial (Level of Evidence: 2). ment of the patient, the nature of patient education Adding MStim to PFMT does not appear to be bene- including the teaching of strategies to control urgency ficial (Grade of Recommendation: C). This hypoth- and prevent stress leakage, the use of reinforcement esis needs to be investigated further with high quality techniques, and the nature of the interactions be- trials, if it is a clinical question of interest to women. tween clinicians and patients. In practice, however, scheduled voiding regimens may share aspects of 6.3. Other LUTS one or more of these features. No trials were identified that analysed the effect of MStim in women with other LUTS alone, i.e., frequency of voiding, urgency and/or nocturia.

85 Table 25 Summary of data on MStim vs no active treatment N Study Duration Author Comparator random- Outcome** population (months) ised Predominantly SUI Tsai 2014 (1) MStim (20) 40 >50% of outcome Cure: NR vs study measured Improvement: NR Sham MStim sample at 4.5 QoL: OQB-q: N = 30, mean difference (20) had SUI months 23.80 lower, 95% CI 36.28 to 11.32 alone lower QoL via Urge-UDI: N = 30, mean difference 1.80 lower, 95% CI 2.64 to 0.96 lowerAdverse effects: no events observed All types of UI Wallis 2012 MStim (62) 122 SUI, UUI or 3 Cure: NR (2) vs MUI Self-reported improvement: 26/46 vs Sham MStim 18/50 (60) QoL via BFLUTS (median change in score, IQR): UI symptoms, N = 101, 1.5 (3.0) vs 1.5 (3.0), p = 0.80; Sexual function, N = 101, 0 (0.75) vs 0 (0), p = 0.80; Quality of life, N = 101, 1.0 (5.0) vs 0 (5.0), p = 0.28 Adverse effects: discomfort of the garment Note: For modality details or parameters, see Table 24. NR = not reported; BFLUTS = Bristol Female Lower Urinary Tract Symptoms (higher scores indicate more severe comlaints); OAB-q = Overactive Bladder Questionnaire (higher scores indicating worse condition); Urge-UDI = Urge-Urinary Distress Inventory (higher scores indicate worse condition); **Source of cure and improvement outcome: women’s self-report was given priority but for studies in which it was not reported, quantification of outcomes based on diaries, pad tests or any other definitions chosen by trialists was used as a proxy. 1. Tsai PY, Wang CP, Hsieh CY, Tsai YA, Yeh SC, Chuang TY. Long-term sacral magnetic stimulation for refractory stress urinary incontinence. Archives of Physical Medicine & Rehabilitation. 2014;95(12):2231- 8. 2. Wallis MC, Davies EA, Thalib L, Griffiths S. Pelvic static magnetic stimulation to control urinary incontinence in older women: a randomized controlled trial. Clinical Medicine & Research. 2012;10(1):7-14.

Bladder training increasing the “reserve capacity” of the lower urinary tract system (177-179). Bladder training (also referred to as bladder drill, bladder discipline, bladder re-education, and bladder re- Timed voiding training) involves a programme of patient education along with a scheduled voiding regimen with gradu- Timed voiding is a fixed voiding schedule that re- ally progressive voiding intervals. Specific goals of mains unchanged over the course of treatment (176). bladder training are to correct faulty habit patterns of The goal is to prevent UI by providing regular oppor- frequent urination (if present), improve control over tunities for bladder emptying prior to exceeding blad- bladder urgency, prolong voiding intervals, increase der capacity. Timed voiding has been recommended bladder capacity, reduce incontinent episodes, and for patients who cannot participate in independent toi- restore patient confidence in controlling bladder func- leting (180). It also has applicability for use in outpa- tion. The underlying mechanism of how bladder train- tient settings with incontinent women who have infre- ing achieves its effect is poorly understood. Several quent or irregular voiding patterns (181) and men who hypotheses have been proposed including improved are independent in their voiding function (182). cortical inhibition over detrusor contractions; im- Habit training proved cortical facilitation over urethral closure during bladder filling; improved central modulation of affer- Habit training is a toileting schedule matched to the ent sensory impulses; altered behaviour resulting individual’s voiding pattern based on their voiding di- from better individual awareness of the lower urinary ary. The toileting schedule is assigned to fit a time in- tract function and circumstances that cause UI, and terval that is shorter than the person’s normal voiding

86 pattern and precedes the time when incontinent epi- 7.1. Prevention sodes are expected. Thus, the voiding interval may be lengthened or shortened throughout the day de- Previously no trials had examined scheduled voiding pending on the patient’s voiding pattern with the goal regimens as a sole intervention in the prevention of to pre-empt UI. Habit training is usually implemented UI. In this update, no further trial was identified. by caregivers in institutional settings with cognitively 7.2. Treatment and/or physically impaired adults, but it has also been tested in homebound older adults (183). It is poten- Previous recommendations related to bladder train- tially useful for adults without cognitive or physical im- ing as treatment for UI were based on review of indi- pairment, who have a consistent pattern of UI (181). vidual published trials and three systematic reviews that provided descriptive synthesis with evidence Prompted voiding grading (180, 185, 186), including a published Prompted voiding refers to a caregiver education pro- Cochrane review last updated in 2006 (185). In this gramme in combination with a scheduled voiding reg- ICI update, 5 additional trials were identified and are imen, typically every two hours. It is used to teach described in the relevant sections below (98, 102, people with or without cognitive impairment to initiate 156, 187, 188). their own toileting through requests for help and pos- 7.2.1 What Is the Most Appropriate Bladder itive reinforcement from caregivers when they do so Training (BT) Protocol? (184). It has been used primarily in institutionalized settings with cognitively and physically impaired older Previously no trials were identified that compared two adults. or more methods of bladder training (BT), and none were identified for this update. In the absence of trials Section Overview comparing two or more approaches, a content analy- This section examines evidence for the use of blad- sis was performed looking at the protocols used in tri- der training and timed voiding for the prevention and als investigating the effects of BT. It has been up- treatment of UI in non-institutionalised women of all dated for this edition. Twenty-two trials on BT involv- ages without cognitive or mobility impairment. How- ing a total of 3194 women have been identified, in- ever, the majority of evidence available pertains to the cluding 5 trials involving 732 women added since the effects of bladder training, thus this is the focus of this last update (98, 102, 156, 187, 188). Several of the section. trials previously reviewed (189-194) and 2 of the new trials identified for this update (156, 187) provided no Previously, no trials were identified that tested habit or minimal details regarding the specific BT protocol training or prompted voiding for UI in independent-liv- used. In trials that did provide some description, BT ing women without cognitive or physical impairment, protocols were implemented in differing ways. In the and no new trials meeting these criteria were identi- following review of BT protocols, information from fied in this update. Therefore, evidence for use of studies added in this update are integrated with infor- habit training and prompted voiding is not included in mation from previously reviewed studies. this section. All protocols involved some type of patient education, Questions addressed in this section include: namely:  Can scheduled voiding regimens prevent UI?  Brief verbal (191, 192) or written instructions (195, 196)  What is the most appropriate bladder training protocol?  Verbal, written, and audio-visual instruction (177,  Is bladder training better than no treatment, pla- 179, 197) cebo or control treatments?  Introduction to an individual who successfully completed BT (193)  Is bladder training better than other treatments? If specified, the education was provided by nurses  Can any other treatment be added to bladder (177, 179, 189, 190, 198, 199), general practitioners training to add benefit? (200), or physiotherapists (98, 102, 188).  Does the addition of bladder training to other Scheduling of voids varied in the following ways: treatments add any benefit?  Assignment of the initial voiding interval varied  Is timed voiding effective at treating UI? from 30 minutes to two hours, with one hour be-  What is the effect of bladder training on other ing the most common interval based upon the LUTS? participant’s voiding pattern or 30 minutes beyond the participant’s average (195, 201) or  What factors might affect the outcome of bladder longest (102) voiding interval. training?  Adjustments to the voiding interval varied from 15 to 30 minutes, with 30 minutes the most com-

87 mon interval. Increases were made daily for in- Typical Bladder Training Regimen patient regimens (193), after 48 hours of dryness (202), every four to five days (201), or weekly if  An initial voiding interval typically beginning at schedule was well-tolerated (102, 177, 179). one hour during waking hours, which is increased by 15 to 30 minutes per week depending  Goals for optimal voiding interval varied from on tolerance of the schedule (such as fewer in- three to four hours. continent episodes than the previous week, minimal interruptions to the schedule, and the pa-  Voiding was ‘mandatory’ with restriction of void- tient’s feeling of control over urgency and confi- ing in between assigned toileting times even if UI dence to expand the voiding interval), until a two occurred (193), a scheduled voiding regimen to three hour voiding interval is achieved. A that allowed interruptions in the schedule if ur- shorter initial voiding interval, e.g. 30 minutes or gency became unbearable (177, 179, 199), or less, may be necessary for patient whose base- self-scheduling of voiding with a target goal to line micturition patterns reveal an average day- reach (195). time voiding interval of less than one hour.  Voids were not scheduled (allowed) during  Education about normal bladder control and sleeping hours (193); none of the other protocols methods to control urgency such as distraction identified how voids were handled during sleep- and relaxation techniques and PFM contraction. ing hours.  Self-monitoring of voiding behaviour using dia- In some protocols the scheduled voiding regimen was ries or logs in order to determine adherence to supplemented by specific strategies to control ur- the schedule, enhance self-awareness, evaluate gency and/or stress leakage, including distraction progress, and determine whether the voiding in- and relaxation (102, 177, 179, 188, 195, 199), and terval should be changed (for example, see Fig- pelvic floor muscle contraction (102, 177, 190, 196). ure 7). In other studies there was encouragement to suppress urge but it was not clear what strategies were  A supervising health care professional to monitor used (192, 200, 203). Feedback techniques included progress, suggest adjustments to the voiding in- self-monitoring (177, 179, 191, 194, 199), goal setting terval, and provide positive reinforcement to pa- with feedback of progress (198), and positive rein- tient undergoing BT at least weekly during the forcement (177, 179, 201). training period. Several protocols included use of adjunctive treat-  If there is no improvement after three weeks of ments: BT, re-evaluation is warranted and other treatment options would be considered. Inpatient BT  Fluid and caffeine adjustments (197, 199, 203) programmes may follow a more rigid scheduling  Fluids allowed up to a certain level (1,500 ml) regimen with progression of the voiding interval (202) on a daily basis.

 No fluid modifications (102, 177, 179, 195, 200) Summary  Advice on constipation prevention (199) There is still no trial evidence to suggest the most effective method or specific BT parameters. For those Both in and outpatient BT programmes have been undertaking BT it is likely that more health profes- used. Early inpatient BT programmes involved five to sional contact will be better than less, based on the 13 days of hospitalisation to ensure strict protocol ad- developing evidence for PFMT, which like BT, re- herence (193). Outpatient programmes are more quires behavioural change (Level of Evidence: 4). commonly described and the amount of health pro- The literature suggests several areas that could be fessional contact ranged from weekly visits for six investigated in future trials, including the instructional weeks with fortnightly telephone calls for six addi- approach, supervisory intensity, strategies for con- tional weeks (177), to weekly visits (179, 188, 189), trolling urgency, scheduling parameters, frequency of fortnightly visits (102, 201), and monthly visits (202). schedule adjustments, length of treatment, and use A “simplified” BT treatment with minimal to no health of adjunctive treatments. professional contact (instructions given to patients on a one page instruction sheet) has also been tested Recommendations (195, 196). Clinicians and researchers are advised to refer to the Overall, there is a lack of consistency in BT protocols. operant conditioning and educational literature to pro- Based on the protocols described, a reasonable out- vide a rationale for their choice of their approach to patient BT protocol, based on expert opinion is shown BT. The ICS Consensus statement and review pain Figure 7.

88 pers on Adherence and PFMT could be useful to cli- Summary nicians and researchers (24-27). (Grade of Recom- mendation: D). The few available trials (reviewed previously) were small and of variable quality. No new trials were iden- Clinicians should provide the most intensive BT su- tified in this update. pervision that is possible within service constraints (Grade of Recommendation: D). There is limited Level 1 evidence that BT may be an effective treatment for women with UUI, SUI, and MUI More research is needed to investigate which BT pa- (Level of Evidence: 1). rameters, supervisory intensity, and adjunctive treatments are most effective. Future trials should include Recommendations outcomes that matter to patients, including the length and frequency of supervisory contact. BT should be recommended as first line conservative therapy for UI in women (Grade of Recommenda- 7.2.2 Is BT Better than No Treatment, Pla- tion: A). cebo or Control Treatments? Additional high quality studies are needed that examine the effect of BT versus no treatment in treatment BT as sole therapy has been used in the treatment of of women with UUI, SUI, and MUI. DO, urodynamic SUI, MUI, UUI, UUI with a stable bladder, and OAB (also called urgency-frequency syndrome). 7.2.3 Is BT Better than Other Treatments? No new trials were identified that addressed this This section, considered trials which compared BT question. Previously 5 RCTs reporting on 515 women alone versus another active therapy. were identified that compared the effect of BT to no For the comparison of BT versus PFMT see II.2.3.3. treatment or control (179, 189, 194, 199, 200). Of 4 trials with relevant analysable data (179, 189, 194, The only other comparison for which trials were found 200), 3 reported improvements in the BT group com- was BT versus drug therapy. One additional trial was pared to the control group (179, 194, 200). The trial identified for this update (98). Previously, two small quality and detailed results were presented in previ- trials were identified that randomized 131 women with ous editions of this chapter. UUI to BT or drug treatment (193, 201). Kafri et al. (98) randomized 184 women with UUI to one of four groups: BT, PFMT, tolterodine extended release 4 mg or combined behavioural therapy (BT and PFMT). Results from the BT and drug treatment group comparison are included here, while results of

Figure 7 Bladder diary

89 the BT vs PFMT comparison are included in section older women undergoing “routines of care” which in- II.2.3.3. cluded BT and Kegel exercises (156). As BT was combined with pelvic floor exercise treatment in both Quality of data groups and no details were provided about the BT Kafri et al. (98, 201) was a single-blinded trial in which protocol, it was not possible to determine the effects 83 women were randomized to drug treatment or BT of BT alone, and it is not considered further here. The and followed up for 12 months. A power calculation third trial tested the addition of oxybutynin, imipra- and adequate random allocation concealment were mine or placebo to BT in 282 men and women with reported. Drug tolerability and adverse events were DO who had failed at least 8 weeks treatment of BT assessed, but side effects were reported in a compo- alone (187). However, very limited information is site outcome only. Intent-to-treat principles were fol- available on the methods and results (abstract only). lowed. Drop out at 3 months was higher in the drug This trial was also not considered further. treatment compared to BT group (36% vs 5%) (98). Prior editions of this chapter identified trials address- Results ing the additional benefit of caffeine reduction, PFMT, and drug therapy when combined with BT. The caf- In Kafri (98), the number of self-reported UUI epi- feine reduction trial was included in the Lifestyle In- sodes in the past week significantly decreased at 3 terventions (II.1) section (203), and the trial address- and 12 months in both the BT and drug treatment ing additional benefit of PFMT when added to BT was groups. Between group comparisons suggested no included in the PFMT section (II.2.3.4) (177). Three difference in outcomes. Differential dropout sug- trials previously identified tested the added benefit of gested drug treatment was associated with more ad- drug therapy to BT treatment. One of these tested a verse effects, but this was not clearly reported. These drug that is no longer available (terodiline), and is not results were similar to one previously reviewed trial in considered further (192). The remaining trials in- short-term outcomes (201), but in that trial symptoms cluded one that randomized 34 women and men with recurred more often after drug treatment than after BT DO to BT or BT plus imipramine therapy (202), and at 6 months follow-up. Another older trial found BT another that randomized 60 women and men to BT superior to drug treatment at 3 months follow-up and placebo or BT plus immediate release oxybutynin (193). Both previously reviewed trials found adverse (191). No difference was seen in incontinence out- effects common in the drug treatment groups, but not comes between the groups in either trial, although in BT groups. subjective improvement was greater with addition of a drug in one (191). Data quality and detailed results Summary from these trials were reviewed in past editions of this chapter. Despite an additional trial, it remains unclear whether BT or drug therapy is more effective for women with Summary DO or UUI (Level of Evidence: 1). No new trials contributed evidence in this area. In two This result is consistent with the findings of the small trials (reviewed in previous editions) comparing Cochrane review (185), which concluded that there BT (or BT plus placebo) versus BT plus drug in DO, was not enough evidence to determine whether first there was a suggestion that the effect of BT might be line therapy should be BT or anticholinergic drugs. enhanced by active drug (Level of Evidence: 2). Recommendations However, both trials were small, conducted in gender-mixed sample populations, and outcomes were When considering BT and anticholinergic drug for not common to both trials. Thus, there is insufficient women with DO or UUI, either may be effective evidence to derive a conclusion related to the effec- (Grade of Recommendation: B). tiveness of augmenting BT with drug therapy. BT may be preferred by women and clinicians be- Recommendations cause it is not associated with the drug related side effects (Grade of Recommendation: D). Direct comparisons of BT versus BT with drug for UI treatment are needed to address the question of 7.2.4 Can Any Other Treatment Be Added to whether the effect of BT can be augmented by drug BT to Add Benefit? therapy. No recommendation is possible (Grade of recommendation D). To be included here, trials needed to investigate the effects of BT versus BT plus therapy A to address the 7.2.5 Does the Addition of BT to Other Treat- additive benefit of therapy Three new trials were iden- ment Add Benefit? tified that tested the additional benefit of PFMT, drug therapy and tibial nerve transcutaneous electrical To be included here, trials needed to investigate the stimulation when added to BT (102, 156, 187). One effects of Therapy A versus Therapy A plus BT to as- trial addressed the added benefit of PFMT to BT (102) sess the added benefit of BT over Therapy A alone. and is considered in the section on PFMT (II.2.3.4). A A search for trials that investigated the effects of 2010 randomized trial tested the additive effects of PFMT alone versus PFMT plus BT, and drug therapy tibial nerve transcutaneous electrical stimulation in 51

90 alone versus drug therapy plus BT was performed. these, a small, double-blind crossover study com- No new trials focused on this question were identified. pared timed voiding plus anticholinergic drug therapy (terodiline) to timed voiding plus placebo (205). As Three trials were reviewed in previous editions of this terodiline is no longer available, this study was not chapter, including one older study that compared BT, considered further. The second study was a small PFMT and combination therapy in 204 community- case series of 20 women with mild UI treated with dwelling women with SUI and/or DO (177). Two larger timed voiding and followed from 6 weeks to 8 months RCTs compared a “simplified” BT (administered by after treatment (206). A successful outcome (not ob- providing participants with a 1 page written instruc- jectively quantified) was reported in 79% of partici- tions on BT) plus drug therapy (tolterodine (2mg twice pants. daily) (190, 195) and solifenacin (5/10mg daily)(195)) to drug therapy alone in women and men with OAB Summary with or without UUI. UI outcomes were secondary endpoints in these trials. There are no RCTs, or high quality observational studies, providing evidence on the effects of timed Summary voiding for UI in cognitively intact, community-dwelling women, and no new trials considering this were A single trial found combining BT with PFMT im- identified. Based upon the data from one small un- proved short-term outcomes compared to PFMT controlled study, it seems a two-hour timed voiding alone, but the added benefit did not persist three schedule may be beneficial in treating women with months later (Level of Evidence: 2). mild UI, infrequent voiding patterns, and stable blad- There is no evidence for an added benefit of combin- der function (Level of Evidence: 3). ing brief written BT instructions with tolterodine (2mg Recommendations twice daily) or solifenacin (5/10mg daily) compared to drug therapy alone for urgency incontinence (Level Timed voiding with a two-hour voiding interval may be of Evidence: 2), although these trials were likely un- considered as a sole intervention for women with mild derpowered to study UI outcomes. UI or infrequent voiding patterns (Grade of Recom- mendation: C) Timed voiding may also be consid- Recommendations ered as an adjunct to other treatment. With no new trials contributing evidence in this update, the available evidence (single trial) supports the 7.3. Other LUTS use of BT as a supplement to PFMT in improving Two new trials were identified that reported frequency short-term outcomes, but not longer-term results outcomes. Kafri et al. (98) compared BT to drug treat- (Grade of Recommendation: C). ment and reported on frequency outcomes (others re- Limited evidence suggests the addition of written in- viewed in section II.7.2 above). A second trial ad- formation on BT for women with OAB taking an anti- dressed the added benefit of PFMT to BT (102) and muscarinic drug does not further improve UI (Grade reported frequency as a secondary outcome. This of Recommendation: B). trial is considered in the section on PFMT (II.2.3.4). More research is needed using an appropriately su- Older trials (from past editions of this chapter), which pervised BT programme combined with drug therapy also contributed evidence related to other LUTS out- versus drug alone. comes, compared BT to drug therapy (190, 193, 201), BT to no treatment (179, 189, 194), BT plus placebo 7.2.6 Is Timed Voiding Effective at Treating to BT and drug treatment (191), and additive effects UI? of a simplified BT program to drug treatment (195, 196). The Cochrane review on timed voiding for management of UI in adults was last updated in 2010 (204). Quality of data Ostaszkiewicz (2010) considered randomized and The quality of data in the new trial that tested BT vs quasi-randomised trials only and identified two trials drug treatment was described earlier (II.VII.2c). comparing timed voiding combined with additional interventions (including medications) to usual care. Results Both trials were conducted in nursing homes and i) Urgency: No new trials reported on urgency out- most participants were elderly women with cognitive comes. In previously reviewed trials, urgency results impairment. Neither study recruited participants that were conflicting. An older trial suggested BT was su- met criteria for inclusion here. perior to drug treatment (193), but another reported No new trials were identified that tested timed voiding greater improvement in urgency after drug treatment for UI in women without cognitive or physical disabil- or combined drug and BT treatment compared to BT ity. Previously, two older non-randomised studies, ex- alone (190). Two larger trials found no additional im- cluded from the Cochrane review, reported findings provement in urgency when a simplified BT treatment on the effects of timed voiding in women with UUI, was added to drug treatment for OAB (195, 196). stable bladders with UUI, and MUI (205, 206). One of

91 ii) Daytime (diurnal) frequency: Recommendations A newly identified trial found frequency improved at 3 Insufficient evidence exists to support either the use and 12 months after treatment with BT, drug treat- of BT to treat urgency in women with UI and/or OAB, ment (tolterodine extended release 4 mg), or a com- or to guide the choice of BT vs drug treatment as ini- bined BT and PFMT group (98). On average, subjects tial treatment aimed at urgency (Grade of Recom- reported 2 to 4 fewer voids per 24 hours after treat- mendation: D). ment. Between group differences were not reported. Routine addition of a “simplified” BT to drug therapy In previously reviewed trials, frequency improved to a does not provide additional improvement in urgency, greater extent in BT groups compared to no treatment but should be considered in treatment of voiding fre- (179, 189, 194). In trials testing BT compared to drug quency in women with UI and OAB (Grade of Rec- treatment, frequency improved in both groups simi- ommendation: B). larly (190, 201), or to a greater extent in the BT group (193). In one trial testing the additive effect of drug BT should be offered as treatment for urinary fre- treatment to BT, a greater reduction in frequency was quency and nocturia (Grade of recommendation: reported in the BT plus drug treatment group com- B), but in choosing between BT and an anticholiner- pared to BT plus placebo group (191). Lastly, two tri- gic drug for women with frequency, either may be ef- als found that “simplified” BT significantly augmented fective (Grade of Recommendation: B). the effect of drug alone. BT with tolterodine had greater improvement in voiding frequency compared 7.4. Factors Affecting Outcome to drug alone (33% versus 25% improvement, re- Few trials on BT have examined predictors of treat- spectively; p<0.001) and BT with solifenacin reduced ment response. New trials reviewed for this update the number of voids in 24 hours to a greater extent did not contribute additional information related to compared to drug alone (2.8 versus 2.1 fewer voids predictors of timed voiding or BT treatment response. in 24 hours; p<0.001) (195, 196). iii) Nocturia: No new trials were identified reporting on 8. COMPLEMENTARY AND nocturia. Three previously reviewed trials reported ALTERNATIVE MEDICINES data in comparisons of BT with no treatment, and found reductions in nocturia after BT but not in the There is limited, but growing evidence that comple- control group (179, 189, 194). Four previously re- mentary and alternative medicines (CAMs) may influ- viewed trials compared BT with drug therapy. Two ence physiological function and/or health outcomes. smaller trials found BT superior compared to drug CAMs include those therapies that are not part of the treatment in treating nocturia (193, 201). Another trial traditional biomedical model, such as meditation, im- compared BT plus placebo versus BT plus drug (191) agery, hypnosis, acupuncture and naturopathic and and found no difference in nocturnal micturition fre- herbal remedies. quency. Song (2006) reported that nocturia improved similarly in women with OAB treated with BT (56.1%), This section reviews the current evidence for the ef- tolterodine (65.4%), and combined treatment (66.3%) fects of CAMs on UI in cognitively intact, community- (190). dwelling women. Studies focused on UI in women with neurogenic aetiologies (for example, UI associ- Summary ated with stroke or multiple sclerosis) have been excluded. A summary of the search strategy used for No trials were identified which tested the effective- selecting studies for review is provided in section I. ness of BT compared to no treatment or a control for urgency. It remains unclear whether BT or drug treat- In the 5th ICI, a single RCT focused on hand acupunc- ment is more effective in treating urgency (Level of ture therapy was reviewed here. Given the lack of ev- Evidence: 2). idence, no recommendations related to CAMs were made for women with UI. Since then, additional RCTs Several trials, including one newer study, suggest BT and a Cochrane review have been published. Many is effective at improving frequency, (Level of Evi- of these trials tested various acupuncture techniques dence: 1), but it is unclear whether BT or drug treat- for UI; these are reviewed and summarized below. ment is more effective (Level of Evidence: 2). Two larger, higher quality trials show that the addition of Small individual trials testing other types of CAM ther- “simplified” BT to drug treatment does not provide ad- apy (for example, meditation and relaxation therapies ditional benefit in treating urgency, but does improve and natural supplements) for UI were identified in the frequency (Level of Evidence: 1). A few small, ran- current search, many labelled as pilot trials. As rec- domized trials of variable quality suggest BT is effec- ommendations cannot be made based on the minimal tive at treating nocturia (Level of Evidence: 2). Addi- evidence available for any individual type of CAM tional small trials do not find that BT (or BT added to therapy, these are not formally reviewed here. As in- drug treatment) is more effective at treatment of noc- terest in CAM therapies grows and more evidence ac- turia than drug treatment, but evidence is limited crues, additional types of non-traditional therapies (Level of Evidence: 2). may be reviewed in future editions.

92 Acupuncture and UI (<5%). Adverse effects were reported in 1 reviewed trial (207). Acupuncture and related techniques have been performed as part of Eastern medicine for thousands of Results years. Related therapies may include body acupuncture, hand acupuncture, electroacupuncture and acu- More women reported improvement in the acupunc- pressure, amongst others. Their mechanisms of ac- ture group (73% vs 33%; risk ratio 2.2 (95% CI 1.3- tion are not fully understood, but discussions of these 3.8)) compared to drug treatment (207). Cure rates mechanisms related to treatment of UI from tradi- were low and not different between the groups (13% tional Chinese medicine and other perspectives are vs 7%; risk ratio 2.0 (95% CI 0.4, 10.1)). Adverse ef- available (207, 208). fects were reported in the drug group only (including headache, dizziness, and thirst). 8.1. Prevention In the trial comparing acupressure, sham acupres- In the 5th ICI, no trials examined acupuncture for the sure vs PFMT alone, numbers of incontinence epi- prevention of UI. In this update, no new trials were sodes were unchanged and did not differ between identified. groups, although self-reported severity of leakage improved to a greater extent in the acupressure com- 8.2. Treatment pared to the sham acupressure group (p=0.04) and 8.2.1 Acupuncture for Treatment of SUI to PFMT alone (p=0.01) (211). Adverse effects were not reported. Previously, one low-quality RCT that studied the effect of hand acupuncture vs control (no treatment) for Summary female SUI in 52 women (209) was identified. In this update, a recently-published Cochrane review of ac- A single trial suggests acupuncture may be better upuncture and SUI was reviewed (207), which in- than drug treatment in SUI (Level of Evidence: 2), cluded one trial of acupuncture vs drug treatment for but the low quality and atypical comparator group lim- SUI (210). An additional trial of acupressure for the its the impact of these results. treatment of is included in this update (211). A single trial suggests mixed results for acupressure The 2013 Cochrane review considered evidence for compared to sham treatment or usual care in treat- acupuncture interventions from traditional Chinese ment of SUI (Level of Evidence: 2), but the low qual- medicine (body acupuncture, scalp acupuncture, ity and small numbers limit these findings. There is electroacupuncture, warm acupuncture, fire needle inadequate evidence to know if acupuncture or acu- and elongated needle), and did not consider trials of pressure treatments are effective for SUI in women. other interventions, such as hand acupuncture and Recommendations acupressure (207). The authors searched and reviewed English and Chinese language publications. No recommendations can be made regarding the use Only 1 trial was included (many were excluded be- of acupuncture or acupressure for SUI in women. cause they combined acupuncture with other treat- High-quality RCTs are needed. ments or compared different types of acupuncture). The single eligible trial randomized 60 women to ac- 8.2.2 Acupuncture for Treatment of OAB, UUI upuncture vs a drug treatment (midrodine) and re- and Mixed UI ported subjective improvement and cure and adverse effects at 6 to 12 weeks (207). In this update, two recently-published reviews were identified which focused on acupuncture for women Chang et al. completed an RCT of 81 women with with OAB, UUI or Mixed UI, as well as an additional urodynamic SUI, randomized to 3 groups including trial (not included in the reviews) (208, 212, 213). acupressure, sham acupressure and “usual care” including PFMT (taught at baseline) (211). The treat- Paik et al. (208) performed a systematic review of ment and sham groups underwent 3 weekly treat- RCTs published in English or Korean language jour- ment sessions for 10 weeks. No information was pro- nals testing acupuncture and acupressure treatments vided on performance of PMFT (or other treatments) for UI. Four small RCTs were found eligible for that in the “usual care” group (or the other treatment review. One of these studied acupressure for SUI and groups). Outcomes were assessed after the 10-week is discussed in the section above (211), but 3 are con- treatment period. The primary outcome was pelvic sidered in this section, all published before 2010 floor muscle strength. Numbers of urine leak epi- (214-216). A second review focused on acupuncture sodes from a 4-day diary and a subjective assess- and OAB in non-comparative and comparative trials ment of urine incontinence severity were also as- (212). This review identified 4 RCTs relevant to this sessed. section (comparative trials of acupuncture for women with UI), 3 of which were also included in the review Quality of data by Paik et al. (214-217). Thus, in total 5 RCTs (four included in the recently published reviews and 1 ad- Randomization concealment was inadequate in both ditional trial) are included in this edition. Study popu- trials (207, 211). No blinding or partial blinding was lations included women with OAB and UUI (214, performed. Few participants withdrew from the trials

93 215), OAB with or without UUI (217), MUI and UUI (214). In the trial comparing combination treatment (216) and MUI (213). with electroacupuncture and tolterodine to electroacupuncture alone for MUI, both groups had signifi- Two trials compared acupuncture to sham or placebo cant reductions in overall UI episodes and in urine acupuncture (215, 216). One was a pilot RCT which leakage volume on pad test, but these outcomes did tested acupuncture vs sham acupuncture (sham nee- not differ between groups. Significantly more women dles which did’nt puncture the skin were used) in 9 in the combination treatment group had a more than women with mixed or urgency UI (216). The second 50% improvement in number of UI episodes (76 vs trial randomised 74 women with OAB including UUI 59%, p<0.05) (213). to acupuncture vs acupuncture performed at relaxation points (215). Two trials compared acupuncture to No serious adverse effects occurred. Minor bruising drug treatment, including an older trial that tested ac- and bleeding as well as some discomfort with needle upuncture vs oxybutynin 5 mg twice daily in 39 placement were described (214-217). women with OAB with UUI (214). A recent, larger trial randomized 272 women with OAB with or without UUI Summary to acupuncture or tolterodine 2 mg twice daily (217). Last, Jin et al. studied combination therapy with elec- Two trials (small to medium size) of varying quality troacupuncture plus tolterodine 2 mg twice daily ver- found no or limited benefit in testing acupuncture to sus electroacupuncture alone in 71 women with MUI various sham treatments for MUI and OAB\UUI in (213). women (Level of Evidence: 2). Acupuncture techniques and regimens differed Several factors limit these results, including small widely among the studies, including different acu- sample size, risk of bias and heterogeneity of active puncture points treated. Treatments occurred 1 to 3 and sham treatments. Two trials comparing acupunc- times weekly, lasted 4 to 8 weeks and outcomes were ture to drug therapy in women with OAB, UUI and assessed immediately after treatment or 2 to 12 MUI found no difference between treatments (Level weeks later. of Evidence: 2). One low quality trial suggested combination therapy of electroacupuncture plus drug for Quality of data MUI was superior to electroacupuncture alone (Level of Evidence: 2). Randomization concealment was considered adequate in 2 of 5 trials (215, 217), one testing acupunc- Recommendations ture vs placebo acupuncture and one comparing acupuncture to drug treatment (tolterodine). Blinding of When choosing between acupuncture and anticholin- participants and assessors was done in the trials test- ergic drug for women with OAB, UUI and MUI, either ing acupuncture with a sham or placebo treatment may be effective (Grade of Recommendation: B). (215, 216). Participants were not blinded in the 3 trials There is insufficient evidence to make a recommen- involving acupuncture and drug treatments (213, 214, dation related to the effectiveness of augmenting 217), and assessors were blinded in only 1 of these electroacupuncture with drug therapy (Grade of rec- (217). Sample size calculations were reported in only ommendation: D). High-quality RCTs are needed, one trial (215). Adverse effects were described in 4 of including standardized acupuncture treatment regi- 5 trials (214-217). mens. Results Conclusions Overall, no differences were seen in primary outcomes when comparing acupuncture with sham or Limited evidence is available for the use of acupunc- placebo treatment and when comparing acupuncture ture techniques for the treatment of UI in women. with drug treatment. One small trial comparing acu- Challenges in this area (in performing research as puncture vs sham acupuncture in women with MUI well as in interpreting the literature) include a lack of reported greater reductions in overall incontinence consistency in acupuncture techniques and regimens episodes and urgency incontinence episodes in the and controversy regarding the best comparator acupuncture group after 4 weeks of treatment (67.5% (sham or placebo acupuncture) (218). Also, much of vs 16.7% and 75.2% vs 24.9%, respectively), but sta- the literature is not published in English language tistical significance was not reached (216). Another journals. Given the limitations to the quality and the trial of acupuncture vs sham acupuncture (relaxation heterogeneous nature of evidence available, few for- points) in women with OAB and with UUI found a non- mal recommendations are possible related to use of significant reduction in incontinence episodes (59 and acupuncture for SUI, UUI, MUI or overall UI in 40% in the treatment and placebo groups) (215). For women. More rigorously-conducted RCTs are acupuncture treatment compared to drug therapy (ox- needed in this area. ybutynin or tolterodine) in women with OAB, no difference was seen in improvement in UUI episodes between treatment groups (214, 217). In one, frequency and urgency improved but UUI was unchanged and in the other, frequency, urgency and UUI improved, but no between group differences were reported 94 Where a population approach is used, the ‘best’ evi- 9. FUTURE RESEARCH DIRECTIONS dence to date suggests the following: (a) an intervention comprising of a daily home PFMT and weekly 9.1. Summary physiotherapist-led exercise classes for 12 weeks, starting at 16-24 weeks’ gestation for pregnant Even with the number of reasonable trials on con- women, and (b) an individually taught strengthening servative management of UI in women, the standards PFMT programme that incorporates adherence strat- of trial conduct and reporting varied considerably. It is egies for postpartum women who have had a forceps strongly recommended that future RCTs on con- delivery or a vaginal delivery of a large baby (4000g servative management include power and sample or more) (Grade of recommendation: C). size calculation, account for potential risk of bias, intervention content and intensity, and choice of out- i) Other women PFMT come measures prior to conducting the trial. Prevention: 9.2. Recommendations for practice There is preliminary evidence that PFMT may help While some recommendations are underpinned by prevent UI in older women. (Grade of Recommen- good and consistent evidence of effects, there re- dation: C New). mains a need for further testing because of insufficient Level 1 and Level 2 evidence. Treatment: 9.2.1 Lifestyle Intervention Supervised PFMT should be offered as first line conservative therapy to women urinary incontinence Treatment: (Grade of Recommendation: A). Nonsurgical weight loss should be considered a first Clinicians should provide the most intensive health line treatment to reduce UI in obese and overweight professional-led PFMT program possible within ser- women (Grade of Recommendation: A). vice constraints because programs that are thaught and supervised by health-professionals are better Moderate exercise may help in decreasing the inci- than those with little or no supervision (Grade of Rec- dence of UI: this effect may be mediated by weight ommendation A). control (Grade of Recommendation: C). Although studies are limited, there does not appear to Caffeine reduction may help in improving inconti- be clear benefit for adding other modalities (i.e. motor nence symptoms (Grade of Recommendation: B). learning, abdominal- or hip-muscle training, intra-vag- Minor decrease of fluid intake (by 25%) may be rec- inal resistance device) to PFMT (Grade of Recom- ommended provided baseline consumption is not mendation: B) less than 30 ml/Kg a day (Grade of Recommenda- There is no clear benefit from adding clinic- (Grade tion: B). of Recommendation B) or home-based BF (Grade 9.2.2 PFMT (Principal Recommendation) of Recommendation B) to a PFMT program. The use of clinic or home-based BF should remain a ther- PFMT in the prevention of UI in childbearing apist/individual decision based on individual’s needs women: Continent, pregnant women should be of- and service constraints. fered a supervised (including regular health professional contact) and intensive strengthening antepar- ii) Other women PFMT vs other intervention tum PFMT programme to prevent antepartum and For women with SUI: postpartum UI (Grade of Recommendation: A).  PFMT and VC are both effective as conservative PFMT in the treatment of UI in childbearing therapy, although PFMT is superior (Grade of women: PFMT should be offered as first line con- Recommendation: B). servative therapy to women with persistent UI symptoms three months after delivery (Grade of Recom-  PFMT is better than EStim as first line conserva- mendation: A). tive therapy (Grade of Recommendation: B). An ‘intensive’ PFMT programme (in terms of supervi-  PFMT is better than BT as first line conservative sion and exercise content) is likely to increase the therapy (Grade of Recommendation: B). treatment effect (Grade of Recommendation: B).  PFMT and drug therapy are both effective as first PFMT in the prevention/treatment of UI in line therapy, PFMT avoids the adverse effects childbearing women: Health providers should care- experienced with drug therapy (Grade of Rec- fully consider the cost/benefit of population based ap- ommendation: B). proaches to health professional taught antepartum or postpartum PFMT, that is, health professional instruc-  Surgery is more effective than PFMT, but poten- tion to all pregnant or postpartum women regardless tial benefit should be weighed against potential of their current or prior continence status (Grade of adverse events. PFMT should be offered as first Recommendation antepartum PFMT: A). (Grade line conservative therapy (Grade of Recom- of Recommendation postpartum PFMT: B). mendation: B New). 95  PFMT and continence pessary are both effective by both VC and EStim appears to limit their utility in as first line conservative therapy (Grade of Rec- clinical practice (Grade of Recommendation: B). ommendation: B New). If this combined intervention proves to be of interest For women with SUI or MUI: to women, then adequately powered studies are needed to confirm or refute the advantages of adding  VC does not appear to be better than PFMT in VCs to PFMT (No Recommendation). the treatment of UI. PFMT should be recommended as first-line conservative therapy 9.2.4 EStim (Grade of Recommendation: B). EStim might be better than no treatment to improve VCs with supervised training sessions by a trained symptoms and quality of life in SUI women (Grade of health professional can be offered to women who can Recommendation: B). and are prepared to use them (Grade of Recommen- EStim may be considered for treatment to improve dation: B) symptoms for UUI (Grade of Recommendation: B). VC may be inappropriate for some women due to in- For women with SUI maximal clinic-based EStim ability to insert or retain the cone or because of side might be better than daily low-intensity home-based effects and discomfort. EStim in improving symptoms (Grade of Recom- For women with UUI or MUI: mendation: B).  PFMT and BT should be offered as effective first- Based on current limited evidence, EStim could be line conservative therapies (Grade of Recom- considered as an alternative to medical treatment. mendation: B). (Grade of Recommendation: B).  PFMT is better than oxybutynin as first line ther- The addition of EStim to PFMT or BF-assisted PFMT apy (Grade of Recommendation: B). programmes does not appear to add benefit (Grade of Recommendation: B) For women with UUI 9.2.5 PTNS PFMT and BT should be offered as effective first line conservative therapy (Grade of Recommendation: For women with UUI or OAB, PTNS may be more ef- B). fective than no active treatment in symptom control (Grade of Recommendation: C New). iii) Other women PFMT + other intervention vs PFMT There is no significant difference between PTNS and In women using VC, it does not appear to help to add tolterodine in terms of quality of life, however PTNS PFMT (Grade of Recommendation: C). may be considered as both may improve quality of life For the treatment of SUI, UUI, or MUI in women, con- and PTNS is less invasive (Grade of Recommenda- sider a combination of PFMT and BT rather then BT tion: B New). alone (Grade of Recommendation: C). PTNS may be considered for women as it is associ- For treatment of UUI (tolterodine) but not for SUI (du- ated with fewer and less bothersome adverse effects loxetine), consider adding PFMT to drug therapy than those from drug treatment (Grade of Recom- (Grade of Recommendation: B), mendation: B New). When treating women with SUI and vaginal atrophy, PTNS may be considered for symptom control when consider combining PFMT and intravaginal oestrogen chosen in combination interventions such as with over estrogen alone may be better than estrogen PFMT plus bladder training, or drug treatment by alone (Grade of Recommendation: C New). women with UUI or OAB (Grade of Recommenda- tion: B New). 9.2.3 Cones 9.2.6 Magnetic Stimulation For women with SUI, VC with supervised training sessions by a trained health professional may be offered No recommendation is possible based on current as a first-line conservative therapy to those who can conflicting evidence (Grade of recommendation: and are prepared to use them (Grade of Recommen- D). dation: B). No recommendation is possible regarding optimum VC may be inappropriate in some cases due to ina- type of MStim (Grade of Recommendation: D). bility to insert or retain the cone or because of side The addition of MStim to PFMT in treatment of SUI effects and discomfort. Trained health professional does not appear to be beneficial (Grade of Recom- assessment is recommended (Grade of Recom- mendation: C). mendation: D). VC and EStim seem equally effective in the treatment of SUI and MUI. Side effects and discomfort caused

96 9.2.7 Bladder Training When choosing between acupuncture and anticholinergic drug for women with OAB, UUI and MUI, either BT should be recommended as first line conservative may be effective (Grade of Recommendation: B). therapy for UI in women (Grade of Recommenda- tion: A). There is insufficient evidence to make a recommendation related to the effectiveness of augmenting Clinicians should provide the most intensive BT su- electroacupuncture with drug therapy (Grade of rec- pervision that is possible within service constraints ommendation: D). (Grade of Recommendation: D). High-quality RCTs are needed, including standard- Clinicians and researchers are advised to refer to the ized acupuncture treatment regimens. operant conditioning and educational literature to provide a rationale for their choice of their approach to 9.3. Future Research Direction BT. (Grade of Recommendation: D). All future trials must be designed, implemented and When considering BT and anticholinergic drug for reported in ways that maximise their usefulness in women with DO or UUI, either may be effective practice; this includes being well powered, with longer (Grade of Recommendation: B). term follow up, with evaluation of cost-effectiveness and planned secondary analysis of trial data to inves- BT may be preferred by women and clinicians be- tigate predictors of effectiveness. Readers are re- cause it is not associated with the drug related side ferred to the revised CONSORT statement for guid- effects (Grade of Recommendation: D). ance (219). The available evidence (single trial) supports the use 9.3.1 Lifestyle Modification Intervention of BT as a supplement to PFMT in improving short- term outcomes, but not longer-term results (Grade of Prevention should be an area for future research in- Recommendation: C). vestment and comprise robust economic evaluation to determine the benefits of lifestyle modification Limited evidence suggests the addition of written in- strategies in women with UI. formation on BT for women with OAB taking an anti- muscarinic drug does not further improve UI (Grade With limited new data on the effects of physical activ- of Recommendation: B). ity (moderate or strenuous), smoking cessation, diet modification and constipation, their association with Timed voiding with a two-hour voiding interval may be UI development or symptom reduction should be in- considered as a sole intervention for women with mild vestigated further. UI or infrequent voiding patterns (Grade of Recom- mendation: C) Timed voiding may also be consid- Separate investigation of the impact of lifestyle modi- ered as an adjunct to other treatment. fication interventions on nocturia, diurnal frequency, urgency and UI should be undertaken. Insufficient evidence exists to support the use of BT to treat urgency in women with UI and/or OAB, or to 9.3.2 PFMT in Antenatal and Postnatal guide the choice of BT vs, drug treatment as initial women treatment aimed at urgency (Grade of Recommen- dation: D).  Additional trials with longer-term follow-up (greater than 12 months postnatal) are needed Routine addition of a “simplified” BT to drug therapy to determine long-term benefits of antenatal does not provide additional improvement in urgency, PFMT. but should be considered in treatment of voiding frequency in women with UI and OAB (Grade of Rec-  Large and good-quality RCTs are needed to in- ommendation: B). vestigate the effect of antepartum PFMT on preventing postpartum UI in multiparous women. BT should be offered as treatment for urinary frequency and nocturia (Grade of recommendation:  Large and good-quality RCTs are needed to in- B), but in choosing between BT and an anticholiner- vestigate the effect of intensive post-partum gic drug for women with frequency, either may be ef- PFMT on preventing postpartum UI in women. fective (Grade of Recommendation: B). Attention must be given to high risk subgroups. 9.2.8 Complementary and Alternative Medi-  There is a need for at least one large, pragmatic, cine well-conducted and explicitly reported trial with long term follow-up (five plus years) of postpar- Acupuncture: tum PFMT that investigates the long term effect The limited evidence available to date does not sup- of ‘intensive’ PFMT for persistent postnatal UI. port the use of acupuncture when compared to sham 9.3.3 PFMT in Women (Others) or placebo treatment for women with OAB, UUI and MUI. The effect of PFMT in the prevention of UI in women should be studied further.

97 Larger, good quality trials are needed to address and in the 5th was recommended as first-line treat- comparisons between PFMT regimens, PFMT and ment for prolapse. other modalities or PFMT + another modality and PFMT alone, if these are of interest to women. In The aims of conservative treatment in the manage- planning comparisons, researchers should consider ment of POP include: carefully power calculation, intervention intensity and  preventing the prolapse becoming worse; the potential impact of different levels of supervisory intensity between groups, particularly in comparisons  decreasing the frequency or severity of symp- of conservative therapies. toms caused by prolapse (pelvic pressure, vaginal bulging, backache, urinary, bowel and sexual 9.3.4 Vaginal Cones dysfunction); Larger, good quality trials are needed to address  aversion or delaying of the need for surgery. comparisons between VC and other modalities, and VC+ other modalities compared to VC, if these are of In this section the evidence is presented for the use interest to women. In planning comparisons re- of conservative management of POP, utilising infor- searchers should consider carefully power calcula- mation from Cochrane systematic reviews (220-222), tion, intervention intensity and the potential impact of and literature identified via a search strategy summa- different levels of supervisory intensity between rised in Section I. For included trials, risk of bias groups, particularly in comparisons of conservative (RoB) was assessed using the Cochrane Collabora- therapies. tions RoB tool (223). For other types of study, bias was assessed using the Critical Appraisal Skills Pro- 9.3.5 EStim/PTNS/MStim gramme (CASP) (224) checklist item relating to ad- Larger, good quality trials are needed to address justment in analysis for potential confounding risk fac- comparisons between EStim/PTNS/MStim and no tors. treatment, EStim/PTNS/MStim and other modalities, EStim/PTNS/MStim + other modalities and ES- 1. LIFESTYLE MODIFICATION tim/PTNS/MStim, if these are of interest to women. In INTERVENTION planning comparisons researchers should consider carefully power calculation, intervention intensity and Lifestyle modification interventions include weight the potential impact of different levels of supervisory loss, reducing exacerbating activities (e.g. lifting, intensity between groups, particularly in comparisons coughing) and treating constipation, and are intended of conservative therapies. to avoid worsening of the prolapse by decreasing in- 9.3.6 Scheduled Voiding Regimen tra-abdominal pressure. Larger, good quality trials are needed to address 1.1. Prevention comparisons between BT and no treatment, BT and 1.1.1 Quality of Data other modalities, BT + other modalities and BT, if these are of interest to women. In planning compari- No trials of lifestyle modification interventions to pre- sons researchers should consider carefully power vent prolapse were identified in this update. However, calculation, intervention intensity and the potential im- 12 observational studies (retrospective reviews, pro- pact of different levels of supervisory intensity be- spective surveys and case-control studies) that ex- tween groups, particularly in comparisons of con- amined the association between modifiable lifestyle servative therapies. factors such as occupation (involving heavy lifting/strenuous physical activity), bodyweight, smok- PELVIC ORGAN PROLAPSE ing, constipation, nutrition and the development of prolapse have been published since last review, (POP) which add to the 33 previously described (9 in 5th edition, 15 in 4th edition and 9 in 3rd edition). These are Pelvic organ prolapse (POP) refers to the loss of sup- reported and summarised in Table 26. Studies often port for the uterus, bladder, colon or rectum leading examined multiple risk factors and contribute evi- to prolapse of one or more of these organs into the dence to more than one section below. If modifiable vagina. Prolapse is characterised by a variety of pel- lifestyle factors are established as risk factors for pro- vic floor symptoms (2), the most commonly reported lapse these could be targeted with a view to preven- being a sensation of bulging into the vagina. Treat- tion. ment depends on the severity of the prolapse and its symptoms, and the woman's general health (2). Con- 1.1.1.1 Association Between POP and Oc- servative treatment is generally considered for those cupation and Physical Activity with a lesser degree of prolapse, those who wish to To date, the potential relationship between occupa- have more children, those with frailty or those unwill- tion and physical activity on prolapse has received ing to undergo surgery. Conservative management, scant attention (5 studies in 3rd edition (225), 5 in 4th defined here as lifestyle modification interventions, edition(226), 3 in 5th edition(2)). A recent narrative lit- PFMT and pessaries, was first reviewed in the 3rd ICI

98 erature review (227) investigated the effect of physi- matched, to assess the impact of lifetime physical ac- cal activity and occupation on the development of tivity on the risk of prolapse (229). Women were re- prolapse. The review authors identified 10 studies, all cruited from primary care or community advertise- of which have been reviewed in–depth in previous ments, and were not seeking treatment for prolapse. editions of this chapter. They concluded that the evi- Physical activity was measured using the Lifetime dence linking physical activity and occupation with an Physical Activity Questionnaire (LPAQ), which as- increased risk for POP was weak and inconclusive. sesses physical activity over four periods (start of menstruation to age 21, 22-34, 35-50, and 51-65 Our systematic search identified a further 5 additional years), and includes leisure activity, outdoor work, studies, included (228-232). The key characteristics and housework. The Occupation Questionnaire (OQ), of these studies are described in Table 26. part of the LPAQ, was also used. Multivariable mod- Akmel et al. (2012) conducted a retrospective study els were fitted to assess the effect of physical activity of women in one Ethiopian hospital treated surgically on prolapse indicating a low risk of bias for this study for prolapse over three years (230). Medical interns (Table 26). extracted data from the women’s medical records on 1.1.1.2 Association Between POP and Body demographics, medical, surgical and obstetric his- Weight tory, symptoms and risk factors. Only univariate analysis was performed and therefore the study was There has been little agreement on whether an in- judged to be at high risk of bias (Table 26). crease in weight (or increased waist circumference or higher BMI) is linked with an increased risk of POP. Bathla et al. 2014 (232) carried out a retrospective re- Seventeen studies have evaluated the effects of bod- view of case notes for women who presented with yweight on POP since the previous editions (4 in 5th prolapse at mobile surgical camps in a remote village edition, 5 in 4th edition, 6 in 3rd edition) (228, 233-248). of Himachal Pradesh between 2009- 2013. Women In addition, two reviews have recently been published underwent POP-Q assessment prior to surgery. The on this topic (249, 250) (Table 26). The former litera- aim was to study the epidemiological risk factors for ture review included a study by Kudish and col- prolapse, including occupation and nutrition (see be- leagues (2009) – which had not previously been in- low). The analysis was mainly descriptive although cluded in earlier editions of this chapter – it is there- the authors report they did “correlation regression”, fore included here (251). but it was not clear what this was, and therefore the results were difficult to interpret and at high risk of 1.1.1.3 Association Between POP and bias (Table 26). Smoking Gumanga et al. (2014) undertook a prospective study Three studies on the relationship between smoking of women presenting with prolapse at gynaecological and prolapse were identified (228, 232, 252). In addi- outpatients at a hospital in Ghana over two years tion, Kudish (2009) investigated smoking as a risk (2010-2011), describing occupational, socio-cultural factor (251). practices and obstetric characteristics and how they related to severity of prolapse (231). Data collected Estanol and colleagues (252) studied female smok- included demographics, reproductive history and pel- ers and non-smokers with and without prolapse in a vic examination findings. Women were included if cross-sectional study in one centre. Smokers and they complained of a mass in the vagina and had de- non-smokers without prolapse were age matched to monstrable descent on pelvic examination. There smokers and non-smokers with prolapse. The aim was no control group so analysis of risk factors was was to examine the impact of smoking on collagen not possible, thus the study was at high risk of bias markers (MMP-9) and vitamin C, in those with and (Table 26). without prolapse. Women with prolapse had a bulge at or beyond the hymen on examination, and an- Lonnée-Hoffmann et al. (2015) recently reported on a swered positively to two Pelvic Floor Distress Inven- cross sectional study they undertook in women 30 tory (PFDI) questions on the feeling of a bulge or a years or older in one Nordic county in the Nord- visible bulge in the last 12 months. A univariate anal- Trøndelag Health Study (228). Outcome measures ysis of the associations was carried out, thus a high examined were self-reported prolapse surgery, age at risk of bias was assumed (Table 26). survey, socio-demographic factors, and information on risk factors for prolapse (smoking, chronic obstruc- A significant association between POP and smoking tive pulmonary disease, asthma, constipation a dec- has also been reported in 2 studies above (228, 251). ade previously, and measured body mass index) (see Bathla et al., (2014) also noted that 30-57% of their Table 26). Multivariable logistic regression was used sample were smokers; collectively these studies sug- for analysis thus the study was judged to be at low gest that smoking is a risk factor for POP. risk of bias.

Nygaard et al. (2014) conducted a retrospective case-control study of women with prolapse (descent ≥ 1 cm beyond the hymen) and without prolapse (descent ≤ 1 cm above the hymen), age and site

99 Table 26 Summary of data on lifestyle factors and POP

Author, Study design/ Modality details/methods Notes (side effects, N Study population Outcomes/results Follow up year Comparator employed loss of follow up…) POP and Occupation/Physical activity Akmel Retrospective 129 Patients who were Demographic and medical Type of occupation was None Data was available for 2012 (1) descriptive case note admitted and operated characteristics were strongly associated with stage 129 out of a possible analysis on in the gynaecology extracted based on the initial of prolapse. Housewives were 143 women. No ward in one hospital in medical history and physical more likely to have stage IV multivariate analysis to Ethopia. All the women examination documentation (vs stage III) prolapse allow for other with POP who were in the case records compared to farmers. confounding variables. admitted and treated during the study period were included; no sampling technique was employed Bathla Retrospective 157 157 women who had Epidemiological data and Factors contributing to POP None “Statistical analysis was 2014(2) descriptive case note POP surgery conducted POP-Q data were collected. believed to be “poor nutritional performed using analysis in 5 mobile surgical status (mean weight 41.1 kg), correlation regression camps in Shillai, multiparity (mean 3.5), early analysis”, but limited Himachal Pradesh from marriage (mean age 18.2 data presented. No 2009 to 2013 (“Project years), unassisted home multivariable analysis Prolapse”). deliveries (100%), premature performed. bearing down (23.8%), early postpartum resumption of strenuous activity (54.7%) and smoking (33%)” This was not based on statistical analysis. Gumanga Prospective 118 POP cases, out-patients Questionnaires and physical Main occupations were None Authors report that 2014 (3) observational study seen over 2-year study exam in patients whose trading of produce 66/118 occupational factors period at a teaching main complaints included “a (55.9%) and farming 44/118 might contribute to the hospital in Ghana. mass falling from the vagina” (37.3%) severity of POP in this or “bulging mass” or sample. Analysis “sensation of mass in the undertaken is unclear vagina”. but appears not to be multivariable analysis.

100 Author, Study design/ Modality details/methods Notes (side effects, N Study population Outcomes/results Follow up year Comparator employed loss of follow up…) Lonnee- Cross-sectional 20,285 All women aged ≥ 30 Women were sent POP surgery was reported by None Authors also report that Hoffmann survey years in one Nordic questionnaires and attended 1,123 (5.3 %). Only women constipation reported a 2015 (4) county invited to take screening stations. Risk reporting more lifting in decade prior, above- part factors assessed: smoking, addition to walking had normal BMI, and COPD chronic obstructive significantly higher odds of were significant non- pulmonary disease, asthma, reporting POP surgery obstetric risks for constipation a decade compared with women with prolapse surgery. previously, and measured sedentary occupation in the Multivariable analysis BMI age-adjusted model performed. Nygaard Retrospective case- 382 191 POP cases and 191 Women were asked to No associations between None Authors suggest that 2014 (5) control study age and recruitment-site complete Lifetime Physical odds of POP and overall strenuous activity matched controls, aged Activity and Occupation lifetime physical activity, during teenage years between 39-65 years, not Questionnaires based on lifetime leisure activity, or may have an seeking prolapse self-reported activities lifetime strenuous activity association with POP, treatment and recommend further prospective research in this area Multivariable analysis?? POP and Smoking (see also entries for Bathla 2014, Kudish 2009 and Lonnee-Hoffmann 2015) Estanol Cross-sectional 96 4 groups of women: Fasting blood panel, Comparing women with POP None Small sample size; was 2015 (6) study smokers with POP including plasma and without, there were no not sufficiently large to (n=16), non-smokers procollagen 1-N propeptide significant differences in detect a true difference, with POP (n=16), (P1NP), matrix vitamin C, P1NP and MMP-9 particularly in those smokers without POP metalloproteinase 9 (MMP- levels, independent of with prolapse. (n=32) and non-smokers 9), and vitamin C. These are smoking status. No multivariable without POP (n=32) markers of collagen MMP-9 levels were higher and analysis metabolism. Smokers were vitamin C lower in smokers defined as smoking 1 pack compared to non-smokers. or more a day for at least 1 year; non-smokers were defined as not having smoked for at least 7 years

101 Study design/ Modality details/methods Notes (side effects, Author, year N Study population Outcomes/results Follow up Comparator employed loss of follow up…) POP and Bowel function (see also entries for Kudish 2009 and Lonnee-Hoffmann 2015) Participants referred to Medical history, series of No differences in None No multivariable one tertiary questionnaires (ICIQ-SF; prolapse status between analysis Cross-sectional urogynecology outpatient KHQ; PISQ-12), Cleveland patients with and without Bezerra 2014 survey 172 clinic with self-reported Constipation Scale and anal incontinence and/or (7) symptoms of PFD: POP physical exam using the bowel disorders. [POP-Q ≥ stage II and/or PERFECT scheme. UI] Chronic constipation and Questionnaire included Multivariable analysis chest disease, level of Women attending family items for details of socio- was performed. Authors education, occupation development centres in demographic, obstetrics, recommended Elbiss 2015 (8) birth weight and BMI Cross-sectional 429 United Arab Emirates, medical and surgical history. None additional healthcare were independent risk 29.6% reported POP Presence of a lump in/out of campaigns to raise factors for having POP symptoms vagina was taken as awareness about risk symptoms. presence of POP. factors for POP

POP and Nutrition (see also entries for Estanol, 2015) Navaneethan Prospective case- 120 Women with (n=51) or Medical history and clinical 23/51 had POP alone, None Multivariable analysis 2015 (9) control study without (n=69) PFD on examination (POP-Q) were 9/51 had POP+SUI; performed to adjust for examination when performed. Serum 25- vitamin D levels were not age. attending outpatient clinic hydroxy vitamin D levels significantly associated at one tertiary care were measured in all with POP centre in South India participants

Parker-Autrey Retrospective chart 394 Women over 19 years Diagnosis categorized as Only higher IIQ-7 scores None Multivariable logistic 2012 (10) review old attending a PDF (SUI, UUI, FI, POP) were significantly related regression model used Urogynecology Care (n=268) or general GYN to having insufficient to adjust for age, BMI, Clinic from 2008 to 2010 (n=126). Demographic and vitamin D. ethnicity, and use of vit. who had vitamin D medical characteristics were D supplementation measured. extracted. PFDI-SF 20, IIQ- 7, Medical, Epidemiologic, and Social Aspects of Aging (MESA) questionnaire.

102 Study design/ Modality details/methods Notes (side effects, Author, year N Study population Outcomes/results Follow up Comparator employed loss of follow up…) POP and body weight Myers et al., Secondary data 338 Overweight and obese Women were randomised Increasing BMI was Evaluated for At 6 months, there were 2012(11) analysis of RCT women with UI either into an intensive 6- associated with only prolapse no significant month weight loss or to the "feeling" a vaginal bulge. symptoms at differences in control group (educational (13% of obese women baseline and 6 improvement of self- program) reported feeling vaginal months reported bothersome Symptomatic prolapse was bulging compared with prolapse symptoms in defined as a positive overweight women (0%). women in the weight response to at least 1 loss or the control prolapse subscale question group of the UDI. "Bother" was Multivariable analysis defined as responses of was performed. slight, moderate, or great. Kudish 2009 Secondary analysis 16608 Post-menopausal women Degree of prolapse Prolapse risk in Annual pelvic Being overweight or (12) of a 5-yr clinical trial aged 50-79 yrs assessed using the WHI overweight and obese exam over a 5 obese is associated Prolapse women (compared with year interval with progression of Classification System the women with healthy POP. Authors reported BMI) increased by 32% that “weight loss was and 48% for not significantly cystocele, by 37% and associated with 58% for rectocele, and by regression of POP”, 43% and 69% for uterine and argued that pelvic prolapse, respectively floor damage related to weight gain could be irreversible Multivariable analysis was performed. Abbreviations: PFDI-20: pelvic floor distress inventory short form 20, PFIQ-7: pelvic floor impact questionnaire short form; PFM: pelvic floor muscles; PFMT: pelvic floor muscle training; POP-Q; Pelvic Organ Prolapse Quantification system; PISQ-12: Pelvic organ prolapse/urinary incontinence sexual questionnaire’ POP-SS: pelvic organ prolapse symptom score; QoL: quality of life; RCT: randomized controlled trial; sEMG: surface electromyography; SR: systematic review; UI: urinary incontinence 1. Akmel M, Segni H. Pelvic organ prolapse in jimma university specialized hospital, southwest ethiopia. Ethiop J Health Sci. 2012;22(2):85-92. 2. Bathla S, Verghese G, Kalla V, Sharma TC, Dam S, Agarwal N, et al. Reaching the unreached: Mobile surgical camps in a remote village of Himachal Pradesh. J Midlife Health. 2014;5(3):139-42. 3. Gumanga SK, Munkaila A, Malechi H. Social demographic characteristics of women with pelvic organ prolapse at the Tamale Teaching Hospital, Ghana. Ghana Med J. 2014;48(4):208-13.

103 4. Lonnee-Hoffmann RA, Salvesen O, Morkved S, Schei B. Self-reported pelvic organ prolapse surgery, prevalence, and nonobstetric risk factors: findings from the Nord Trondelag Health Study. Int Urogynecol J. 2015;26(3):407-14. 5. Nygaard IE, Shaw JM, Bardsley T, Egger MJ. Lifetime physical activity and pelvic organ prolapse in middle-aged women. Am J Obstet Gynecol. 2014;210(5):477 e1-12. 6. Estanol MV, Crisp CC, Oakley SH, Kleeman SD, Fellner AN, Pauls RN. Systemic markers of collagen metabolism and vitamin C in smokers and non-smokers with pelvic organ prolapse. Eur J Obstet Gynecol Reprod Biol. 2015;184:58-64. 7. Bezerra LRPS, Vasconcelos Neto JA, Vasconcelos CTM, Karbage SAL, Lima AC, Frota IPR, et al. Prevalence of unreported bowel symptoms in women with pelvic floor dysfunction and the impact on their quality of life. International Urogynecology Journal and Pelvic Floor Dysfunction. 2014;25(7):927-33. 8. Elbiss HM, Osman N, Hammad FT. Prevalence, risk factors and severity of symptoms of pelvic organ prolapse among Emirati women. BMC Urology. 2015;15(1):no pagination. 9. Navaneethan PR, Kekre A, Jacob KS, Varghese L. Vitamin D deficiency in postmenopausal women with pelvic floor disorders. J Midlife Health. 2015;6(2):66-9. 10. Parker-Autry CY, Markland AD, Ballard AC, Downs-Gunn D, Richter HE. Vitamin D status in women with pelvic floor disorder symptoms. Int Urogynecol J. 2012;23(12):1699-705. 11. Myers DL, Sung VW, Richter HE, Creasman J, Subak LL. Prolapse symptoms in overweight and obese women before and after weight loss. Female pelvic medicine & reconstructive surgery. 2012;18(1):55-9. 12. Kudish BI, Iglesia CB, Sokol RJ, Cochrane B, Richter HE, Larson J, et al. Effect of weight change on natural history of pelvic organ prolapse. Obstet Gynecol. 2009;113(1):81-8.

104 1.1.1.4 Association Between POP and India. A clinical and obstetric history was taken, ex- Bowel Function amination performed (POP-Q and stress test) and vitamin D measured in all women. A multivariable lo- Previously there were 8 studies that had examined gistic regression analysis was undertaken to assess the potential impact of bowel disorders/ or bowel dys- the relationship between vitamin D and PFD adjusting th function on prolapse (4 in 5 edition, 4 in 4th edition). for other variables (254). A low risk of bias was as- This update adds a further 4 studies. Two are de- sumed (Table 26). scribed below; two have been described above (Lon- nee-Hoffman et al., 2012; Kudish et al., 2009). Parker-Autry et al. (2012) in Alabama, US, undertook a retrospective review of women who were new Bezerra et al. (2014) undertook a cross-sectional urogynaecology outpatients presenting over a two- study of women with pelvic floor disorders (POP-Q ≥ year period (2008 to 2010) and who then had vitamin stage II and/or urinary incontinence) presenting at a D measured within one year of that visit. 25(OH)D < tertiary urogynaecology clinic in Brazil (253). The aim 15 ng/ml was defined as being deficient in vitamin D, was to report the prevalence, bother, and impact on and 15 to 29 ng/ml as having insufficient vitamin D. quality of life (QoL) of unreported bowel symptoms, to Women were classified into two groups, either PFD improve the patient care. Those with and without def- (having SUI, UUI, FI or POP) or general gynaecology aecatory or anal continence problems were com- (no PFD) (Table 26). Demographic, medical, and la- pared. A number of measures were used depending boratory data were extracted from the initial history on the presence/absence of bowel and urinary incon- and physical examination notes. Questionnaire re- tinence symptoms: ICIQ-SF, King’s Health Question- sponses from the clinic visit were also available naire, PISQ-12, the Wexner score, Cleveland Clinic (PFDI-SF 20 (including the UDI, CRADI, POPDI) and Florida Incontinence Scale, Cleveland Clinic Florida IIQ-7). Multivariable logistic regression was used to Constipation Scale (mild, moderate or severe). All examine the association between vitamin D status women completed the SF-36 and had their pelvic and pelvic floor disorder symptoms, with adjustment floor assessed by a physiotherapist using the for age, BMI, race/ethnicity, and the use of vitamin D PERFECT scheme. Only univariate analysis was per- supplementation. This study was judged to be at low formed, putting the study at a high risk of bias (Table risk of bias. 26). 1.1.2 Results Elbiss and colleagues (238) carried out a cross-sectional study of parous Emirati women 30 years or 1.1.2.1 Association Between POP and Oc- older attending three family development centres in cupation and Physical Activity Al-Ain in one year. The objective was to EStimate the prevalence of prolapse and its risk factors. The study In the Akmel et al. (2012) study, 143 prolapse cases used a self-developed questionnaire, which was pi- were examined and complete data retrieved for 129 loted and revised before use. The questionnaire cov- cases. The mean age of the women was 42 years ered demographic, obstetrics, and medical history, (range 22 to 72 years), mean parity was 6.5 (range 1 prolapse symptoms (lump coming down in the to 14) and all had stage III (56%) or IV (44%) pro- vagina, lump coming out of vagina or lump felt or seen lapse. Significant univariate associations were found outside vagina), other vaginal symptoms and the between age, parity and rurality and stage of pro- need to digitate to empty bladder or bowel. Following lapse. Occupation was significantly associated with univariate analysis, multivariable binary logistic re- prolapse stage in that farmers were more likely to gression analysis (prolapse symptoms versus no pro- have a lower stage and housewives a higher stage of lapse symptoms) was performed including all signifi- prolapse (230). cant variables to determine which were independent In the retrospective case analysis described by risk factors for prolapse (Table 26). This study was Bathla et al. (2014), 192/490 surgical case presenta- judged as having a low risk of bias. tions were gynaecological and 82% of these were 1.1.1.5 Association Between POP and Nu- prolapse (99% involving the anterior compartment). trition These women were aged between 30 and 70 (mean 47 years), had mean parity of 3.5, and all had unas- Two studies were identified in previous editions of the sisted home deliveries. Premature bearing down dur- chapter that examined the effects of anaemia (4th edi- ing delivery was reported for 24%. Women on aver- tion) and vitamin D (5th edition) on prolapse. A further age resumed strenuous physical work and sexual ac- 2 studies relating to vitamin D are included in this up- tivity 20 days after giving birth. 33% were smokers date (Table 26). and 55% lifted heavy weights uphill. The authors reported non-significant correlations between POP-Q Navaneethan et al (2015) conducted a case-control stage and parity, resumption of physical activity and study of postmenopausal women with and without weight (232). pelvic floor disorders, in particular, symptoms of urinary incontinence or prolapse. It was not clear how Out of 4403 outpatients, 118 cases (2.7%) of pro- the women were selected but the study took place in lapse were seen over the study period in the a department of obstetrics and gynaecology in South Gumanga (2014) paper, with a mean age of 46 years (SD 15 years) and mean parity 4.4 (SD 1.7). For 57%

105 of women their deliveries had been exclusively at prolapse increased in overweight and obese women home. Women were predominantly farmers (37%) or compared to post-menopausal women with normal traders (56%) (231). BMIs (Table 26). When the data was adjusted for women diagnosed with prolapse at baseline and Of all women in the county 30 years or older, 20,285 baseline BMI, a 10% weight loss was associated with (50.3%) were included in the Lonnee-Hoffman (2012) minimal change in overall POP (251). This study was study; of which 1,123 (5.5%) reported having under- judged to be at low risk of bias. gone prolapse surgery. In the multivariable models, the odds of having prolapse surgery were shown to 1.1.2.3 Association Between POP and be significantly greater for women with BMI of 25 or Smoking greater, and for those who reported “marked” constipation (as compared to no or mild constipation). In the Estanol (2015) study, a total of 96 women were There was no relationship with smoking, asthma or recruited: 32 with prolapse (16 smokers, 16 non- occupation (228). smokers) and 64 without prolapse (32 smokers, 32 non-smokers) (252). Smokers with prolapse had One hundred and ninety-one cases and an equal lower levels of vitamin C and higher levels of MMP-9, amount of matched controls were recruited in the Ny- compared to non-smokers with prolapse, but this re- gaard et al. (2014) study. The mean age of women lationship was not statistically significant. However, was 50 years (SD 7). Cases tended to have a higher comparing smokers without prolapse to non-smokers BMI than controls, and had significantly greater parity without prolapse, there were significant differences in and more vaginal deliveries. Vaginal bulge was more both vitamin C and MMP-9, suggesting an impact of prevalent in the cases than controls, but there was no smoking on these collagen markers. Ignoring smok- difference in other pelvic floor symptoms. The authors ing status, vitamin C and MMP-9 levels in women with found no evidence that physical activity (lifetime over- prolapse were similar to those without prolapse. The all activity, leisure or strenuous physical activity) was authors suggest that the damage to connective tissue associated with risk of prolapse in multivariable mod- that leads to prolapse may be different from the harm els. The exception was that strenuous physical activ- caused by cigarette smoking. ity in the teenage years showed a significant relationship with the odds of prolapse, indicating an in- Lonnee-Hoffman et al. (2015) found no association creased risk of prolapse for those reporting 21 between smoking and whether a woman reported hours/week or more of strenuous physical activity having had prolapse surgery or not (228). Bathla et when a teenager (229). al. (2104) commented that 33% of the women having prolapse surgery were smokers (232). Elbiss (2015) found nature of occupation to be a significant predictor of prolapse symptoms (see below); 1.1.2.4 Association Between POP and women with non-physical jobs were more likely to re- Bowel Function port symptoms. Bezerra and colleagues reported that a total of 172 1.1.2.2 Association Between POP and Body women with prolapse and/or urinary incontinence Weight participated and initially none of them reported defecatory problems or anal incontinence, although Vergeldt et al. (2015) systematically identified 8 stud- 54.6% reported one or both on interview. Those with ies that examined the association of higher BMI and AI/defecatory problems were no more likely to report POP and presented the results in a narrative synthe- prolapse (sensation of a ball in the vagina) than those sis. The authors reported that 4/8 studies identified without these problems (68% versus 65%). Stage of that higher BMI was a significant risk factor for POP; prolapse was not associated with AI/defecatory prob- 3/8 found no relationship between high BMI and the lems either (253). development of primary POP and 1/8 study found that a “higher BMI slightly protective”. The authors re- Of the 482 eligible women in the Elbiss (2015) study, ported that the relationship between weight and POP 429 (89%) participated, of which 127 (29.6%) re- was inconsistent; thus no conclusions can be reliably ported prolapse symptoms. Those women with and drawn (250). This well-conducted review was judged without prolapse symptoms did not differ in terms of to be at low risk of bias. age. Of the factors significantly associated with prolapse in the univariate analysis (BMI, education level, Kudish and colleagues (2009) explored the relation- nature of occupation, history of chronic chest dis- ship between prolapse progression (or regression) ease, constipation, diabetes, previous instrumental and weight change in 16,608 post-menopausal delivery, maximum birth weight, history of urinary in- women (Table 26). This secondary analysis study continence and previous surgery for urinary inconti- was based on longitudinal data from the WHI E+P nence) six were independent risk factors in the multi- trial; a double-blinded RCT. Women aged between variable analysis: history of constipation, education 50 and 79 years received annual pelvic examinations level, chronic chest disease, nature of occupation, over a 5-year interval. The majority of women enrolled maximum birth weight and BMI. Women were more in the study gained weight, and the overall prolapse likely to have prolapse symptoms if they had chronic prevalence (stage I-III) increased from 40.9% at baseline to 43.8% by year 5. The risk of all types of

106 chest disease or constipation, higher BMI or maxi- in earlier editions (Level of Evidence: 3; No rec- mum birth weight, lower education level, and were a ommendation). housewife or had a non-physical job (238).  Evidence from previous editions regarding the Lonnee-Hoffman et al (2015) found that marked con- association between constipation or straining at stipation (as compared to no or mild constipation) was stool and prolapse was conflicting. Two new, low significantly associated with having prolapse surgery risk studies, which adjusted for covariates, con- (228). cluded that constipation was associated with both prolapse symptoms and having prolapse 1.1.2.5 Association Between POP and Nu- surgery, contributing more evidence of an asso- trition ciation (Level of Evidence: 3; Grade of Recom- Navaneethan et al (2015) reported that of the 120 mendation: C New; Majority evidence of an women who participated: 42.5% had PFD (54.9% association). prolapse alone, 27.4% SUI alone, 17.6% SUI and  Two new, low risk studies on vitamin D supported prolapse) and 57.5% had no PFD. Multivariable anal- previous findings (1 study, 5th Edition) of no as- ysis showed that, after adjustment for age, having any sociation with prolapse (Level of Evidence: 3; PFD was associated with having lower vitamin D lev- Grade of Recommendation: C Majority evi- els and being five years or more post-menopausal. dence of no association). However having prolapse was only associated with the latter, and there was no significant relationship 1.2. Treatment with vitamin D (254). Previously no studies that evaluated the effective- Of 550 women potentially eligible to participate in the ness of lifestyle modification interventions in the treat- Parker-Autry (2012b) study, 394 were included (268 ment of women with POP had been identified. Sub- in the PFD group and 126 in the no PFD group). The sequently we have identified one trial, which meas- prevalence of vitamin D insufficiency was 51% in the ured prolapse outcomes after weight loss pro- PFD group. Vitamin D insufficiency was inde- grammes (243). Not all trial participants had prolapse pendently associated only with IIQ-7 score. That is, however, which limits the usefulness of this evidence there was no relationship with prolapse symptoms (see Table 26). (255). Quality of data Summary Myers and colleagues (2012) performed a secondary  There remain no trials of lifestyle modification in- analysis of data from the PRIDE trial of an intensive terventions to prevent prolapse. Some new ob- weight loss programme versus an educational pro- servational studies have added to our knowledge gramme for urinary incontinence in overweight and of potentially helpful ways to modify lifestyle risk obese women (243). Prolapse symptoms were meas- factors. ured at baseline and six months (post-intervention) using the prolapse items from the UDI. Any positive  Two new good quality observational studies response to any questionnaire items indicated the (228, 229) suggested occupation and physical presence of prolapse symptoms. A subgroup of activity are not risk factors for prolapse surgery women agreed to have urodynamics and POP-Q as- or prolapse 1cm or more beyond the hymen on sessment before and after intervention. Women were examination. A third study however found classified as overweight, obese or severely obese ac- women who were housewives or in a non-physi- cording to their BMI. Women who had prolapse symp- cal occupation were more likely to report pro- toms at baseline were analysed at 6 months for dif- lapse symptoms (238). The vast majority of stud- ferences between the randomised groups. The risk of ies reported in previous editions supported an bias was judged as low in relation to selection, detec- association between current heavy occupational tion and attrition for the original PRIDE trial (6) in a lifting and prolapse, overall therefore the evi- recent Cochrane review (5). dence seems to be conflicting, and this may be due to different ways of defining prolapse (Level Results of Evidence: 3; Grade of Recommendation: D Myers et al., (2012) reported that 338 women were Conflicting therefore no recommendation). randomised and 110 had urodynamics and POP-Q  Recent evidence on the relationship between (243). Of the women randomised, 16% were over- prolapse and bodyweight is conflicting. (Level of weight, 58% obese and 26% severely obese. At Evidence: 3; Grade of Recommendation: D baseline 53% reported at least one prolapse symp- Conflicting therefore no recommendation). tom, but this did not differ across BMI groups. A significantly greater proportion of obese women reported  Smoking was found not to be associated with feeling vaginal bulge compared to overweight women prolapse in two studies: a matched case control (10% severely obese vs 14% obese vs 0% over- study (Estanol) (high risk) and large cross-sec- weight), but the mean number of prolapse symptoms tional survey with multivariable analysis (Lon- did not differ across the BMI groups. The proportion nee-Hoffman) (low risk). No studies were found with prolapse beyond the hymen was 17% overall but

107 did not differ across the BMI groups (24% overweight 2.1. Prevention group, 17% obese group, 13% severely obese group). Post-intervention the intensive weight loss Quality of data group had lost significantly more weight than the con- Previously there were no trials of PFMT for prevention trol group (mean 7.8kg vs 1.5kg). Over 70% reported of prolapse, only evidence from cross-sectional stud- cure or improvement in prolapse symptoms, although ies of a possible association between pelvic floor this was no different between the intensive weight muscle function and risk of prolapse. Two RCTs have loss group and the educational group. There was also now been identified that evaluate the role of PFMT in no significant difference in prolapse beyond the hy- prevention of prolapse (257, 258). A summary of the men between weight loss intervention and educa- setting, design and study population of prevention tri- tional groups. The findings would suggest a lack of als is presented in Table 27. Blinding of patients is relationship between weight loss and improvements generally not possible in trials of PFMT so this is not in prolapse. However, the trial was designed for taken into consideration in the risk of bias analysis. women with urinary incontinence rather than prolapse, and the post-intervention comparison was not Bo et al (2013) (258) carried out a trial in primiparous randomised. women after vaginal delivery of a singleton infant after 32 weeks gestation, comparing supervised PFMT Summary with written advice to do PFMT (control), to prevent and treat prolapse. All women had received written Currently, there is evidence from secondary analysis recommendations to perform PFMT in the delivery of one robust trial regarding the role of weight loss in ward. The PFMT group attended a weekly PFMT the treatment of POP. The trial however was in over- class for 4 months starting at 6-8 weeks postpartum, weight and obese women with UI, some of whom had and performed home-based exercise. The control prolapse. It would appear that weight loss in both group had no further supervision or follow-up. Out- groups led to an improvement in prolapse, however comes measured at 6 weeks and 6 months post-par- there was no relationship between degree of weight tum included POP-Q stage, bladder neck position loss (intensive vs normal weight loss programme). (transperineal ultrasound) and symptom of vaginal Any weight loss may improve prolapse in overweight bulge (ICIQ-VS). The trial was judged as low risk for or obese women and UI. (Level of Evidence: 2; selection bias, performance and detection bias, attri- Grade of Recommendation: D). tion bias and reporting bias. However, the trial was judged as unclear for other types of bias (Table 27).

The PREVPROL trial, undertaken by Hagen and colleagues, compared PFMT versus lifestyle advice leaf- 2. PELVIC FLOOR MUSCLE let as secondary prevention for women with pre-clini- TRAINING cal signs of prolapse (257). Participants were women originally enrolled in a longitudinal follow-up of post- The pelvic floor muscles play a critical role in giving natal incontinence (at 3 months, 6 and 12 years) after structural support to the pelvic organs and pelvic giving birth in 1993/94. Those, who had not sought openings. It is hypothesized that improving pelvic treatment for prolapse, but who showed early signs of floor muscle function may improve this structural sup- prolapse (POP-Q stage I, II or III) were invited to take port for the pelvic organs. part. They were randomised to receive either a programme of PFMT (individualised physiotherapy ap- A programme of supervised PFMT includes assess- pointments, maintenance via Pilates-based classes ment of the woman’s pelvic floor muscles and her (progressive Pilates-based exercises, pelvic floor ability to contract these muscles; education about the muscle training as given in the one-to-one interven- pelvic floor muscles and how they support the pelvic tion, core exercises from the “Pelvicore Technique organs; instruction in how to correctly perform pelvic with Kari Bo” DVD) and annual one-to-one check- floor muscle exercises and “the Knack” (pelvic floor ups) or a prolapse prevention lifestyle advice leaflet. muscle bracing against increased intra-abdominal Women were followed up at 1 and 2 years in terms of pressure, for example when coughing and sneezing) their prolapse symptoms (POP-SS), prolapse-related (256). An individualised exercise programme is pre- QoL, uptake of prolapse treatment, symptoms of uri- scribed for the woman to follow. Adjuncts to PFMT nary incontinence (ICIQ-UI SF), anorectal or sexual (such as biofeedback) or other physical therapies dysfunction (PISQ-12), perceived health benefit, and (such as neuromuscular EStim) may be used. These cost-effectiveness. therapies aim to improve PFM strength, endurance, coordination and function. Other forms of physical The trial was judged as low risk for selection bias, per- therapy involving diaphragmatic aspiration are formance and detection bias, attrition bias, reporting emerging. bias and other types of bias (Table 27). Results One hundred and seventy-five primiparous women were randomised in the Bo (2013) study, 87 to PFMT and 88 to control, mean age 30. At 6 months post-

108 partum there was no difference between groups in 2.2.1 PFMT Alone the prevalence of symptoms of bulging either inside or outside the vagina. There was no difference be- Thirteen trials now exist in this area, six of which were tween groups in the prevalence of stage II prolapse previously reported in previous consultations. Key at 6 months, or in any of the individual POP-Q meas- characteristics of the seven new trials, and new pub- urements, or the bladder neck position (258). lished information about two earlier trials, are summarised in Table 28. Narrative descriptions of these trials In the PREVPROL trial 414 women were randomised, are reported below. 207 to PFMT and 207 to control, mean age 46 years (257). There was a significantly lower POP-SS score Alves and colleagues carried out a trial in post-men- at 2 years in the PFMT group compared to the control opausal women in which they compared a general fit- group, indicating fewer symptoms. Women in the ness programme (control) with a PFMT programme control group were significantly more likely to have (intervention) (47). Both groups performed a fitness sought treatment for prolapse symptoms by 2 years. programme twice weekly for 6 weeks including global No significant difference was found between the muscle stretching, endurance and functional exer- groups in the percentage who experienced any urine cises. In addition, the intervention group took part in leakage at 2 years, although there was a significant a physiotherapist-supervised PFMT programme in difference in favour of the intervention group in the groups of seven, 30 minutes twice weekly for 6 ICIQ-UI short form score. Faecal urgency and leak- weeks. The control group women were taught about age, and sexual function were not significantly differ- pelvic floor muscles and how to contract them cor- ent between the groups at 2 years. Women in the in- rectly, but without any training. Outcomes were as- tervention group were more likely to say they felt a sessed after the intervention at 6 weeks. The trial was health-related benefit from the study compared to the generally judged as low risk for performance and de- control women. tection bias, attrition bias, reporting bias and other types of bias. However it was unclear if opaque enve- Summary lopes were used in the allocation process, and an intention to treat analysis was not performed (Table Currently, there is evidence from two robust trials re- 28). garding the role of PFMT in the prevention of POP. The trials were in different populations of women Culligan and colleagues compared a standardised however (postnatal women, 13% of whom had stage PFMT programme with a standardised Pilates pro- II prolapse, and middle-aged women with mild programme in community women with POP-Q stage I lapse, 55% with stage II or greater, who had not prolapse to see if they provided similar improvements sought treatment) and drew differing conclusions. It in PFM strength (260). The PFMT programme, made would appear that in younger post-natal women up of twice weekly 1-hour sessions for 12 weeks, in- PFMT does not influence the development of pro- cluded computerized biofeedback, vaginal manipula- lapse by 6 months, whereas in older women, more tion, neuromuscular re-education, and manual ther- than 12 years after giving birth, there was a significant apy. Participants completing 20 or more of the 24 benefit of PFMT in terms of fewer prolapse symptoms possible sessions were defined as “successful”. The after 2 years and less uptake of treatment. PFMT can control group attended a Pilates programme with the prevent symptoms of prolapse which develop in the same pattern of sessions where they were taught full- longer term after childbirth but not immediately after body exercises focusing on the “core muscles”, and giving birth. the pelvic floor in particular. Women were assessed at baseline and after the intervention at 12 weeks in Postnatal: Level of Evidence: 1; Grade of Recom- terms of PFM strength and pelvic floor symptoms and mendation: B New; majority evidence from RCT impact (PFDI-20), Short Form Pelvic Floor Impact of no effect for postnatal. Questionnaire (PFIQ-7). The trial was judged as low risk for selection bias, attrition bias, reporting bias and 12 years post-childbirth: Level of Evidence 1; other types of bias. However, performance and de- Grade of Recommendation: B New; Majority evi- tection bias was judged as unclear risk of bias as it dence from RCT of effect in the long term after was unclear if outcome assessors were blinded (see childbirth. Table 28). 2.2. Treatment Due et al. (2016) undertook a trial in women with symptomatic prolapse of stage II or greater, compar- Quality of data ing a structured lifestyle advice programme plus Evidence from trials now exists relating to the role of PFMT (combined group) with lifestyle programme PFMT in the treatment of prolapse. The role of PFMT alone (lifestyle group) (261). The PFMT included both as an adjunct to surgery or pessary has also been the group PFMT (6 sessions over 12 weeks) and individ- subject of randomised studies. A Cochrane review ual home training after an assessment and individual specifically addressing this question was first pub- instruction. Women were assessed using the Patient lished in 2004 (259), and updated in 2011 (220) and Global Index of Improvement scale (PGI-I), POP-Q, 2016 (in preparation) (221). PFDI-20, PFIQ-7 and PISQ-12 at baseline, immediately post-intervention (3 months) and at 6 months.

109 The trial was judged as low risk for selection bias, per- measured at 6 and 12 months were prolapse symp- formance and detection bias, attrition bias, reporting toms (POP-SS), prolapse-related QoL, need for fur- bias and other types of bias (261) (see Table 28). ther prolapse treatment, bladder, bowel and sexual symptoms. POP-Q was measured at baseline and 6 Frawley et al. (2012) carried out a trial in four Austral- months. The trial was judged as low risk for selection ian centres involving women with symptomatic pro- bias, performance and detection bias, attrition bias, lapse stage I to III, randomised to PFMT or a lifestyle reporting bias and other types of bias (Table 28). advice leaflet (control) (262). This trial protocol was based on that of the UK POPPY trial (263) described Kashyap and colleagues reported a single-centre trial below with some slight adjustments and additional in parous women with stage I to III prolapse, which measurements. Women were assessed at baseline, compared taught pelvic floor muscle training plus a 6 and 12-month follow-up for PFM function (mano- self-instruction manual (SIM) with the SIM alone as metric and digitally-assessed strength and endurance the control intervention (265). One person delivered (ICS scale)), prolapse symptoms (POP-SS) and se- the training to all intervention women. This included verity/type (POP-Q). The trial was judged as low risk explanation of anatomy and function of pelvic floor for selection bias, performance and detection bias, muscles, PFMT training, plus exercise practice and reporting bias and other types of bias. However, the checking for correct PFM contractions by vaginal ex- trial was considered to have high risk of attrition bias. amination. Pressure manometry was included for bi- Communication with the author provided explana- ofeedback and motivation. A home exercise pro- tions about the dropout observed (Table 28). gramme was prescribed for 3 times daily. Women were instructed to perform the intervention for 24 Giraudo et al’s (2011) trial compared PFMT plus neg- weeks. After the initial training session, there were 6 ative pressure abdominal work (intervention group) follow up visits (at weeks 1, 3, 6, 12, 18 and 24). The with PFMT plus abdominal hollowing exercise (con- content of the manual was not described and there- trol group) in women with untreated stage I or II pro- fore what written instruction the control group re- lapse (264). Both groups were taught correct contrac- ceived is unclear. Women were assessed at baseline tion of the pelvic floor muscles and given tailored life- and at 1, 3, 6, 12, 18 and 24 weeks using the POP- style advice on ways of reducing intra-abdominal SS, a visual analogue scale and the PFIQ-7. The trial pressure, as well as a standardised lifestyle advice was judged as low risk for selection bias. However, sheet. There was no description of 1) the negative the trial was judged as high risk for performance and pressure abdominal work or 2) abdominal hollowing detection bias, attrition bias, reporting bias and other exercises but they are described elsewhere as 1) hy- types of bias (see Table 28). popressive exercises, which are thought to result in negative pressure in the thoracic cavity and involun- Stupp/Resende carried out a three-arm trial compar- tary contraction of the pelvic floor and abdominal wall, ing PFMT versus hypopressive exercises (diaphrag- and 2) pulling the belly button in towards the spine. matic breathing) plus voluntary pelvic floor muscle The intervention duration was 24 weeks, with individ- contraction (PFMT+HE) versus lifestyle advice only ual supervision twice weekly for one hour each ses- (266, 267). Participating women were attending a sion during the first 3 months, followed by once a urogynaecology service in Sao Paulo, Brazil, and had week for the last 3 months. Women were assessed at untreated stage II anterior or posterior prolapse. The baseline and at 24 weeks for prolapse symptom se- PFMT intervention involved three physiotherapy ap- verity (P-QOL), prolapse severity (POP-Q) and PFM pointments to learn how to perform PFMT correctly strength (Oxford scale). The trial was judged as low (weeks 0, 1, 2). Then a 12-week home exercise pro- risk for selection bias (randomisation only), perfor- gramme was prescribed consisting of three sets of mance and detection bias and reporting bias. The trial exercises daily. The PFMT+HE group had three ap- was judged as unclear risk of bias for selection bias pointments (weeks 0, 1, 2) during which they had in- (allocation concealment), attrition bias, and other struction in correctly performing hypopressive exer- types of bias (264) (Table 28). cises and pelvic floor muscle contractions, and how to do these simultaneously. They practiced the exer- Hagen and colleagues carried out a multicentre trial cises for 12 weeks. Both of these groups completed (POPPY trial) comparing an individualised PFMT pro- an exercise diary, had a telephone call from the phys- gramme (including lifestyle advice) with a lifestyle ad- iotherapist every two weeks and a monthly appoint- vice leaflet (control group) (263). Participants were ment (weeks 6, 10, 14). The control group received outpatients attending with newly-diagnosed, sympto- one appointment, were given lifestyle advice and matic stage I, II, or III prolapse. The PFMT interven- were instructed how to perform PFM contractions. tion which was delivered by a specialist women’s However, it was unclear with whom the appointment health physiotherapist over 16 weeks in 5 sessions was held, or whether the instruction was verbal or in- included teaching of anatomy and function of pelvic cluded a digital assessment. A standardised lifestyle floor muscles, the correct exercise technique and 'the advice sheet was given to all women containing Knack', and a prescription of a home exercise pro- global stretching exercises and advice on weight loss, gramme. A standardised lifestyle advice leaflet and constipation, coughing and avoidance of heavy lifting. tailored lifestyle advice were given. The control group received the lifestyle advice leaflet by post. Outcomes

110 Table 27 Summary of data on POP prevention

Author, Comparator N Study population Modality details or parameters Outcomes/results Follow up Notes (side effects, loss of follow year up…)

Bø Parallel group 175 Primiparous All women: written advice in delivery Stage of POP and Assessed at 6 Randomisation was computer 2013 RCT; 2 groups, women, vaginal ward to perform PFMT. 2 trained bladder neck position. weeks (pre- generated and opaque sealed (1) PFMT vs control delivery, singleton physiotherapists taught PFM Secondary outcomes: test) and 6 envelopes were used. Outcome infant, ≥ 32 weeks contraction and assessed PFM symptoms of POP months’ assessors were blinded and the groups function before randomisation by (sensation of bulging) (post-test) were comparable at baseline. Intention- observation of the perineum and using International postpartum to-treat analysis was employed and vaginal palpation. Consultation on dropouts were clearly reported. The trial PFMT group: weekly PFMT classes Incontinence Vaginal was judged as potentially having an delivered by trained Symptoms unclear risk of bias as the authors noted physiotherapists for 4 months questionnaire that "because of ethical reasons the starting at 6-8 weeks postpartum. control group was not discouraged from Performed home-based exercise, 3 performing PFMT on their own, but they sets of 8-12 close to maximum PFM were asked to follow the prescription for contractions per day. the group they were randomised to." Control group: no further supervision or follow-up during the intervention period. Hagen Parallel group 407 Primi and Intervention: 1:1 PFMT taught using POP-SS, Prolapse- Data Randomisation was generated using a 2014 RCT; 2 groups, multiparous digital palpation delivered in 5 related QoL, uptake of collected at computer program located at the Trial (2) PFMT versus women assessed physiotherapy appointments over 16 prolapse treatment, baseline, 1 Office on a password-protected PC. control (lifestyle for pelvic floor weeks, followed by Pilates-based symptoms of urinary and 2 years Outcome assessors were blinded and advice leaflet) dysfunction 12 classes, including PFMT. Classes incontinence, post the groups were comparable at years after an were carried out in 6-week block anorectal or sexual randomisation baseline. Intention-to-treat analysis was index birth; women (one class per week) and each dysfunction, women’s employed and dropouts were involved in the woman was offered two 6-week perceived health adequately reported. ProLong cohort blocks. Exercise DVD was provided benefit, and cost- Questionnaire response rate: 81% year study (3) who have for home use. Women offered a 1:1 effectiveness 1, 86% year 2 follow-up. POP-Q stage I, II physiotherapy annual review Attendance at annual review or III); no previous appointment at 1 and 2 years after appointments: 52% and 46% at year 1 treatment for randomisation. and year 2 respectively, and uptake of prolapse (surgery, Control group: sent a lifestyle advice classes in the UK was 33% and 17% at pessary, PFMT) leaflet containing advice on weight 1st and 2nd block respectively. By year loss, and avoidance of constipation, 2, 77% in the intervention group heavy lifting, coughing and high reported they had done PFM exercises impact exercise. in the last 4 weeks

111 Abbreviations: PFM: pelvic floor muscles; PFMT: pelvic floor muscle training; POP-SS: pelvic organ prolapse symptom score; QoL: quality of life; RCT: ran-domized controlled trial 1. Bo K, Hilde G, Tennfjord MK, Jensen JS, Siafarikas F, Engh ME. Randomized controlled trial of pelvic floor muscle training to prevent and treat pelvic organ prolapse in postpartum primiparous women. Neurourology and Urodynamics. 2013;32(6):806-7. 2. Hagen S, Glazener C, McClurg D, MacArthur C, Herbison P, Wilson D, et al. A multicentre randomised controlled trial of a pelvic floor muscle training intervention for the prevention of pelvic organ prolapse (PREVPROL). Neurourology and Urodynamics. 2014;33(6):852-3. 3. Glazener C, Elders A, Macarthur C, Lancashire RJ, Herbison P, Hagen S, et al. Childbirth and prolapse: long-term associations with the symptoms and objective measurement of pelvic organ prolapse. BJOG. 2013;120(2):161-8.

Table 28 Summary of data on PFMT vs no active treatment for POP

Study Notes (side effects, loss of Author, year design/ N Study population Modality details or parameters Outcomes/results Follow up follow up…) Comparator Alves 2015 (1) Parallel 30 Women ≥5 years Intervention: PFMT taught using digital POP-Q, ICIQ-VS, Assessed Generally low ROB. group RCT, 2 postmenopausal, palpation. Groups of 7 women; 12 ICIQ-UI SF, ICIQ- at baseline Randomisation by draw, with each groups, with sessions, twice weekly for 30 minutes, OAB, PFM and post participant blindly drawing a PFMT vs urogynecological over six weeks. Include pelvic mobility assessment (digital intervention sealed envelope containing a pre- control complaints exercises, stretching, strengthening and palpation, sEMG), (6 weeks). printed card; however it was relaxation in supine, sitting, on Gym treatment unclear whether opaque Ball, squatting and standing, along with satisfaction (VAS, envelopes were used. Outcome PFM contractions, consisting of 4 sets of 0-10) assessors were blinded. Sample 10 fast contractions and 4 sets of 10 size calculation was reported and sustained contractions, 8 seconds hold the groups were comparable at followed 16 seconds relaxation baseline. ITT not employed but dropouts were adequately All women: performed a Fitness reported. High attrition rate for the Program based on global muscle intervention group; 30/42 women stretching, endurance and functional completed the trial. exercises for the elderly Culligan 2010(2) Parallel 62 Non pregnant Intervention: 1:1 PFMT, 24 1-hour PFDI-20; PFIQ-7, Assessed Generally low ROB. Randomised group RCT; 2 community women sessions, twice weekly for 12 weeks. PFM strength at baseline using blocked random assignment, groups, with POP-Q stage Included computerized biofeedback, (vaginal pressure) and post- and allocation concealment used standardised I, with or without vaginal manipulation, neuromuscular re- intervention sequentially numbered opaque PFMT complaint of pelvic education, and manual therapy. Authors at 12 weeks sealed envelopes. Unclear program vs floor dysfunction imply women taught contraction used whether participants and outcome standardised digital palpation. assessors were blinded. Groups were comparable at baseline.

112 Study Notes (side effects, loss of Author, year design/ N Study population Modality details or parameters Outcomes/results Follow up follow up…) Comparator Pilates Control: Pilates, 24 1-hour sessions, Although intention-to-treat analysis program twice weekly for 12 weeks. Taught full- was not employed, dropouts were body exercises designed to emphasize adequately reported. Authors the “core muscles” and pelvic floor. reported that seven of the eight participants who dropped out of the PFMT arm reported doing so because they found the treatments “unpleasant in some way”. All Pilates and PFMT sessions were free to the participants, who were also paid up to $75 for completion of the study protocol. Due 2016 (3) Parallel 109 Women ≥18 years Intervention: group PFMT and home POP-Q; PFDI-20; Assessed Low ROB for selection bias, group RCT; 2 with POP training after a digital assessment and PFIQ-7; PISQ-12; at baseline, performance and detection bias, groups, symptoms and a individual instruction. 6 group sessions Patient Global immediately attrition bias, reporting bias and Combined POP-Q ≥ stage II delivered over 12 weeks. Include the Index of post- other types of bias group (PFMT Knack, home exercise 5 days a week (3 Improvement scale intervention "Majority of the women in the plus sets of up to 10, 10 second (PGI-I) (3 months) lifestyle advice group (p≤0.001) structured contractions). Plus lifestyle advice and 6 had sought further treatment at six lifestyle program as below. months months follow-up, mainly as advice Control group: Lifestyle advice program, PFMT" program) and 6 1-hour group teaching sessions lifestyle including information on POP-promoting group factors, reducing pelvic floor pressure, (structured bladder and bowel function, diet, body lifestyle image and physical activity. No PFMT. advice program alone)

Frawley 2012 (4) Parallel 168 Women with This trial used the same protocol as the POP-Q; POP-SS; Data Generally low ROB. group RCT; 2 symptomatic POP UK POPPY trial with some slight PFM manometric collected at Randomisation was computer groups, of stage I, II or III adjustments and additional strength and baseline generated and participants were PFMT vs measurements (see Hagen 2010 below) endurance; digitally and 6 and allocated to the trial using a control assessed PFM 12 month remote randomisation service. (lifestyle strength/endurance follow up Outcome assessors were blinded. (ICS scale) Potentially high risk for attrition

113 Study Notes (side effects, loss of Author, year design/ N Study population Modality details or parameters Outcomes/results Follow up follow up…) Comparator advice bias. Dropouts were accounted for leaflet) but no explanations given for high attrition rates at 6 and 12 month follow up. Attrition was greater in the PFMT group at 6 months (14.3% vs 10.7%) but greater in the lifestyle group at 12 months (5.6% vs 21.3%).

Giraudo 2011 (5) Parallel 44 Women with Both interventions: PFMT delivered over Prolapse symptom Assessed ROB was judged as low or group RCT; 2 untreated stage I or 24 weeks, 1:1 supervised sessions severity measured at baseline unclear. Randomisation was groups, II prolapse. No twice-weekly for 1 hour for 3 months, via P-QOL and at 24 computer generated, but allocation PFMT plus mention of whether once a week for last 3 months. Women Questionnaire. weeks concealment was not described. negative women had taught to contract the PFMs correctly by Prolapse severity The physician involved in the trial pressure symptoms. vaginal palpation. Tailored lifestyle (POP-Q) and PFM was blinded to study group abdominal advice given on reducing intra- strength (Oxford allocation, but there was work vs abdominal pressure plus standardised scale). ICIQ-UI SF insufficient information about the PFMT plus lifestyle advice sheet (advice on weight also used but not blinding of participants. The abdominal loss, constipation, avoidance of heavy reported. groups were comparable at hollowing lifting, coughing and high-impact baseline and all pre-specified exercise exercise). outcomes were reported. However Groups: individualised PFM contraction although dropouts were accounted with either abdominal hollowing for, it was not clear when the exercises or negative pressure women were excluded, and which abdominal work. group they had originally been allocated to.

Hagen 2014(6) Parallel 447 Female outpatients Intervention: Anatomy/function of PFMs POP-SS; POP-Q; Assessed Low ROB. Randomisation was group RCT; 2 with newly- explained, taught to correctly contract women’s perceived at 12 computer generated, and groups, diagnosed, PFMs (vaginal palpation, PERFECT) change in prolapse; months, university-based trial coordinator individualised symptomatic stage and pre-contracting against increases in interference due to except accessed the web-based programme I, II, or III POP abdominal pressure. 5 physiotherapy prolapse; days of POP-Q application and then informed the of PFMT vs appointments delivered over 16 week symptoms; uptake assessed at woman, and the physiotherapist as prolapse period (weeks 0, 2, 6, 11 & 16). Home of prolapse 6 months necessary, of the allocated group. lifestyle exercise prescribed - 3 sets of exercises treatment; ICIQ UI- Outcome assessors and advice leaflet per day: 10 maximum voluntary SF; PISQ; SF-12 investigators (who were contractions held for up to 10s, with gynaecologists at trial sites), were

114 Study Notes (side effects, loss of Author, year design/ N Study population Modality details or parameters Outcomes/results Follow up follow up…) Comparator 4srest between; after 1 minute rest, up masked to group allocation; the to 50 rapid contractions. Standardised statistician was masked until after lifestyle advice leaflet and tailored data analysis. ITT analysis was lifestyle advice given. employed and dropouts were Control group: received the same adequately reported. lifestyle advice leaflet by post.

115 Study Notes (side effects, loss of Author, year design/ N Study population Modality details or parameters Outcomes/results Follow up follow up…) Comparator Kashyap 2013 (7) Parallel 140 Parous women Intervention: anatomy/function of PFMs POP-SS; Assessed Low risk for selection bias as the group RCT; 2 attending explained. 1:1 PFMT training, plus symptoms at baseline trialists had used a block groups, gynaecology exercise practice and checking for measured by visual and at 1, 3, randomisation method for PFMT and a outpatient correct contractions of PFM by vaginal analog scale (VAS), 6, 12, 18 randomisation, and a physician, self- department, aged examination. Pressure manometry for PFIQ-7, POP-Q and 24 who was not involved in the study, instruction 20–70 years, who biofeedback and motivation. Given SIM. weeks performed the patient allocation. manual (SIM) were willing to After training session, 6 follow up visits Potential for high risk of bias, as vs SIM alone attend follow-up (weeks 1, 3, 6, 12, 18 and 24). Home “the study design did not include visits exercise programme of 3 times/daily set blinding”. There was a crossover of exercises comprising 10 voluntary of 4 participants from group B to contractions, held for 10s each with 10s group A after randomisation. rest in between. Participants performed Details of attrition were not well the intervention as described for 24 reported: it was indicated that weeks. analysis was performed on all 140 Control: SIM plus 3 follow-up visits participants however there was (weeks 6, 18 and 24). Home exercise also reference to individuals prescribed as above. “missing” and “lost to follow-up”. Self-completed exercise diary but unclear how well adherence was actually delivered as planned.

Resende 2008/ 3-arm RCT; 58 Women with stage PFMT: Session 1 (week 0), anatomy/ POP-Q; symptom Baseline Low or unclear ROB. Stupp 2011 PFMT II POP and not function of PFMs explained, training on severity (Prolapse evaluation Randomisation was computer (8, 9) training vs undergoing surgery PFM contractions, observation on Quality of Life (P- and generated, but allocation hypopressive to correct it during correct performance. Follow up sessions QoL)); PFM immediately concealment was not described.

116 Study Notes (side effects, loss of Author, year design/ N Study population Modality details or parameters Outcomes/results Follow up follow up…) Comparator exercises the study. Not on at weeks 1 (vaginal palpation), 2, 6, 10 function after The main investigator was blind to plus PFMT hormonal and 14. 12 week home exercise assessment intervention the study groups and not involved vs lifestyle treatment, not had programme of 3 sets of exercises daily: (strength and (at 3 in delivering the intervention. advice & previous PFMT or 1 set of 8-12 maximum contractions held endurance), months) Groups were comparable at instruction on pelvic floor surgery. 6-10s, double time rest between each electrical activity baseline and dropouts were PFM Recruited during contraction, then 3-5 fast contractions. measured (sEMG) accounted for. Attrition was contraction routine Bi-weekly telephone calls. evident only in the control group (1 consultation. PFMT+HE: Hypopressive exercises plus where 16/21 (76%) were analysed. appointment) PFMT (3 sessions). Monthly Single trial reported over multiple appointments, 12 week home exercise papers (confirmed by author). programme. Bi-weekly telephone calls. Reports do not agree about number of participants, age. Control group: 1 appointment for lifestyle Authors state: “The women who advice and instruction in PFM took part in the present study also contraction. participated in a randomized, controlled trial on PFMT to reduce All: Standardized lifestyle advice sheet POP, the results of which are not about weight loss, fluid intake, available yet. Therefore, in this constipation and avoidance of heavy study, we considered only the lifting. results regarding PFM function.” Wiegersma 2014 Two parallel 287 Women aged at Intervention: Explanation of the function Assessed 3 months Data Low ROB. Block randomisation; (10) RCTs. least 55 years with of pelvis/pelvic floor/ pelvic floor from the start of collected at independent statistician generated POPPS 1 symptomatic mild dysfunctions, taught “the Knack”. treatment (or from baseline the allocation sequence. Research (described POP identified via Weekly visits to pelvic physiotherapist randomisation for and at 3, physicians and pelvic here) 2 screening survey in until able to correctly contract and relax control group): 12, and 24 physiotherapists were blinded to groups; 15 GP practices PFMs (assessed by digital palpation), PFDI-20, PFIQ-7, months all answers on the participant PFMT vs then 2-3 weeks intervals. Home exercise POP-Q, SF-12, after the completed questionnaires, and watchful 3-5 times a week, 2 or 3 times a day. PISQ-12, patients’ start of research physicians were blinded waiting Advice on lifestyle (diet, body weight) perceived change treatment to the outcomes of the previous and toilet habits. in symptoms (VAS) POP-Q measurements and Control: no treatment or , PFM function (ICS previous evaluations of pelvic floor recommendations method) muscle function. ITT analysis employed; dropouts and missing data were accounted for.

117 Abbreviations: PFDI-20: pelvic floor distress inventory short form 20, PFIQ-7: pelvic floor impact questionnaire short form; PFM: pelvic floor muscles; PFMT: pelvic floor muscle training; POP-Q; Pelvic Organ Prolapse Quantification system; PISQ-12: Pelvic organ prolapse/urinary incontinence sexual questionnaire’ POP-SS: pelvic organ prolapse symptom score; QoL: quality of life; RCT: randomized controlled trial; sEMG: surface electromyography 1. Alves FK, Riccetto C, Adami DB, Marques J, Pereira LC, Palma P, et al. A pelvic floor muscle training program in postmenopausal women: A randomized controlled trial. Maturitas. 2015;81(2):300-5. 2. Culligan PJ, Scherer J, Dyer K, Priestley JL, Guingon-White G, Delvecchio D, et al. A randomized clinical trial comparing pelvic floor muscle training to a Pilates exercise program for improving pelvic muscle strength. Int Urogynecol J. 2010;21(4):401-8. 3. Due U, Brostrom S, Lose G. Lifestyle advice with or without pelvic floor muscle training for pelvic organ prolapse: a randomized controlled trial. Int Urogynecol J. 2016;27(4):555-63. 4. Frawley HC, Hagen S, Sherburn M, Neumann P, Herbison P, Hay-Smith J, et al. Changes in prolapse following pelvic floor muscle training: A randomised controlled trial. Neurourology and Urodynamics. 2012;31(6):938-9. 5. Giraudo D, Beccaria N, Lamberti G. Pelvic floor muscle training, negative pressure abdominal exercise and pelvic organ prolapse symptoms: A randomized clinical trial. Neurourology and Urodynamics. 2011;30(6):1009-11. 6. Hagen S, Stark D, Glazener C, Dickson S, Barry S, Elders A, et al. Individualised pelvic floor muscle training in women with pelvic organ prolapse (POPPY): a multicentre randomised controlled trial. Lancet. 2014;383(9919):796-806. 7. Kashyap R, Jain V, Singh A. Comparative effect of 2 packages of pelvic floor muscle training on the clinical course of stage I-III pelvic organ prolapse. Int J Gynaecol Obstet. 2013;121(1):69-73. 8. Stupp L, Magalhaes Resende AP, Oliveira E, Castro RA, Castello Girao MJB, Ferreira Sartori MG. Pelvic floor muscle training for treatment of pelvic organ prolapse: An assessor-blinded randomized controlled trial. International Urogynecology Journal and Pelvic Floor Dysfunction. 2011;22(10):1233-9. 9. Resende APM, Stupp L, Bernardes BT, Oliveira E, Castro RA, Girao MJBC, et al. Can hypopressive exercises provide additional benefits to pelvic floor muscle training in women with pelvic organ prolapse? Neurourology and urodynamics. 2012;31(1):121-5. 10. Wiegersma M, Panman CM, Kollen BJ, Vermeulen KM, Schram AJ, Messelink EJ, et al. Pelvic floor muscle training versus watchful waiting or pessary treatment for pelvic organ prolapse (POPPS): design and participant baseline characteristics of two parallel pragmatic randomized controlled trials in primary care. Maturitas. 2014;77(2):168-73.

118 Women were assessed at baseline and immediately 2 in the Pilates arm withdrew for health reasons un- after intervention (at 3 months) using POP-Q (blinded related to the trial. All women could correctly contract assessment); PFM strength and endurance (Oxford their pelvic floor muscles (as assessed by perineom- scale), electrical activity (sEMG) and symptom sever- etry at baseline) and most did not have pelvic floor ity and impact (P-QoL). The trial was judged as low symptoms at baseline. Both groups demonstrated im- risk for selection bias (randomisation) and unclear proved PFDI-20 and PFIQ-7 scores and pelvic floor risk for allocation concealment. The trial was also muscle strength, at 12 weeks, but these improve- judged at low risk for performance and detection bias, ments were not significantly different between the attrition bias, reporting bias and other types of bias groups. The authors concluded that it is feasible for a (266, 267) (Table 28). Pilates exercise program to strengthen the pelvic floor muscles of women with stage I prolapse. Wiegersma and team conducted a trial in primary care (POPP1) of women aged ≥55 years with symp- In the Due (2016) trial, 109 women were randomised, tomatic mild prolapse (above the hymen) (268). 56 to the combined group and 53 to the lifestyle group Women in the trial were randomised to either PFMT (261). Of these 43% had POP stage III and 57% had or watchful waiting. Women in the PFMT group were POP stage II. Follow-up at 3 and 6 months was 82% given an explanation of pelvic floor anatomy and pel- and 78% complete respectively. Significantly more vic floor dysfunction, were taught “the Knack”, and women in the combined group indicated improvement given lifestyle advice. They had weekly visits with the in the PGI-I at 3 and 6 months. There was no differ- pelvic physiotherapist initially. The intervals between ence between groups in any PFDI-20 or PFIQ-7 appointments were extended when they could cor- scores at 3 months. Both groups improved signifi- rectly contract and relax their PFMs. Home exercise cantly in the total PFDI-20 score and its subscores by was recommended three to five times a week, twice 3 months, except the lifestyle group had no improve- or three times each day. Data were collected at base- ment in the POPDI sub-score. Significant improve- line and at 3, 12, and 24 months after the start of ment in PFIQ-7 could only be found in the lifestyle ad- treatment. The primary outcomes were change in vice group. It was unclear if there were any differ- bladder, bowel, and pelvic floor symptoms, as meas- ences between groups in PFDI/PFIQ at 6 months. ured by the PFDI-20, assessed at three months. Sec- The PISQ-12 and objective prolapse stage did not im- ondary outcomes were condition specific and general prove significantly. Significantly more women (68% quality of life, sexual functioning, and degree of pro- vs 28%) in the lifestyle advice group had sought fur- lapse, PFM function, and patients’ perceived change ther treatment (mainly PFMT) at 6 months follow-up. in symptoms from the start of the study. The POPP1 The authors concluded that there was a small benefit trial was judged as low risk for selection bias, perfor- of both lifestyle advices alone or combined with PFMT mance and detection bias, attrition bias, reporting for women with stage II or III prolapse. There was a bias and other types of bias (Table 28). significant difference in the primary outcome (PGI-I) in favour of the combined group, but this was not re- Results ally highlighted. In the Alves (2015) trial, 42 women were randomised, In the Frawley (2012) trial, 168 women were random- 21 in each group, however there was high attrition ised to either the PFMT group (n=84) or the control and only data from 18 and 12 women in the interven- group (n=84), with 12 and 9 lost to follow-up by 6 tion and control group respectively were analysed. months and a further 4 and 16 by 12 months respec- Two in the intervention group did not complete the tively, 19% and 30% in total (262). 82% of women in treatment due to health or family problems, while 9 in the PFMT group attended 4 or 5 of the 5 physiother- the control group refused to complete the final physi- apist appointments. The POP-SS score was signifi- cal exam: these women were excluded from the sam- cantly lower, indicating fewer symptoms, in the PFMT ple. There was a significant difference in favour of the group compared to the control group at both 6 and 12 intervention group in terms of decreased anterior pro- months. There was no difference in POP-Q stage be- lapse, but no difference in posterior prolapse or the tween groups at 6 or 12 months, although there was ICIQ-VS score. There was a significant difference be- some evidence of a difference at 12 months between tween the groups in the ICIQ-OAB and ICIQ-UI SF groups, in the posterior wall POP-Q measurements, scores after treatment, with improvement occurring Ap and Bp, in favour of the PFMT group. Digital mus- only in the intervention group. There was a greater cle strength was significantly stronger in the PFMT increase in PFM contractility (measured both by digi- group compared to the control group at 6 months but tal palpation and sEMG) in the PFMT intervention not at 12 months. There were no significant differ- group compared to the general fitness control group ences between groups in manometry outcomes, ex- at 6 weeks. Satisfaction with the treatment was re- cept total work performed was higher in the PFMT ported to be greater in the intervention group (47). group at 6 months. The intervention was concluded Sixty-two women were randomised in the Culligan to be beneficial immediately following the interven- (2010) trial; 32 to PFMT and 30 to Pilates (control), of tion, and after a further 6 months. which 8 and 2 withdrew respectively (260). Seven of In a conference abstract, Giraudo and colleagues re- the 8 who dropped out of the PFMT arm did so be- ported 47 women were randomised but 3 were ex- cause they found the treatment unpleasant, while the cluded because they could not contract their pelvic

119 floor muscles, leaving 44, 23 intervention (PFMT + to PFMT, 21 to PFMT plus hypopressive exercise, 21 negative pressure abdominal work) and 21 control to lifestyle group. Five women in the lifestyle group (PFMT + abdominal hollowing) (264). Improvement in discontinued leaving 58 women at follow-up. The prolapse symptoms and POP-Q values from baseline PFMT group women were more likely than the control to 24 weeks was significantly greater in the interven- group to have an improvement in their POP-Q stage tion group compared with the control group. There of prolapse, both anterior and posterior. The domains was evidence of an improvement in PFM strength and and symptoms scores from the P-QoL were com- endurance in both groups, but no significant differ- pared before and after in each group separately, but ence between the groups. The authors concluded no between-group comparisons were made. How- there may be benefit of adding negative pressure ever, the authors concluded erroneously that since core exercises to PFMT for women with stage I and II the PFMT group scores improved significantly and prolapse. the control groups did not, the PFMT group had ben- efitted more. Pelvic floor muscle assessment out- Of the 447 women enrolled in the POPPY trial, 225 comes were compared between the three groups women were randomised to intervention and 222 to (266). Significant differences were found between the control (263). 84% and 66% of women completed PFMT and PFMT+HE groups when compared to the questionnaires at 6 and 12 months respectively, with control group, in favour of the PFMT groups, in terms no differential drop-out. POP-Q re-assessments at 6 of Oxford score, contraction endurance and muscle months were obtained for 75% (168) and 77% (171) activity (SEMG). No significant difference was found of women respectively. 80% in the intervention group between the PFMT and PFMT+HE groups in terms of attended 4 or 5 of the 5 appointments. Women in the Oxford score and muscle activity (SEMG), however intervention group had a significantly greater reduc- PFMT group did significantly better than PFMT+HE in tion in prolapse symptoms at 6 and 12 months than terms of contraction endurance. those in the control group, although they were no more likely than control women to have a reduced se- Wiegersma trialists screened 4,465 women to identify verity of prolapse at 6 months. At 12 months, signifi- those with mild prolapse (n=365) (268). Of these, 287 cantly more women in the control group than the in- women were randomised, 145 to PFMT and 142 to tervention group had received further treatment. In watchful waiting, of which 250 (87%) completed the particular significantly more women in the control trial follow-up. Women in the intervention group im- group (27% vs 1%) had had a referral for PFMT. At 6 proved their PFDI-20 score significantly more than months, all aspects of daily life, and sexual, bladder, those in the watchful waiting group, and were also and bowel function (except for faecal incontinence), more likely to report overall symptom improvement were significantly better in the intervention group (57% vs 13%). There were no other significant differ- compared to the control group. This was not sus- ences between the trial groups including improve- tained at 12 months. It was concluded that one-to-one ment in POP-Q stage. The authors concluded that the PFMT is effective for improving symptoms in women difference in PFDI-20 may not be clinically significant with stage I to III prolapse in the medium term. and that more studies are needed to examine the factors affecting success. In the Kashyap trial, 140 women were randomised, 70 per group, although four women transferred from Summary the control group to the training plus manual group and the group in which these women were analysed Results from an additional seven new trials are now was unclear (265). Improvements were reported in available and more complete reports for two of the six POP-SS, VAS and PFIQ-7 scores in both groups from earlier trials. baseline to week 24. There were significant differences between groups in the change in POP-SS with There have been different types of control groups the intervention group reporting greater symptom im- used in the trials to date: minimal intervention control provement. There was also significantly more im- e.g. lifestyle leaflet or watchful waiting; other type of provement in the intervention group in terms of the exercise as control e.g. Pilates or general fitness; life- VAS scores and the PFIQ-7 scores. Five women in style intervention as control; other form of delivery of the intervention group had an improved POP-Q stage PFMT as control e.g. self-instruction PFMT manual. compared to one woman in the control group. The au- One small trial compared PFMT of two different types thors concluded that one-to-one PFMT plus the SIM (PFMT with negative pressure abdominal work and led to more symptom improvement than the SIM PFMT with abdominal hollowing). alone. Most evidence (8 trials) exists for PFMT versus a min- The Stuppe-Resende trial was reported over multiple imal intervention control, and it can now be concluded publications (4 conference abstracts, 2 papers and a more confidently that PFMT significantly reduces pel- trial register entry) (confirmed by personal communi- vic floor symptoms in women with stage I to III pro- cation with an author), reporting on different compar- lapse. Evidence of effectiveness for PFMT relating to isons amongst the three trial groups, and not all re- the specific symptom of a vaginal bulge or something ports agreed on participant details (e.g. the number coming down associated with prolapse was Level 1, of participants, age) making it difficult to interpret the but less consistent. findings. 63 women were randomised in the trial: 21

120 6 months (PFDI-UDI), and prolapse symptoms Six of these trials reported change in POP-Q stage, (PFDI-POPDI) and anatomical failure (descent on the but a beneficial effect on the stage was reported in POP-Q or retreatment) at 24 months after surgery only 1 of these, and thus there was evidence of no (Table 29). effect of PFMT on prolapse stage. The OPTIMAL trial was judged as low risk for selec- In the other categories, four out of six trials providing tion bias, performance and detection bias, attrition data were small and very likely underpowered alt- bias, reporting bias and other types of bias (104) (Ta- hough they were otherwise at low risk of bias (Alves ble 29). n=42, Culligan n=62, Giraudo n=47, Stupp n=63). McClurg and colleagues undertook a pilot trial of pre- Of the remaining two, Due (n=109) found PFMT plus and post-operative PFMT for women undergoing a lifestyle advice to be superior to a lifestyle advice pro- primary prolapse repair (274). Women were random- gramme in terms of overall improvement on the Pa- ised to either the intervention group receiving PFMT tient Global Index of Improvement Scale, but not pro- or the control group receiving usual care. Prior to sur- lapse symptoms or POP-Q severity, where both gery all women were seen by a physiotherapist to groups improved [46]. Kashyap (n=140) found a complete baseline outcome measures. Those in the taught course of PFMT plus self-instruction manual intervention group were also seen once pre-opera- was better than a self-instruction manual alone in im- tively by another physiotherapist to be taught pelvic proving prolapse symptoms [50]. The difference in floor muscle exercises and the Knack. Women were findings might in part be due to different PFMT inter- advised to do three sets of exercise per day. After 6 ventions: group PFMT delivered in the Due trial ver- weeks the intervention group women were seen for 5 sus one-to-one PFMT in the Kashyap trial. visits over 16 weeks and an individualised home ex- Based on previous studies and new evidence there is ercise programme was prescribed and advice given. now evidence of benefit that PFMT is effective in re- Control group women received a lifestyle advice leaf- ducing pelvic floor symptoms in women with prolapse let. Both groups completed outcome measures at 6 (Consistent Level of Evidence: 1, Grade of recom- and 12 months post-surgery: POP-SS (primary), mendation: A). There is some evidence of benefit ICIQ-UI SF, ICIQ-BS, PISQ-12, SF-12, and PFM as- showing that PFMT is effective in alleviating specific sessment (PERFECT and modified Oxford scale) prolapse symptoms (e.g. vaginal bulge) (Majority (Table 29). The feasibility trial was generally judged Level of Evidence: 1, Grade of recommendation: as low risk for selection bias, attrition bias, reporting C). There is no evidence that PFMT is effective in re- bias and other types of bias. However, there was po- ducing severity of prolapse based on POP-Q stage tential for performance and detection bias (see Table (Consistent Level of Evidence: 1, Grade of recom- 29) (274). mendation: B). Pauls et al. (2011) (271, 272) undertook a trial in women undergoing vaginal reconstruction, compar- 2.2.2 PFMT and Surgery ing PFMT as an adjunct to surgery with standard Two trials have been reported in previous editions of care. Women were having planned surgical correc- this chapter (269, 270). There have been an addition, including a native tissue vaginal repair with or tional 4 trials published since the last edition (104, without a vaginal hysterectomy or suburethral sling. 271-274). Three trials compared surgery plus PFMT PFMT consisted of one appointment 2 weeks before with surgery alone, and one trial compared surgery their scheduled surgery date, and five post-opera- with PFMT (275). The latter did not report on prolapse tively, each with a specialist pelvic floor physiothera- outcomes and is therefore not discussed further. Sa- pist. The control group attended appointments with lient features of each trial are described in Table 29. physician assessment alone at all the same post-operative intervals. There was a final follow-up assess- Quality of data ment at 24 weeks (Table 29). The primary outcome Barber and colleagues carried out the OPTIMAL 2x2 was quality of life measured using the WHOQOL- factorial trial of PFMT as an adjunct to vault prolapse Bref. Secondary outcomes included PFDI-20, PFIQ- 7, POP-Q assessment, the Female Sexual Function surgery (104). This trial compared two methods of suspending the vaginal vault in women undergoing Index (FSFI), PISQ-12, a modified Oxford scale for pelvic floor strength and contraction, short form Gen- surgery for prolapse. Additionally, participants were randomised either to adjunctive post-operative PFMT eral Health Survey (SF-12), and a 24-hour voiding diary (271, 272). The trial was judged as low risk for or routine care, to assess whether such adjunct therapy improves both anatomical and symptomatic out- selection bias, performance and detection bias, attrition bias, reporting bias and all other types of bias comes two years after surgery. (Table 29). The adjunct intervention consisted of one pre-operative visit and four post-operative visits with a behavioural interventionist for PFMT and education in behavioural strategies. Routine care was usual peri-operative teaching and post-operative instructions. The authors’ primary outcomes were urinary symptoms at

121 Table 29 Summary of data on PFMT + surgery for POP

Notes (side effects, loss of Author, year Comparator N Study population Modality details or parameters Outcomes/results Follow up follow up) Barber 2014 (1) 2 x 2 factorial 374 Women having Women were randomised to Outcomes for PFMT: 6, 12 and 24 RCT: comparing surgical repair for both surgery type and PFMT. long-term improvement months Low ROB for selection bias, to surgery types apical or uterine Surgery: 1) sacrospinous in anatomic outcomes performance and detection +/- PFMT POP of stage II or ligament fixation, or 2) (POP-Q) and prolapse bias, attrition bias, reporting greater, who also uterosacral vaginal vault symptoms (POPDI bias and other types of bias have SUI suspension subscale of the PFDI); Perioperative PFMT: 1) 1:1 short term (6 months) PFMT (1 pre-operative + 4 post- improvement in urinary operative visits for PFMT symptoms (UDI subscale (2,4,6,8 and 12 weeks), of PFDI). examination at each visit, and exercise and education in behavioural strategies), or 2) usual care (usual peri-operative teaching and post-operative instructions) McClurg 2010 (2) Parallel group 57 Women attending Intervention: 1 pre-operative + 5 POP-SS; ICIQ-UI SF; All outcomes Generally judged as low ROB. RCT; 2 groups, the post-operative appointments ICIQ-BS; PISQ-12; SF- were measured Randomisation generation and treatment group gynaecological within a period of 12 weeks. 12; PFM assessment at baseline, 6 allocation concealment were (surgery and clinic and for Pre-operatively, anatomy and (PERFECT, modified and 12 months judged as adequate. Outcome PFMT sessions) whom primary function of PFMs, types of Oxford scale). assessors reported as being vs a control surgery was prolapse and the surgical blinded but researchers group (usual recommended procedure discussed with involved in the trial were not care) due to their POP information about blinded and it is unclear what symptoms recovery/return to normal direct involvement they may activities. Women taught by have had with each digital palpation to contract participant. Groups were PFMs and 'the Knack'. Home comparable at baseline, but exercise, 3 sets per day of 10 there were more complaints of maximum contractions (up to bowel dysfunction in the 10s hold) with 4s rest between, treatment group. Dropouts 1-min rest followed by 10 fast were clearly accounted for and contractions. all pre-specified outcomes were adequately reported. Control: received the same lifestyle advice leaflet.

122 Notes (side effects, loss of Author, year Comparator N Study population Modality details or parameters Outcomes/results Follow up follow up) Pauls 2013 Parallel group 49 Women aged > 18 Intervention: physiotherapy PFDI-20, PFIQ-7, POP- Assessments Low ROB. Randomisation (3) RCT; 2 groups, years having appointment 2 weeks before Q, Female Sexual undertaken at generation and allocation PFMT vs surgical correction surgery, and 2, 4, 6, 8, and 12 Function Index (FSFI), baseline (some concealment were judged as standard care in to include a native weeks postoperatively, in PISQ-12, SF-12, WHO- limited adequate. Outcome assessors women tissue vaginal conjunction with a physician QOL Bref, modified assessment at were blinded appropriately. undergoing repair +/- a assessment. Sessions covered Oxford scale for PFM appointments Groups were comparable at vaginal vaginal bladder and bowel function, pain strength and contraction, occurred at 2, 4, baseline, dropouts were all reconstruction hysterectomy or management, breathing and 24-hour voiding diary 6, 8 weeks), and accounted for and all pre- suburethral sling relaxation, core exercises, scar follow-up specified outcomes were tissue mobilization, increased assessments at reported. strengthening and training over 12 and 24 time. weeks The trial is fully reported in two main papers. Control: attended appointments (biweekly until 12 weeks postoperatively) with physician assessment alone at intervals as above.

Abbreviations: PFM: pelvic floor muscles; PFMT: pelvic floor muscle training; POP: pelvic organ prolapse; POP-SS: pelvic organ prolapse symptom score; QoL: quality of life; RCT: randomized controlled trial; ROB: risk of bias; SUI: stress urinary incontinence; UI: urinary incontinence 1. Barber MD, Brubaker L, Burgio KL, Richter HE, Nygaard I, Weidner AC, et al. Comparison of 2 transvaginal surgical approaches and perioperative behavioral therapy for apical vaginal prolapse: the OPTIMAL randomized trial. JAMA. 2014;311(10):1023-34. 2. McClurg D. A two group, single-blind, randomised controlled study to assess the feasibility of physiotherapy following surgery for prolapse to avoid recurrence. http://isrctnorg/ISRCTN08203452010. 3. Pauls RN, Crisp CC, Novicki K, Fellner AN, Kleeman SD. Impact of physical therapy on quality of life and function after vaginal reconstructive surgery. Female Pelvic Med Reconstr Surg. 2013;19(5):271-7.

123 Results severity (POP-Q), and perceived change in prolapse since pessary fitted) measured at 7 months. The au- In the OPTIMAL trial, 408 women were randomised, thors aimed to randomise 50 women from 4 centres 34 withdrew prior to surgery leaving 374 women ran- to inform the development of a larger trial. The domised to PFMT (n=186) or usual care (n=188). PEPPY trial was judged as low risk for selection bias, There was no significant difference at 6 months or 24 performance and detection bias, attrition bias, report- months between the PFMT and usual care groups in ing bias and all other types of bias (276, 277). the prolapse scores or POP-Q. The routine use of PFMT was concluded to be unnecessary (104). ii) PFMT vs PFMT + pessary McClurg (2010) randomised 57 women from three Cheung and colleagues carried out a trial in women sites, 28 to PFMT and 29 to control, with the majority with symptomatic stage I to III prolapse with no previ- of women (n=27, 47%) coming from one site. By 6 ous treatment for prolapse, who were randomised to months there was significant improvement in both either a vaginal ring pessary+ PFMT, or PFMT alone groups on prolapse, bladder, bowel and general (278). Both groups were taught and encouraged to do health measures, but no difference between groups. PFMT standardized pelvic floor exercise training Analysing 12 month data from the highest recruiting course which included a teaching session within 2 site where longer follow-up was possible (there were weeks after the first consultation and three individual significant recruitment and logistical issues at other training sessions at 4, 8, and 16 weeks. Women were sites which caused delays and limited the follow-up), advised to practice daily with at least two sets of 8– there were significant differences between groups in 12 preset exercise repetitions per day, with 8– 10 ex- POP-SS and SF-12: prolapse symptoms and general ercises per session at least two times per week. health were more improved in the PFMT groups Change in urinary symptoms was measured using (274). the PFDI (including PFDI-UDI) before, 6 months and 12 months after the treatment. The PFIQ was also In Pauls (2013) trial, a total of 57 women were ran- completed at 12 months. The trial was judged as low domised, 29 to physiotherapy and 28 to control, and risk for selection bias, performance and detection 49 completed the study (24 and 25 respectively). Im- bias, attrition bias, reporting bias and other types of provement over baseline was found in both groups in bias (Table 30). quality of life, PFDI and PFIQ measures, but there were no differences between groups. The PFMT iii) Colpexin sphere +PFMT vs PFMT group had better muscle strength after 12 weeks, but at 24 weeks this was no longer evident. (271, 272). Manonai and colleagues (2012) studied the use of the Colpexin sphere in women with Stage I or II prolapse (279). The Colpexin sphere is an intra-vaginal device Summary similar to a pessary except it requires the woman to Although there were three new randomised studies actively contract her pelvic floor muscles to keep the reporting prolapse outcomes, only one trial device in place. Women were randomised to either (OPTIMAL) was both at low risk of bias and of ade- PFMT alone (control) or PFMT along with a Colpexin quate size (104). It found no evidence of an effect on sphere (study group). All women were instructed to prolapse symptoms or stage at 2 years of adding perform three sets of exercises daily. Those in the PFMT to surgery in women having vault prolapse re- Colpexin group exercised with the device in situ. The pair (Level of Evidence: 1). The other two small trials intervention duration was 16 weeks. Pelvic floor mus- provided no evidence of an effect of PFMT. cle strength was measured using the Colpexin pull test and digitally using the Brink scale, at baseline, 4, Peri-operative PFMT does not improve prolapse 8, 12 and 16 weeks after treatment. There were no symptoms in women undergoing surgery for vault specific prolapse outcomes reported. Participants prolapse (Grade of Recommendation: B New). were excluded if compliance with the daily pelvic floor exercise was less than 80%. The trial was judged as 2.2.3 PFMT and Pessary low risk for selection bias, performance and detection Three RCTs have been published since the 5th Edi- bias, attrition bias, reporting bias and other types of bias (Table 30). tion when previously there had been none. Details of the categories and treatment components of the ac- Results tive interventions are provided in Table 30. Of the 31 eligible women recruited to the PEPPY trial, i) Pessary + PFMT vs pessary 16 were randomised, eight to the pessary alone group and eight to the pessary plus PFMT group. The Hagen and colleagues carried out a pilot trial in which women with stage I to IV prolapse of any type who mean age of women was 63 years (SD 14); 25% had stage I prolapse, 50% stage II, and 25% stage III. had successfully been fitted with a pessary were randomised to have PFMT or not. PFMT was delivered Compliance with the intervention was good: 75% of intervention women attended 4 or 5 appointments. by a specialist women’s health physiotherapist in 5 appointments over 16 weeks (276) (Table 30). The With such a small sample size statistical analysis was pessary was removed at 6 months and outcomes not carried out. (POP-SS, prolapse-related quality of life, prolapse

124 Table 30 Summary of data on PFMT + pessary / other medical devices for POP

Author, Comparator N Study Modality details or Outcomes/results Follow up Notes (side effects, loss of follow year population parameters up…)

Cheung Parallel group 276 Women had Intervention: vaginal pessary, Total scores on PFDI Data collected at baseline Low ROB for selection bias, 2016(1) RCT; 2 groups, symptomatic (ring pessary was used). and PFIQ. Change of and 6 months (PFDI-UDI performance and detection bias, Vaginal ring prolapse (Stage Women also received regular urinary symptoms only) and 12 months after attrition bias, reporting bias and other pessary + PFMT I- Stage III), PFMT. Home exercise measured by PFDI- treatment. types of bias vs PFMT had received no performed at least 3 times a UDI and subscales previous week and 2 times each day. (obstructive, irritative, treatment Control: PFMT alone. stress). Hagen Feasibility pilot 16 Women with PFMT was delivered at 5 POP-SS, prolapse- Data collected at baseline Low ROB for performance and 2010(2) RCT; 2 groups, prolapse of any appointments over 16 weeks (as related quality of life, (after pessary fitted but detection bias, attrition bias, PFMT in type, of stage I per Hagen 2010 in Table 28) prolapse severity before randomization), 6 reporting bias and all other types of conjunction with to IV, with a with the pessary in place. (POP-Q), and months post- bias. pessary pessary newly perceived change in randomization management vs successfully Control: pessary management prolapse since (with pessary in situ, then Poor recruitment was a key issue pessary fitted (still in alone. pessary fitted. pessary removed), and 7 within the trial – target was to management place after 2 months post- randomise 50 women alone weeks) randomization without pessary. Manonai Parallel group 91 Women aged All participants: taught about ICIQ-VS (Thai POP-Q and ICIQ-VS at Low ROB. Random sequence 2012(3) RCT; 2 groups, 20 years+, home-based practice with version), POP-Q, baseline and 16 weeks. generation was computer generated Colpexin sphere prolapse stage I booklet of PFM. Visual PFM strength PFM data collected at and information for decoding with PFMT vs or II inspection of contraction. Home (Colpexin pull test, baseline, 4, 8, 12 and 16- randomisation was kept secure and PFMT alone exercise involved tightening the digital test using week after starting opaque sealed envelopes were used PFMs, holding for 10s, relaxing Brink scale) treatment. during the allocation process. 10s and doing 10 repetitions 3 Although ITT analysis was not times a day for a period of 16 employed, all dropouts were clearly weeks. accounted for. The groups were Colpexin+PFMT: exercised as comparable at baseline and the above with the Colpexin sphere authors reported a sample size in situ. calculation. Abbreviations: PFM: pelvic floor muscles; PFMT: pelvic floor muscle training; POP-SS: pelvic organ prolapse symptom score; RCT: randomized controlled trial 1. Cheung RY, Lee JH, Lee LL, Chung TK, Chan SS. Vaginal Pessary in Women With Symptomatic Pelvic Organ Prolapse: A Randomized Controlled Trial. Obstet Gynecol. 2016. 2. Hagen S. Pessary Plus Physiotherapy for Pelvic Organ Prolapse (PEPPY). http://ClinicalTrialsgov/show/NCT011368892010. 3. Manonai J, Harnsomboon T, Sarit-apirak S, Wattanayingcharoenchai R, Chittacharoen A, Suthutvoravut S. Effect of Colpexin Sphere on pelvic floor muscle strength and quality of life in women with pelvic organ prolapse stage I/II: a randomized controlled trial. Int Urogynecol J. 2012;23(3):307-12. 125 However, from observing the mean scores it was ap- PFMT + pessary may be as effective as PFMT alone parent that symptoms in both groups were worse a in reducing symptoms. (Grade of Recommenda- month after the pessary had been removed (month 7) tion: B New; however; some caution since the two compared to baseline (when the pessary had been in trials considered two very different devices). place for 2 weeks), and there was no indication of a symptom or objective benefit for those women who had received PFMT on any of the outcomes. In both groups 2 out of 7 women said their prolapse was the 3. PESSARIES same or worse at 6 months after the pessary had been inserted. Recruitment was problematic in the A pessary is defined as a device that is inserted into trial, and it was concluded that this would need to be the vagina to provide structural support to one or addressed before moving to a larger trial (276, 277). more of descending vaginal compartments, i.e., the Of the initial 311 women recruited to the Cheung trial, uterus, anterior vaginal wall (and bladder), posterior 276 were randomised, 137 to PFMT treatment and vaginal wall (and rectum) and/or vaginal apex (with or 139 to vaginal pessary plus PFMT. Authors reported without small intestine after a prior hysterectomy in one abstract an intention-to-treat analysis of the dif- (280). They offer a non-surgical option for the treat- ference between the groups in PFDI-UDI and its sub- ment of urinary incontinence and pelvic organ pro- scores, and on the prevalence of women with SUI, lapse (POP). This section, discusses evidences for UUI and voiding difficulties (278). Prolapse outcomes use of pessary to prevent or treat POP; evidence for were not reported. Although there was significant im- use of pessaries to prevent or treat UI will be covered provement on all outcomes at 12 months, there was in chapter 20. no significant difference between the groups. In a A range of vaginal pessaries (Figure 8) exist which second abstract the authors reported a non-random- can be broadly divided into two types: support and ised comparison of those women who continued suc- space-filling pessaries. Support pessaries lie along cessfully with their pessary until 12 months (n=78) the vaginal axis, with the posterior component sitting and those who continued successfully with pelvic in the posterior fornix and the anterior component floor exercises only (n=118) (278). At 12 months the coming to rest just under the symphysis pubis, thus pessary group had significantly better scores that the providing a supportive shelf for the descending pelvic PFMT group on all the PFDI and PFIQ scores and organs. As there is no evidence to support the use of subscores. This would seem to suggest that generally a specific type of pessary, choice is based on clinical pessary and pelvic floor exercises are equally effec- experience and trial and error. It is generally accepted tive at improving urinary symptoms in women with that the ring pessary should be tried first because of prolapse, however in the select subgroup of women ease of insertion and removal, and if this fails, other who adhere to treatment, the pessary provides extra pessaries can be used (281). benefit in reducing pelvic floor symptoms (278). A recent review of data obtained from public use files A total of 91 women were randomised in the Manonai from the Centres for Medicare and Medicaid Services (2012) trial, 45 in the study group and 46 in the control in the United States over a 10-year period from 1999 group. Eighty-five women (93%, equal in both to 2009 showed that the rates of pessary insertion groups) completed the 16-week assessment with were consistent at 11-13% over the period (282). In 80% compliance to daily exercise. There was no sig- the United Kingdom, a postal survey demonstrated nificant difference in improvement in pelvic floor mus- that 87% of consultants use vaginal pessaries for cle strength between the groups at 16 weeks; either management of POP (283). The likely candidates for measured using the pull test or the digital assessment vaginal pessaries are those with co-morbid medical (279). conditions, those who still wish to bear children, as interim relief prior to surgery and for those who prefer Summary Three new trials were found, one comparing pessary alone versus pessary plus PFMT and two comparing pessary plus PFMT versus PFMT alone. The first was a pilot trial with recruitment difficulties and a very small sample, which did not contribute to the evidence (276). The other two trials were larger and concluded no difference between pessary plus PFMT and PFMT alone in terms of muscle strength at 4 months (279) or prolapse symptoms at 12 months (non-randomised comparison) (278). Combined pessary and PFMT and PFMT alone can be equally effective (Level of Evidence: 1)

Figure 8 Pessaries

126 non-surgical treatment (284). Other indications in- of a ring pessary with support in women with stage 3 clude vaginal laxity, neonatal prolapse mainly seen in and 4 prolapse. Of the 74% who were initially suc- association with neural tube defects such as spina bi- cessfully fitted with the pessary, 10% failed to retain fida and prolapse during pregnancy (285). the pessary at 3 months. Their findings contradict the manufacturer’s recommendations regarding use of Factors that predict the type of treatment chosen for this type of pessary in early prolapse. Of interest, POP have been evaluated in various studies. 82.7% women who were of the median age of 70 Younger women (286) and those with a higher inci- years were able to manage the pessary themselves dence of stress incontinence (286) are more likely to indicating that with proper counselling and encour- refuse pessary use. Age greater than 65 years at the agement hospital care can be minimised. time of pessary insertion and more severe prolapse (Stage III-IV) were more predictive for pessary dis- The cube pessary seemed to be a viable option for continuation at one year (287-289). Ko et al. (2011) sexually active women to self-manage their pessary found that substantially older women or post meno- as Nemeth et al. (292), were able to demonstrate a pausal women opted for a pessary rather than sur- significant improvement in general wellbeing in the gery, and more sexually active women expressed a 78% women who were still using the pessary at 12 significantly greater preference for surgery. In addi- months, 85% rated pessary care use as easy or very tion to opting for surgery over pessary use, younger easy. sexually active women are more likely to change from conservative to surgical treatment over a one-year Brazell et al. (293) demonstrated a significant im- period (290). provement in both bowel related symptoms and quality of life. Patients who completed the 12-month fol- 3.1. Prevention of POP with Pessaries low-up were significantly older and more likely to have stage 3 and 4 prolapse compared to stage 2. Previously no trials had examined pessaries as an intervention in the prevention of POP. In this update, no 3.2.2 Pessary Versus no Treatment further trials were identified. No new studies were identified. 3.2. Treatment of POP with Pessaries 3.2.3 Pessary and PFMT 3.2.1 Pessary Alone  Pessary versus PFMT Three new studies were included. Study details are presented in table 31. Ding et al. (291) evaluated 81 One new study was identified. A RCT comparing women with stage 3 and 4 prolapse who were suc- PFMT to pessary treatment in 160 women (PFMT cessfully fitted with a ring pessary with support after n=79, pessary n=81) aged ≥55 years with advanced 3 months. Subjective evaluation was carried out us- POP. A pessary was fitted successfully in 47/81 (58 ing non-validated questions and prolapse was objec- %)women (296). Only those women in whom a pes- tively assessed using POP-Q. sary was fitted successfully, were compared. Risks of bias were high as both participant and evaluator To evaluate if the cube pessary can be used as a first were not blinded and an ITT analysis was not per- line treatment, Nemeth et al. (292) prospectively eval- formed. uated 78% of women who had a cube pessary inserted after one year. Subjective outcome was estab- In women aged ≥55 years with an advanced symp- lished by using a non-validated questionnaire as a tomatic POP, PFMT resulted in a significant but not validated questionnaire was not available in Hungar- clinically relevant improvement of pelvic floor symp- ian. As one of the aims of the study was to evaluate toms after 3 months (PFDI-20). There was no differ- if it was well tolerated, the authors rated the process ence between PFMT and pessary treatment. PFMT of pessary insertion and general wellbeing on a nu- was more effective in improving anterior wall POP meric rating scale and also on a patient global im- than pessary treatment on POP-Q (Table 31). provement scale.  Pessary plus PFMT versus PFMT Brazell et al (293) reported findings from a secondary Refer to section III.2.3.2 analysis of a study that sought to evaluate if pessary use was associated with improvement in bulge symp-  Pessary plus PFMT versus Pessary toms and improvement in body image (294). They focused on bowel symptoms using the Colorectal Anal Refer to section III.2.3.2 Distress Inventory, a subscale of PFDI-20 and Colo- 3.2.4 Pessary Versus Surgery rectal Anal Impact Questionnaire, a subscale of PFIQ-7. The study had a high attrition rate as only 43 One new study was identified (297), making a total of women of the initial 104 had complete data at 12 three studies comparing pessary versus surgery (Ta- months. ble 32). All were prospective, observational cohort case controlled. The new study by Lone et al. (297) Results evaluated 133 women who opted for surgery and 154 The study by Ding et al. (295) found improvement in who opted for surgery one year after pessary treat- prolapse and bladder symptoms 3 months after use ment using the ICIQ-VS and ICIQ-UI (297). Women

127 who had surgery were older with no difference in There is no agreement as to what constitutes a suc- characteristics like body mass index, parity, ethnicity, cessful fitting of a pessary. Some consider success if history of hysterectomy and prolapse surgery. 69% a pessary was perceived comfortable by a patient women who used pessary and 67% women who had when retained during Valsalva and voiding at the ini- surgery completed the questionnaire at one year. The tial visit, while others consider it success if a patient non-randomised design of the study with approxi- continues to use the pessary until the following visit mately 30% attrition rate suggested a high risk of to the doctor. Thus quoted rates of successful fitting bias. vary widely with differing follow-up times. (Table 33). Reasons for failure range from expulsion due to com- Three studies (285, 297, 298) compared patient re- plications such as vaginal discharge, erosion, de lated outcomes after pessary use and surgery. novo SUI, pain, voiding difficulty and constipation Abdool demonstrated a significant improvement in (Table 33). The risk factors for failure also vary, mak- prolapse, urinary, bowel and sexual function in both ing it difficult to draw conclusions. treatment arms but no difference between the two groups. Using the ICIQ-VS and ICIQ-UI, Lone et al. Complications found a statistically significant vaginal, sex, QOL and urinary symptoms score improvement in both groups Minor complications after pessary insertion range but no statistically significant difference was noted from vaginal discharge, erosion, de novo SUI, bleed- between the surgery and pessary groups. However, ing, pain and constipation (Table 33). A recent study Barber (2006) found that subjects in the surgery by Collins (300) has shown that women who have a group had significantly greater improvement in each pessary are more likely to be bothered by discharge of the scales of the PFDI and the prolapse and urinary (30.0% vs 2.1%, p < .001) and this develops early and scales of the PFIQ than did the pessary group. may be due to an inflammatory process in the vagina (300). Using the cube pessary appears to be compli- 3.2.5 Comparison of One Pessary to Another cation free, probably because it has to be removed on a regular basis. No new trial was identified for this update, making a total of one trial included in this section comparing Rarely major complications may occur. Neglected pessary to another device. Cundiff (299) conducted pessaries present with more serious complications the largest multi-centre crossover trial, comparing a namely fistula formation and peritonitis. Erosion into ring with support and a Gellhorn pessary for the treat- the bowel or bladder and dense adhesions to other ment of symptomatic stage II or greater symptomatic pelvic structures have been reported. Unusual com- prolapse in 134 women. There were no significant dif- plications of cervical entrapment, small bowel incar- ferences between groups in baseline characteristics. ceration, and hydronephrosis have also been re- Participants were fitted with one of the pessaries for ported (301). three months, and with the second for a further three months. During each three-month period, data was Conclusion collected at one, six and twelve weeks from women who had a successful fit. Outcomes were measured As in the most recent Cochrane review (222) and 1 at enrolment, three and 12 months, and included ob- recent RCT, there is no good quality evidence from jective assessment using POP-Q and subjective as- randomised controlled trials on which to base the sessment using PFDI, PFIQ, and a sexual function management of POP using pessaries. questionnaire. Allocation was by computer-generated Prospective case controlled cohort studies suggest random numbers using permuted blocks of variable that pessaries are a viable option for women who size. Opaque, sealed envelopes were used to store complain of symptomatic prolapse. (Level of Evi- the random allocation. Participants and clinicians dence: 3). were not blind to the allocation, but data was coded such that the analysis was conducted blind. Those There appears to be no advantage of pessary use women who were successfully fitted were asked to over PFMT, from one high risk of bias RCT (Level of wear the pessary for three months, but if they discon- Evidence: 3). tinued prior to three months’ data collection was ac- One single randomised study, with a high attrition rate celerated. Attrition rates in the study were high with of 40% found no significant difference between ring only 85 of the 134 women completing the study lead- pessaries with support and the Gellhorn pessary in ing to high risk of bias. However, the trial was not un- PFDI and PFQI scores (Level of Evidence: 2). derpowered due to the cross over design. Cundiff et al. (299) found a statistically and clinically significant improvement in the majority of the PFDI scales and many of the PDIQ scales with both pessaries but no difference between the ring or Gellhorn pessary. Approximately 60% of women offered a pessary continued treatment in the long term irrespective of the type of device. Success rates

128 Table 31 Summary of data on PFMT vs pessary for POP

Author/ year Study Comparison group Participants Type of pessary Subjective Objective Length of Improvement in design assessment assessment follow-up symptoms

Panman CM, RCT PFMT (n=79): 160 women aged > First ring, then Primary outcome: Secondary: change 3 months No sign diff. in PFDI 2014 No standard protocol 55 years with self ring with support PFDI-20 in POP-Q stage POP-Q Anterior – individual identified POP and then Shaatz compartment: 26.5% adaptation in line symptoms (on or Gellhorn >=1 stage change in with normal practice screening) PFMT v 7.1% pessary (included being able POP at or beyond p=0.013 to use electrical hymen (POPQ) * No ITT, only 47 of 81 stimulation) (58%) with successful vs fitting included in PESSARY (n 81): analysis Fitting – opted for 2 week try of pessary with refit at 2 weeks if necessary and max of 3 refits.Fitted by ‘trained research physician’ Foot notes: RCT- Randomised Controlled study, PFDI- Pelvic Floor Distress Inventory, PFMT- Pelvic floor muscle training, POP- pelvic organ prolapse

Table 32 Summary of data on pessary for POP

Subjective and Author/ Type of Objective Length of Study design Comparison group Participants Objective Improvement in symptoms year pessary assessment follow-up assessment Abdool et Prospective Pessary treatment 359 Ring, Sheffield prolapse Baden-Walker 12 months Awareness of lump, prolapse coming out al. 2011 (1) observational compared to Gellhorn, questionnaire of vagina, dragging pain in lower cohort case surgery Cube, Donut abdomen, low back pain, voiding difficulty, controlled need to push prolapse to void, urinary urgency, fecal urgency, sexual satisfaction, interference with physical activity and quality of life

129 Subjective and Author/ Type of Objective Length of Study design Comparison group Participants Objective Improvement in symptoms year pessary assessment follow-up assessment Barber et Prospective Pessary treatment Pessary (n= Ring and PFDI and PFIQ POP-Q 3 months Significant improvement in prolapse and al. 2006 (2) observational (3 months) 42) Surgery Gellhorn for pessary urinary scales of the PFDI. No change in cohort case compared to (n=64) the colorectal scale. No change in the controlled surgery (6 months) PFIQ scales for pelvic organ prolapse Brazell et Prospective N/A 43 Ring with PFDI-20 POP-Q 12 months CRADI-8 mean scores decreased by 6.9 al. 2014 (3) observational support and PFIQ-7 and CRAIQ-7 decreased by 8.1 cohort Gellhorn Clemons et Prospective N/A 100 Ring and Not validated POP-Q 2 months Bulge (90% to 3%) al. 2004 (4) observational Gellhorn Pressure (49% to 3%) cohort Discharge (12% to 0%) Splinting (14% to 0%) SUI 45% UI 46% Voiding difficulty 53% Cundiff et Randomised Ring with support to 134 Ring with PFDI,PFIQ, Sexual POP-Q 6 months Statistically and clinically significant al. 2007 (5) cross-over Gellhorn support to Function improvements in majority of the PFDI and Gellhorn Questionnaire many PFIQ scales in both pessaries, but no clinically significant differences between the two pessaries Ding et al. Prospective N/A 81 with Stage Ring with Not validated POP-Q 3 months Improved bulging (90.4% to 23.3%) 2015 and observational III and IV support Decreased pelvic pressure (64.4% to 2016 (6, 7) study 13.7%) Improved urinary symptoms as follows Voiding – 97.8% Splinting- 100% Urge urinary incontinence- 76.9% Stress urinary incontinence- 58.1%

130 Subjective and Author/ Type of Objective Length of Study design Comparison group Participants Objective Improvement in symptoms year pessary assessment follow-up assessment Fernando et Prospective N/A 203 Ring, Sheffield prolapse Baden Walker 4 months Awareness of lump (71%), prolapse al. 2006 (8) observational Gellhorn, questionnaire coming out of vagina (52%), vaginal cohort Cube, Donut soreness (21%), dragging sensation in lower abdomen 24%), lower back ache (30%), difficulty emptying bladder (40%), push prolapse to void (29%), urinary urgency (38%), urge urinary incontinence (29%), stress urinary incontinence (40%), incomplete emptying of bowels (28%), rectal digitation to empty bowels (12%), vaginal digitation to empty bowel (7%), faecal urgency (30%), urge faecal incontinence (20%), frequency of sexual intercourse (16%), sexual satisfaction (11%)

Jones et al. Prospective N/A 90 Ring, PFDI POP-Q 3 months Improvement in the overall PFDI scale 2008 (9) observational Incontinence and all subscales with the exception of cohort ring, colorectal distress inventory Gellhorn, Oval Komesu et Prospective Compare PF 64 Choice of PFDI-20 POP-Q 6-12 In the continuation group final PFDI-20 al. 2007 observational symptoms in pessary left months total, bladder and prolapse scale scores (10) cohort patients who to discretion were better than the discontinuation continue and of the group. discontinue pessary provider use

Kuhn et al. Prospective N/A 73 Cube Female Sexual POP-Q 3 months Improvement in feeling of bulge, 2009 (11) observational Function Index, improvement in stool outlet problems, cohort Sheffield overactive bladder symptoms. questionnaire, Improvement in sexual desire, orgasm, Kings Health lubrication and satisfaction after therapy Questionnaire

131 Subjective and Author/ Type of Objective Length of Study design Comparison group Participants Objective Improvement in symptoms year pessary assessment follow-up assessment Lone et al. Prospective Pessary treatment 269 Ring, ICIQ-VS POP-Q 12 months Statistically significant vaginal, sex, QOL 2015 (12) observational versus surgery Gellhorn, ICIQ-UI and urinary symptoms score improvement study Cube, Donut in both groups. There was no statistically significant difference was noted between the surgery and pessary groups. Nemeth et Prospective N/A 78 Cube Non-validated POP-Q 12 months Improved general wellbeing score al. 2013 observational pessary questionnaire (13) study (Hungarian) Patel et al. Prospective N/A 75 Ring, Ring Body Image Scale POP-Q 3 months Improvement in body image scale scores, 2010 (14) observational with support, (BIS) and PFDI-20, PFDI-20 scores, PFIQ scores cohort Gellhorn PFIQ, Prolapse subscale of PFIQ

Foot notes: N/A-Not applicable, SUI- Stress Urinary Incontinence, UUI-Urge urinary incontinence, RCT- Randomised Controlled study, P- Prospective Observational study, PFDI- Pelvic Floor Distress Inventory, UDI -Urinary Distress Inventory, ICIQ-VS- International Consultation on Incontinence- Vaginal Symptoms, ICIQ- UI- Interna- tional Consultation on Incontinence- Urinary Incontinence, CES-D – The center for Epidemiological Depression Measures, MOS Scores- Medical Outcome Study (MOS) Social Support Survey 1. Abdool Z, Thakar R, Sultan AH, Oliver RS. Prospective evaluation of outcome of vaginal pessaries versus surgery in women with symptomatic pelvic organ prolapse. Int Urogynecol J. 2011;22(3):273-8. 2. Barber MD, Walters MD, Cundiff GW, Group PT. Responsiveness of the Pelvic Floor Distress Inventory (PFDI) and Pelvic Floor Impact Questionnaire (PFIQ) in women undergoing vaginal surgery and pessary treatment for pelvic organ prolapse. Am J Obstet Gynecol. 2006;194(5):1492-8. 3. Brazell HD, Patel M, O'Sullivan DM, Mellen C, LaSala CA. The impact of pessary use on bowel symptoms: one-year outcomes. Female Pelvic Med Reconstr Surg. 2014;20(2):95-8. 4. Clemons JL, Aguilar VC, Tillinghast TA, Jackson ND, Myers DL. Patient satisfaction and changes in prolapse and urinary symptoms in women who were fitted successfully with a pessary for pelvic organ prolapse. Am J Obstet Gynecol. 2004;190(4):1025-9. 5. Cundiff GW, Amundsen CL, Bent AE, Coates KW, Schaffer JI, Strohbehn K, et al. The PESSRI study: symptom relief outcomes of a randomized crossover trial of the ring and Gellhorn pessaries. Am J Obstet Gynecol. 2007;196(4):405 e1-8. 6. Ding J, Chen C, Song XC, Zhang L, Deng M, Zhu L. Successful use of ring pessary with support for advanced pelvic organ prolapse. Int Urogynecol J. 2015;26(10):1517-23. 7. Ding J, Chen C, Song XC, Zhang L, Deng M, Zhu L. Changes in Prolapse and Urinary Symptoms After Successful Fitting of a Ring Pessary With Support in Women With Advanced Pelvic Organ Prolapse: A Prospective Study. Urology. 2016;87:70-5.

132 8. Fernando RJ, Thakar R, Sultan AH, Shah SM, Jones PW. Effect of vaginal pessaries on symptoms associated with pelvic organ prolapse. Obstet Gynecol. 2006;108(1):93-9. 9. Jones K, Yang L, Lowder JL, Meyn L, Ellison R, Zyczynski HM, et al. Effect of pessary use on genital hiatus measurements in women with pelvic organ prolapse. Obstet Gynecol. 2008;112(3):630-6. 10. Komesu YM, Rogers RG, Rode MA, Craig EC, Gallegos KA, Montoya AR, et al. Pelvic floor symptom changes in pessary users. Am J Obstet Gynecol. 2007;197(6):620 e1-6. 11. Kuhn A, Bapst D, Stadlmayr W, Vits K, Mueller MD. Sexual and organ function in patients with symptomatic prolapse: are pessaries helpful? Fertil Steril. 2009;91(5):1914-8. 12. Lone F, Thakar R, Sultan AH. One-year prospective comparison of vaginal pessaries and surgery for pelvic organ prolapse using the validated ICIQ-VS and ICIQ- UI (SF) questionnaires. Int Urogynecol J. 2015;26(9):1305-12. 13. Nemeth Z, Nagy S, Ott J. The cube pessary: an underEStimated treatment option for pelvic organ prolapse? Subjective 1-year outcomes. Int Urogynecol J. 2013;24(10):1695-701. 14. Patel M, Mellen C, O'Sullivan DM, LaSala CA. Impact of pessary use on prolapse symptoms, quality of life, and body image. Am J Obstet Gynecol. 2010;202(5):499 e1-4.

Table 33 Summary of data on pessary success rates and risk factors for failure

Follow-up Success rate Author/ year Number Types of pessaries Study design Reason for failure Risk factors period n (%) Abdool et al. 2011 554 Ring, Gellhorn, Prospective 12 months 243 (68%) N/A N/A (1) Cube, Donut observational case controlled cohort Brazell et al. 2014 104 Ring with Prospective 12 months 34(41%) N/A N/A (2) observational study Ding et al. 2015 81 Ring with support Prospect 3 months 73 (67%) Feeling of discomfort and pressure, a No specific risk factors and 2016 (3, 4) observational cohort desire for surgical correction, extrusion like stage or type of of pessary, bothersome de novo stress prolapse identified incontinence Fernando et al. 203 Ring, Gellhorn, Prospective 2 weeks 153 (75%) Failure to retain pessary, pain/ Increasing parity, previous 2006 (5) Cube, Donut observational cohort bleeding/discomfort, worsening hysterectomy symptoms

133 Follow-up Success rate Author/ year Number Types of pessaries Study design Reason for failure Risk factors period n (%) Handa et al. 2002 56 Ring, Donut, Prospective 3 months 36 (64.3%) Discomfort, expulsion - (6) Gellhorn, Cube observational cohort Jones et al. 2008 90 Ring, Incontinence Prospective 3 months 42 (47%) Failure to retain, Inadequate relief of Large baseline (7) ring, Gellhorn, Oval observational cohort symptoms measurement of the perineal body at rest Large levator hiatus Komesu et al. 2007 64* Choice of pessary Prospective 6-12 64 (56%) failure to retain, uncomfortable Prolapse score decrease (8) left to discretion of observational cohort months to 77% of baseline the provider Kuhn et al. 2009 (9) 73 Cube Prospective 12 months 32 (44%) Pessary expulsion, desire for surgery, N/A observational cohort bothersome de novo SUI, inability to remove or insert pessary, pain or feeling of discomfort, unspecified Lone et al. 2011 246 Ring, Gellhorn, Prospective 5 years 53 (28.3%) Expulsion, excoriation/bleeding, N/A (10) Cube, Donut observational cohort pain/discomfort, constipation

Patel et al. 2010 75 Ring, Prospective 3 months 54 (79%) Failure to retain, ineffective N/A (11) Ring with support, observational cohort Gellhorn

Nemeth et al. 2012 78 Cube Prospective 12 months 62 (79%) Stress incontinence Parity, previous (12) observational cohort Vaginal discomfort hysterectomy and/or colpoperineorrhaphy, Difficult insertion Wu et al. 1997 (13) 110 Ring with and Prospective Initial visit 81 (74%) Failure to sustain support of the Younger women, previous without diaphragm, observational cohort prolapse, intolerable urinary pelvic surgery, history of Cube incontinence, vaginal discharge, pelvic stress incontinence prior pain, vaginal abrasions and erosions to pessary insertion Foot notes: N/A = not applicable; N/R = not reported,*Includes patients with incontinence and/or prolapse 1. Abdool Z, Thakar R, Sultan AH, Oliver RS. Prospective evaluation of outcome of vaginal pessaries versus surgery in women with symptomatic pelvic organ prolapse. Int Urogynecol J. 2011;22(3):273-8. 2. Brazell HD, Patel M, O'Sullivan DM, Mellen C, LaSala CA. The impact of pessary use on bowel symptoms: one-year outcomes. Female Pelvic Med Reconstr Surg. 2014;20(2):95-8. 134 3. Ding J, Chen C, Song XC, Zhang L, Deng M, Zhu L. Successful use of ring pessary with support for advanced pelvic organ prolapse. Int Urogynecol J. 2015;26(10):1517-23. 4. Ding J, Chen C, Song XC, Zhang L, Deng M, Zhu L. Changes in Prolapse and Urinary Symptoms After Successful Fitting of a Ring Pessary With Support in Women With Advanced Pelvic Organ Prolapse: A Prospective Study. Urology. 2016;87:70-5. 5. Fernando RJ, Thakar R, Sultan AH, Shah SM, Jones PW. Effect of vaginal pessaries on symptoms associated with pelvic organ prolapse. Obstet Gynecol. 2006;108(1):93-9. 6. Handa VL, Jones M. Do pessaries prevent the progression of pelvic organ prolapse? Int Urogynecol J Pelvic Floor Dysfunct. 2002;13(6):349-51; discussion 52. 7. Jones K, Yang L, Lowder JL, Meyn L, Ellison R, Zyczynski HM, et al. Effect of pessary use on genital hiatus measurements in women with pelvic organ prolapse. Obstet Gynecol. 2008;112(3):630-6. 8. Komesu YM, Rogers RG, Rode MA, Craig EC, Gallegos KA, Montoya AR, et al. Pelvic floor symptom changes in pessary users. Am J Obstet Gynecol. 2007;197(6):620 e1-6. 9. Kuhn A, Bapst D, Stadlmayr W, Vits K, Mueller MD. Sexual and organ function in patients with symptomatic prolapse: are pessaries helpful? Fertil Steril. 2009;91(5):1914-8. 10. Lone F, Thakar R, Sultan AH. One-year prospective comparison of vaginal pessaries and surgery for pelvic organ prolapse using the validated ICIQ-VS and ICIQ- UI (SF) questionnaires. Int Urogynecol J. 2015;26(9):1305-12. 11. Patel M, Mellen C, O'Sullivan DM, LaSala CA. Impact of pessary use on prolapse symptoms, quality of life, and body image. Am J Obstet Gynecol. 2010;202(5):499 e1-4. 12. Nemeth Z, Nagy S, Ott J. The cube pessary: an underEStimated treatment option for pelvic organ prolapse? Subjective 1-year outcomes. Int Urogynecol J. 2013;24(10):1695-701. 13. Wu V, Farrell SA, Baskett TF, Flowerdew G. A simplified protocol for pessary management. Obstetrics and gynecology. 1997;90(6):990-4.

135 4.2. Future Research Directions 4. RECOMMENDATIONS 4.2.1 Lifestyle Interventions There is growing attention being paid to the effective- 1. Trials of interventions for constipation are ness of conservative interventions for POP. There are needed to assess their effectiveness in prevent- encouraging signs of more rigorous research in the ing/treating prolapse. area. 2. Studies to fully investigate the association be- 4.1. Recommendations for Practice tween occupation/heavy lifting, and bodyweight 4.1.1 Lifestyle Modification and prolapse are needed as current evidence is conflicting. These studies should ensure that: 1. Constipation is associated with development of prolapse (Grade of Recommendation: C New). i. Occupation, physical activity and diet are assessed rigorously, using instruments with 2. Smoking cessation, while generally recom- sound psychometric properties. mended, cannot be recommended specifically for the avoidance of prolapse development ii. Potential confounding variables are consid- (Grade of Recommendation: D). ered. Attempts are made to overcome recall bias inherent in assessing lifetime occupational 3. Vitamin D deficiency is not associated with de- history, and healthy worker bias, which is a velopment of prolapse (Grade of Recommen- problem when attempting to compare prolapse dation: C New). in women currently employed in heavy labour type jobs versus others. 4.1.2 Pelvic Floor Muscle Training (PFMT) iii. Outcome measures used are valid and relia- 1. PFMT does not influence the development of ble, and are consistent across studies; pro- prolapse post-natally (Grade of Recommenda- lapse symptoms should be the primary out- tion: B New). come within studies, followed by prolapse ana- 2. PFMT intervention delivered 12 years+ after tomical severity. childbirth can reduce symptoms of prolapse 4.2.2 Pelvic Floor Muscle Training (PFMT) which develop in the longer term (Grade of Rec- ommendation: B New). 1. Further studies are needed to confirm the role of physical therapies in the prevention of POP. 3. There is evidence of benefit that PFMT is effective in reducing pelvic floor symptoms (Grade of 2. Further trials are needed to add to the evidence recommendation: A New) regarding: 4. There is some evidence of benefit showing that i) The effectiveness of PFMT for different stages PFMT is effective in alleviating specific prolapse and types of prolapse. symptoms (e.g. vaginal bulge) (Grade of recommendation: C New) ii) The role of PFMT as an adjunct to surgery for anterior and posterior prolapse. 5. There is no evidence that PFMT is effective in reducing severity of prolapse based on POP-Q 3. More trials needed to improve the evidence re- stage (Grade of recommendation: B New). lating to the following comparisons: 6. Peri-operative PFMT does not improve prolapse i) Low versus high intensity supervision of PFMT symptoms in women undergoing surgery for (taught PFMT vs self-instruction manual already vault prolapse (Grade of Recommendation: B trialled) New). ii) Individual versus group PFMT (group PFMT vs 7. Combined pessary and PFMT and PFMT alone group lifestyle already trialled) can be equally effective in reducing symptoms iii) PFMT versus surgery (anterior/posterior repair and increasing muscle strength and should be vs PFMT already trialled) considered for treatment (Grade of Recommen- dation: B New). iv) PFMT versus pessary. (PFMT vs PFMT plus pessary already trialled) 4.1.3 Pessaries The assessment and measurement of POP and the 1. In a choice between the Gellhorn pessary and a assessment of prolapse symptoms need to be made ring with support, offer either to improve prolapse in a standardised fashion using a validated outcome symptoms and reduce their impact (Grade of measure (such as the POP-Q examination). A single Recommendation: B). validated symptom tool was not apparent in new studies, but the PFDI, PFIQ and POP-SS tools were most commonly used and may provide a useful basis for comparisons across trials in future.

136 4.2.3 Pessarie for Prolapse EStim and MStim for non-prostatectomy related incontinence or other LUTS. Studies on PTNS are pre- Although the use of pessaries has been common clin- sented in section V.1. ical practice and has been used for many centuries robust evidence of their use is lacking. There is a A literature search of relevant systematic reviews and pressing need for well-designed randomised studies reports of RCTs and quasi-RCTs was updated. No using validated measures for subjective and objective other types of study designs were considered. One assessment. Areas that need focus are: systematic review, an update of the Cochrane systematic review on conservative management of post-  Pessary versus no treatment prostatectomy incontinence, was identified (304). For  Pessary versus PFMT this review, 14 new published trials (305-318) and eight abstracts (319-326) were identified. Table 34  Pessary versus surgery provides summary information on the 22 trials added in this review. One study previously included only as  Risk benefit of the use of local oestrogen in con- an abstract (327) is now included as a full publication junction with a pessary (328). This study is included in the table as in the full  Progression or regression of prolapse using in publication with additional information that was not women using pessaries previously included in the abstract. The published peer reviewed journal publication (329) of a previ-  Optimal management protocols for pessary us- ously included report of two parallel trials (RP and age e.g. indications, interval between pessary TURP arms) was added to the references which pre- changes, complications and their treatment and viously included an abstract and health technology which pessary is indicated for a specific type of report (330, 331). One study previously excluded as prolapse. it was a study in progress (332) was included this time as a peer reviewed publication of a completed trial (307). Nine ongoing trials were identified from trial URINARY INCONTINENCE IN registries, but not included in this review. MEN 1. LIFESTYLE As in earlier consultations, UI in men remains under- reported and under-studied in comparison to studies Lifestyle recommendations such as smoking cessa- of women. Pooled prevalence of UI in community tion, healthy eating, appropriate body weight, avoid- based men ranges from 4.81-32.17%. (302). UI and ing excessive caffeine or alcohol are all part of a pri- other LUTS in men increase with age, with variation mary care approach and are intended to be preven- in prevalence rates reflecting different study popula- tative in the onset of obesity, cardiovascular disease, tions, definitions of incontinence and methods (303). diabetes. Up until the last ICI edition, no trial had ad- Despite the prevalence of UI and LUTS in older men, dressed the topic of lifestyle interventions alone in the only aspect which continues to receive systematic men with UI. In this edition, a few new trials have consideration with respect to conservative manage- been added. ment is post-prostatectomy urinary incontinence after radical prostatectomy (RP). The primary conservative 1.1. Weight Loss by Obese or Overweight approach for prevention and treatment of UI after RP, Men or transurethral resection (TURP), remains PFMT, with or without some form of biofeedback (BF). One new trial was identified on weight loss in over- PFMT, in combination with anal EStim, BF or weight/obese men (7, 333) and has been added in transcutaneous electrical nerve stimulation (TENS), this update. This was the partner study to Phelan, Ka- MStim, and novel therapies such as dyadic planning naya et al. (2012) conducted to determine the effect and concentration have been utilised for UI in men. of an intensive weight loss programme over 4 years previously outlined in the look AHEAD trial - on a sub- The PFMT, EStim, MStim and PTNS interventions, set of male participants (n=1910). Men were random- and other combinations of PFMT with general exer- ised to an intensive weight loss programme or diabe- cise and other approaches, in the current review were tes support and education group. Self-report of incon- kept in the same organizational format as the previ- tinence, nocturia and daytime voiding frequency were ous consultation to reflect evolving evidence and recorded at baseline and 1 year. The odds of preva- emerging directions of research. All new studies of lent UI at one year were reduced by 38% in the inten- EStim and MStim in men undergoing or post prosta- sive lifestyle modification intervention group com- tectomy combined these modalities with PFMT and pared to the support and education group, with UI de- thus were included under the PFMT section. Penile creasing from 11% to 9% in men. As reported for vibratory stimulation (PVS) was added as a novel women there was uncertainty over the risk of selec- technology category. The study on PVS involved men tion bias for a number of key parameters including: incontinent after RP, but did not include PFMT, so random sequence generation and allocation conceal- was kept in a separate section, as were studies of ment. Performance bias was unclear as the blinding of participants and staff was not undertaken. Both

137 studies did ensure the blinding of outcome assess- Weight loss: Where weight loss through lifestyle ment but neither provided complete outcome data changes should be recommended to obese and over- and both provided only selective reporting indicating weight men with UI, particularly those with type 2 di- possible reporting bias (Table 2). abetes. (Grade of Recommendation: B New) 1.2. Smoking Smoking abstinence should be recommended for men with UI. (Grade of Recommendation: C New) One new trial was found however data was not presented separately for women and men. (Refer to sec- A reduction in caffeine intake is recommended for tion II.1.2.4 for more details) those with incontinence symptoms; evidence suggest the equivalent of 2 cups of coffee a day (250mg) is 1.3. Dietary Modification in Men associated with urinary incontinence in both men and In a separate sample (from the study by Davis, women. (Grade of Recommendation: C) Vaughan et al. 2013 presented in Lifestyle interven- Larger RCTs to assess effect of lifestyle modification tion in women’s section II.1.2.5) 3960 men over 20 interventions are important. were included (16). The authors found that the highest level of caffeine intake was associated with having moderate to severe UI (1.72, 95% 1.18-2.49 and 2.08, 95% 1.15- 3.77) respectively. All parameters in terms of risk of bias reporting were unclear or high risk. The blinding of participants and personnel and blinding of outcome assessment were at high risk of bias and the likelihood of incomplete outcome data and selective reporting was also high. A sample of 683 men aged 40-75 who completed a food frequency questionnaire and the Consultation on Incontinence short Form (ICI-SF) as part of the Hirayama study (18). The data showed a slight increase in the risk of UI at the highest level of caffeine consumption (similar to the US data), but this was not significant after adjusting for confounding factors with OR: 95% CI, 1.36 (0.65-2.88) in the male participants. The risk of bias reporting was largely unclear in this study with blinding of participants and personnel and blinding of outcome assessment, incomplete outcome data and selective reporting all of high risk. Re- sults of sex stratified analysis of the data from the sample of Japanese men and women did not show an association between caffeine and UI, the authors suggest the need for further larger samples to explore any association further (Table 5).

Summary Evidence from 1 RCT supports lifestyle modification interventions promoting weight loss as a tool to reduce urinary incontinence in men who are overweight or obese. (Level of Evidence: 2; 1 new RCT). Evidence from a small new RCT indicates that urinary frequency may be improved by smoking abstinence, (Level of Evidence: 3). Caffeine consumption is likely to play a role in exacerbating UI in men. New epidemiological evidence from a large cross sectional study supports this conclusion (Level of Evidence: 3). Recommendations

138 Table 34 Summary of data on conservative management in male urinary incontinence

Study Modality details or Notes (side effects, Author, year Comparator N Outcomes/results Follow up population parameters loss of follow up…) Pelvic Floor Muscle Training (PFMT) Ahmed et al. 3 group N= 90 men Men undergoing Treatments started one Primary outcome 24 hour Continence Dropouts: none 2012 (1) comparison randomized, N=80 RP for clinically week after catheter pad test, secondary was assessed at reported PFME vs PFMT completed trial localized removal, twice weekly quality of life assessed baseline, 6, 12 Sample size analysis plus EStim vs Group 1 n= 26 prostate cancer for 12 weeks with IIQ-7. and 24 weeks reported (n=80) PFMT plus Group 2 n=26 Group 1 PFME Mean leakage significantly Dropouts EStim plus Group 3 n= 28 (control) lower in Group 3 (EStim Group 1= 4 (2 biofeedback Randomization by Group 2 PFMT plus plus biofeedback) at 6 radiotherapy, 2 computer generated EStim starting one through 24 weeks (p < postoperative random number list in week after catheter 0.05) complications sealed envelopes. removal, twice weekly Significant differences in Group 2 = 2 (both Surgeons blinded to for 12 weeks continence at weeks 12 radiotherapy) randomization. Blinding Group 3 PFMT plus and 24 with Group 3 Group 3 = 4 (2 of outcomes assessor EStim plus having more continent radiotherapy, 2 not indicated. biofeedback starting patients, followed by declined followup) one week after Group 2. Concluded early No intention to treat catheter removal, twice EStim and biofeedback analysis. weekly for 12 weeks decrease duration and EStim – electrodes on degree of skin over sacrum postprostatectomy UI Biofeedback - surface electrodes on abdomin and perineum Baroni et al. Intervention N= 40 men Men with SUI or Intervention: 5 Outcomes included Initial briefing by Dropouts: none 2013(2) (individual Divided into two groups mixed UI after individual training adherence to training therapist, final reported. PFMT and (randomization not RP (time after sessions with a program (exercise at evaluation one No sample size Abstract only group described). surgery not physical therapist home), self report of month after analysis reported. treatment) vs Intervention: n = 16 reported) followed by small change in continence rehabilitation. control patients mean age 61.8 groups sessions (total using VAS scale and (individual years 15 sessions at the number of pads, quality of PFMT) Control n = 24 mean rehabilitation centre), life ICIQ-SF, cost age 67.5 home exercises. effectiveness (therapist Assessor blind to Control: individual time). Non significant allocation until training sessions with difference in home assignment. physical therapist. practice (intervention 69% vs 58% controls). No

139 Study Modality details or Notes (side effects, Author, year Comparator N Outcomes/results Follow up population parameters loss of follow up…) difference between groups on VAS or pads used per week. Authors conclude group treatment allows cost effective improvement. Burkert et al. Intervention N = 112 prostatectomy Men undergoing All patients received Main outcome: self Questionnaires at Dropouts 2011(3) (dyadic PFME patients and their laprascopic RP standard care including reported dyadic planning 2 days, 2 weeks, Intervention n=5 planning) vs partner (dyads). who had a written information on and PFME. 1, 3 and 6 months Control: dyadic one of three 2x2 mixed design. partner willing to PFME 1 day post Self reported dyadic post-surgery. nutrition planning n=2 controls Couples randomized in participate surgery (PFME 3x day PFME planning increased Control individual blocks of four to one of for 10 minutes), with both the dyadic and PFME planning n=7 four groups Intervention physical therapist individual PFME planning Control: n= 28, introduced PFME on session. No effect on individual nutrition Controls n= 29, 29 and day 3 or 4. Discharge PFME was found. planning n= 4 26 day all participated in Continence was not an Sample size analysis Research assistants single 30 minute outcome. required 112 couples blinded to allocation planning session (dyads). Mean age of patients intervention with Intention-to-treat 62.8 years, partners completion of planning analysis completed. 59.3 years. sheet for health behaviours (in dyad or individually). Intervention: dyad PFME planning Control: (dyadic nutrition planning, Control: individual PFME planning, Control: individual nutrition planning Collado, Serra Intervention N= 193 patients Men with Intervention: starting 3 Primary outcome: Degree No detailed Dropouts –detail only 2013 (4) PFMT plus BF recruited localized weeks before surgery of continence description. Pad for intervention group pre-op vs Randomization and prostate cancer included weekly improvement compared to tests reported for n=5 (2 perineal pain, 2 Abstract only Control PFME blinding detail not scheduled for assisted BF sessions week 1 during followup week 1, week 6, post RP complication, after RP provided. RP (surface electrodes) (continence not defined) month 3, month 6 1 left treatment. with periodic, rapid, and 1 year.

140 Study Modality details or Notes (side effects, Author, year Comparator N Outcomes/results Follow up population parameters loss of follow up…) N= 179 included in final intense and maximal Secondary outcomes 24 No sample size analysis strength contractions hour pad test and ICIQ-UI calculation provided. Intervention = 87 and transversus SF score No intention–to-treat Control = 92 abdominis activation Significant difference in analysis. plus daily PFMT continence improvement performed at home reported for week 6, with written instruction month 3, month 6 and 1 (no detail on whether it year favouring intervention was continued post- group. Difference in 24 operatively) hour pad test significant Control: oral instruction only at 3rd month, on PFME post surgery, favouring intervention structured program of group. BF, transversus abdominis activation and PFMT 3 months post surgery Dijkstra-Eshuis Intervention N=122 patients Men undergoing All participants Primary outcome: urinary Questionaires, 24 Dropouts et al. 2015 (5) PFMT plus recruited, N= 121 laparascopic RP assessed pre-op by continence defined as no hour diary and Intervention: language biofeedback randomized. Pre- (one surgeon), physical therapist leakage on 24 hour pad pad test at 6 barrier (1), lost to pre-op vs operatively. prostate cancer Intervention: once test and self report by weeks, 3, 6, 9 followup (4), Contol PFME Mean age 63.7 years. state T1 or T2. weekly 30 minute KHQ, IPSS and PeLFI and 12 months discontinued post-op Randomization by session of PFMT with (pelvic floor inventories) post-operatively. intervention (4) computer generated biofeedback pre-op x 4 20.8% of patients PeLFT and Control: random numbers (block weeks provided by continent at 6 weeks, examination of Lost to followup (4), randomization, variable physical therapist with 43.6% at 3 months, 61.5% the pelvic floor discontinued block size). Therapists twice daily practice at at 6 months, 72.3 % at 9 pre-operatively intervention (6) and participants blinded home. Told to restart months, 77.2% at one and 1 year post- Sample size analysis to randomization until immediately after year. operatively. was 124 patients in first visit. catheter removal. No difference in SUI or each group (n=248), Intervention n= 65 Written instructions for QoL between intervention interim analysis Control n= 56 2 sets of 30 and control at any time planned at 122 contractions. points. patients. Trial halted Control: standard care Authors suggested that as interim analysis of written PFME post-prostatectomy UI showed no benefit. instructions on catheter likely due to intrinsic removal (7-10 days sphincter deficiency that post surgery)

141 Study Modality details or Notes (side effects, Author, year Comparator N Outcomes/results Follow up population parameters loss of follow up…) cannot be treated with exercise. Geraerts et al. Intervention N= 180 men Men undergoing Intervention: 30 minute Primary outcome: Pre-operative Dropouts: 2013 (6) PFMT plus Computer generated open or robot session of therapist incidence of contience and baseline and 1, 3, Intervention n=6 died biofeedback randomization using assisted guided PFMT with time to continence. 6 and 12 months (1), stroke (1), pre-op vs permuted blocks for laproscopic RP biofeedback weekly Continence defined as 3 post-operatively transport problems Control PFMT strata (age <65/≥65, (3 surgeons) starting 3 weeks prior consecutive days of 0gm (3), refused further post-op surgical approach to surgery plus home urine loss on 24 hour pad participation (1) open/robot assisted program of 60 test. Other measures Control n= 4 transport laproscopic contractions per day. were: 1 hour pad test, probems (2), refused Intervention n= 91 Restarted PFME day 4 VAS, IPSS, KHQ (QoL). further participation Control n=89 post-op with catheter in No difference between (2) situ groups on duration of UI, Sample size analysis Control: PFMT after pad test, VAS or IPSS. required 166 for catheter removal Median time continence power. Both groups: Post-op was 30 (control) and 31 weekly session to (intervention) days. discuss bladder diary Intervention group scored and have guided better on impact of session with digital or incontinence at 3 and 6 EMG biofeedback plus months home program. PFMT continued until continent Ghanem et al. Intervention N=100 men Men with Intervention: Continence defined as Timepoints for Dropouts: not 2013 (7) PFME pre-op vs randomized. localized PFME protocol for 2 using 0-1 pads. Also measurement not described Abstract only Control PFME Randomization prostate cancer weeks prior to surgery completed the ICIQ SF clear, last dated post-op technique not undergoing (detail not provided) male. reported at 54 Sample size analysis described. radical with post-operative More intervention patients weeks post-op. not described. Intervention N=50 prostatectomy PFME program. continent at 14 weeks than Control N=50 Control: postoperative controls (p< 0.05). 70% of PFME program only patients in both groups continent at 18 weeks, 85 % by 54 weeks. Hou et al 2013 Intervention N= 66 randomized. Men undergoing Intervention: PFME Outcome: early recovery Baseline and 1, 4 Dropouts n= 5 (group (8) post-op PFME Randomization TURP for benign after removal of of bladder function post and 12 weeks assignment not vs control catheter day 2 post-op. TURP using the IPSS and post-op identified) delay in

142 Study Modality details or Notes (side effects, Author, year Comparator N Outcomes/results Follow up population parameters loss of follow up…) technique not prostatic Instructed to contract SF-36. Q-max (maximum catheter removal prior described. hyperplasia. PFM for 5 seconds, urinary flow), voided to discharge (2), lost N= 61 completed study relax for 10 seconds. volume and PVR (post to followup (3). (intervention n= 32; Surface EMG to void residual volume). Sample size analysis control n= 29) confirm PFM At 12 weeks, intervention not provided. No contraction. Home group had significantly intention-to-treat program 5 minutes less severe LUTS analysis. practice 3 x per day. compared to control Weekly telephone (p<0.001), Qmax reminders. (p=0.026) and QoL Control: not described physical (p=0.029) and mental health domains (p= 0.005). No significant differences in voided volume or PVR. Kakihara et al. Intervention N=20 men from a Men with urinary Intervention: UI measured with 1 hour Measures at Dropouts 2007 (9) (PFMT with single urology clinic, incontinence Physical therapy taught pad test (incontinence < 2 baseline, visit 2, Intervention n=2 EStim) vs mean age 64.3 years post RP functional PFMT with gm), visual analogue and 3, 6 and 12 discharged at 3rd and PFME only Randomly divided into (minimum of 6 EStim. scales for incontinence months. 6th months, reason two groups (detail on months post-op, PFME started with 2s patient perception of the not provided. randomization had undergone contractions increasing problem, and pad use Control n=3 technique not provided) urodynamic daily by 1suntil 10s Intervention group: 4 discharged at 3rd, 6th Intervention group testing) reached. Patients were patients UUI, 6 had SUI, and 12th as regained n=10, control n=10 instructed to do 90 mean time post surgery continence N=18 in final analysis contractions/day at 12.3 months. Control Length of time EStim for pad test and visual home (divided equally group: 5 had UUI, 5 SUI, continued for the analogue data : into 3 times per day). meantime post surgery intervention group intervention n=8, Also had EStim with 16.8 months unclear. control n=10 endo-anal electrode Significant decrease in No sample size or weekly. UUI – 8 Hz urine loss on pad test and intention-to-treat increasing to 10 Hz self report baseline to 12 analysis reported. after 3 months, for SUI months in both groups. No 35Hz increasing to 50 difference between groups Hz after 3 months. Control: Physical therapy taught functional PFMT only

143 Study Modality details or Notes (side effects, Author, year Comparator N Outcomes/results Follow up population parameters loss of follow up…) Kongtragul et Intervention N= 138 men recruited Men 17-80 Intervention: Starting 3 Primary outcome was Measurements at Dropouts al. 2014 (10) (PFME with from a single hospital years of age weeks post- urinary incontinence baseline (3 weeks Intervention n=1 concentration N= 135 cases in final with prostate operatively, PFME with (incontinence ≥ 2 gm urine post surgery) and (refused treatment) therapy) vs analysis (n= 68 cancer who had concentration therapy on 1 hour pad test) weekly for up to 3 Control n=2 (refused control (PFME intervention, n= 67 no incontinence (concentrating on the 65/68 in intervention group months treatment) only) control) prior to radical exercise and compared to 48/67 Needed 69 cases in Randomization prostatectomy eliminating other controls achieved each group for power. technique not given. surgery issues or use of the continence (p< 0.001 No intentiontotreat Sample stratified for rehabilitation health More men in intervention analysis. time of catheter spa rock - not clearly group 66/68 practiced removal (prior to or described). PFME was regularly than in control after discharge) and practiced 240 times group 34/67 (p<0.001) type of surgery (no daily by holding and further detail provided). relaxing muscle tension around the anus. Control: PFME only Laurienzo et al. 3 groups N= 58 men recruited. Men with Intervention Primary outcome: Measurements at Dropouts 2013 (11) Pre-op PFMT Randomization by prostate cancer Group 1: 10 pre-op continence assessed by 1 1, 3 and 6 months N=9 (detail on group plus EStim vs computer generated stage T2 waiting physiotherapy sessions hour pad test (> 2 gm post surgery. not provided). pre-op PFMT vs list. All intervention RP of EStim using rectal incontinence). Secondary Desistance (sic)(2), control provided by a single probe. Tonic fibers – outcomes: ICIQ-SF, SF-36 adjuvant radiotherapy therapist 20HZ, 700 (1), urethral stenosis N= 49 completed study microseconds, work No significant differences (1), urinary fistula (1), (n= 17 PFME plus time 6s, rest 6s. Phasic between groups on pad surgical risk (1), EStim, n= 17 PFME, n= fibers – 65 HZ, 150 test at any time points. No inadequate followup 15 control) microseconds, work differences in QoL (SF-36) (1) time 6s, rest 18s. Five No sample size exercises for PFM analysis described. contraction. No intention-to-treat. Group 2: 10 pre-op physiotherapy sessions with only the PFME. Control Pre-op information on prostate anatomy only

144 Study Modality details or Notes (side effects, Author, year Comparator N Outcomes/results Follow up population parameters loss of follow up…) No post-op PFM rehabilitation for any group Laurienzo et al 3 groups N=123 men incontinent Men with more Assisted PFMT Outcomes Incontinence Patients Dropouts not 2015 (12) PFMT plus after RP than 3 gram loss described as two and quality of life using the evaluated pre- discussed. Abstract only EStim vs PFMT Randomization by of urine on a 1 series of ten exercises ICQ-SF, erectile operatively and Sample size analysis vs control computor, sealed hour pad test taught by verbal dysfunction by IIEF-5, then at 1, 3, and 6 not provided. envelopes. one month after command IPSS, 1 hour pad test, months post RP Group 1 n=40 control, RP EStim protocol was 20 evaluation of pelvic floor no treatment minute sessions twice with perineometer. Group 2 n=41 assisted weekly over 7 weeks. No differences between PFMT Frequency 35 HZ groups in quality of life, Group 3 n=42 EStim erectile dysfunction or with assisted PFME PFM strength at any time No details on blinding. point post-surgery. Pad test data not provided. Martini et al Intervention N= 70 men recruited Men undergoing Intervention: 5 Outcome: focus on PFI as Measurement Dropouts: not 2011 (13) pre-op PFMT vs N= 49 with pelvic floor laparoscopic RP sessions of physical a potential factor in post pre-op baseline, identified. Abstract only control impairment identified for localised therapist guided PFMT RP UI. Continence defined and 1, 3 and 6 Sample size analysis on exam randomized to prostate cancer 2-3 weeks before by not wearing a pad. months post-op. not provided. intervention or control. T1-T3 surgery, PFME Other: 24 hour pad test, Intervention n=24 continued post-op pad use, 7 day bladder Control n=25 Control: usual care – diary, QoL. Randomization details written instruction on Patients with PFI used not provided. PFME provided post- more pads and had more operatively SUI episodes. PFI independent predictor of UI at 3 and 6 months. For those with PFI, UI on getting up and squating was significantly lower in intervention group at 1 month (p=0.006) but not other time points. Morihoro et al Intervention N= 34 men Men who had Intervention: PFME Outcome UI, defined as Measurement at Dropouts: not 2011 (14) post-operative Randomized post- undergone and EStim 2x daily for not needing a pad to keep 1, 3, 6, and 12 described. Abstract only PFME with operatively to 2 groups. laparoscopic RP 15 minutes for one clothing dry. months.

145 Study Modality details or Notes (side effects, Author, year Comparator N Outcomes/results Follow up population parameters loss of follow up…) EStim vs control Randomization month after catheter Recovery of urinary No sample size PFME technique not removal, starting on continence significantly analysis based on described. day 5 after surgery. favored intervention group multivariate or other Intervention n=20 Control: PFME alone at 12 months (p = 0.007). analysis provided. Control = 14 Multivariate analysis showed EStim associated with continence recovery at 6 and 12 months. Ng 2014 (15) Intervention N=66 men Men undergoing Intervention: PFMT Primary outcome: urine Measures at 1,2, Dropouts: no Abstract only pre-op PFMT vs Randomization radical started 3 weeks pre- loss on 24 hour pad test. 3 and 6 months information provided. control technique not given. prostatectomy operatively from Secondary outcomes: advanced practice incontinence impact and nurse. potency satisfaction. Control: Standard care Earlier return to (no detail provided). continence in intervention group at 3 months compared to controls (p= 0.002). Intervention group also had higher score on potency at 3 months (p = 0.005). Ocampo-Trujillo Intervention N=16 men, mean age Men > 40 years Intervention: Intensive 24 hour pad test for Measurements Dropouts: none et al. 2014 (16) (pre-op PFMT 58 years waiting RP for PFMT 3 times daily for continence, histological (pad test, health with Randomized to 2 prostate cancer 4 weeks prior to analysis of muscle tissue related quality of biofeedback) groups (T<3 N0M0 surgery. Biofeedback from external sphincter of life) at week 0 No histiology samples vs control Intervention n=8, PSA<20) audible and visual urethra. Health related prior to surgery taken prior to start of (standard pre- control n=8 signals, anal pressure quality of life assessed and week 8 post- intervention. op teaching Randomization –single probe. with UCLA-PCI and SF 12. op. only) blind, centralized. Control: standard pre- Continence defined as 3 Histology Researchers blind to operative diet and consecutive days of no samples taken at group assignment prior general health teaching urine loss on 24 hour pad the end of the to enrollment. test, results not reported. intervention and 6/8 intervention vs 4/8 of on day of surgery. controls did not need pads at 8 weeks. Self report of symptoms favoured improvement in intervention group (no p

146 Study Modality details or Notes (side effects, Author, year Comparator N Outcomes/results Follow up population parameters loss of follow up…) values provided except for erectile dysfunction). Intervention patients had higher values in cross sectional area of external sphincter muscle than controls (p=0.03) Park et al. 2012 Intervention N= 66 men 65 years or Men who had Intervention: Exercise Continence a secondary Measurements at Dropouts (17) (combined older recruited. undergone (resistance, pelvic outcome 1 week prior to Intervention: Previously only exercise with Random allocation by laparoscopic flexibility and Kegel Measured by 24 hour pad surgery, 3 weeks n=7 (other unrelated in abstract form. PFME vs random number RP. exercises) initiated test. Urinary continence post-op and after surgery, TURP for control (PFME generator, sealed post-op week 3 for 12 was < 1 gm the intervention urethral stricture, non only) envelopes weeks (twice weekly). Significant improvement (15 weeks post- compliance, new N= 39 completed study. Control : Kegel (p= 0.033) in favour of op) employment) (n= 26 in intervention exercises only intervention group at time Control: group, n= 25 in control) Detail of Kegal (PFME) of last visit (15 weeks n=8 (other unrelated Participants not exercises not provided. post-op). surgery, adjuvant blinded; surgeons and Exercise program Concluded intervention radiotherapy, non research assistant developed by sport group had an earlier return compliance, new completing evaluation science experts, detail to continence. employment) questionnaires not of who supervised the No intention-to-treat. blinded. intervention not provided. Pedriali et al. Intervention N= 69 Men one month Intervention: Outcomes: continence on Time points of Dropouts: no details 2014 (18) post-op pilates after RP 10 sessions mat 24 hour pad test, number measurement not provided. Abstract only vs control Randomization Pilates exercises with of pads used, ICIQ-SF provided, 3 No information on PFME plus technique not certified Pilates Significant reduction in months of sample size EStim described physical therapist urine loss/24 hours and treatment calculation provided. N=54 completed study Control: 10 sessions of pads used, improvement mentioned. ( n=26 intervention, n= PFME plus EStim with in QoL in both groups. 28 control) physical therapist. 58% of intervention and Both: Written 50% of controls achieved instructions for home continence as measured exercises by no pad use (p=0.57) corresponding to Authors concluded Pilates treatment. as efficacious as control condition and may reduce health care costs.

147 Study Modality details or Notes (side effects, Author, year Comparator N Outcomes/results Follow up population parameters loss of follow up…) Serda et al. Intervention N= 69 men with Men with no Intervention: UI assessed by 20 Data collected at Dropouts 2014 (19) post-op prostate cancer (any prior Progressive strength minutes pad test. Type of baseline and 24 Intervention: n=3 combined stage) who had incontinence program in 3 incontinence measured by weeks. (medical reason, exercise with undergone recruited after consecutive stages: UI-4, intensity and Calculated cognitive problem, PFME vs prostatectomy +/- treatment for global posture re- frequency of urine loss by sample size metastasis) control no hormone or hormone prostate cancer education, PFMT, Sandvik scale and VAS-UI achieved. Control: None intervention and radiotherapy (time period not exercises to radiate and FACT-P (urinary No intention-to-treat. randomized to 2 identified). muscle strength. related QoL). Also groups. Details of Duration 24 weeks, 16 included data on obesity, n= 36 intervention randomization weeks of supervised muscle resistance, n= 33 control approach not work, 8 of autonomous constipation andactivity. Mean age 71 in both provided. exercise. 33.33% of intervention groups. PFMT performed with group had stress UI N=66 completed music, focus on compared to 36.36% biofeedback (sense of controls. The rest had awareness). Slow urgency UI or mixed UI. twitch (≤1s) and fast Significant improvement twitch (5s) on VAS-UI scale regarding contractions. intensity UI symptoms Control group: watchful favouring intervention waiting by telephone group. QoL improvement contact greater in those with greater improvement in UI. Pad test dated/number achieving continence not reported. Tienforti et al. Intervention N= 34 men screened Men undergoing Intervention: Pre- Recovery of continence Measurements at Dropouts 2012 (20) (pre-operative and recruited pre-op, standard open operative biofeedback (self report on ICIQ-UI) baseline, 1, 3 and Intervention: n=1 and post- computer generated retropubic RP (anal probe, reference primary outcome. 6 months. (intolerence to rectal operative PFMT randomization for localized electrode on anterior Significant difference in probe insertion) with schedule. prostate cancer superior iliac spine) achievement of continence Control: n=1 biofeedback) N= 32 men in the final at one centre. with teaching on PFM favoring intervention group (intraoperative vs analysis (n= 16 in the day prior to at 1, 3 and 6 month surgical control group intervention, age 60-74; surgery. After catheter followups. 10/16 complications) (post-op PFME n= 16 in control age 52- removal, oral and intervention vs 1/16 No intention-to-treat. only) 74). written PFME with control patients continent No significant structured exercise at 6 months. differences between program (3 daily 10

148 Study Modality details or Notes (side effects, Author, year Comparator N Outcomes/results Follow up population parameters loss of follow up…) intervention and control minutes sets 5 second Patients in intervention groups in age, disease contractions followed group reported higher QoL stage, intra or post- by 5 seconds (IPSS QoL), difference not operative features. relaxation) lying/sitting/ significant standing, recorded in Concluded pre-operative diary. Monthly biofeedback with PFME supervised and monthly assisted biofeedback and sessions improves motivation sessions recovery to continence. Control: oral and written instructions on home PFME post-op after catheter removal, no diary, monthly followup Pre-op session provided by a dedicated caregiver. Zopf et al. 2015 Intervention N = 85 men who had Men recruited 6- Intervention: 15 month Multiple secondary Measures for UI Dropouts (21) (combined undergone 12 weeks after supervised multimodal outcomes including urinary at 3 time points, exercise with prostatectomy for prostatectomy exercise program incontinence as measured detail not Intervention: n=6, PFME vs prostate cancer. +/- radiation. (aerobic, resistance by 20 minute pad test. provided. reasons not related to control (no n= 56 in intervention Multicentre, and PFME) with one of Primary endpoint was intervention intervention) group partially 4 community sports physical fitness. Authors identify n= 29 in control group randomized groups. Exercise 60 Significant improvement in poor compliance Control: dropouts not N= 70 (50 intervention, controlled trial. minutes per week UI for intervention group with pad testing described. 20 control) completed Patients who Details of PFME not between baseline and post was a limitation. followup consented to provided. testing (p= 0.005). No Intention-to-treat randomization Control: No significant difference Calculated analysis performed. were to be intervention between intervention and sample size not randomized, control. achieved, limited patients who Significant improvement of power to detect refused self reported urinary differences randomization symptoms (EORTC-QLQ between groups. were to receive PR 25) in favour of intervention of intervention group. their choice. As all patients

149 Study Modality details or Notes (side effects, Author, year Comparator N Outcomes/results Follow up population parameters loss of follow up…) indicated a group preference, trial was not randomized. Penile Vibratory Stimulation (PVS) Fode 2015 (22) Intervention N= 39 randomized by Men with UI > 1 12 week study. Primary outcome: Measures at Drop outs: PVS vs control computer generated list year post RP. All Intervention: PVS for difference in leakage baseline prior to Group 1 early delayed PVS 1 :1 ratio patients had first 6 weeks of study. measured by 24 hour pad inclusion and 6 intervention n=4 lung N= 31 in analysis previously Stimulation of ventral test and 72 hour voiding and 12 weeks. infection (1), pain on (Group 1 early received PFMT surface of glans once diary. Subjective stimulation (2), non intervention n= 19; with RP. daily; 10 seconds of assessment by ICIQ-SF, compliance (1) Group 2 delayed stimulation followed by IPSS and satisfaction Group 2 delayed intervention n=20 10 second pause questionnaire. intervention n= 5 repeated 10 times. Early intervention group: untreated diabetes Control: delayed 12/15 had reduction in (1), urinary tract intervention – no leakage on pad test infection (2), change intervention in first 6 baseline to 6 weeks (p = in drinking habit (1) weeks, PVS in second 0.021); maintained in 8/12 discontinued after 6 6 weeks at 12 weeks (p= 0.04) weeks non Delayed intervention compliance (kept in group: reduction (not analysis) (1) significant) with treatment Sample size weeks 6-12. No significant calculation was n=50, difference between groups enrollment stopped at 6 weeks. Pooled early as sample size analysis: Significant deemed adequate for overall median decline in purpose (pilot) urine loss on 24 hour pad test (p= 0.07) Authors suggest sufficient evidence to move to larger trial. Scheduled Voiding Regimes Burgio 2011 Combined N=143 men from 2 Men with Intervention: 8 weeks Primary outcome: post Baseline Intention-to-treat (23) behavioural sites. continued OAB (4 clinic visits) treatment 24 hour voiding assessment and analysis undertaken.

150 Study Modality details or Notes (side effects, Author, year Comparator N Outcomes/results Follow up population parameters loss of follow up…) treatment Randomized using symptoms after combined behavioural frequency on 7 day measure, post (PFME, urge sealed envelopes after treatment with treatment composed of bladder diary. Other OAB treatment Dropouts reported: suppression, stratification for voiding an alpha blocker elements of bladder symptoms (nocturia, measures. n=9 in behavioural delayed frequency for prostatic training (PFME, urge urgency, incontinence) therapy group, n= 10 voiding) vs drug Behaviour treatment obstruction. suppression also collected on bladder in drug therapy group. therapy n=73 techniques and diary. Secondary Drug therapy n=70 delayed voiding). outcomes: global Verbal instruction by perception of improvement nurse practitioners, and patient satisfaction. and PFME guided Results practice using verbal Frequency: both groups feedback based on had statistically significant anal palpation. Daily improvement from practice of 45 baseline to after treatment. exercises (3 sessions Equivalence analysis of 15 exercises). Also showed post treatment used fluid voiding frequency management. equivalent between behaviour and drug Control: drug therapy therapy groups (p< 0.001) with oxybutynin 5- Behaviour therapy group 30mg. had greater reduction of nocturia episodes, but urgency scores were lower in the drug therapy group. No difference in reduction in incontinence episodes. BF – biofeedback; RP – radical prostatectomy; TURP – transuretheral prostatectomy; PFME – pelvic floor muscle exercises; PFMT – Pelvic floor muscle training; PVS – Penile vibratory stimulation; ICIQ-SF – International Consultation on Incontinence Questionnaire Short Form; IPSS – International Prostate Symptom Score 1. Ahmed MT, Mohammed AH, Mansour AA. Effect of pelvic floor electrical stimulation and biofeedback on the recovery of urinary continence after radical prostatectomy. Turkish Journal of Physical Medicine and Rehabilitation. 2012;58(3):171-7. 2. Baroni M, Lorenzetti R, Renzi C, Brizzi A, Branchini W, Altavilla MG, et al. Approach HTA (health technology assessment) to treat urinary incontinence after radical prostatectomy (Abstract number 23). Neurourology and urodynamics. 2013;32:S20. 3. Burkert S, Scholz U, Gralla O, Roigas J, Knoll N. Dyadic planning of health behavior change after prostatectomy: a randomized controlled trial planning intervention. Social Science and Medicine. 2011;73:783-92.

151 4. Collado Serra A, Pellicer Cabo M, Ramirez Backhaus M, Dominguez-Escrig J, Rubio-Briones J, Gomez-Ferrer A, et al. Intensive preoperatory Pelvic Floor Muscle Training reduce duration and severity of stress urinary incontinence after radical prostatectomy: A randomized controlled trial (Abstract number 1007). European Urology Supplements. 2013;12(1):e1007-e8. 5. Dijkstra-Eshuis J, Van den Bos TWL, Splinter R, Bevers RFM, Zonneveld WCG, Putter H, et al. Effect of preoperative pelvic floor muscle therapy with biofeedback versus standard care on stress urinary incontinence and quality of life in men undergoing laparascopic radical prostatectomy: a randomized control trial. Neurourol- ogy and urodynamics. 2015;34:144-50. 6. Geraerts I, Van Poppel H, Devoogdt N, Joniau S, Van Cleynenbreugel B, De Groef A, et al. Influence of preoperative and postoperative pelvic floor muscle training (PFMT)compared with postoperative PFMT on urinary incontinence after radical prostatectomy: a randomized controlled trial. European urology. 2013;64:766-72. 7. Ghanem A, Khallaf M, Assem A, Hassan A. Does preoperative pelvic floor muscle exercise improve post prostatectomy urinary incontinence (Abstract 695)? . Conference: Annual Meeting of the International Continence Society ICS, Barcelona, Spain. 2013. 8. Hou CP, Chen TY, Chang CC, Lin YH, Chang PL, Chen CL, et al. Use of the SF-36 quality of life scale to assess the effect of pelvic floor muscle exercise on aging males who received transurethral prostate surgery. Clinical Interventions in Aging. 2013;8:667-73. 9. Kakihara C, Sens Y, Ferreira U. Effect of functional training for the pelvic floor muscles with or without electrical stimulation in cases of urinary incontinence following radical prostatectomy. Revista Brasileira de Fisioterapia. 2007;11(6):481-6. 10. Kongtragul J, Tukhanon W, Tudpudsa P, Suedee K, Tienchai S, Leewansangtong S, et al. Effects of adding concentration therapy to Kegel exercise to improve continence after radical prostatectomy, randomized control. Journal of the Medical Association of Thailand. 2014;97(5):513-7. 11. Laurienzo CE, Sacomani CAR, Rodrigues TR, de Cassio Zequi SGGCLA. Results of preoperative electrical stimulation of pelvic floor muscles in the continence status following radical retropubic prostatectomy. International Brazilian Journal of Urology. 2013;39:182-8. 12. Laurienzo C, Magnabosco W, Jabur F, Gameiro M, Yamamoto H, Guerra R, et al. Post-prostatectomy urinary incontinence and erectile dysfunction: The role of pelvic floor rehabilitation (Abstract number 527). Conference: Annual Meeting of the International Continence Society, ICS Montreal, QU, Canada. Neurourology and urodynamics. 2015;34:S449-S50. 13. Martini M, Bernardini S, Blanc E, Piretta K, Tappero R. Relationship between integrity of pelvic floor function and recovery of continence after laparoscopic prostatectomy and effects of preventive pelvic floor muscle training in males with pelvic floor weakness (Abstract 14). Conference: Annual Congress of the Italian Urody- namics Society, Turin, Italy. Neurourology and urodynamics. 2011;30(SUPPL 1):11-2. 14. Morihoro N, Masatsugu I, Shinji K, Kenichi T, Kazumasa M, Shiro B. Effectiveness of sacral surface therapeutic electrical stimulation (SSTES) on early recovery of urinary incontinence after laparoscopic radical prostatectomy: a prospective study. Neurourology and urodynamics. 2011;30(6):889-90. 15. Ng SI. A Randomised Controlled Trial Study of the Efficacy of Intensive Pre-Operative Pelvic Floor Muscle Training to Decrease Post-Prostatectomy Urinary Incon- tinence (Abstract number OP.4.7Dec.38). International Journal of Urology. 2014;21:A169. 16. Ocampo-Trujillo A, Carbonell-Gonzalez J, Martinez-Blanco A, Diaz-Hung A, Munoz CA, Ramirez-Velez R. Pre-operative training induces changes in the histomorphometry and muscle function of the pelvic floor in patients with indication of radical prostatectomy. Actas Urologicas Espanolas. 2014;38(6):378-84. 17. Park SW, Kim TN, Nam JK, Ha HK, Shin DG, Lee W, et al. Recovery of overall exercise ability, quality of life, and continence after 12 week combined exercise intervention in elderly patients who underwent radical prostatectomy: A randomized controlled study. Urology. 2012;80:299-306.

152 18. Pedriali F, Gomes C, Soares L, Urbano M, Moreira E, de AS. The efficacy of pilates compared to pelvic floor muscle training associated with electrical stimulation in the recovery of post-prostatectomy urinary incontinence: A randomized controlled trial (Abstract number 306). Neurourology and urodynamics. 2014;33(6):742- 3. 19. Serda BC, Marcos-Gragera R. Urinary incontinence and prostate cancer: a progressive rehabilitation program design. Rehabil Nurs. 2014;39(6):271-80. 20. Tienforti D, Sacco E, Marangi F, D’Addessi A, Racioppi M, Galino G, et al. Efficacy of an assisted low-intensity programme of perioperative pelvic floor muscle training in improving the recovery of continence after radical prostatectomy: a randomized controlled trial. BJU international. 2012;110:1004-11. 21. Zopf EM, Bloch W, Machtens S, Zumbe J, Rubben H, Marschner S, et al. Effects of a 15-month supervised exercise program on physical and psychological outcomes in prostate cancer patients following prostatectomy: The ProRehab study. Integrative Cancer Therapies. 2015;14(5):409-18. 22. Fode M, Sonksen J. Penile vibratory stimulation in the treatment of post-prostatectomy incontinence: a randomized pilot study. Neurourology & Urodynamics. 2015;34(2):117-22. 23. Burgio KL, Goode PS, Johnson TM, Hammontree L, Ouslander JG, Markland AD, et al. Behavioral vs. drug treatment for overactive bladder in men: The Motive Trial. Journal of the American Geriatrics Society. 2011;59(12):2209-16.

153 Results 2. PELVIC FLOOR MUSCLE TRAINING (PFMT) Ocampo-Trujillo and colleagues (2013) did not report the results of the 24-hour pad test data but did report that 6/8 (75%) of the intervention group compared to Almost all studies of PFMT in men focused on UI as- 4/8 (50%) of the control did not require pads at 8 sociated with prostate surgery. In the 5thICI, improve- weeks. The authors also report a statistically signifi- ment in the quality of trials included was found when cant difference in objective measures of muscle tis- compared to earlier reports, although heterogeneity sue from the external sphincter of the urethra favour- and varying outcome measures continued to affect ing the intervention group. As no histology samples the ability to compare trial findings. Studies in the first were taken prior to the intervention, the claim that the part of this section remain grouped into three types: difference is due to the intervention cannot be sup- pre-operative interventions, mixed pre- and post-op- ported. Subjective report of symptoms reportedly fa- eratives’ studies of all men undergoing prostatectomy voured improvement in the intervention group. Lauri- and studies of post-operative interventions for men enzo (2013) reported no differences between inter- with established incontinence. A total of 20 new RCTs vention and control groups at any time point (1, 3 and of PFMT alone or combined with other conservative 6 months post-surgery) on objective measure of in- interventions focused on prostate surgery were continence (one-hour pad test). There were no differ- added to the 38 studies from the previous ICI reports. ences between the two interventions and one control Nineteen of the new trials involved men pre or post group on subjective report of quality of life. Collado RP (N= 1597 randomized), one trial involved men Serra (2013) reported improvement in continence at post TURP (N= 66 randomized). Recommendations week 6, month 3, month 6 and 1-year favouring inter- are at the end of the section. vention group, but did not clearly define its primary 2.1. Pre-Operative RP PFMT outcome. Significant difference on objective 24-hour pad test was only at 3 months favouring the interven- Trials in this section compared pre-operative inter- tion group. ventions. Three new trials of pre-operative PFMT intervention (313, 314, 320) were identified and added 2.2. Pre-Operative and/or Post-Operative to the two previously included trials. Ocampo-Trujillo RP PFMT, Post RP Continence Status (314) compared pre-operative PFMT with BF (n=8) to Not Established Prior to Intervention standard pre-operative teaching (n=8). Laurienzo (313) randomized to three groups, reporting group Trials in this section addressed the effect of PFMT in- numbers for only the 49 of 58 men that completed the itiated pre-operatively and/or post-operatively but be- study: pre-operative PFME plus EStim (n= 17), pre- fore post-op continence/incontinence was estab- operative PFME (n=17) and control which received lished. In these studies, men were recruited pre-op- only pre-operative information on prostate anatomy eratively or immediately post-operatively and the in- (n=15). Collado-Serra (320) also reported group num- vestigators compared a treatment group that received bers for those completing (179 or 193) and compared the intervention pre-operatively and/or post-opera- an intervention group receiving pre-operative PFMT tively RP to a group receiving an intervention only plus BF (n=87) to a control group (n=92) receiving post-operatively. Nine new studies (305-307, 309, only oral PFME instruction post-surgery followed by 315, 321, 323, 324, 326) were added to the 10 previ- PFMT plus BF 3 months later. ously included trials, for a total of 19 trials in the category, over three subsections. Quality of data 2.2.1 Pre-Operative PFMT Instruction with Ocampo-Trujillo (2013) used a centrally randomised, Post-Operative Home PFMT Versus single blind approach, with researchers blind to the Control group assignment prior to enrolment. No information on sample size analysis or post hoc power analysis Four new trials (307, 321, 323, 326) were added to was provided. This was a very small study that was three previously included studies. In the previous likely underpowered. There were no dropouts from consultation, studies varied in support of an earlier re- the study. Laurienzo (2013) described randomization turn to continence with PFME plus or minus biofeed- using a computer-generated list. Detail of blinding back. Dijkstra-Eshuis (2015) compared an interven- was not provided and a single therapist provided all tion group that received preo-perative PFMT plus BF interventions. There was no report of sample size cal- with home practice and post-operative home PFME culation, post hoc power analysis or intention-to-treat (n=65) to a control group that received written post- analysis. Nine dropouts were reported, but the group operative PFME instruction (n=56). Ghanem (2013), from which they withdrew was not given. Collado available only in abstract, compared an intervention Serra (2013) study was available only as an abstract group with a PFME programme two weeks pre-oper- and did not report how participants were randomized atively and continued post-operatively (n=50) to a or if there was any blinding. There was no information control group that received only a postoperative on sample size, and no intention-to-treat analysis re- PFME programme (n=50). Detail of the PFME pro- ported. Only dropouts from the intervention group gramme was not provided. Ng (2014), also only avail- (n=5) were described. able in abstract, compared an intervention (of PFMT starting 3 weeks pre-operatively to a control group

154 that received standard care. A total of 66 men partic- up. Geraerts (2013) compared the intervention (n=91) ipated, but numbers randomized to each group and of PFME plus BF and a home program starting 3 details of the standard care were not provided. Martini weeks prior to surgery with PFME starting 4 days (2011), again in abstract only, described an interven- post-operatively to a control group (n=89) that started tion group (n=24) that received five physiotherapy PFMT post-operatively. Both groups received weekly guided PFMT sessions 2-3 weeks pre-operatively, post-operative treatment that included BF and blad- with continued PFME post-operatively. The control der diary review, with PFMT continuing until conti- group (n=25) received post-operative written instruc- nence achieved. tion on PFME. Quality of data Quality of data Tienforti (2012) used a computer-generated random- Dijkstra-Eshuis and colleagues (2015) reported ade- isation schedule. Information on blinding to group as- quate random allocation concealment. Participants signment and outcome assessment was not pro- and therapists were blinded to randomisation until vided. In this small study, no sample size calculation first visit. They also reported results of follow up were or intention-to-treat was reported. Dropouts in each send to clinic of the single surgeon performing the group, one in each intervention and control (6% from RP, but blinding of outcome assessment is not made each group) were described. Geraerts (2013) used explicit. Sample size analysis was provided, inten- computer generated randomisation using stratification-to-treat was not addressed. Dropouts in both in- tion for age and surgical approach, blinding to group tervention (n=9/65, 13.6% and control groups assignment and outcome assessment was not pro- (n=10/56, 18%) were described. Three studies (321, vided. In this large trial, sample size calculation re- 323, 326) were available only as abstracts with ran- cruited an adequate sample to power the study. In- domisation, blinding, sample size calculation and tention-to-treat analysis was not discussed. There dropouts not described. were n=6/91 (7%) dropouts in the intervention group, and n=4/89 (5%) in the control, reasons were pro- Results vided. In all new studies, the primary outcome was conti- Results nence measured by either the pad test or number of pads used, secondary outcomes were self-report of The newer studies continue to report variable results. symptoms and/or quality of life on standardised ques- Tienforti (2012) found a significant difference in tionnaires. Dijkstra-Eshuis et al. (2015) found no dif- achieving continence at one, three and six months fa- ferences between intervention and control at any time vouring the intervention group, but this was by self- points on the primary objective outcome (24-hour pad report of symptoms, with no objective measure of test) or on secondary self-reported symptom and continence included as an outcome. In the Geraerts quality of life measures but the trial was stopped prior (2013) study, which was larger and of good quality, to full recruitment after a planned interim analysis there were no differences between the groups with showed no demonstrated benefit, with the authors regards to objective duration of UI measured by 24- concluding that post RP UI is likely due to intrinsic hour pad test or self-reported symptoms at any time sphincter injury and not treatable with exercise. Two point. The intervention group scored better on impact other studies (321, 326) reported an earlier postoper- of incontinence at 3 and 6 months, but not at 12 ative return to continence (at 14 weeks and 3 months months follow up. respectively) favouring the intervention group, but the effect was not demonstrated later time points. Simi- 2.2.3 Post-Operative PFMT Immediately Be- larly, Martini (2011), who included only men with fore or After Catheter Removal (No Pre- demonstrated pelvic floor impairment prior to surgery, Operative Instruction) reported significantly less UI on getting up and squat- Three new trials (305, 306, 324) were found compar- ting (no quantification provided) in the intervention ing post-operative PFMT immediately before or after group at 1 month but not at later time points. catheter removal, bringing the total to six. In the pre- 2.2.2 Pre-Operative PFMT Instruction Fol- vious consultation, there was considerable heteroge- lowed by supervised Post-Operative neity in intervention and findings. Ahmed (2012), PFMT Versus Post-Operative PFMT compared three groups, reporting on group assignment only for those completing the trial. Group 1, the Two new trials (309, 315) were added to the previous control group (n=26), received PFME post-opera- four studies in this subsection, making a total of six tively. Group 2 (intervention, n=26) started PFME studies included. Tienforti (2012) randomized 34 men plus EStim after catheter removal for 12 weeks, and undergoing RP, but reported only on group assign- Group 3 (intervention, n=28) started PFME plus ES- ment for those completing the study. The intervention tim and BF after catheter removal for 12 weeks. Mori- (n=16) was pre-operative and post-operative PFMT horo (2011) was available only in abstract form, and with BF, started the day before surgery and continu- compared an intervention group (n=20) that received ing monthly, with daily home practice, during follow- PFMT with EStim for one month after catheter re- up. The control group (n=16) received oral and written moval to a control group (n=14) assigned to PFME instructions post-operatively with no monthly follow-

155 alone. Burkert (2011), using a 2x2 mixed design, ran- of sphincter function using imaging techniques would domized patients and their partners to one interven- be helpful in understanding the subpopulations where tion group (n=28) that received dyadic PFME plan- conservative treatments are most helpful. There is ning or to 3 control groups (n= 29, 29 and 26) that modest evidence but inconsistent evidence that ther- received dyadic nutrition planning, individual PFME apist delivered PFMT with or without BF or EStim be- planning or individual nutrition planning. fore or after surgery may support an earlier return of continence after RP in some men up to 3-6 months Quality of data post-surgery, however this difference is not signifi- Morohiro (2011) available only as an abstract, did not cant at 12 months post-surgery (Level of Evidence: describe randomisation, blinding, sample size calcu- 2). lation or dropouts. Ahmed (2012) randomised using Variation in outcomes measurement remains prob- computer generated random numbers in sealed en- lematic. As concluded in the previous consultation, it velopes. Surgeons were blinded to randomisation but is possible that the emphasis on quantitative out- blinding of the outcomes assessor was not clear. comes (i.e. pad test) is not meaningful to participants Sample size was met, and dropouts from each group as men appear to find therapy helpful and value the described: control n= 4/26+4 (13%), intervention direction provided by a therapist. Studies comparing Group 2 n= 2/26+2 (7%) and intervention Group 3 the effectiveness of pre- versus post-operative n=4/28+4 (12.5%). No intention-to-treat analysis was PFMT, and the number of sessions required, are undertaken. In the Burkert (2011) study, research as- needed so that practitioners may advise men about sistants were blinded to allocation. Sample size cal- pre-operative preparation and budget conscious culation and intention-to-treat were not discussed. health services can make informed decisions on pro- Dropouts for the intervention group n=5/28 (18%) and gramme funding. In designing such studies, the natu- three control groups (n= 2/29 (7%), 7/29 (24%), 4/26 ral history of UI after radical prostatectomy must be (15%) were presented but no reasons for withdrawal considered because the spontaneous recovery rate were provided. means that sample sizes must be large to detect any Results differences between protocols. As well, attempts to identify which men might benefit more, such as in the Ahmed (2012) reported more men continent at 12 study by Martini (2011) who focused the intervention and 24 weeks post RP in the two intervention groups on those with identified pelvic floor impairment, need (PFMT plus EStim and PFMT plus EStim and BF) to be undertaken. compared to PFMT alone using the objective 24-hour pad test as primary outcome. Mean leakage was sig- Recommendations nificantly lower in intervention Group 3 at 6 through Some pre-operative instruction or immediate post-op- 24 weeks. Morihoro (2012) found recovery of urinary erative instruction in PFMT for men undergoing RP continence favoured the intervention group (PFMT may be helpful in earlier recovery of continence with EStim) at 12 months. Incontinence was meas- (Grade of Recommendation: B Clarified). ured by subject report of requiring a pad to keep clothing dry. Burkurt reported only adherence to PFMT Use of BF to assist PFMT should remain a thera- with no continence data provided. There were no dif- pist/patient decision based on economics and prefer- ferences between the groups. ence. (Grade of Recommendation B).

Summary 2.3. Post-Operative RP PFMT for Incontinent Men A total of 9 new trials are now included which apply a variety of pre- or immediately post-operative or post- Trials in this section addressed the effect of PFMT catheter removal PFMT based interventions (or a post-operatively after incontinence was established. combination of both, plus or minus BF and EStim). As In these studies, incontinent men were recruited at in earlier consultations, differences between interven- variable lengths of time after surgery. tion and control groups were modest and short term, and often reflect self-reported but not objective data 2.3.1 PFMT with Digital Rectal Feedback such as the pad test. Few differences were sustained (DRE) After Radical Prostatectomy up to 12 months post-surgery. Many of the studies re- No new trials were identified to add to the previous viewed were small, varied in design and quality, and five trials in which PFMT was taught using digital rec- had different outcome measures. The two larger stud- tal feedback (DRE) to men incontinent after RP. ies that were of better quality (307, 309) found no difference between intervention and control groups. The authors of one of larger study (307) suggested post RP UI is likely often due to a mechanism such as intrinsic sphincter deficiency that is not amenable to exercise based treatment. It is possible that this is true for some men with persistent UI post RP, and they may require surgical rather than conservative intervention. Studies that include evaluation of the extent

156 men receiving PFMT with EStim compared to those Summary who received only PFMT. Men had undergone urody- There are no new trials to clarify whether PFMT namic testing, and both those with UUI and SUI were taught by DRE offers any benefit over and above ver- included. Pedriali (2014) reported 58% of the Pilates bal or written instruction, (Level of Evidence: 2). group and 50% of the EStim plus PFME achieved continence, as measured by use of no pads. The dif- Recommendations ference was not statistically significant and the au- The recommendation is unchanged (Grade of Rec- thors concluded that Pilates, which includes pelvic ommendation: B). floor strengthening, is as efficacious as EStim plus PFME. Laurienzo (2015) reported no differences be- 2.3.2 PFMT with BF After Radical Prostatec- tween controls and two intervention groups (PFMT, PFMT with EStim) at any time points, objective pad tomy test results not provided, subjective reports of quality No new trials were found under this category to add of life and symptoms only. to the six previously included trials. Summary Recommendations Data suggest no further benefit of EStim when added The use of BF in clinic, over and above home PFMT, to PFMT over PFMT alone for men with incontinence should remain a therapist/individual decision based post RP (Level of Evidence: 2). on economics and preference. (Grade of Recom- Recommendations mendation: B). There does not appear to be any benefit of adding 2.3.3 PFMT Plus or minus BF with EStim or EStim to a PFMT programme for men with persistent MStim after Radical Prostatectomy post-prostatectomy incontinence. (Grade of Recom- mendation: B). Three new trials (311, 322, 325) were added to the eight previously included in this section. As in the ear- 2.3.4 PFMT Compared to or Plus Other Inter- lier consultation, a problem of heterogeneity of study ventions After Radical Prostatectomy samples in this category was noted. Kakihara (2007) compared an intervention of physiotherapy lead Four new trials (312, 316, 318, 319) were identified PFMT with EStim (n=10) to PFMT alone (n=10) in that combined PFMT post-operatively with other men a minimum 6 months post-surgery (mean 12.3 novel interventions in men with post prostatectomy in- months for intervention group, 16.8 months for con- continence, adding to the two previously included trol). Pedriali (2014) recruited incontinent men one- studies. The previous consultation found variation in month post RP and assigned them to post-operative intervention and time since surgery. Zopf (2015) re- pilates (n=26) versus PFMT plus EStim (n=28), but cruited men 6-12 weeks post RP, and assigned them reported group assignment in only the 54/69 (78%) of to either an intervention of a multimodel exercise pro- men completing the study Laurienzo (2015), available gramme with PFME (n=50) or a control group that re- only in abstract, also included men incontinent one- ceived no intervention (n=20). Serda (2014), using a month post RP, randomising to three groups: Group two group design, randomised men post RP to a com- 1 control (no treatment) (n=40), Group 2 PFMT bined exercise and PFME intervention group (n=36) (n=41) and Group 3 PFMT plus EStim (n=42). or control with no intervention (n=33). Time after surgery was not reported. Kongtragul (2014) randomised Quality of data 138 incontinent men recruited in hospital after RP. Kakihara (2007) did not provide information on ran- Group assignment was reported only for those com- domisation or blinding. No sample size calculation pleting the study; 68 men were in the intervention was given. Dropout numbers for each group were group with PFME and concentration therapy (no de- given, but the reasons for dropping out of the inter- tail provided) starting 3 weeks post-surgery. The con- vention group were not provided. Three participants trol group (n=67) received only PFME. Baroni (2013) were discharged from the control group at 3, 6 and 12 was available only in abstract. Men with SUI or mixed months as continence was achieved and were in- UI after RP (time after surgery not reported) were as- cluded in the final analysis. Pedriali (2014), available signed to an intervention of individual PFMT and only as an abstract, did not provide details of random- group treatment (n= 16). Controls (n=24) received isation, blinding, sample size calculation or dropouts. only individual PFMT. Laurienzo (2015) was only available in abstract, but Quality of data did describe randomisation approach, but not blinding, sample size calculation or dropouts. The Zopf (2015) study had intended to randomise those who consented, but as all participants refused Results randomisation and gave a group preference, blinding In the Kakihara (2007) trial, there was improvement of outcome measurement was not described. The au- in urine loss on 1-hour pad test in both groups be- thors indicated intention–to-treat analysis was under- tween baseline and 12 months, but no difference in taken but the study did not achieve the calculated

157 sample size and was thus underpowered. Dropouts 2.5. Pre-Operative and/or Post-Operative from the intervention group only (n=6/50 12%) were TURP PFMT described, but reasons were not provided. Serda (2014) randomised participants but did describe how No new trials were identified. this was done. Blinding to outcome assessment was not reported. Calculated sample size was achieved, Summary number of dropouts (intervention group (n=3/36 8%; In the absence of sufficient data from rigorous and control 0/33) and reason for withdrawal was provided. well-reported trials it is not known if PFMT reduces UI Baroni (2013) was available only in abstract and had following TURP. More systematic investigation of the no detail on randomisation but reported the assessor natural history of UI after TURP is probably needed, was blinded until assignment. Sample size calcula- to establish the potential cost/benefit of intervention, tion and dropouts were not described. Kongtragul before further trials are initiated. (2014) did not describe the approach to randomisation or blinding, but dropouts (intervention n=1/68+1, 2.6. Post-Operative TURP PFMT for 1%; control n=2/67+2, 3%) were described. The study did not achieve the number in each group re- Incontinent Men quired for power to detect a difference between One new study (310) was added. The previous con- groups. sultation reported limited evidence in this area. The Results larger well designed study previously included showed no benefit in terms of objective measures of Zopf (2015) reported a significant improvement in UI, incontinence. Hou (2013) reported group assignment as measured by the 20-minute pad test, from base- only for those completing the study. The intervention line to post testing, but no significant difference be- group (n=32) was given instruction in PFME two days tween intervention (exercise plus PFMT) and controls after surgery when the catheter was removed, and in- (no intervention). However, a significant difference structed to practice daily at home, with weekly tele- between the groups on reported urinary symptoms phone reminders. The control group (n=29) activity was found. Serda (2014) reported a significant im- was not described. provement in favour of the intervention group (combined exercise and PFME) on self-report of the inten- Quality of data sity of urinary symptoms, but did not report the results Approach to randomisation, blinding, sample size cal- of the 20-minute pad test. Baroni (2013) reported no culation and dropouts in the smaller, newly included significant difference between intervention and con- trial (Hou 2013) were not described. trol on the VAS, number of pads used or home practice. In the study of concentration exercises plus Results PFME (Kongtragul 2014), significant differences were Hou (2013) found a statistically significant difference found between intervention and control groups on in self-report of symptom severity favouring the inter- achieving continence measured by the 1-hour pad vention group at 12 weeks. No objective measure of test. incontinence, such as the pad test, was reported. Summary Summary Interventions were varied: general exercise plus There continues to be limited evidence on the benefit PFME (316, 318) PFME with concentration therapy of PFME post TURP. The new smaller trial suggests (not defined by author) (312) and individual PFMT a potential benefit on perception of symptom severity compared to individual and group therapy (319). of UI, but no objective data were reported. There was also heterogeneity of samples in terms of time since surgery. There is some early evidence that novel interventions, such as general exercise pro- 2.7. Factors Affecting Outcome grammes, use of support groups and concentration Based on the current evidence, it appears that time exercises added to PFME after RP for incontinent from surgery to implementation of exercises does af- men may be beneficial, but this is very limited. Some fect outcome and that by three months after surgery of the trials are small and of low quality. Larger stud- less improvement is noted. Future trials should con- ies of high quality are needed to determine the bene- sider analysis to evaluate the effect of sphincter insuf- fits of these adjuncts to PFME. ficiency, pelvic floor dysfunction, urethral length, length of time from prostatectomy, co-morbid condi- 2.4. Pre-Operative TURP PFMT tions, prior pelvic surgery, medications, other risk factors (including smoking and alcohol use) on treatment No new trials were identified. Little research has been outcome. dedicated to UI after TURP as the incidence of UI after TURP is reported to be very low.

158 2.8. PFMT for Other LUTS had been included in previous consultations. Details of these were provided in the 5th consultation (Moore 2.8.1 PFMT for Post-Micturition Dribble 2013): insufficient data exist to establish if EStim is (PMD) better than either intervention. Both studies combined No new trials were identified to add to the two previ- male and female data. ously included trials. 3.2.4 Does the Addition of EStim to Other Treatments Add Benefit? Recommendations All of the studies combining EStim with other treat- Offer men with PMD instruction to do a strong PFM ments included the PFMT plus EStim combination, contraction immediately after voiding, or urethral and are reviewed under the Pelvic Floor Muscle massage to empty the urethra (Grade of Recom- Training section. mendation: C). 3.3. Other LUTS

No new studies were identified. No previous studies of EStim alone for other LUTS were included in pre- 3. ELECTRICAL STIMULATION vious consultations. (ESTIM) 3.4. Factors Affecting Outcome

An extensive overview of EStim in men is reported in The previous consultation identified age, type of in- the 5th consultation(2). Rectal or surface electrodes continence (SUI post prostatectomy versus OAB and are most common; surface electrodes are positioned UUI) and other factors (types of electrodes, frequency over the perineal region. EStim can also be applied and duration of treatment) as potentially affecting out- via the posterior tibialis nerve (PTN). Trials of EStim comes. No trials have investigated these factors in in men post-prostatectomy combined this treatment males. with PFME, and are included in the sections on post- prostatectomy treatment above. 4. MAGNETIC STIMULATION (MSTIM) 3.1. Prevention of UI Although MStim has been used to treat UI after RP No new trials were identified. There remain no studies and to inhibit DO, no new studies were identified. on the effect of EStim for prevention of non-post pros- 4.1. Prevention of UI tatectomy UUI or SUI in men. No trials investigating the primary or secondary pre- 3.2. Treatment of UI vention effects of MStim for men with UI were found. No new trials were added to the nine previously in- 4.2. Treatment of UI cluded RCTs, four of which included PFME plus ES- tim post prostatectomy. New trials that combined ES- No new trials to add to the three trials (two published, tim with PFME were reported in the PFMT section one abstract without data) identified and described in above as the separate effects of EStim alone cannot the 5th consultation were found. Only the two pub- be assessed. lished trials were included in the previous consultation. 3.2.1 Is EStim Better than No Treatment, Pla- cebo or Control Treatments? 4.2.1 Is MStim Better Than No Treatment, Placebo or Control Treatment? No new trials were identified to add to the two previously included trials. Both previously included trials No studies were found addressing this question. did not separate male and female results. In the continued absence of sufficient data from rigorous and 4.2.2 Is One Approach to MStim Better Than well-reported trials it is not known if EStim, as a stand- Another? alone treatment for male UUI or SUI, is better than no No studies were found addressing this question. treatment, placebo or control treatments. 4.2.3 Is MStim Better Than Other Treat- 3.2.2 Is One Approach to EStim Better Than ments? Another? No new studies were identified. No new trial was identified. No studies comparing ES- tim protocols were included in previous consultations. 4.3. Other LUTS 3.2.3 Is EStim Better Than Other Treat- No studies were found. ments? 4.4. Factors Affecting Outcome No new trials were identified. Three studies, two comparing EStim to Mstim and one EStim to medication, The relationship between age (or any other factor, such as treatment parameters, treatment adherence,

159 or diagnosis) and the outcome of MStim has yet to be Conclusion determined. This was a small pilot study of a novel intervention. Further well designed studies are needed to under- 5. PENILE VIBRATORY stand the place for PVS in treatment of male UI. No STIMULATION (PVS) recommendation can be made. 5.3. Other LUTS This is a novel conservative treatment, based on the stimulation of the pudendal nerve via penile vibratory No trials were identified. stimulation (PVS) to treat SUI in men post prostatectomy. 6. SCHEDULED VOIDING REGIMENS PVS has been shown to increase external sphincter pressure in men with spinal cord injury (334, 335), Scheduled voiding regimens include bladder training, and vibratory stimulation applied to the perineum in timed voiding, habit training and prompted voiding. healthy women also increased external urethral pres- They are frequently combined to achieve maximum sure (336). benefits. Although there is evidence to suggest that scheduled voiding regimens, especially bladder train- 5.1. Prevention of UI ing and timed voiding, are commonly used in the No trials of prevention were identified. treatment of men with UI and other LUTS, there has been substantially less research that addresses their 5.2. Treatment of UI use in men compared to the literature on their use in women, leaving insufficient evidence to comment on One new trial was identified (308). Fode (2015) ran- effectiveness. domised 39 men with UI over one-year post RP to two groups. Group 1 (n= 19) was the early intervention 6.1. Prevention of UI group who received PVS using a commercially available hand held personal vibrator for 6 weeks. Stimu- No trials investigating the preventive effects of sched- lation was performed on the ventral surface of the pe- uled voiding regimens for men with UI were found. nis once daily with 10 seconds of stimulation followed 6.2. Treatment of UI by a ten second pause, repeated 10 times. Group 2 was the delayed treatment control group, who had no 6.2.1 Bladder Training intervention the first 6 weeks, but performed PVS in the second 6 weeks. One new trial of a mixed behavioural treatment approach composed of elements of bladder training Quality of data compared to drug therapy in men with OAB was identified (317) and added to the five previously included Computer generated randomisation technique was trials of bladder training that have included men. Bur- described, but blinding of outcome measurement was gio (2011) compared a combined behavioural treat- not (308). A sample size calculation was provided, but ment (PFME, urgency suppression, delayed voiding) the authors report stopping recruitment early as the to drug therapy (n=73) with drug therapy (individually study was a pilot and the sample achieved was titrated extended release oxybutynin) for OAB (n = thought to be adequate. Drop outs in both early 70. This category previously included studies that (n=4/19, 21%) and delayed (n=5/20, 25%) interven- have included BT plus caffeine reduction, BT plus pla- tion groups were described, with one of five dropouts cebo or anticholinergic drug therapy in the delayed intervention group kept in the analysis, although others were not. The participant kept in had Quality of data a pattern of non-compliance over the study, and discontinued after 6 weeks. Burgio and colleagues (2011) used sealed envelopes to randomise, blinding of person(s) gathering data on Results outcomes was not described. Sample size analysis was not given, but intention-to-treat analysis was un- Primary outcome was difference in leakage meas- dertaken. Dropouts from each group (combined be- ured by 24 pad test. Participants also completed void- havioural therapy n=9/73, 12%; drug therapy ing diary. Results showed a significant improvement n=10/70, 14%) were reported, reasons were not pro- in the early intervention group from baseline to 6 vided. weeks, but not in the delayed intervention group. Pooled analysis results from both early and delayed Results intervention groups showed significant overall median decline in urine loss on 24-hour pad test (p= Primary outcome was 24-hour voiding frequency rec- 0.07). orded on a seven-day bladder diary. Secondary outcomes were other OAB symptoms, perception of improvement and satisfaction. Results showed that treatment with behavioural strategies (bladder training) compared to pharmacological treatment was equivalent. There was no significant 160 difference between incontinence episodes between 7.2.1 What Is the Most Effective Acupunc- the behaviour and drug therapy groups. ture Protocol? 6.2.2 Timed Voiding No new trials were identified. No new trials were identified. 7.2.2 Acupuncture Versus No Treatment, Sham Acupuncture or Any Other Treat- 6.3. Other LUTS ment One new study was identified for treatment of other No new trials were identified. LUTS in men. The newly included study, described above (317) also measured OAB symptoms, with the 7.3. Other LUTS primary outcome being frequency of voiding. No new trials were identified. Quality of data 7.4. Factors Affecting Outcome Refer to C.VI.2a above. No new trials were identified. Results Primary outcome was 24-hour voiding frequency rec- 8. SUMMARY orded on a seven-day bladder diary. Secondary outcomes were other (non UUI) OAB symptoms includ- Despite the prevalence of UI and LUTS in older men, ing nocturia and urinary urgency; perception of im- research continues to focus mainly on men following provement and satisfaction. Treatment with behav- radical prostatectomy. Overall, the effect of conserva- ioural strategies (bladder training) compared to phar- tive treatment (lifestyle interventions, physical thera- macological treatment was equivalent. The behaviour pies, schedule voiding regimes, complementary ther- therapy group had greater reduction in nocturia epi- apies) for men has received much less research at- sodes, but urgency scores were lower in the drug tention compared to women. therapy group. 8.1. Recommendations for Practice Summary There is generally insufficient Level 1 or 2 evidence In men, behavioural treatment may be just as effec- on which to base recommendations for practice, and tive for some LUTS, including UI, as pharmacological most recommendations are, in effect hypotheses, therapy. Further studies of high quality are needed that need further testing in research. before a recommendation for practice can be sup- Lifestyle interventions ported. Weight loss through lifestyle changes should be rec- 6.4. Factors Affecting Outcome ommended to obese and overweight men with UI, particularly those with type 2 diabetes. (Grade of Age: No new trials were found addressing age as a Recommendation: Grade B New) factor affecting outcome. Smoking abstinence should be recommended for Other: No studies were identified on other factors af- men with UI (Grade of Recommendation: C New) fecting outcome of BT or prompted voiding in men. A reduction in caffeine intake is recommended for 7. COMPLEMENTARY AND those with incontinence symptoms. (Grade of Rec- ommendation: C New) ALTERNATIVE MEDICINES Pelvic floor muscle training (PFMT) Therapies include acupuncture, relaxation, medita-  Some pre-operative or immediate post-operative tion, imagery, hypnosis, naturopathic and herbal rem- instruction in PFMT for men undergoing radical edies. In previous consultations, only trials of acu- prostatectomy may be helpful in earlier recovery puncture therapy were found in men with UI. Studies of continence. (Grade of Recommendation: B were small, objective measured of UI were missing Clarified). and long-term follow-up was lacking, leaving only limited evidence on the effectiveness of acupuncture for  In men with persistent post-prostatectomy incon- men with UI. tinence, PFMT taught by digital rectal examination (DRE) may be undertaken but it is unclear 7.1. Prevention of UI whether this offers any benefit over and above No new trials were identified on the preventative role verbal or written instruction in PFMT (Grade of of complementary therapies in men. Recommendation: B Clarified). 7.2. Treatment UI  Use of BF in clinic, over and above home PFMT should remain a therapist/patient decision based No new trials were identified. on economics and preference (Grade of Rec- ommendation: B).

161  There does not appear to be any benefit of add-  A comparison of pre-operative versus post-oper- ing EStim to a PFMT programme for men with ative PFMT verbal and written feedback to re- persistent post-prostatectomy incontinence duce prevalence and severity of UI following rad- (Grade of Recommendation: B). ical prostatectomy is needed.  Use of a strong pelvic floor muscle contraction  Methods of PFMT instruction and supervision re- immediately after voiding, or urethral massage to quire further investigation. Two areas of re- empty the urethra, should be offered for symp- search interest are: toms of post-micturition dribble (Grade of Rec- ommendation: C). o Whether PFMT taught by DRE offers any benefit over and above verbal or written in- Electrical Stimulation (EStim) struction. For men with persistent post-prostatectomy inconti- o The effect of group exercise as peer support nence there does not appear to be any benefit of add- may be helpful to a healthy recovery. ing EStim to a PFMT programme (Grade of Recom- mendation: B).  Future studies should focus on identification of men more likely to benefit from conservative in- 8.2. Future Research Directions terventions, such as those with pelvic floor dysfunction, and screening those with potential in- Despite recognition that there is much scope for re- trinsic sphincter deficiency post-surgery. This search on the effects of conservative therapies for UI th might be established using imaging. Men with and LUTS in men in the 6 consultation, this remains extensive sphincter deficiency should be re- largely unexplored. Research that is urgently needed, ferred for urological intervention. Studies of con- in the opinion of committee members, is highlighted servative measures including PFMT should in- with the use of italics. The committee continues to clude those with milder damage in order to de- support the recommendations that apply to all future termine efficacy in this group. studies in men, namely:  More systematic investigation of the natural his-  All future intervention studies must be designed tory of UI after TURP is needed, to establish the to allow standardised and comprehensive re- potential cost/benefit of intervention, before fur- porting of results based on the ICS and ther trials are initiated. CONSORT recommendations.  The relationship between age, or any other factor  The natural history of UI after radical prostatec- (urethral length, PFM function), and the outcome tomy must be considered in study design as the of PFMT for UI in men, and predictors of success spontaneous recovery rate means that sample needs to be investigated. sizes must be large to detect any differences between protocols. 8.2.3 Electrical Stimulation (EStim) and Mag- netic Stimulation (MStim)  More research is needed to find out what are the most important outcomes for men with UI, so  It is not known if pre- or post-operative EStim or such measures can be incorporated as the pri- MStim has a role in reducing UI after radical mary outcome measures in further trials. prostatectomy.  Data is needed to establish to the cost, and cost  RCTs in larger samples, with long-term follow up, effectiveness, of conservative therapies in men are needed to investigate all aspects of the ef- with UI. fectiveness of EStim and MStim as a treatment for UI in men, including:  Surgical approaches with laparoscopy or robot- ics offer promising improvements in visualisation o Either type of stimulation versus no treat- for nerve-sparing procedures; further research ment, sham stimulation or other control con- should address continence and erectile function ditions. after these newer surgical procedures. o Comparisons of both EStim and MStim pro- 8.2.1 Lifestyle Interventions tocols.  The effects of interventions such as weight, caf- o EStim versus MStim. feine reduction, constipation and physical activ- o Either type of stimulation versus medication. ity are priorities for future research. o Whether the addition of either type of stimu- 8.2.2 Pelvic Floor Muscle Training (PFMT) lation to other treatments adds benefit, in  Further studies to test the hypothesis that pre- particular, the addition of stimulation to operative proprioceptive training plus PFMT is PFMT. more effective than PFMT alone to prevent UI in o The effect of age, and other factors, on out- men undergoing radical prostatectomy. come of stimulation. Older men may have more co-morbid conditions than young men 162 and a more pragmatic ap-proach to inclu- Evidence overview sion in EStim studies is needed. A total of 7 RCTs of posterior tibial nerve stimulation 8.2.4 Scheduled Voiding Regimens (PTNS) in adults were identified (337-343) and one Cochrane systematic review of anticholinergic drugs  RCTs of voiding regimens and combinations of versus non-drug active therapies for non-neurogenic conservative treatments in men with OAB and overactive bladder syndrome in adults (141), in which UUI are needed. Further comparison of these PTNS trials with adults of both sexes and men only conservative approaches to drug therapy would were included. Only one randomised trial of PTNS in be helpful for clinicians in recommending treat- men has been reported (344), which addressed post– ment to patients. stroke neurogenic bladder dysfunction and therefore did not meet the inclusion criteria for the conservative URINARY INCONTINENCE IN management chapter. No trials of non-neurogenic MEN AND WOMEN bladder dysfunction have been reported for men only. The trial data is summarised in tables 35 to 38. 1. POSTERIOR TIBIAL NERVE Three randomised trials compared PTNS with a sham STIMULATION (PTNS) intervention (337-339). Four trials were comparative: Posterior tibial nerve stimulation (PTNS) is a form of peripheral neuromodulation targeted towards symp-  Two compared different anticholinergics (340, tom relief of OAB and UUI (141). Evidence for the use 341) of PTNS for the prevention and treatment of UI in men  Two compared different stimulation protocols or in adults (study with men and women with results (342, 343) combined) is presented below. Percutaneous PTNS was used in 5 randomised trials Questions addressed are: (337, 339-342)  Can PTNS prevent UI? Transcutaneous PTNS was used in 2 (338, 343).  Is PTNS better than no treatment, placebo or No direct comparisons of percutaneous and control treatments for UI? transcutaneous PTNS for treatment of UI have been  Is PTNS better than other treatments for UI? undertaken in the adult population or in men only. On the basis of the included studies, percutaneous  Does the addition of PTNS to other treatments PTNS seems to involve a standard protocol with re- add any benefit? gard to stimulation parameters of frequency and ses-  What is the best programme of PTNS for UI in sion-duration using the Urgent PCTM stimulator. How- adults? ever, the number of individual sessions, overall duration of programme and timing of delivery protocols  What is the effect of PTNS on LUTS other than may vary. There is also variation between percutane- UI? ous and transcutaneous stimulation parameters and one transcutaneous device study (343) did not report  What factors might affect the outcome of PTNS? stimulation parameters, preventing comparison with A literature search for reports of relevant systematic other results. This variability reflects the limited un- reviews and reports of RCTs and quasi-RCTs was derstanding of the mechanisms of PTNS, which can- performed (see section I). Trial data reported in con- not be assumed to be identical for both percutaneous ference abstracts as well as full text papers were in- and transcutaneous routes. Further investigation is cluded. Since this is a new section for the Conserva- required for both percutaneous and transcutaneous tive Management chapter no date restrictions were PTNS to determine the most effective type of stimu- applied. lation and treatment protocols. Eligibility criteria: 1.1. Prevention of UI 1. Reports of RCTs or quasi-randomised trials of There have been no studies on the effect of PTNS for PTNS, percutaneous or transcutaneous. prevention of UUI/OAB in adults or in men only. 2. Adult men and women with UI and/or OAB (with 1.2. Treatment of UI or without urgency incontinence) presentation of data for men only or combined for men and 1.2.1 Is PTNS Better Than No Treatment, Pla- women. Trials including women only are pre- cebo or Control Treatments for UI? sented in section II.5. Three randomised controlled trials address this question in adult men and women. One RCT (337) is adequately powered whereas Vohra (339) and Booth (338) are both pilot RCTs.

163 Quality of evidence pilot trial (339) reported significant improvements in quality of life (QoL questionnaires not specified and Computerised randomisation was used in all three SF-36) but provided no data or figures to support this. studies with adequate allocation concealment reported in two (337, 338) and unclear in one (339). A prospective study to assess long-term outcomes Subjects and outcome assessors were blinded and determine frequency of top-up stimulation ses- throughout in two studies (337, 338), with blinding un- sions required was reported (345). Fifty responders clear in one(339). Two studies reported intention-to- to the original trial underwent a fixed 14-week tapered treat analysis (337, 338); the type of analysis was not stimulation protocol, followed by a personal treatment reported in one (339). There was no apparent ine- plan aimed at maintaining improvements. Twenty- quality in loss to follow-up across the groups: in one nine of 50 (58%) completed the outcomes. 77% of trial (345) seven subjects were lost to follow-up in the these maintained moderate or marked improvement PTNS and 5 in the sham group thus 94% and 95% in OAB symptoms at three years with a median 1.1 respectively were analysed. Two of the 30 subjects treatments each month. (6.6%), both from the sham group, discontinued in one trial (338) and one subject (4.5%) from the control Summary group discontinued in the other trial (339). Mild or moderate treatment related adverse events were re- The results of two trials, one rigorous, well-reported ported by 6 PTNS subjects in one trial (337). They trial and one low quality pilot trial, showed that percu- included ankle bruising (1 of 110, 0.9%) discomfort at taneous PTNS is a safe and more effective interven- needle site (2 of 110, 1.8%) bleeding at the needle tion than a sham treatment for improving urgency in- site (3 of 110, 2.7%) and tingling in the leg (1 of 110, continence in adults with OAB/UUI. 0.9%). No subjects in the transcutaneous PTNS trial There is evidence that, with regular treatment, effects (338) reported adverse events and presence of ad- are sustained for up to 3 years. (Level of Evidence: verse events was unclear in one study(339). Two tri- 1). als were reported as pilots with no sample size calculation (338, 339) and one was adequately powered One small high quality pilot trial indicates that (337) however the payment of subjects for time and transcutaneous PTNS is safe and may be more effec- expense was a potential limitation and the efficacy tive than sham stimulation for reducing UI and urinary achieved may not be equivalently reflected in trans- symptoms in older adults in institutional care. Further lating to real world practice. Long term follow-up to 3 studies are needed to determine the effectiveness of years post-initial treatment was reported by Peters transcutaneous PTNS. (Level of Evidence: 2). (345). Overall risk of bias was low in two trials (337, Recommendation 338) and high in one (339). Percutaneous PTNS can be offered to men and Results women (adults) with UUI/OAB who do not achieve No study reported cure rates for UI, however all three satisfactory results from first line lifestyle and behav- trials reported improvements. In the SUmiT trial(337) ioural intervention and pharmacological therapy. 37.9% PTNS subjects reported moderate or markedly (Grade of Recommendation: B New). improved urgency incontinence compared to 22.1% In this population, Transcutaneous PTNS may be a sham subjects (p=0.02). Voiding diary analysis useful option to test in the adult with UUI/OAB earlier showed PTNS to be statistically superior to sham in in the algorithm, following lifestyle and behavioural in- reducing urge incontinence episodes (p=0.002) from terventions and before more invasive therapies are a median of 3.0 episodes accompanied by moderate considered. to severe urgency per day at baseline, to a median of 0.3 episodes at 13 weeks (p<0.0001). In the two pilot trials Booth (338) reported improved ICIQ-UI SF in 10 1.2.2 Is PTNS Better Than Other Treatments of 15 (67%) transcutaneous PTNS group and 6 of 13 for UI? (46%) sham group (p=0.132, NS); Vohra reported One RCT compared PTNS with another treatment for significantly reduced UI in 7 of 11 PTNS subjects but UI. The other treatment was extended release tolter- no estimates of effect size were provided; data was odine 4mg daily (341). There are no trials comparing not differentiated for men and women. In one large transcutaneous PTNS with other treatments. trial (337) PTNS subjects reported statistically significant improvements in overall bladder symptoms with Quality of evidence 54.5% reporting moderate or markedly improved Global Response Assessment (GRA) from baseline 1:1 randomisation to percutaneous PTNS or tolterodine was implemented using a random block design compared to 20.9% of sham subjects (p<0.001). A significant difference between the groups in favour of stratified by investigational site however the success of allocation concealment was unclear. Blinding of PTNS was found for the OAB-q quality of life scores subjects, clinicians or assessors was not possible (p=0.006) and the SF-36 general health survey quality of life scores significantly improved between base- given the different nature of the interventions. An intention to treat analysis was not undertaken. Seven line and 13 weeks for the PTNS group in the physical (p=0.002) and mental (p=0.049) domain scales. One subjects (14%) withdrew from the drug group and 9 (18%) from the percutaneous PTNS group; none due

164 to adverse effects. Adverse effects were mild or mod- Recommendation erate in both groups with 14.3% (7 subjects) from the tolterodine arm reporting moderate adverse effects Percutaneous PTNS can be offered as an alternative and 16.3% (8 subjects) from the percutaneous PTNS to tolterodine for OAB/UUI in adult men and women. group. Percutaneous PTNS related adverse events (Grade of Recommendation: B New). included leg cramps, intermittent foot/toe pain, gener- alised swelling, headache, haematuria, inability to tol- 1.2.3 Does the Addition of PTNS to Other UI erate stimulation, worsening incontinence and vaso- Treatment Add Any Benefit? vagal response to needle placement. A sample size One RCT was identified (340) that compared percu- calculation was provided and the study was ade- taneous PTNS with percutaneous PTNS and oxy- quately powered for a non-inferiority margin of 20% butynin for treatment of patients with overactive blad- in number of voids per 24 hours. Long-term follow-up der (with or without incontinence) and urodynamically of percutaneous PTNS subjects for nine months after diagnosed detrusor overactivity (DO). initial treatment completion was reported (346). The overall risk of bias was high. Quality of evidence Results Randomisation following urodynamic studies into percutaneous PTNS group or percutaneous PTNS plus Data was reported for the whole group with no differ- 5mg daily oxybutynin hydrochloride was reported but entiation by sex. The global response assessment no description of method of randomisation. Adequacy (GRA) demonstrated that subjective assessment of of allocation concealment and blinding of subjects, bladder symptom change compared to baseline was clinicians or assessors was not reported nor was type statistically significant with 79.5% (35) of the percuta- of analysis, which was unclear. It appears that no neous PTNS group reporting cure (1) or improvement subject dropped out and reported adverse events (34) and 54.8% (23) of the tolterodine group reporting were mild: percutaneous PTNS plus drug group - cure (2) or improvement (21) (p=0.01). Both groups seven reported dry mouth, one blurred vision; percu- had improved significantly. Symptoms of UUI re- taneous PTNS group – one reported a small haema- ported in the voiding diaries improved significantly in toma, one local tenderness. There was no sample both groups however there was no significant differ- size calculation, no long-term follow-up and overall ence between the groups for these measures. Quality risk of bias was high. of life scores showed statistically significant improvements for both treatment groups (P<0.001) but be- Results tween group differences were not statistically significant. The percutaneous PTNS group reported statis- Four of five subjects receiving percutaneous PTNS tically significantly less dry mouth than the tolterodine only and all five subjects receiving percutaneous group (p=0.01) and a non-significant lower rate of PTNS and oxybutynin, with urgency UI reported cure constipation. on voiding diary. The numbers were too small for statistical analysis. Overall treatment response was de- Follow-up to determine duration of effect up to 12 fined as patient-reported improvement in OAB symp- months from baseline was offered to percutaneous toms of frequency, urgency and urge incontinence by PTNS responders (those who reported a successful > 35% and occurred in 61.6% percutaneous PTNS response to GRA after 12 weeks) (346). Thirty-three group and 83.2% percutaneous PTNS plus oxy- of the 35 responders chose to continue. The GRA butynin group. The between group difference was not showed sustained improvements in 96% at 12 statistically significant. months with a mean of 21 days between treatment sessions. Summary

Summary The evidence is limited to a single low quality trial (Level of Evidence: 2) which indicates there may be Evidence from a single RCT indicates that percutane- additional effects if oxybutynin is added to a pro- ous PTNS may be as effective as tolterodine for ur- gramme of percutaneous PTNS. Further trials are gency UI with an improved side effect profile; how- needed to establish whether the addition of an anti- ever, design limitations suggest caution and further cholinergic drug enhances the effectiveness of percu- studies are recommended to establish the effects of taneous PTNS in adults with UUI/OAB and which percutaneous PTNS and transcutaneous PTNS in drugs provide the greatest effect. comparison to other pharmacological treatments (Level of Evidence: 2). As there were no trials comparing transcutaneous PTNS added to another active treatment with the ac- As there were no trials comparing transcutaneous tive treatment alone studies are also needed to deter- PTNS with another active treatment trials are also mine the effects of adding transcutaneous PTNS to needed to determine the effects of transcutaneous first line lifestyle and behavioural treatments for OAB PTNS compared to common anticholinergic drugs and UUI including bladder training and percutaneous used to treat UUI/OAB in adults. PFMT, with these first line lifestyle and behavioural treatments alone.

165 Recommendation Recommendation Percutaneous or transcutaneous PTNS should be de- Oxybutynin may be considered in addition to percuta- livered at least once weekly and the protocol deter- neous PTNS in adults with DO. (Grade of Recom- mined by patient preference. (Grade of Recommen- mendation: B New). dation: B New). This hypothesis needs to be investigated further with high quality trials 1.3. What is the Effect of PTNS on LUTS Other Than UI? 1.2.4 What is the Best PTNS Protocol for UI in Adults? No trials were identified that analysed the effect of PTNS in adults or men with other LUTS alone i.e. fre- Two RCTs compared different stimulation protocols quency of voiding, urgency, nocturia and integrated (342, 343). Details are provided in table 38. reporting of UI and other LUTS was a feature of all studies. For percutaneous PTNS compared with Quality of the evidence sham percutaneous PTNS, in one large well reported Randomisation methods were not described for either trial, the percutaneous PTNS group reported statisti- study. Allocation concealment could not be deter- cally significant improvements in voiding diary symp- mined from the reports and blinding of subjects or cli- toms of frequency, night-time voids and voids with nicians was not possible with this design. Neither moderate to severe urgency, compared to the sham study reported the type of analysis undertaken. Pri- group (337). A small pilot trial reported reduced day mary outcomes were reported at the end of the (342) and night time frequency and urgency after 12 weekly treatment protocol. No long-term follow up was re- 30 minute percutaneous PTNS sessions by 63% of ported. Withdrawals were high (29.2 %) in the Seth percutaneous PTNS subjects, although no EStimates (343) pilot trial, with 8 of the 14 withdrawals being de- of effect size were provided. The elimination of detru- vice related, although no significant adverse events sor overactivity on repeat urodynamic testing was occurred. There were no withdrawals in the Finazzi- also shown (339). Agro trial and no adverse events reported. For transcutaneous PTNS compared with sham Results transcutaneous PTNS one trial in care home residents (338) reported statistically significantly im- In the Finazzi-Agro (342) trial 4 of 11 (36%) subjects proved total American Urological Association Symp- with UI in the weekly percutaneous PTNS group and tom Index AUASI urinary symptom scores for 87% of 5 of 11 (45%) subjects with UI in the 3 X weekly per- the transcutaneous PTNS group compared to 31% cutaneous PTNS group reported complete cure after sham group (p<0.001). treatment. Overall success of >50% reduction in micturition episodes/24 hours or (if incontinent) UI epi- When comparing percutaneous PTNS with another sodes/24 hours was confirmed for 11 of 17 (63%) treatment one trial (341) reported significant reduc- subjects in the weekly percutaneous PTNS group and tion in bladder diary reports of void frequency, noctu- 12 of 18 (67%) subjects in the 3 times weekly percu- ria, moderate to severe urgency episodes in both taneous PTNS group. Subjective improvement was groups and no between group differences. reported after 6-8 sessions, regardless of frequency One trial comparing two percutaneous PTNS proto- of delivery. In the Seth (343) trial 18 of 34 (54%) sub- cols (342) reported a statistically significant reduction jects who completed the 12week protocol were rein frequency (p=0.01) for both once and three times sponders, who rated their improvement as moderate weekly treatment regimes. to significant on the General Response Assessment (GRA). Statistically significant improvements in ICIQ- OAB (p=0.001) and ICIQ LUTSqol (p=0.000) were reported for both daily and weekly stimulation groups. There were no statistically significant differences in ICIQ or bladder diary parameters between those with idiopathic and OAB of neurogenic origin.

Summary Two small trials indicate that no additional benefit is conferred by a more than once weekly stimulation protocol for percutaneous or transcutaneous PTNS. However, it is possible that symptom improvement may be more rapid with a more frequent delivery protocol (Level of Evidence: 2). Further rigorous and well-reported trials are needed to establish the most effective timing and duration of PTNS protocols.

166 Table 35 Summary of data on PTNS vs no treatment in man and women

Study Comparator groups N Study population Modality details or parameters Outcomes Follow up Percutaneous posterior tibial nerve stimulation Peters Treatment group: Percutaneous 220 Adults (>18) with OAB Pulse width: not reported Primary outcome: Week 13 (2010) (1) PTNS symptoms Frequency: 20 Hz Moderate/marked improvement Active electrode: Males: 46 Intensity: in 7 level GRA at week 13 34 gauge needle provides Females: 174 0.5-9mA according to sensory & Secondary outcomes: sensation. Results reported together motor response. Change in individual GRA Inactive electrode: calcaneal Duration: 30 minutes symptoms surface electrode. Number sessions: 12 3-day voiding diary parameters Control group: Sham Programme length: 12 weeks OABq scores Percutaneous PTNS SF-36 QoL scores Placebo needle with sensation of insertion. Inactive Vohra (2002) Treatment group: Percutaneous 22 Adults with urgency Pulse width: not reported Micturition diary 12 weeks (2) PTNS frequency syndrome of > 6 Frequency: 20 Hz QOL questionnaires Active electrode: months and Intensity: Repeat urodynamics 34 gauge needle provides urodynamic detrusor 0.5-10mA according to sensory & SF-36 sensation. overactivity motor response. Inactive electrode: calcaneal Males: not reported Duration: 30 minutes . surface electrode. Females: not reported Number sessions: 12 Control group: Sham Results reported together Programme length: 12 weeks Percutaneous PTNS Described as ‘PTNS treatment without nerve stimulation’ Transcutaneous posterior tibial nerve stimulation Booth 2013 (3) Treatment group: Transcutaneous 30 Older adults (>65) resident Pulse width: 200 μS AUASI Week 6 PTNS in care homes with urinary Frequency: 10 Hz ICIQ-UI SF Active: Surface electrode medial symptoms and/or Intensity: 1-50mA according to PVR malleolus provides sensation. incontinence sensory & motor threshold and Inactive: surface electrode 10cm Males: 6 subject comfort. proximal to medial malleolus, Females: 24 Duration: 30 minutes Control group: Sham Results reported together Number sessions: 12 Transcutaneous PTNS Programme length: 6 weeks Surface electrodes X2 positioned below lateral malleolus and 10cm

167 proximal to avoid posterior tibial nerve current applied 2 mA Footnotes; GRA : Global Response Assessment, UUI : urinary urge incontinence, OABq : Overactive Bladder Questionnaire, SF-36 : 36-Item Short Form Health Survey; AUASI : American Urological Association Symptom Index; ICIQ-UI SF : International Consultation on Incontinence Questionnaire on Urinary Incontinence Short Form; PVR : post-void residual urine volume; PTNS: Posterior tibial nerve stimulation 1. Peters KM, Carrico DJ, Perez-Marrero R, Khan AU, Wooldridge LS, Davis GL, et al. Randomized trial of percutaneous tibial nerve stimulation versus Sham efficacy in the treatment of overactive bladder syndrome: results from the SUmiT trial. The Journal of urology. 2010;183(4):1438. 2. Vohra AK, Britchford A, Neale E, Husain I, Waterfall N, editors. The efficacy of stoller afferent nerve stimulation in frequency/urgency syndrome: A randomised control trial. Proceedings of the International Continence Society (ICS) 32nd Annual Meeting, Aug 28-30; 2002. 3. Booth J, Hagen S, McClurg D, Norton C, MacInnes C, Collins B, et al. A feasibility study of transcutaneous posterior tibial nerve stimulation for bladder and bowel dysfunction in elderly adults in residential care. Journal Of The American Medical Directors Association. 2013;14(4):270.

Table 36 Summary of data on PTNS vs other active treatments in men and women

Author, year Comparator groups N Study population Modality details or Outcomes/results Follow up parameters

Peters (2009) (1) Treatment group: Percutaneous 100 Adults with urinary Pulse width: not reported Primary outcome: mean reduction in 12 weeks PTNS frequency of at least 8 Frequency: 20 Hz number of urinary voids in 24 hours Active electrode: 34 gauge needle voids/24 hours Intensity: Secondary outcomes: change in 24 provides sensation. Males: 6 0.5-9mA according to sensory hour UUI episodes, number of voids Inactive electrode: calcaneal Females: 94 & motor response. causing waking, daily voided volume, surface electrode. Results reported Duration: 30 minutes episodes of urgency. Comparator treatment group: together. Number sessions: 12 OABq Investigator & subject OAB Daily tolterodine ER 4mg Programme length: 12 weeks ratings using GRA. Footnotes; GRA : Global Response Assessment, UUI : urinary urge incontinence, OABq : Overactive Bladder Questionnaire ER : extended release PTNS : Posterior tibial nerve stimulation 1. Peters KM, MacDiarmid SA, Wooldridge LS, Leong FC, Shobeiri SA, Rovner ES, et al. Randomised trial of percutaneous tibial nerve stimulation versus extended- release tolteridine: Results from the overactive bladder innovative thereapy trial. Journla of Urology. 2009;182(3):1055-61.

168 Table 37 Summary of data on PTNS + another active vs PTNS

Author, year Comparator groups N Study population Modality details or parameters Outcomes/results Follow up Karademir Treatment group: 43 Adults with > 6 month Pulse width: 200 μS Outcomes measured with Bristol 8 weeks 2005 (1) Percutaneous PTNS plus oral history of OAB symptoms Frequency: 20 Hz Urinary Questionnaire and voiding oxybutynin hydrochloride 5mg daily and DO on UDS Intensity: diary. Active electrode: 0.5-9mA according to sensory & 34 gauge needle provides Males: 5 motor response. Improvements in symptoms by sensation. Females: 38 Duration: 60 minutes >70%, 35-70% and <35% Inactive electrode: calcaneal Results reported together Number sessions: 8 represented complete remission, surface electrode. Programme length: 8 weeks partial remission and no response. Comparator treatment group: Percutaneous PTNS Active electrode: 34 gauge needle provides sensation. Inactive electrode: calcaneal surface electrode. Posterior tibial nerve stimulation: PTNS 1. Karademir K, Baykal K, Sen B, Senkul T, Iseri C, Erden D. A peripheric neuromodulation technique for curing detrusor overactivity: Stoller afferent neurostimulation. Scandinavian journal of urology and nephrology. 2005;39(3):230.

169 1.4. Factors Affecting Outcomes Only two small trials investigated transcutaneous None of the included percutaneous PTNS trials ad- PTNS but the promising results indicate that further dressed the effect of age or any other factor on pre- well-designed and reported trials would allow deci- diction of outcome of PTNS. Effectiveness of sions to be made about the place of transcutaneous transcutaneous PTNS in older adults resident in care PTNS in the treatment algorithms for OAB/UUI in men homes was the focus of one study where the mean and women. age was 84.2 years (338). The adherence to the Health economic information is required to establish transcutaneous PTNS was 100% for all participants, the cost effectiveness of the different forms of PTNS, with no adverse effects reported and a response to particularly in comparison to pharmacotherapy. therapy similar to that found with younger groups. A Recommendations for practice: prospective study (347) of prognostic factors for successful percutaneous PTNS showed that gender, In adults with OAB/UUI percutaneous PTNS is better age, weight, body mass index, indication for percuta- for improving UUI than no treatment or sham and neous PTNS, duration of symptoms, number and should be offered to adults with UUI/OAB who do not type of previous treatments, number of UI epi- achieve satisfactory results from first-line lifestyle and sodes/24 hours, voiding frequency/24 hours and total behavioral interventions or drug therapy. (Grade of IQoL scores were all unrelated to the success or not Recommendation: B New) of percutaneous PTNS in men and women with OAB, non-obstructive urinary retention or chronic pelvic At least weekly PTNS sessions should be offered dur- pain. A low Mental Component Summary Score on ing an active treatment program with regular top-ups the SF-36 was a negative predictive factor for suc- provided to sustain benefits for up to three years. cess of percutaneous PTNS, both subjectively and (Grade of Recommendation: B New) objectively. Additionally patients with detrusor over- Transcutaneous PTNS is a safe treatment option and activity had poorer outcomes than those without, as may be offered to frail older adults with UI or urinary did those with low bladder capacity at baseline (143). symptoms however definitive evidence of effective- This means that there is no reason to exclude older ness is needed. (Grade of Recommendation: C adults, those with a long symptom history, weight dif- New) ficulties, severe symptoms or failure of previous treatments and they should be offered PTNS where indi- Percutaneous PTNS can be offered as an alternative cated, except where recognised contraindications to to tolterodine for OAB/UUI in adult men and women. PTNS, such as a cardiac pacemaker are present. (Grade of Recommendation: B New). Only in those patients with poor mental health and/or DO and/or low capacity bladders at baseline should Oxybutynin may be considered in addition to percuta- the possibility of limited success be considered. neous PTNS in adults with DO. (Grade of Recom- mendation: B New). New factors are emerging which may influence understanding of potential effects for different diagnostic Future research directions: groups. In one pilot study (343) urinary neurotrophin Currently available evidence compares percutaneous levels (nerve growth factor and brain derived neu- PTNS with older antimuscarinics. Future rigorous tri- rotrophic factor) were measured. Results indicated als should compare percutaneous PTNS with other that for idiopathic OAB higher levels of nerve growth commonly used antimuscarinics and beta 3 adrener- factor at baseline may predict poor response to gic agonists for efficacy and adverse effect profiles, transcutaneous PTNS. Level of brain derived neu- in men and women with OAB/UUI. rotrophic factor significantly reduced over the treatment course in those who responded to transcutane- The effectiveness of adding drug therapy to percuta- ous PTNS, which suggests a potential biomarker for neous PTNS should be investigated in high quality tri- response in idiopathic OAB but the reduction was not als with adults with OAB/UUI. seen in responders with neurological disease. Further Definitive evidence of the effectiveness of transcuta- investigation is required to fully understand the influ- neous PTNS to treat OAB/UUI in adults is needed ence of these patient factors. and its place in the treatment algorithm defined. Summary Research comparing transcutaneous PTNS with all types of drug therapy is required. The evidence on which to base recommendations for best practice in the use of PTNS to treat OAB/UUI in Direct comparison of percutaneous PTNS and men and women is sparse, for both percutaneous transcutaneous PTNS to treat OAB/UUI in adults and transcutaneous PTNS. However it is sufficiently should be investigated. robust to support the use of percutaneous PTNS when less intensive and invasive behavioral treat- Further rigorous and well-reported trials are needed ment options have failed (Level of Evidence: 1) and to establish the most effective dose of percutaneous there is the suggestion that percutaneous PTNS may PTNS and transcutaneous PTNS, including the tim- be as effective as some drug therapy, making it a vi- ing and duration of PTNS protocols. able alternative (Level of Evidence: 2).

170 Table 38 Summary of data PTNS protocols comparisons

Author, year Comparator N Study population Modality details or parameters Outcomes Follow up

Percutaneous posterior tibial nerve stimulation

Finazzi-Agro Treatment group A: weekly 35 Adults with refractory OAB Pulse width: not reported 24 hour BD 4 weeks (3X 2005 (1) Percutaneous PTNS syndrome Frequency: 20 Hz IQoL weekly PTNS) Active electrode: Males : 7 Intensity: SF36 34 gauge needle provides sensation. Females: 28 0.5-9mA according to sensory & UDS 12 weeks Inactive electrode: calcaneal surface Results reported together motor response. Success defined as those (1 X weekly electrode. Duration: 30 minutes who reported PTNS) Comparator treatment group B: 3 x Number sessions: 12 micturition weekly Percutaneous PTNS Programme length: 12 weeks episodes/24hours or Active electrode: (Group A); incontinence episodes/24 34 gauge needle provides sensation. 4 weeks (Group B) hours reduced by >50% Inactive electrode: calcaneal surface electrode. Transcutaneous posterior tibial nerve stimulation

Seth Treatment group 48 Adults with idiopathic (24 Pulse width: not reported Outcomes: Responder 12 weeks 2014 (2) Transcutaneous PTNS using GekoTM subjects) or neuropathic Frequency: not reported defined as those rating device weekly (24 subjects) OAB Intensity: not reported moderate to significant Comparator treatment group Males NR Duration: 30 minutes improvement on GRA at Transcutaneous PTNS using GekoTM Females NR Number sessions: 12 (weekly 12 weeks device daily Results reported together group), 84 (daily group) Change in individual Programme length: 12 weeks GRA symptoms ICIQ-OAB, ICIQ-LUTSqol BD parameters Footnotes; GRA : Global Response Assessment, UDS : Urodynamic studies; SF-36 :36-Item Short Form Health Survey; ICIQ-OAB : International Consultation on Incon- tinence Questionnaire on Overactive Bladder; ICIQ-LUTSqol : International Consultation on Incontinence Quality of Life questionnaire; BD parameters : bladder diary parameters; PTNS: Posterior tibial nerve stimulation 1. Finazzi Agrò E, Campagna A, Sciobica F, Petta F, Germani S, Zuccalà A, et al. Posterior tibial nerve stimulation: is the once-a-week protocol the best option? Minerva urologica e nefrologica = The Italian journal of urology and nephrology. 2005;57(2):119. 2. Seth J, Gonzales G, Haslam C, Ochulor J, Elneil S, Vashisht A, editors. Single centre randomised pilot study of two regimens (30mins daily or 30 mins weekly for 12 weeks) of Transcutaneous Tibial Nerve Stimulation using an adhesive skin patch for the treatment of Overactive Bladder (OAB) Symptoms Proceedings of the 44th Annual Meeting of the International Continence Society (ICS), 2014 Oct 20-24; 2014.

171 11. McGrother CW, Donaldson MMK, Thompson J, REFERENCES Wagg A, Tincello DG, Manktelow BN. Etiology of overactive bladder: A diet and lifestyle model 1. Bo K, Frawley H, Haylen B, Abramov Y, for diabetes and obesity in older women. Almeida F, Berghmans B, et al. An international Neurourology and urodynamics. urogynecological association (IUGA) / 2012;31(4):487-95. international continence society (ICS) joint 12. Jorgensen S, Hein HO, Gyntelberg F. Heavy report on the terminology for the conservative lifting at work and risk of genital prolapse and and non-pharmacological management of herniated lumbar disc in assistant nurses. female pelvic floor dysfunction (Accepted). Occupational medicine. 1994;44(1):47-9. N&U. 2016. 13. Nygaard IE, Shaw JM, Bardsley T, Egger MJ. 2. Moore K, Dumoulin C, Bradley C, Burgio K, Lifetime physical activity and female stress Chambers T, Hagen S, et al. Adult urinary incontinence. American Journal of Conservative Management. In: Abrams P, Obstetrics and Gynecology. 2015;213(1). Cardozo L, Khoury S, Wein AJ, editors. 5th International Consultation on Incontinence. 5th 14. Wyman J, Allen A, Hertsgaard L, Overson E, ed. Paris: ICUD-EAU; 2013. p. 1101-227. Allen S, Hatsukami D. Effect of Smoking Cessation on Overactive Bladder Symptoms in 3. Group CI. Cochrane Incontinence Group [cited Adults: A Pilot Study. Neurourology and 2016 24 August 2016]. Available from: urodynamics. 2014;33(6):866-7. http://incontinence.cochrane.org/our-reviews 15. Wells MJ, Jamieson K, Markham TC, Green 4. Grant AM, Cody DJ, Glazener CMA, Hay‐Smith SM, Fader MJ. The effect of caffeinated versus J, Herbison P, Lapitan MC, et al. About The decaffeinated drinks on overactive bladder: a Cochrane Collaboration (Cochrane Review doubleblind, randomized, crossover study. Groups (CRGs)2012. Journal of wound, ostomy, and continence 5. Imamura M, Williams K, Wells M, McGrother C. nursing : official publication of The Wound, Lifestyle interventions for the treatment of Ostomy and Continence Nurses Society. urinary incontinence in adults. Cochrane 2014;41(4):371-8. Database Syst Rev. 2015;12:CD003505. 16. Davis NJ, Vaughan CP, Johnson TM, Goode 6. Subak LL, Wing R, West DS, Franklin F, PS, Burgio KL, Redden DT, et al. Caffeine Vittinghoff E, Creasman JM, et al. Weight loss Intake and its Association with Urinary to treat urinary incontinence in overweight and Incontinence in United States Men: Results obese women. N Engl J Med. 2009;360(5):481- from National and Nutrition Examination 90. Surveys 2005-2006 and 2007-2008. Journal of Urology. 2013;189(6):2170-4. 7. Phelan S, Kanaya AM, Subak LL, Hogan PE, Espeland MA, Wing RR, et al. Weight Loss 17. Gleason JL, Richter HE, Redden DT, Goode Prevents Urinary Incontinence in Women With PS, Burgio KL, Markland AD. Caffeine and Type 2 Diabetes: Results From the Look urinary incontinence in US women. AHEAD Trial. Journal of Urology. International Urogynecology Journal. 2012;187(3):939-44. 2013;24(2):295-302. 8. Auwad W, Steggles P, Bombieri L, Waterfield 18. Hirayama F, Lee AH. Is caffeine intake M, Wilkin T, Freeman R. Moderate weight loss associated with urinary incontinence in in obese women with urinary incontinence: a Japanese adults? J Prev Med Public Health. prospective longitudinal study. Int Urogynecol J 2012;45(3):204-8. Pelvic Floor Dysfunct. 2008;19(9):1251-9. 19. Kegel AH. Progressive resistance exercise in 9. Pinto AM, Subak LL, Nakagawa S, Vittinghoff the functional restoration of the perineal E, Wing RR, Kusek JW, et al. The effect of muscles. Am J Obstet Gynecol. weight loss on changes in health-related quality 1948;56(2):238-48. of life among overweight and obese women 20. Delancey JO. Structural aspects of with urinary incontinence. Quality of Life Re- urethrovesical function in the female. search. 2012;21(10):1685-94. Neurourology and urodynamics. 10. Vissers D, Neels H, Vermandel A, De Wachter 1988;7(6):509-19. S, Tjalma WAA, Wyndaele JJ, et al. The effect 21. Miller J, Ashton-Miller J, DeLancey J. The of non surgical weight loss interventions on uri- Knack: use of precisely timed pelvic muscle nary incontinence in overweight women: a contraction can reduce leakage in SUI. systematic review and meta analysis. Obesity Neurourology and urodynamics. Reviews. 2014;15(7):610-7. 1996;15(4):392-3.

172 22. Godec C, Cass AS, Ayala GF. Bladder 32. Fritel X, de TR, Bader G, Savary D, Gueye A, inhibition with functional electrical stimulation. Deffieux X, et al. Preventing Urinary Urology. 1975;6(6):663-6. Incontinence With Supervised Prenatal Pelvic Floor Exercises: A Randomized Controlled 23. Burgio KL, Whitehead WE, Engel BT. Urinary Trial. Obstetrics & Gynecology. incontinence in the elderly. Bladder-sphincter 2015;126(2):370-7. biofeedback and toileting skills training. Ann Intern Med. 1985;103(4):507-15. 33. Hilde G, Staer-Jensen J, Siafarikas F, Ellstrom EM, Bo K. Postpartum pelvic floor muscle 24. McClurg D, Frawley H, Hay-Smith J, Dean S, training and urinary incontinence: a Chen SY, Chiarelli P, et al. Scoping review of randomized controlled trial [Erratum appears adherence promotion theories in pelvic floor in: Obstet Gynecol. 2014 Sep;124(3):639]. muscle training - 2011 ICS state-of-the-science Obstetrics & Gynecology. 2013;122(6):1231-8. seminar research paper i of iv. Neurourology and urodynamics. 2015;34(7):606-14. 34. Kim EY, Kim SY, Oh DW. Pelvic floor muscle exercises utilizing trunk stabilization for treating 25. Dumoulin C, Alewijnse D, Bo K, Hagen S, Stark postpartum urinary incontinence: randomized D, Van Kampen M, et al. Pelvic-Floor-Muscle controlled pilot trial of supervised versus Training Adherence: Tools, Measurements and unsupervised training. Clinical Rehabilitation. Strategies-2011 ICS State-of-the-Science 2012;26(2):132-41. Seminar Research Paper II of IV. Neurourology and urodynamics. 2015;34(7):615-21. 35. Ko PC, Liang CC, Chang SD, Lee JT, Chao AS, Cheng PJ. A randomized controlled trial of 26. Hay-Smith J, Dean S, Burgio K, McClurg D, antenatal pelvic floor exercises to prevent and Frawley H, Dumoulin C. Pelvic-floor-muscle- treat urinary incontinence. Int Urogynecol J. training adherence "modifiers": A review of 2011;22(1):17-22. primary qualitative studies-2011 ICS State-of- the-Science Seminar research paper III of IV. 36. Kocaoz S, Eroglu K, Sivaslioglu AA. Role of Neurourology and urodynamics. pelvic floor muscle exercises in the prevention 2015;34(7):622-31. of stress urinary incontinence during pregnancy and the postpartum period. Gynecologic & Ob- 27. Frawley HC, McClurg D, Mahfooza A, Hay- stetric Investigation. 2013;75(1):34-40. Smith J, Dumoulin C. Health professionals' and patients' perspectives on pelvic floor muscle 37. Miquelutti MA, Cecatti JG, Makuch MY. training adherence-2011 ICS State-of-the- Evaluation of a birth preparation program on Science Seminar research paper IV of IV. lumbopelvic pain, urinary incontinence, anxiety Neurourology and urodynamics. and exercise: A randomized controlled trial. 2015;34(7):632-9. BMC Pregnancy and Childbirth. 2013;13(154). 28. Dumoulin C, Hay-Smith J, Frawley H, McClurg 38. Pelaez M, Gonzalez-Cerron S, Montejo R, D, Alewijnse D, Bo K, et al. 2014 consensus Barakat R. Pelvic floor muscle training included statement on improving pelvic floor muscle in a pregnancy exercise program is effective in training adherence: International Continence primary prevention of urinary incontinence: a Society 2011 State-of-the-Science Seminar. randomized controlled trial. Neurourology & Neurourology and urodynamics. Urodynamics. 2014;33(1):67-71. 2015;34(7):600-5. 39. Stafne S, Salvesen K, Romundstad P, Torjusen 29. Ahlund S, Nordgren B, Wilander EL, Wiklund I, I, Morkved S. Does regular exercise including Friden C. Is home-based pelvic floor muscle pelvic floor muscle training prevent urinary and training effective in treatment of urinary anal incontinence during pregnancy? A ran- incontinence after birth in primiparous women? domised controlled trial. BJOG: an In- A randomized controlled trial. Acta Obstetricia ternational Journal of Obstetrics & Gy- et Gynecologica Scandinavica. naecology. 2012;119(10):1270-80. 2013;92(8):909-15. 40. Boyle R, Hay-Smith EJ, Cody JD, Morkved S. 30. Barakat R, Pelaez M, Montejo R, Luaces M, Pelvic floor muscle training for prevention and Zakynthinaki M. Exercise during pregnancy treatment of urinary and faecal incontinence in improves maternal health perception: A antenatal and postnatal women. Cochrane randomized controlled trial. American journal of Database Syst Rev. 2012;10:CD007471. obstetrics and gynecology. 2011;204(5):402. 41. Diokno AC, Sampselle CM, Herzog AR, 31. Bo K, Haakstad LA. Is pelvic floor muscle Raghunathan TE, Hines S, Messer K, et al. training effective when taught in a general Prevention of urinary incontinence by fitness class in pregnancy? A randomised behavioral modification program: a controlled trial. Physiotherapy. 2011;97(3):190- randomized, controlled trial among older 5. women in the community. The Journal of urology. 2004;171(3):1165-71.

173 42. Berzuk K, Shay B. Effect of increasing study. Medicine 20 Conference; 2014: JMIR awareness of pelvic floor muscle function on Publications Inc., Toronto, Canada. pelvic floor dysfunction: a randomized controlled trial. International urogynecology 53. Abdulaziz K, Hasan T. Role of pelvic floor journal. 2015;26(6):837-44. muscle therapy in obese perimenopausal females with stress incontinence: A 43. Lucio AC, Perissinoto MC, Natalin RA, randomized control trial. The Internet Journal of Prudente A, Damasceno BP, D'ancona CAL. A Gynecology and Obstetrics. 2012;16(2):34-42. comparative study of pelvic floor muscle training in women with multiple sclerosis: its 54. Tosun OC, Mutlu EK, Ergenoglu A, Yeniel A, impact on lower urinary tract symptoms and Tosun G, Malkoc M, et al. Does pelvic floor quality of life. Clinics. 2011;66(9):1563-8. muscle training abolish symptoms of urinary incontinence? A randomized controlled trial. 44. Tak EC, van Hespen A, van Dommelen P, Clinical rehabilitation. 2015;29(6):525-37. Hopman-Rock M. Does improved functional performance help to reduce urinary 55. Dumoulin C, Sran M, Lieblich P, Wilson P. incontinence in institutionalized older women? Physiotherapy significantly reduces leakage in A multicenter randomized clinical trial. BMC postmenopausal women with osteoporosis and geriatrics. 2012;12(1):1. urinary incontinence: result of a parallel randomised controlled trial. Neurology & Uro- 45. Voorham J, De Wachter S, Van Den Bos T, dynamics. 2011;30(6):985. Putter H, Lycklama à Nijeholt G, Voorham-Van Der Zalm P. The effect of EMG biofeedback 56. Ferreira S, Ferreira M, Carvalhais A, Santos assisted pelvic floor muscle therapy on PC, Rocha P, Brochado G. Reeducation of symptoms of the overactive bladder syndrome pelvic floor muscles in volleyball athletes. in women: A randomised controlled trial 2015. Revista da Associação Médica Brasileira. 2014;60(5):428-33. 46. Yang EJ, Lim J-Y, Rah UW, Kim YB. Effect of a pelvic floor muscle training program on 57. McLean L, Varette K, Gentilcore‐Saulnier E, gynecologic cancer survivors with pelvic floor Harvey MA, Baker K, Sauerbrei E. Pelvic floor dysfunction: a randomized controlled trial. muscle training in women with stress urinary Gynecologic oncology. 2012;125(3):705-11. incontinence causes hypertrophy of the urethral sphincters and reduces bladder neck 47. Alves FK, Riccetto C, Adami DB, Marques J, mobility during coughing. Neurourology and Pereira LC, Palma P, et al. A pelvic floor muscle urodynamics. 2013;32(8):1096-102. training program in postmenopausal women: A randomized controlled trial. Maturitas. 58. Kim H, Yoshida H, Suzuki T. The effects of 2015;81(2):300-5. multidimensional exercise treatment on community dwelling elderly Japanese women 48. Singh A, Kumari S, Jain V. Why behavior with stress, urge, and mixed urinary therapy for urinary incontinence has been incontinence: a randomized controlled trial. ignored by doctors/women? . Climacteric2011. International journal of nursing studies. 2011;48(10):1165-72. 49. Kang H, Hong G-RS. Effect of Muscle Strength Training on Urinary Incontinence and Physical 59. Leong B, Mok N. Effectiveness of a new Function: A Randomized Controlled Trial in standardised Urinary Continence Long-term Care Facilities. Journal of Korean Physiotherapy Programme for community Academy of Nursing. 2015;45(1). dwelling older women in Hong Kong. Hong Kong Medical Journal 2015;21(1):30-7. 50. Pereira VS, Correia GN, Driusso P. Individual and group pelvic floor muscle training versus 60. Rutledge TL, Rogers R, Lee S-J, Muller CY. A no treatment in female stress urinary pilot randomized control trial to evaluate pelvic incontinence: a randomized controlled pilot floor muscle training for urinary incontinence study. European Journal of Obstetrics & among gynecologic cancer survivors. Gynecology and Reproductive Biology. Gynecologic oncology. 2014;132(1):154-8. 2011;159(2):465-71. 61. Assis LC, Dias A, Barbosa AMP, Santini ACM, 51. Jahromi MK, Talebizadeh M, Mirzaei M. The Sousa VO, Vianna LS. Contribution of early Effect of Pelvic Muscle Exercises on Urinary intensive prolonged pelvic floor exercises. Incontinency and Self Esteem of Elderly American Journal of Epidemiology. Females With Stress Urinary Incontinency, 2011;(Abstract number 782-S). 173(Suppl 2013. Global journal of health science. 11):S196. 2015;7(2):71. 62. Siva Priya R, Kokila V, Malai K, Kumar S. 52. Asklund I, editor Treatment of stress urinary Effectiveness of Antenatal Motor Relearning incontinence via a smartphone application. Approach ofDiaphragm, Deep Abdominal and Report from an ongoing randomised controlled Pelvic Floor MusclesVersus Kegels Exercises

174 on Postpartum Pelvic FloorMuscle Strength. 72. Delgado D, White P, Trochez R, Drake MJ. A Indian Journal of Physiotherapy & pilot randomised controlled trial of the pelvic Occupational Therapy. 2014;8(1):203-7. toner devicein female stress urinary incontinence. nt Urogynecol J 2013;24:1739- 63. Bo K, Hilde G, Staer-Jensen J, Braekken IH. 45. Can the Paula method facilitate co-contraction of the pelvic floor muscles? A 4D ultrasound 73. Ong TA, Khong SY, Ng KL, Ting JRS, Kamal study. Int Urogynecol J. 2011;22(6):671-6. N, Yeoh WS, et al. Using the Vibrance Kegel Device With Pelvic Floor Muscle Exercise for 64. Kamel D, Thabet A, Tantawy S, Radwan M. Stress Urinary Incontinence: A Randomized Effect of abdominal versus pelvic floor muscle Controlled Pilot Study. Urology. exercises in obese Egyptian women with mild 2015;86(3):487-91. stress urinary incontinence: A randomised controlled trial. Hong Kong Physiotherapy 74. Hirakawa T, Suzuki S, Kato K, Gotoh M, Journal 2013;31(1):12-8. Yoshikawa Y. Randomized controlled trial of pelvic floor muscle training with or without 65. Cruz C, Riesco ML, Zanetti M. Supervised biofeedback for urinary incontinence. pelvic floor muscle training to treat urinary International Urogynecology Journal. incontinence during pregnancy: A randomized 2013;24(8):1347-54. controlled trial. Neurourology and urodynamics. 2014;33(6(Abstract#403)):867-8. 75. Fitz FF, Stupp L, Costa TF, Sartori MG, Girao MJ, Castro RA. Supervised versus non- 66. Lamb SE, Pepper J, Lall R, Jorstad-Stein EC, supervised pelvic floor muscle training for Clark MD, Hill L, et al. Group treatments for stress urinary incontinence: Randomized sensitive health care problems: a randomised controlled trial. Int Urogyn J. controlled trial of group versus individual 2015;26(Suppl1):145-6. physiotherapy sessions for female urinary incontinence. BMC Womens Health. 2009;9:26. 76. Jordre B, Schweinle W. Comparing resisted hip rotation with pelvic floor muscle training in 67. Liebergall-Wischnitzer M, Paltiel O, Lavy Y, women with stress urinary incontinence: A pilot Shveiky D, Manor O, Hochner-Celnikier D. study. J Womens Health Phys Therap. Long term efficacy of Paula method as 2014;38(2):81-9. compared with pelvic floor muscle training for stress urinary incontinence in women. Journal 77. Konstantinidou E, Kalaitzi M, Mytilekas K, of wound, ostomy, and continence nursing : Mikos T, Ioannides E, Hatzichristou D. Is there official publication of The Wound, Ostomy and a role for training of the transversus abdominis Continence Nurses Society. 2013;40(1):90-6. muscles in the physiotherapy schemes applied in the treatment of female urinary inconti- 68. Junginger B, Metz M, Baessler K. Comparison nence? . Proceedings of the 43rd Annual of a bladder neck effective pelvic floor Meeting of the International Continence Society rehabilitation program and EMG biofeedback (ICS). 2013. augmented pelvic floor muscle training: a randomized controlled trial Neurourology and 78. Donahoe-Fillmore B, Chomy W, Brahler CJ, urodynamics. 2014;33(6):970-1. Ingley A, Kennedy J, Osterfeld V. A comparison of two pelvic floor muscle training programs in 69. Kashanian M, Ali SS, Nazemi M, Bahasadri S. females with stress urinary incontinence: A pilot Evaluation of the effect of pelvic floor muscle study. The Journal of Applied Research. training (PFMT or Kegel exercise) and assisted 2011;11(2):73-83. pelvic floor muscle training (APFMT) by a resistance device (Kegelmaster device) on the 79. Prudencio C, Barbosa A, Derobio AL, Anezio A, urinary incontinence in women "comparison Vesentini G, Almeida AP, editors. Comparison between them : a randomized trial". European of three physiotherapy methods for treatment of Journal of Obstetrics & Gynecology and stress urinary incontinence: impact in quality of Reproductive Biology. 2011;159(1):218-23. life and muscle function. Proceedings of the 44th Annual Meeting of the International 70. Galea M, Tisseverasinghe S, Sherburn M. A Continence Society (ICS), 2014 Oct 20-24, randomised controlled trial of transabdominal 2014; 2014; Rio de Janeiro, Brazil. ultrasound biofeedback for pelvic floor muscle training in older women with urinary inconti- 80. Shin J, Sul C, Na Y, Song K, Lim J, Yun C, et nence. Australian and New Zealand al. Effectiveness of perinometer biofeedback Continence Journal. 2013;19(2):38-44. pelvic floor muscle exerciser with ExTT-101tm in female stress urinary incontinence. 71. Ferreira M, Clara P, Duarte JA, Rodrigues R. Proceedings of the 42nd Annual Meeting of the Exercise programmes for women with stress International Continence (ICS) (Beijing, China). urinary incontinence. Primary Health Care. 2012. 2012;22(3):24-7.

175 81. Manonai J, Kamthaworn S, Petsarb K, incontinence in women. J Womens Health Wattanayingcharoenchai R. Development of a (Larchmt). 2009;18(3):377-85. pelvic floor muscle strength evaluation device Neurourology and urodynamics. 91. Delgado D, Drake M. A randomized study to 2013;32(6):657-8. compare the pelvic toner device against standard pelvic floor exercises in the treatment 82. Marques S, Haddad J, Passaro A, Silveira S, of stress urinary incontinence in women Baracat E, Ferreira E. Effectiveness of the (Abstract 486). 39th annual meeting of he strengthening of pelvic floor muscles, International Continence Society; Sept 29th to adductors of hip, gluteus maximus and gluteus October 30rd; San Francisco, California2009. medius in the treatment of stress urinary incontinence: blind randomized clinical trial - 92. Herbison GP, Dean N. Weighted vaginal cones partial results. Proceedings of the 44th Annual for urinary incontinence. Cochrane Database Meeting of the International Continence Society Syst Rev. 2013;7:CD002114. (ICS). 2014. 93. Harvey MA, Johnston SL. A randomized, single 83. Dumoulin C. Physiotherapy compared to blind, prospective trial comparing pelvic floor individual physiotherapy to treat urinary physiotherapy with biofeedback versus incontinence in aging women: A randomized weigthted vaginal cones in the treatment of controlled trial. Ref ID: 60909 Trials registry female genuine stress urinary incontinence: a number(s): NCT02039830. 2012. pilot study. Int J Urogyn J. 2006;17(Suppl 2):S235-S6. 84. Buen M. Clinical trial: influence of the practice of Pilates on the incidence of urinary 94. Pereira VS, de Melo MV, Correia GN, Driusso incontinence, perineal strength low back pain in P. Vaginal cone for post menopausal women the third trimester. Ref ID: 64504 Trials registry with stress urinary incontinence: randomized, number(s): RBR-4wkr8y. 2014. controlled trial. Climacteric. 2012;15(1):45-51. 85. Navarro B, MT. Randomized Clinical Trial on 95. Labrie J, Berghmans BL, Fischer K, Milani AL, the Effectiveness of Hipopressive Exercises van der Wijk I, Smalbraak DJ, et al. Surgery Versus Classical Perineal Physiotherapy in versus physiotherapy for stress urinary Women With Pelvic Floor Dysfunction. Ref ID: incontinence. N Engl J Med. 64750 Trials registry number(s): 2013;369(12):1124-33. NCT02259712. 2013. 96. Kenton K, Barber M, Wang L, Hsu Y, Rahn D, 86. Rao B, Nayak S. Prevalence and Whitcomb E, et al. Pelvic floor symptoms Physiotherapy intervention for Pelvic floor improve similarly after pessary and behavioral dysfunction in women of Udupi Taluk. Ref ID: treatment for stress incontinence. Female 47889 Trials registry number(s): Pelvic Med Reconstr Surg. 2012;18(2):118-21. CTRI/2012/12/003226. 2012. 97. Richter HE, Burgio KL, Brubaker L, Nygaard IE, 87. Hagen S. OPAL: A multicentre randomised trial Ye W, Weidner A, et al. Continence pessary of the effectiveness and cost effectiveness of compared with behavioral therapy or combined basic versus bio feedback mediated intensive therapy for stress incontinence: a randomized pelvic floor muscle training for female stress or controlled trial. Obstet Gynecol. mixed urinary incontinence. Ref ID: 64519 2010;115(3):609-17. Trials registry number(s): ISRCTN57746448; 98. Kafri R, Deutscher D, Shames J, Golombp J, UKCRN15841 2014. Melzer I. Randomized trial of a comparison of 88. Haruna M, Asai Y. Effect of postpartum pelvic rehabilitation or drug therapy for urgency floor muscle training with ultrasound urinary incontinence: 1-year follow-up. Int biofeedback on recovery of pelvic floor muscle Urogynecol J. 2013;24(7):1181-9. function: a randomized controlled trial. Ref ID: 99. Golmakani N, Khadem N, Arabipoor A, Kerigh 66324 Trials registry number(s): JPRN- BF, Esmaily H. Behavioral Intervention UMIN000015878. 2014. Program versus Vaginal Cones on Stress 89. Bertotto A. Evaluating the effectiveness of Urinary Incontinence and Related Quality of pelvic Floor muscle training with and without Life: A Randomized Clinical Trial. Oman Med J. EMG biofeedback and quality of life in peri and 2014;29(1):32-8. postmenopausal women with stress urinary 100. Pereira VS, de Melo MV, Correia GN, Driusso incontinence. Ref ID: 67535 Trials registry P. Long term effects of pelvic floor muscle number(s): NCT02275728. 2014. training with vaginal cone in post-menopausal 90. Liebergall-Wischnitzer M, Hochner-Celnikier D, women with urinary incontinence: a Lavy Y, Manor O, Shveiky D, Paltiel O. randomized controlled trial. Neurourology and Randomized trial of circular muscle versus urodynamics. 2013;32(1):48-52. pelvic floor training for stress urinary

176 101. Furst MC, Mendonca RR, Rodrigues AO, treatments for women with stress urinary Matos LL, Pompeo AC, Bezerra CA. Long term incontinence. Health Technology Assessment results of a clinical trial comparing isolated (Winchester, England). 2010;14(40):1-506. vaginal stimulation with combined treatment for women with stress incontinence. Einstein. 111. Grant AM, Cody DJ, Glazener CMA, Hay-Smith 2014;12(2):168-74. J, Herbison P, Lapitan MC, et al. Cochrane Incontinence Group. About the Cochrane 102. Kaya S, Akbayrak T, Gursen C, Beksac S. Collaboration (Cochrane Review Groups Short-term effect of adding pelvic floor muscle (CRGs)). The Cochrane Library [serial on-line]. training to bladder training for female urinary 2007(4). incontinence: a randomized controlled trial. International Urogynecology Journal. 112. Beuttenmüller L, Cader SA, Macena RHM, 2015;26(2):285-93. Araujo NDS, Nunes EFC, Dantas EHM. Floor muscles contraction in women with stress 103. Capobianco G, Donolo E, Borghero G, Dessole urinary incontinence underwent to F, Cherchi PL, Dessole S. Effects of exercisesand electric stimulation therapy: a intravaginal estriol and pelvic floor rehabilitation randomized study [Portuguese]. Fisioterapia e on urogenital aging in postmenopausal women. Pesquisa. 2011;18(3):210-6. Archives of gynecology and obstetrics. 2012;285(2):397-403. 113. Correia GN, Pereira VS, Hirakawa HS, Driusso P. Effects of surface and intravaginal electrical 104. Barber MD, Brubaker L, Burgio KL, Richter HE, stimulation in the treatment of women with Nygaard I, Weidner AC, et al. Comparison of 2 stress urinary incontinence: randomized transvaginal surgical approaches and controlled trial. European Journal of Obstetrics perioperative behavioral therapy for apical and Gynecology and Reproductive Biology. vaginal prolapse: the OPTIMAL randomized 2014;173(1):113-8. trial. JAMA. 2014;311(10):1023-34. 114. Lopès P, Rimbault F, Scheffler M, Andre C, 105. Peattie AB, Plevnik S, Stanton SL. Vaginal Cappelletti MC, Mares P. [Multicentric cones: a conservative method of treating prospective randomized and controlled study genuine stress incontinence. British Journal of assessing effectiveness of intravaginal Obstetrics and Gynaecology. electrostimulation at home compared to usual 1988;95(10):1049-53. care in female patients with urinary incontinence and prior perineal reeducation]. 106. Hahn I, Milsom I, Ohlsson BL, Ekelund P, [French]. Gynecologie, Obstetrique & Fertilite. Uhlemann C, Fall M. Comparative assessment 2014;42(11):779-86. of pelvic floor function using vaginal cones, vaginal digital palpation and vaginal pressure 115. Pereira VS, Bonioti L, Correia GN, Driusso P. measurements. Gynecologic and Obstetric [Effects of surface electrical stimulation in older Investigation. 1996;41(4):269-74. women with stress urinary incontinence: a randomized controlled pilot study]. [Spanish]. 107. Porta Roda O, Simo Gonzalez M, Reula Blasco Actas Urologicas Espanolas. 2012;36(8):491- MC, Diaz Lopez MA, Diaz Bellido P, Vara 6. Paniagua J. Use of a vaginal spheres device in the conservative treatment of stress urinary 116. Terlikowski R, Dobrzycka B, Kinalski M, incontinence: a randomized controlled trial Kuryliszyn-Moskal A, Terlikowski SJ. Neurourology and urodynamics. Transvaginal electrical stimulation with surface- 2013;32(6):661-3. EMG biofeedback in managing stress urinary incontinence in women of premenopausal age: 108. Stupp L, Yamamoto D, Fonseca T, Resende a double-blind, placebo controlled, random- AM, Ploger C, Oliveira E. Proprioception and ized clinical trial. International Urogynecology awareness training prior pelvic floor muscle Journal. 2013;24(10):1631-8. exercises for treatment of urinary incontinence: randomized controlled trial. Int Urogyn J. 117. Hofbauer J, Preisinger F, Nurnberger N. [The 2011;22(Supp 1):S162-S4. value of physical therapy in genuine female stress incontinence]. [German]. Zeitschrift Fur 109. Santos PF, Oliveira E, Zanetti MR, Arruda RM, Urologie Und Nephrologie. 1990;83(5):249-54. Sartori MG, Girao MJ, et al. [Electrical stimulation of the pelvic floor versus vaginal 118. Bo K, Talseth T, Holme I. Single blind, cone therapy for the treatment of stress urinary randomised controlled trial of pelvic floor incontinence]. Rev Bras Ginecol Obstet. exercises, electrical stimulation, vaginal cones, 2009;31(9):447-52. and no treatment in management of genuine stress incontinence in women. BMJ. 110. Imamura M, Abrams P, Bain C, Buckley B, 1999;318(7182):487-93. Cardozo L, Cody J, et al. Systematic review and economic modelling of the effectiveness 119. Castro RA, Arruda RM, Zanetti MR, Santos PD, and cost-effectiveness of non-surgical Sartori MG, Girao MJ. Single-blind,

177 randomized, controlled trial of pelvic floor incontinence. British Journal of Obstetrics & muscle training, electrical stimulation, vaginal Gynaecology. 1987;94(6):575-82. cones, and no active treatment in the management of stress urinary incontinence. 130. Knight S. Evaluation of neuromuscular Clinics (Sao Paulo, Brazil). 2008;63(4):465-72. electrical stimulation in the treatment of genuine stress incontinence. Physiotherapy. 120. Luber KM, Wolde-Tsadik G. Efficacy of 1998;84(2):61-71. functional electrical stimulation in treating genuine stress incontinence: a randomized 131. Wise BG, Haken J, Cardozo LD, Plevnik S. A clinical trial. Neurourology & Urodynamics. comparative study of vaginal cone therapy, 1997;16(6):543-51. cones + Kegel exercises, and maximal electrical stimulation in the treatment of female 121. Laycock J, Jerwood D. Does pre modulated genuine stress incontinence (Abstract number interferential therapy cure genuine stress 76). Neurourology & Urodynamics. incontinence? Physiotherapy. 1993;79(8):553- 1993;12(4):436-7. 60. 132. Arruda RM, Castro RA, Sousa GC, Sartori MG, 122. Brubaker L, Benson JT, Bent A, Clark A, Shott Baracat EC, Girao MJ. Prospective randomized S. Transvaginal electrical stimulation for female comparison of oxybutynin, functional urinary incontinence. American Journal of electrostimulation, and pelvic floor training for Obstetrics & Gynecology. 1997;177(3):536-40. treatment of detrusor overactivity in women. International Urogynecology Journal. 123. Sand PK, Richardson DA, Staskin DR, Swift 2008;19(8):1055-61. SE, Appell RA, Whitmore KE, et al. Pelvic floor electrical stimulation in the treatment of genu- 133. Lin LS, Song YF, Song J, Chen MF. [A clinical ine stress incontinence: a multicenter, placebo- study of pelvic floor electrical stimulation in controlled trial. American Journal of Obstetrics treatment of overactive bladder] [Chinese]. & Gynecology. 1995;173(1):72-9. Chung-Hua Fu Chan Ko Tsa Chih [Chinese Journal of Obstetrics & Gynecology]. 124. Henalla SM, Hutchins CJ, Robinson P, 2004;39(12):801-3. MacVicar J. Non-operative methods in the treatment of female genuine stress 134. Wang AC, Chih SY, Chen MC. Comparison of incontinence of urine. Journal of Obstetrics and electric stimulation and oxybutynin chloride in Gynaecology. 1989;9(3):222-5. management of overactive bladder with special reference to urinary urgency: a randomized 125. Jeyaseelan SM, Haslam EJ, Winstanley J, Roe placebo controlled trial. Urology. BH, Oldham JA. An evaluation of a new pattern 2006;68(5):999-1004. of electrical stimulation as a treatment for urinary stress incontinence: a randomized, 135. Smith JJ, 3rd. Intravaginal stimulation double-blind, controlled trial. Clinical Reha- randomized trial. The Journal of urology. bilitation. 2000;14(6):631-40. 1996;155(1):127-30. 126. Alves PG, Nunes FR, Guirro EC. Comparison 136. Jeyaseelan S, Oldham JA. Can the effects of between two different neuromuscular electrical pelvic floor muscle exercises be enhanced with stimulation protocols for the treatment of a new pattern of electrical stimulation in women female stress urinary incontinence: a with stress incontinence (Abstract). Proceed- randomized controlled trial. Revista Brasileira ings of the World Confederation for Physical de Fisioterapia. 2011;15(5):393-8. Therapy (WCPT), 14th International Congress, 7-12 June, Barcelona. 2003. 127. Huebner M, Riegel K, Hinninghofen H, Wallwiener D, Tunn R, Reisenauer C. Pelvic 137. Patil SP, Nagrale AV, Ganvir SD. Additive floor muscle training for stress urinary effect of interferential therapy over pelvic floor incontinence: a randomized, controlled trial exercises. International Journal of Therapy & comparing different conservative therapies. Rehabilitation. 2010;17(11):596-602. Physiother Res Int. 2011;16(3):133-40. 138. Goode PS, Burgio KL, Locher JL, Roth DL, 128. Maher RM, Crowe L, Caulfield B. Comparison Umlauf MG, Richter HE, et al. Effect of of two methods of electrical muscle stimulation behavioral training with or without pelvic floor training of pelvic floor musculature in the electrical stimulation on stress incontinence in treatment of stress urinary incontinence women: a randomized controlled trial. JAMA. (Abstract). Journal of Women's Health Physical 2003;290(3):345-52. Therapy. 2009;33(1):24. 139. Indrekvam S, Sandvik H, Hunskaar S. A 129. Wilson PD, Al Samarrai T, Deakin M, Kolbe E, Norwegian national cohort of 3198 women Brown AD. An objective assessment of treated with home managed electrical physiotherapy for female genuine stress stimulation for urinary incontinence: Effectiveness and treatment results.

178 Scandinavian Journal of Urology and Transcutaneal tibial nerve stimulation versus Nephrology. 2001;35(1):32-9. long release oxibutinin in the treatment of patients with overactive bladder. A randomize 140. Moore KN. Treatment of urinary incontinence in control trial (Abstract number 013). men with electrical stimulation: is practice International Urogynecology Journal and Pelvic evidence based? Journal of Wound, Ostomy, & Floor Dysfunction. 2013;24(Suppl 1):S14. Continence Nursing. 2000;27(1):20-31. 151. Marques A, Herrmann V, Ferreira N, Bellette P. 141. Rai BP, Cody JD, Alhasso A, Stewart L. Transcutaneous posterior tibial nerve Anticholinergic drugs versus non drug active stimulation in overactive bladder (Abstract therapies for non neurogenic overactive number 471). Proceedings of the 38th Annual bladder syndrome in adults. The Cochrane Meeting of the International Continence Society Database Of Systematic Reviews. (ICS), 2008 Oct 20-24, Cairo, Egypt. 2008. 2012;12:CD003193. 152. Peters K, Carrico D, Perez-Marrero R, Khan A, 142. Slovak M, Chapple CR, Barker AT. Non- Wooldridge L, Davis G, et al. 12 week results invasive transcutaneous electrical stimulation from the SUMIT trial: percutaneous tibial nerve in the treatment of overactive bladder. Asian stimulation vs validated sham in those exposed Journal of Urology. 2015;2(2):92-101. to pharmacologic therapy (Abstract number 143. Vandoninck V, van Balken MR, Finazzi Agro E, 125). Neurourology and urodynamics. Petta F, Micali F, Heesakkers J, et al. 2010;29(6):988-9. Percutaneous Tibial Nerve Stimulation in the 153. Preyer O, Gabriel B, Mailath-Pokorny M, Treatment of Overactive Bladder: Urodynamic Doerfler D, Laml T, Umek W, et al. Peripheral Data. Neurourology and urodynamics. tibial neurostimulation (PTNS) versus 2003;22:227-32. tolterodine in the treatment of women with urge 144. Amarenco G, Ismael SS, Even-Schneider A, urinary incontinence and urge symptoms (Ab- Raibaut P, Demaille-Wlodyka S, Parratte B, et stract number 246). International Uro- al. Urodynamic effect of acute transcutaneous gynecology Journal and Pelvic Floor posterior tibial nerve stimulation in overactive Dysfunction. 2007;18(Suppl 1):S139-S40. bladder. The Journal of urology. 154. Preyer O, Umek W, Laml T, Bjelic-Radisic V, 2003;169(6):2210-5. Gabriel B, Mittlboeck M, et al. Percutaneous 145. Klingler HC, Pycha A, Schmidbauer J, tibial nerve stimulation versus tolterodine for Marberger M. Use of peripheral overactive bladder in women: a randomised neuromodulation of the S3 region for treatment controlled trial. European Journal of Obstet- of detrusor overactivity: a urodynamic based rics, Gynecology, & Reproductive Biology. study. Urology. 2000;56(5):766-71. 2015;191:51-6. 146. Finazzi-Agro E, Rocchi C, Pachatz C, Petta F, 155. Sancaktar M, Ceyhan ST, Akyol I, Muhcu M, Spera E, Mori F, et al. Percutaneous tibial Alanbay I, Mutlu EC, et al. The outcome of nerve stimulation produces effects on brain adding peripheral neuromodulation (stoller activity: study on the modifications of the long afferent neuro-stimulation) to anti-muscarinic latency somatosensory evoked potentials. therapy in women with severe overactive Neurourology and urodynamics. bladder. Gynecological Endocrinology. 2009;28(4):320-4. 2010;26(10):729-32. 147. Apostolidis A. Neuromodulation for intractable 156. Schreiner L, dos Santos TG, Knorst MR, da OAB. Neurourology and urodynamics. Silva Filho IG. Randomized trial of 2011;30(5):766-70. transcutaneous tibial nerve stimulation to treat urge urinary incontinence in older women. 148. Bellette PO, Rodrigues-Palma PC, Hermann V, International Urogynecology Journal. Riccetto C, Bigozzi M, Olivares JM. [Posterior 2010;21(9):1065-70. tibial nerve stimulation in the management of overactive bladder: a prospective and con- 157. Souto SC, Reis LO, Palma T, Palma P, Denardi trolled study]. [Spanish]. Actas Urologicas F. Prospective and randomized comparison of Espanolas. 2009;33(1):58-63. electrical stimulation of the posterior tibial nerve versus oxybutynin versus their combination for 149. Finazzi-Agrò E, Petta F, Sciobica F, Pasqualetti treatment of women with overactive bladder P, Musco S, Bove P. Percutaneous tibial nerve syndrome. World Journal of Urology. stimulation effects on detrusor overactivity 2014;32(1):179-84. incontinence are not due to a placebo effect: a randomized, double blind, placebo controlled 158. Vecchioli-Scaldazza C, Morosetti C, Berouz A, trial. Journal of Urology. 2010;184(5):2001-6. Giannubilo W, Ferrara V. Solifenacin Succinate versus Percutaneous Tibial Nerve Stimulation 150. Manriquez VI, Naser ME, Gomez M, Guzman in Women with Overactive Bladder Syndrome: R, Valdevenito R, Lecannelier J, et al. 179 Results of a Randomized Controlled Cross- 170. Manganotti P, Zaina F, Vedovi E, Pistoia L, over Study. Gynecologic & Obstetric Rubilotta E, D'Amico A, et al. Repetitive Investigation. 2013;75(4):230-4. magnetic stimulation of the sacral roots for the treatment of stress incontinence: a brief report. 159. Barker AT, Freeston IL, Jalinous R, Jarratt JA. Europa Medicophysica. 2007;43(3):339-44. Magnetic stimulation of the human brain and peripheral nervous system: an introduction and 171. Suzuki T, Yasuda K, Yamanishi T, Kitahara S, the results of an initial clinical evaluation. Nakai H, Suda S, et al. Randomized, double- Neurosurgery. 1987;20(1):100-9. blind, sham controlled evaluation of the effect of functional continuous magnetic stimulation in 160. Galloway NT, El-Galley RE, Sand PK, Appell patients with urgency incontinence. RA, Russell HW, Carlan SJ. Extracorporeal Neurourology & Urodynamics. 2007;26(6):767- magnetic innervation therapy for stress urinary 72. incontinence. Urology. 1999;53(6):1108-11. 172. But I, Faganelj M, Sostaric A. Functional 161. Goldberg RP, Sand PK. Electromagnetic pelvic magnetic stimulation for mixed urinary floor stimulation: applications for the incontinence. Journal of Urology. gynecologist. Obstet Gynecol Surv. 2005;173(5):1644-6. 2000;55(11):715-20. 173. Lee JS, Hong JY, Kim MH, Seo JT. 162. But I. Conservative treatment of female urinary Comparative study of the pelvic floor magnetic incontinence with functional magnetic stimulation with BIOCON-2000 (trademark) in stimulation. Urology. 2003;61(3):558-61. female urinary incontinence patients. Korean 163. Quek P. A critical review on magnetic Journal of Urology. 2004;45(5):438-43. stimulation: What is its role in the management 174. Gilling P, Kennett K, Bell D, Wrigley T, of pelvic floor disorders? Current Opinion in Fraundorfer M. A double blind randomised trial Urology. 2005;15(4):231-5. comparing magnetic stimulation of the pelvic 164. Craggs MD, Sheriff MKM, Shah PJR, Fowler floor to sham treatment for women with stress CJ, Petersen T. Response to multipulse urinary incontinence (Abstract). Nerourology & magnetic stimulation of spinal nerve roots Urodynamics. 2001;20(4):432-3. mapped over the sacrum in man (Abstract). 175. Brodak PP, Bidair M, Joseph A, Szollar S, Journal of Physiology. 1995;483(Suppl):127P- Juma S. Magnetic stimulation of the sacral 8P. roots. Neurourology & Urodynamics. 165. Kralj B. Conservative treatment of female 1993;12(6):533-40. stress urinary incontinence with functional 176. Hadley EC. BLadder training and related electrical stimulation. European Journal of therapies for urinary incontinence in older Obstetrics Gynecology and Reproductive people. JAMA. 1986;256(3):372-9. Biology. 1999;85(1):53-6. 177. Wyman JF, Fantl JA, McClish DK, Bump RC. 166. Lindstrom S, Fall M, Carlsson CA, Erlandson Comparative efficacy of behavioral BE. The neurophysiological basis of bladder interventions in the management of female inhibition in response to intravaginal electrical urinary incontinence. American Journal of stimulation. Journal of Urology. Obstetrics and Gynecology. 1998;179(4):999- 1983;129(2):405-10. 1007. 167. Tsai PY, Wang CP, Hsieh CY, Tsai YA, Yeh 178. Wyman JF, Fantl JA. Bladder training in SC, Chuang TY. Longterm sacral magnetic ambulatory care management of urinary stimulation for refractory stress urinary incontinence. Urologic Nursing. 1991;11(3 incontinence. Archives of Physical Medicine & Sep):11-7. Rehabilitation. 2014;95(12):2231-8. 179. Fantl J, Wyman JF, McClish DK, et al. EFficacy 168. Wallis MC, Davies EA, Thalib L, Griffiths S. of bladder training in older women with urinary Pelvic static magnetic stimulation to control incontinence. JAMA. 1991;265(5):609-13. urinary incontinence in older women: a randomized controlled trial. Clinical Medicine & 180. Fantl JA, Newman DK, Colling JC, DeLancey Research. 2012;10(1):7-14. JO, Kees C, Loughery R. Urinary incontinence in adults: acute and chronic management. 169. Fujishiro T, Enomoto H, Ugawa Y, Takahashi Clinical Practice Guideline: Update 1996;2. S, Ueno S, Kitamura T. Magnetic stimulation of the sacral roots for the treatment of stress 181. Wyman JF. Treatment of urinary incontinence incontinence: an investigational study and in men and older women: the evidence shows placebo controlled trial. Journal of Urology. the efficacy of a variety of techniques. 2000;164(4):1277-9. American Journal of Nursing. 2003;Suppl:26- 35.

180 182. Burgio KL, Stutzman RE, Engel BT. Behavioral 193. Jarvis GJ. A controlled trial of bladder drill and training for post-prostatectomy urinary drug therapy in the management of detrusor incontinence. The Journal of urology. instability. Br J Urol. 1981;53(6):565-6. 1989;141(2):303-6. 194. Jarvis GJ, Millar DR. Controlled trial of bladder 183. Colling J, Owen TR, McCreedy M, Newman D. drill for detrusor instability. Br Med J. The effects of a continence program on frail 1980;281(6251):1322-3. communitydwelling elderly persons. Urologic Nursing. 2003;23(2):117-22, 27-31. 195. Mattiasson A, Blaakaer J, Høye K, Wein AJ, The Tolterodine Scandinavian Study G. 184. Eustice S, Roe B, Paterson J. Prompted Simplified bladder training augments the voiding for the management of urinary effectiveness of tolterodine in patients with an incontinence in adults. Cochrane Database of overactive bladder. BJU International. Systematic Reviews. 2000(2). 2003;91(1):54-60. 185. Wallace SA, Roe B, Williams K, Palmer M. 196. Mattiasson A, Masala A, Morton R, Bolodeoku Bladder training for urinary incontinence in J. Efficacy of simplified bladder training in adults. Cochrane Database of Systematic patients with overactive bladder receiving a Reviews. 2004(1). solifenacin flexible dose regimen: results from a randomized study. BJU International. 186. Berghmans LCM, Hendriks HJM, De Bie RA, 2010;105(8):1126-35. Van Waalwijk ESC, Van D, Bø K, et al. Conservative treatment of urge urinary 197. Diokno AC, Ocampo MS, Jr., Ibrahim IA, Karl incontinence in women: a systematic review of CR, Lajiness MJ, Hall SA. Group session randomized clinical trials. BJU International. teaching of behavioral modification program 2000;85(3):254-63. (BMP) for urinary incontinence: a randomized controlled trial among incontinent women. 187. Assassa P, Williams K, Lambert P, Abrams K, International Urology and Nephrology. Turner D, Shaw C, et al. A double blind 2010;42(2):375-81. randomised placebo controlled trial of the effectiveness of bladder training with 198. Dougherty MC, Dwyer JW, Pendergast JF, oxybutynin or imipramine in the management of Tomlinson BU, Boyington AR, Vogel WB, et al. detrusor overactivity (DO) (Abstract number Community based nursing: Continence care for 330). Proceedings of the Joint Meeting of the older rural women. Nursing Outlook. International Continence Society (ICS) and the 1998;46(5):233-44. International Urogynecological Association, 2010 Aug 23-27, Toronto, Canada. 2010. 199. Dougherty MC, Dwyer JW, Pendergast JF, Boyington AR, Tomlinson BU, Coward RT, et 188. Sherburn M, Bird M, Carey M, Bo K, Galea MP. al. A randomized trial of behavioral Incontinence improves in older women after management for continence with older rural intensive pelvic floor muscle training: an women. Research in Nursing & Health. assessor blinded randomized controlled trial. 2002;25(1):3-13. Neurourology & Urodynamics. 2011;30(3):317- 24. 200. Lagro-Janssen A, Debruyne F, AJA S, Van Weel C. The Effects of Treatment of Urinary 189. Yoon HS, Song HH, Ro YJ. A comparison of Incontinence in General Practice. Family effectiveness of bladder training and pelvic Practice. 1992;9(3):284-9. muscle exercise on female urinary incontinence. Int J Nurs Stud. 2003;40(1):45- 201. Colombo M, Zanetta G, Scalambrino S, Milani 50. R. Oxybutynin and bladder training in the management of female urinary urge 190. Song C, Park JT, Heo KO, Lee KS, Choo MS. incontinence: a randomized study. Int Effects of bladder training and/or tolterodine in Urogynecol J Pelvic Floor Dysfunct. female patients with overactive bladder 1995;6(2):63-7. syndrome: a prospective, randomized study. J Korean Med Sci. 2006;21:1060-3. 202. Castleden CM, Duffin HM, Gulati RS. DOUBLEBLIND STUDY OF IMIPRAMINE 191. Szonyi G, Collas DM, Ding YY, Malone-Lee JG. AND PLACEBO FOR INCONTINENCE DUE Oxybutynin with Bladder Retraining for TO BLADDER INSTABILITY. Age and Ageing. Detrusor Instability in Elderly People: A 1986;15(5):299-303. Randomized Controlled Trial. Age and Ageing. 1995;24(4):287-91. 203. Bryant CM, Dowell CJ, Fairbrother G. Cafeine reduction education to improve urinary 192. Wiseman PA, Malone-Lee J, Rai GS. symptoms. Br J Nurs. 2002;11(8):560-5. Terodiline with bladder retraining for treating detrusor instability in elderly people. BMJ. 204. Ostaszkiewicz J, Johnston L, Roe B. Timed 1991;302(6783):994-6. voiding for the management of urinary

181 incontinence in adults. Cochrane Database of 217. Yuan Z, He C, Yan S, Huang D, Wang H, Tang Systematic Reviews. 2004(1). W. Acupuncture for overactive bladder in female adult: a randomized controlled trial. 205. Klarskov P, Gerstenberg TC, Hald T. Bladder World Journal of Urology. 2015;33(9):1303-8. training and terodiline in females with idiopathic urge incontinence and stable detrusor function. 218. Dincer F, Linde K. Sham interventions in Scand J Urol Nephrol. 1986;20(1):41-6. randomized clinical trials of acupuncture—a review. Complementary Therapies in Medicine. 206. Godec CJ. Timed voiding-- a useful tool in the 2003;11(4):235-42. treatment of urinary incontinence. Urology. 1984;23(1):97-100. 219. Schulz KF, Altman DG, Moher D, Group C. CONSORT 2010 Statement: updated 207. Wang Y, Zhishun L, Peng W, Zhao J, Liu B. guidelines for reporting parallel group Acupuncture for stress urinary incontinence in randomised trials. Trials. 2010;11:32. adults. Cochrane Database of Systematic Reviews. 2013(7):Art. No.: CD009408. 220. Hagen S, Stark D. Conservative prevention and management of pelvic organ prolapse in 208. Paik S-H, Han S-R, Kwon O-J, Ahn Y-M, Lee women. Cochrane Database Syst Rev. B-C, Ahn S-Y. Acupuncture for the treatement 2011(12):CD003882. of urinary incontinence: A review of randomized controlled trials. Experimental and Thera- 221. Hagen S, Stark D, Campbell P. Conservative peutic Medicine. 2013;6(3):773-80. prevention and management of pelvic organ prolapse in women. Cochrane Database Syst 209. Kim JH, Nam D, Park MK, Lee ES, Kim SH. Rev. 2016 (in prep). Randomized control trial of hand acupuncture for female stress urinary incontinence. 222. Bugge C, Adams EJ, Gopinath D, Reid F. Acupuncture & Electro Therapeutics Research. Pessaries (mechanical devices) for pelvic 2008;33(3-4):179-92. organ prolapse in women. Cochrane Database Syst Rev. 2013(2):CD004010. 210. Bi W. Clinical study on electro-acupuncture treatment of female stress incontinence. 223. Higgins J, Green S. Cochrane Handbook for Cinese Archives of Traditional Medicine. Systematic Reviews of Interventions: The 2007;25(6):1284-5. Cochrane Collaboration; 2011. 211. Chang KK, Wong TK, Wong TH, Leung AW, 224. CASP. Critical Appraisal Skills Programme Chung JW. Effect of acupressure in treating 2015. urodynamic stress incontinence: a randomized controlled trial. American Journal of Chinese 225. Abrams P. Incontinence, 3rd International Medicine. 2011;39(6):1139-59. Consultation on Incontinence. 3rd ed2004. 212. Forde JC, Jaffe E, Stone BV, Te AE, Espinosa 226. Hay Smith J, Berghams B, Burgio K, Dumoulin G, Chughtai B. The role of acupuncture in C, Hagen S, Moore K, et al. Adult conservative managing overactive bladder; a review of the Management In: Abrams P, Cardozo L, Khoury literature. International Urogynecology Journal. S, Wein A, editors. Incontinence. 4th ed2009. 2016:1-7. 227. Majumdar A, Saleh S, Hill M, Hill SR. The 213. Jin C, Zhou X, Pang R. Effect of impact of strenuous physical activity on the electroacupuncture combined with tolterodine development of pelvic organ prolapse. J Obstet on treating female mixed urinary incontinence. Gynaecol. 2013;33(2):115-9. Journal of Wound, Ostomy, & Continence 228. Lonnee-Hoffmann RA, Salvesen O, Morkved S, Nursing. 2014;41(3):268-72. Schei B. Self reported pelvic organ prolapse 214. Kelleher CJ, Filshie J, Burton G, Khullar V, surgery, prevalence, and nonobstetric risk Cardozo LD. Acupuncture and the treatment of factors: findings from the Nord Trondelag irritative bladder symptoms. Acupuncture in Health Study. Int Urogynecol J. Medicine. 1994;12:9-12. 2015;26(3):407-14. 215. Emmons SL, Otto L. Acupuncture for 229. Nygaard IE, Shaw JM, Bardsley T, Egger MJ. overactive bladder: a randomized controlled Lifetime physical activity and pelvic organ trial. Obstetrics & Gynecology. prolapse in middle aged women. Am J Obstet 2005;106(1):138-43. Gynecol. 2014;210(5):477 e1-12. 216. Engberg S, Cohen S, Sereika SM. The efficacy 230. Akmel M, Segni H. Pelvic organ prolapse in of acupuncture in treating urge and mixed jimma university specialized hospital, incontinence in women: a pilot study. Journal of southwest ethiopia. Ethiop J Health Sci. Wound, Ostomy, & Continence Nursing. 2012;22(2):85-92. 2009;36(6):661-70. 231. Gumanga SK, Munkaila A, Malechi H. Social demographic characteristics of women with

182 pelvic organ prolapse at the Tamale Teaching 242. Johnson P, Larson K, Hsu Y, DeLancey J, Hospital, Ghana. Ghana Med J. Fenner D, Morgan D. Self reported experiences 2014;48(4):208-13. of recurrent prolapse. International Urogynecology Journal and Pelvic Floor 232. Bathla S, Verghese G, Kalla V, Sharma TC, Dysfunction. 2011;22:S865-S6. Dam S, Agarwal N, et al. Reaching the unreached: Mobile surgical camps in a remote 243. Myers DL, Sung VW, Richter HE, Creasman J, village of Himachal Pradesh. J Midlife Health. Subak LL. Prolapse symptoms in overweight 2014;5(3):139-42. and obese women before and after weight loss. Female pelvic medicine & reconstructive 233. Aston BL, Sheehan L, Mawu G, Roberts C, surgery. 2012;18(1):55-9. Barnick C. Do women seeking treatment for pelvic floor dysfunction have higher than 244. Perez A, Palau MJ, Sanchez E, Rodriguez L, average BMI's? International Urogynecology Flores L, Hergueta BN, et al. Long term study Journal and Pelvic Floor Dysfunction. on the effect of weight loss in women with 2011;22:S129. obesity and urinary incontinence. Neurourology and urodynamics. 2013;32(6):541-2. 234. Aytan H, Ertunc D, Tok EC, Yasa O, Nazik H. Prevalence of pelvic organ prolapse and 245. Wein AJ. Re: Impact of surgically induced related factors in a general female population. weight loss on pelvic floor disorders: Editorial Turk Jinekoloji ve Obstetrik Dernegi Dergisi. comment. Journal of Urology. 2014;11(3):176-80. 2013;189(5):1817. 235. Chen Y, Johnson B, Li F, Lin X, Chen J, Chen 246. Whitcomb EL, Horgan S, Donohue MC, Lukacz C, et al. Weight at one year postpartum affects ES. Impact of surgically induced weight loss on the development of pelvic organ prolapse. pelvic floor disorders. International Uro- Reproductive Sciences. 2014;21(3 SUPPL. gynecology Journal and Pelvic Floor 1):221A. Dysfunction. 2012;23(8):1111-6. 236. Cuicchi D, Lombardi R, Cariani S, Leuratti L, 247. Young N, Atan I, Dietz HP. Obesity: How much Lecce F, Cola B. Clinical and instrumental does it matter for female pelvic organ prolapse? evaluation of pelvic floor disorders before and BJOG: An International Journal of Obstetrics after bariatric surgery in obese women. Surgery and Gynaecology. 2015;122:390. for obesity and related diseases : official journal of the American Society for Bariatric Surgery. 248. Awwad J, Sayegh R, Yeretzian J, Deeb ME. 2013;9(1):69-75. Prevalence, risk factors, and predictors of pelvic organ prolapse: a community based 237. Direkvand-Moghadam A, Ghazanfari Z, study. Menopause. 2012;19(11):1235-41. Sayehmiri K. Predictive Factors For Pelvic Organ Prolapse In Iranian Women's, An 249. Ramalingam K, Monga A. Obesity and pelvic Ordinal Logistic Approc. Journal of Clinical and floor dysfunction. Best Pract Res Clin Obstet Diagnostic Research. 2014;8(1):96-9. Gynaecol. 2015;29(4):541-7. 238. Elbiss HM, Osman N, Hammad FT. 250. Vergeldt TF, Weemhoff M, IntHout J, Kluivers Prevalence, risk factors and severity of KB. Risk factors for pelvic organ prolapse and symptoms of pelvic organ prolapse among its recurrence: a systematic review. Int Emirati women. BMC Urology. 2015;15(1):no Urogynecol J. 2015;26(11):1559-73. pagination. 251. Kudish BI, Iglesia CB, Sokol RJ, Cochrane B, 239. Espitia de la Hoz FJ. Risk factors associated Richter HE, Larson J, et al. Effect of weight with female genital prolapse: Case control change on natural history of pelvic organ study. Urologia Colombiana. 2015;24(1):12-8. prolapse. Obstet Gynecol. 2009;113(1):81-8. 240. Glazener C, Elders A, Macarthur C, Lancashire 252. Estanol MV, Crisp CC, Oakley SH, Kleeman RJ, Herbison P, Hagen S, et al. Childbirth and SD, Fellner AN, Pauls RN. Systemic markers of prolapse: long term associations with the collagen metabolism and vitamin C in smokers symptoms and objective measurement of and non-smokers with pelvic organ prolapse. pelvic organ prolapse. BJOG. Eur J Obstet Gynecol Reprod Biol. 2013;120(2):161-8. 2015;184:58-64. 241. Gozukara YM, Akalan G, Tok EC, Aytan H, 253. Bezerra LRPS, Vasconcelos Neto JA, Ertunc D. The improvement in pelvic floor Vasconcelos CTM, Karbage SAL, Lima AC, symptoms with weight loss in obese women Frota IPR, et al. Prevalence of unreported does not correlate with the changes in pelvic bowel symptoms in women with pelvic floor anatomy. International urogynecology journal. dysfunction and the impact on their quality of 2014;25(9):1219-25. life. International Urogynecology Journal and Pelvic Floor Dysfunction. 2014;25(7):927-33.

183 254. Navaneethan PR, Kekre A, Jacob KS, 265. Kashyap R, Jain V, Singh A. Comparative Varghese L. Vitamin D deficiency in effect of 2 packages of pelvic floor muscle postmenopausal women with pelvic floor training on the clinical course of stage I-III disorders. J Midlife Health. 2015;6(2):66-9. pelvic organ prolapse. Int J Gynaecol Obstet. 2013;121(1):69-73. 255. Parker-Autry CY, Markland AD, Ballard AC, Downs-Gunn D, Richter HE. Vitamin D status 266. Resende APM, Stupp L, Bernardes BT, in women with pelvic floor disorder symptoms. Oliveira E, Castro RA, Girao MJBC, et al. Can Int Urogynecol J. 2012;23(12):1699-705. hypopressive exercises provide additional benefits to pelvic floor muscle training in 256. Miller JM, Ashton-Miller JA, Delancey J. A women with pelvic organ prolapse? pelvic floor muscle pre-contraction can reduce Neurourology and urodynamics. cough related urine loss in selected women 2012;31(1):121-5. with SUI. Journal of the American Geriatric Society. 1998;46:870-4. 267. Stupp L, Magalhaes Resende AP, Oliveira E, Castro RA, Castello Girao MJB, Ferreira Sartori 257. Hagen S, Glazener C, McClurg D, MacArthur MG. Pelvic floor muscle training for treatment C, Herbison P, Wilson D, et al. A multicentre of pelvic organ prolapse: An assessor blinded randomised controlled trial of a pelvic floor randomized controlled trial. International muscle training intervention for the prevention Urogynecology Journal and Pelvic Floor of pelvic organ prolapse (PREVPROL). Dysfunction. 2011;22(10):1233-9. Neurourology and urodynamics. 2014;33(6):852-3. 268. Wiegersma M, Panman CM, Kollen BJ, Vermeulen KM, Schram AJ, Messelink EJ, et 258. Bo K, Hilde G, Tennfjord MK, Jensen JS, al. Pelvic floor muscle training versus watchful Siafarikas F, Engh ME. Randomized controlled waiting or pessary treatment for pelvic organ trial of pelvic floor muscle training to prevent prolapse (POPPS): design and participant and treat pelvic organ prolapse in postpartum baseline characteristics of two parallel primiparous women. Neurourology and urody- pragmatic randomized controlled trials in namics. 2013;32(6):806-7. primary care. Maturitas. 2014;77(2):168-73. 259. Hagen S, Stark D, Maher C, Adams E. 269. Jarvis SK, Hallam TK, Lujic S, Abbott JA, Conservative management of pelvic organ Vancaillie TG. Peri-operative physiotherapy prolapse in women. Cochrane Database Syst improves outcomes for women undergoing Rev. 2004(2):CD003882. incontinence and or prolapse surgery: Results 260. Culligan PJ, Scherer J, Dyer K, Priestley JL, of a randomised controlled trial. Australian and Guingon-White G, Delvecchio D, et al. A New Zealand Journal of Obstetrics and Gy- randomized clinical trial comparing pelvic floor naecology. 2005;45:300-3. muscle training to a Pilates exercise program 270. Frawley HC, Phillips BA, Bø K, Galea MP. for improving pelvic muscle strength. Int Physiotherapy as an adjunct to prolapse Urogynecol J. 2010;21(4):401-8. surgery: An assessor blinded randomized 261. Due U, Brostrom S, Lose G. Lifestyle advice controlled trial. Neurourology and with or without pelvic floor muscle training for urodynamics. 2010;29:719-25. pelvic organ prolapse: a randomized controlled 271. Pauls RN, Crisp CC, Novicki K, Fellner AN, trial. Int Urogynecol J. 2016;27(4):555-63. Kleeman SD. Impact of physical therapy on 262. Frawley HC, Hagen S, Sherburn M, Neumann quality of life and function after vaginal P, Herbison P, Hay-Smith J, et al. Changes in reconstructive surgery. Female Pelvic Med prolapse following pelvic floor muscle training: Reconstr Surg. 2013;19(5):271-7. A randomised controlled trial. Neurourology 272. Pauls RN, Crisp CC, Novicki K, Fellner AN, and urodynamics. 2012;31(6):938-9. Kleeman SD. Pelvic floor physical therapy: 263. Hagen S, Stark D, Glazener C, Dickson S, impact on quality of life 6 months after vaginal Barry S, Elders A, et al. Individualised pelvic reconstructive surgery. Female Pelvic Med floor muscle training in women with pelvic Reconstr Surg. 2014;20(6):334-41. organ prolapse (POPPY): a multicentre 273. McClurg D. A two group, single-blind, randomised controlled trial. Lancet. randomised controlled study to assess the 2014;383(9919):796-806. feasibility of physiotherapy following surgery for 264. Giraudo D, Beccaria N, Lamberti G. Pelvic floor prolapse to avoid recurrence. muscle training, negative pressure abdominal http://isrctnorg/ISRCTN08203452010. exercise and pelvic organ prolapse symptoms: 274. McClurg D, Hilton P, Dolan L, Monga A, Hagen A randomized clinical trial. Neurourology and S, Frawley H, et al. Pelvic floor muscle training urodynamics. 2011;30(6):1009-11. as an adjunct to prolapse surgery: a

184 randomised feasibility study. International 286. Powers K, Lazarou G, Wang A, LaCombe J, Urogynecology Journal. 2014;25(7):883-91. Bensinger G, Greston WM, et al. Pessary use in advanced pelvic organ prolapse. Int 275. Eftekhar T, Sohrabi M, Haghollahi F, Shariat M, Urogynecol J Pelvic Floor Dysfunct. Miri E. Comparison effect of physiotherapy with 2006;17(2):160-4. surgery on sexual function in patients with pelvic floor disorder: A randomized clinical trial. 287. Clemons JL, Aguilar VC, Sokol ER, Jackson Iran J Reprod Med. 2014;12(1):7-14. ND, Myers DL. Patient characteristics that are associated with continued pessary use versus 276. Bugge C, Williams B, Hagen S, Logan J, surgery after 1 year. Am J Obstet Gynecol. Glazener C, Pringle S, et al. A process for 2004;191(1):159-64. Decision making after Pilot and feasibility Trials (ADePT): development following a feasibility 288. Ko PC, Lo TS. Delayed onset advanced pelvic study of a complex intervention for pelvic organ organ prolapse after pelvic trauma in a prolapse. Trials. 2013;14(353):353. nulliparous young female: case report. Int Urogynecol J. 2011;22(6):757-9. 277. Hagen S. Pessary Plus Physiotherapy for Pelvic Organ Prolapse (PEPPY). 289. Ko PC, Lo TS, Tseng LH, Lin YH, Liang CC, http://ClinicalTrialsgov/show/NCT0113688920 Lee SJ. Use of a pessary in treatment of pelvic 10. organ prolapse: quality of life, compliance, and failure at 1-year follow-up. J Minim Invasive Gy- 278. Cheung RY, Lee JH, Lee LL, Chung TK, Chan necol. 2011;18(1):68-74. SS. Vaginal Pessary in Women With Symptomatic Pelvic Organ Prolapse: A 290. Sullivan SA, Davidson ER, Bretschneider CE, Randomized Controlled Trial. Obstet Gynecol. Liberty AL, Geller EJ. Patient characteristics 2016. associated with treatment choice for pelvic organ prolapse and urinary incontinence. Int 279. Manonai J, Harnsomboon T, Sarit-apirak S, Urogynecol J. 2016;27(5):811-6. Wattanayingcharoenchai R, Chittacharoen A, Suthutvoravut S. Effect of Colpexin Sphere on 291. Ding J, Chen C, Song XC, Zhang L, Deng M, pelvic floor muscle strength and quality of life in Zhu L. Successful use of ring pessary with women with pelvic organ prolapse stage I/II: a support for advanced pelvic organ prolapse. Int randomized controlled trial. Int Urogynecol J. Urogynecol J. 2015;26(10):1517-23. 2012;23(3):307-12. 292. Nemeth Z, Nagy S, Ott J. The cube pessary: an 280. Haylen BT, Maher CF, Barber MD, Camargo S, underestimated treatment option for pelvic Dandolu V, Digesu A, et al. An International organ prolapse? Subjective 1-year outcomes. Urogynecological Association (IUGA) / Int Urogynecol J. 2013;24(10):1695-701. International Continence Society (ICS) Joint Report on the Terminology for Female Pelvic 293. Brazell HD, Patel M, O'Sullivan DM, Mellen C, Organ Prolapse (POP). Neurourology and uro- LaSala CA. The impact of pessary use on dynamics. 2016;35(2):137-68. bowel symptoms: one year outcomes. Female Pelvic Med Reconstr Surg. 2014;20(2):95-8. 281. Oliver R, Thakar R, Sultan AH. The history and usage of the vaginal pessary: a review. Eur J 294. Patel M, Mellen C, O'Sullivan DM, LaSala CA. Obstet Gynecol Reprod Biol. 2011;156(2):125- Impact of pessary use on prolapse symptoms, 30. quality of life, and body image. Am J Obstet Gynecol. 2010;202(5):499 e1-4. 282. Khan AA, Eilber KS, Clemens JQ, Wu N, Pashos CL, Anger JT. Trends in management 295. Ding J, Chen C, Song XC, Zhang L, Deng M, of pelvic organ prolapse among female Zhu L. Changes in Prolapse and Urinary Medicare beneficiaries. Am J Obstet Gynecol. Symptoms After Successful Fitting of a Ring 2015;212(4):463 e1-8. Pessary With Support in Women With Advanced Pelvic Organ Prolapse: A 283. Gorti M, Hudelist G, Simons A. Evaluation of Prospective Study. Urology. 2016;87:70-5. vaginal pessary management: a UK-based survey. J Obstet Gynaecol. 2009;29(2):129-31. 296. Panman CM, Wiegersma M, Kollen BJ, Berger MY, Lisman - van leeuwen Y, Dekker JH. 284. Thakar R, Stanton S. Management of genital Effects of pelvic floormuscle training and prolapse. BMJ. 2002;324(7348):1258-62. pessary treatment in women ≥55 years with an advanced pelvic organ prolapse. Int Uro- 285. Abdool Z, Thakar R, Sultan AH, Oliver RS. gynecol J 2014;25 ((Suppl 1):S1-S240):S79. Prospective evaluation of outcome of vaginal pessaries versus surgery in women with 297. Lone F, Thakar R, Sultan AH. One-year symptomatic pelvic organ prolapse. Int prospective comparison of vaginal pessaries Urogynecol J. 2011;22(3):273-8. and surgery for pelvic organ prolapse using the validated ICIQ-VS and ICIQ-UI (SF)

185 questionnaires. Int Urogynecol J. in men undergoing laparascopic radical 2015;26(9):1305-12. prostatectomy: a randomized control trial. Neurourology and urodynamics. 2015;34:144- 298. Barber MD, Walters MD, Cundiff GW, Group 50. PT. Responsiveness of the Pelvic Floor Distress Inventory (PFDI) and Pelvic Floor 308. Fode M, Sonksen J. Penile vibratory Impact Questionnaire (PFIQ) in women stimulation in the treatment of post- undergoing vaginal surgery and pessary prostatectomy incontinence: a randomized pilot treatment for pelvic organ prolapse. Am J study. Neurourology & Urodynamics. Obstet Gynecol. 2006;194(5):1492-8. 2015;34(2):117-22. 299. Cundiff GW, Amundsen CL, Bent AE, Coates 309. Geraerts I, Van Poppel H, Devoogdt N, Joniau KW, Schaffer JI, Strohbehn K, et al. The S, Van Cleynenbreugel B, De Groef A, et al. PESSRI study: symptom relief outcomes of a Influence of preoperative and postoperative randomized crossover trial of the ring and pelvic floor muscle training (PFMT)compared Gellhorn pessaries. Am J Obstet Gynecol. with postoperative PFMT on urinary 2007;196(4):405 e1-8. incontinence after radical prostatectomy: a randomized controlled trial. European urology. 300. Collins S, Beigi R, Mellen C, O'Sullivan D, 2013;64:766-72. Tulikangas P. The effect of pessaries on the vaginal microenvironment. Am J Obstet 310. Hou CP, Chen TY, Chang CC, Lin YH, Chang Gynecol. 2015;212(1):60 e1-6. PL, Chen CL, et al. Use of the SF-36 quality of life scale to assess the effect of pelvic floor 301. Vierhout ME. The use of pessaries in vaginal muscle exercise on aging males who received prolapse. Eur J Obstet Gynecol Reprod Biol. transurethral prostate surgery. Clinical 2004;117(1):4-9. Interventions in Aging. 2013;8:667-73. 302. Shamliyan TA, Wyman JF, Ping R, Wilt TJ, 311. Kakihara C, Sens Y, Ferreira U. Effect of Kane RL. Male urinary incontinence: functional training for the pelvic floor muscles prevalence, risk factors, and preventative with or without electrical stimulation in cases of interventions. Reviews in Urology. urinary incontinence following radical 2009;11(3):145-66. prostatectomy. Revista Brasileira de 303. Milsom I, Altman D, Cartwright R, Lapitan MC, Fisioterapia. 2007;11(6):481-6. Nelson R, Sillen U, et al. Epidemiology of 312. Kongtragul J, Tukhanon W, Tudpudsa P, urinary incontinence (UI) and other lower Suedee K, Tienchai S, Leewansangtong S, et urinary tract symptoms (LUTS), pelvic organ al. Effects of adding concentration therapy to prolapse (POP) and anal incontinence (AI). In: Kegel exercise to improve continence after Abrams P, Cardozo L, Khoury S, editors. 5th radical prostatectomy, randomized control. International Consultation on Incontinence. 5th: Journal of the Medical Association of Thailand. ICUD-EAU; 2013. 2014;97(5):513-7. 304. Anderson CA, Omar MI, Campbell SE, Hunter 313. Laurienzo CE, Sacomani CAR, Rodrigues TR, KF, Cody JD, Glazener CMA. Conservative de Cassio Zequi SGGCLA. Results of management for postprostatectomy urinary preoperative electrical stimulation of pelvic floor incontinence. Cochrane Database of muscles in the continence status following Systematic Reviews. 2015;1(Art. No.: radical retropubic prostatectomy. International CD001843). Brazilian Journal of Urology. 2013;39:182-8. 305. Ahmed MT, Mohammed AH, Mansour AA. 314. Ocampo-Trujillo A, Carbonell-Gonzalez J, Effect of pelvic floor electrical stimulation and Martinez-Blanco A, Diaz-Hung A, Munoz CA, biofeedback on the recovery of urinary Ramirez-Velez R. Preoperative training continence after radical prostatectomy. Turkish induces changes in the histomorphometry and Journal of Physical Medicine and muscle function of the pelvic floor in patients Rehabilitation. 2012;58(3):171-7. with indication of radical prostatectomy. Actas 306. Burkert S, Scholz U, Gralla O, Roigas J, Knoll Urologicas Espanolas. 2014;38(6):378-84. N. Dyadic planning of health behavior change 315. Tienforti D, Sacco E, Marangi F, D’Addessi A, after prostatectomy: a randomized controlled Racioppi M, Galino G, et al. Efficacy of an trial planning intervention. Social Science and assisted low intensity programme of Medicine. 2011;73:783-92. perioperative pelvic floor muscle training in 307. Dijkstra-Eshuis J, Van den Bos TWL, Splinter improving the recovery of continence after R, Bevers RFM, Zonneveld WCG, Putter H, et radical prostatectomy: a randomized controlled al. Effect of preoperative pelvic floor muscle trial. BJU international. 2012;110:1004-11. therapy with biofeedback versus standard care 316. Zopf EM, Bloch W, Machtens S, Zumbe J, on stress urinary incontinence and quality of life Rubben H, Marschner S, et al. Effects of a 15-

186 month supervised exercise program on surface therapeutic electrical stimulation physical and psychological outcomes in (SSTES) on early recovery of urinary prostate cancer patients following incontinence after laparoscopic radical prostatectomy: The ProRehab study. prostatectomy: a prospective study. Integrative Cancer Therapies. 2015;14(5):409- Neurourology and urodynamics. 18. 2011;30(6):889-90. 317. Burgio KL, Goode PS, Johnson TM, 325. Pedriali F, Gomes C, Soares L, Urbano M, Hammontree L, Ouslander JG, Markland AD, et Moreira E, de AS. The efficacy of pilates al. Behavioral vs. drug treatment for overactive compared to pelvic floor muscle training bladder in men: The Motive Trial. Journal of the associated with electrical stimulation in the American Geriatrics Society. recovery of post prostatectomy urinary 2011;59(12):2209-16. incontinence: A randomized controlled trial (Abstract number 306). Neurourology and 318. Serda BC, Marcos-Gragera R. Urinary urodynamics. 2014;33(6):742-3. incontinence and prostate cancer: a progressive rehabilitation program design. 326. Ng SI. A Randomised Controlled Trial Study of Rehabil Nurs. 2014;39(6):271-80. the Efficacy of Intensive Pre-Operative Pelvic Floor Muscle Training to Decrease Post- 319. Baroni M, Lorenzetti R, Renzi C, Brizzi A, Prostatectomy Urinary Incontinence (Abstract Branchini W, Altavilla MG, et al. Approach HTA number OP.4.7Dec.38). International Journal (health technology assessment) to treat urinary of Urology. 2014;21:A169. incontinence after radical prostatectomy (Abstract number 23). Neurourology and urody- 327. Park SW, Park CS, Kim TN, Lee W, Nam JK, namics. 2013;32:S20. Lee SD, et al. The effects of a 12-week’s combined exercise intervention on physical 320. Collado Serra A, Pellicer Cabo M, Ramirez function and mental health after radical Backhaus M, Dominguez-Escrig J, Rubio- prostatectomy in elderly patients with prostate Briones J, Gomez-Ferrer A, et al. Intensive cancer: A prospective, randomized controlled preoperatory Pelvic Floor Muscle Training study. (Abstract 1309). Conference: 2011 reduce duration and severity of stress urinary Annual Meeting of the American Urological incontinence after radical prostatectomy: A Association, AUA Washington, DC United randomized controlled trial (Abstract number States. Journal of Urology. 2011;185(4 SUPPL. 1007). European Urology Supplements. 1):e524. 2013;12(1):e1007-e8. 328. Park SW, Kim TN, Nam JK, Ha HK, Shin DG, 321. Ghanem A, Khallaf M, Assem A, Hassan A. Lee W, et al. Recovery of overall exercise Does preoperative pelvic floor muscle exercise ability, quality of life, and continence after 12 improve post prostatectomy urinary week combined exercise intervention in elderly incontinence (Abstract 695)? . Conference: patients who underwent radical prostatectomy: Annual Meeting of the International Continence A randomized controlled study. Urology. Society ICS, Barcelona, Spain. 2013. 2012;80:299-306. 322. Laurienzo C, Magnabosco W, Jabur F, 329. Glazener C, Boachie C, Buckley B, Cochran C, Gameiro M, Yamamoto H, Guerra R, et al. Post Dorey G, Grant A, et al. Urinary incontinence in prostatectomy urinary incontinence and men after formal one-to-one pelvic floor muscle erectile dysfunction: The role of pelvic floor training following radical prostatectomy or tran- rehabilitation (Abstract number 527). surethral resection of the prostate (MAPS): two Conference: Annual Meeting of the parallel randomised controlled trials. Lancet. International Continence Society, ICS 2012;379(9814):328-37. Montreal, QU, Canada. Neurourology and urodynamics. 2015;34:S449-S50. 330. Glazener C, Boachie C, Buckley B, Cochran C, Dorey G, Grant A, et al. A randomised 323. Martini M, Bernardini S, Blanc E, Piretta K, controlled trial of conservative treatment (pelvic Tappero R. Relationship between integrity of floor muscle training and bladder training) for pelvic floor function and recovery of continence urinary incontinence in men after prostate sur- after laparoscopic prostatectomy and effects of gery (MAPS) (Abstract 200).Conference: Joint preventive pelvic floor muscle training in males Annual Meeting of the International Continence with pelvic floor weakness (Abstract 14). Society, ICS and International Conference: Annual Congress of the Italian Urogynecological Association, IUGA Toronto, Urodynamics Society, Turin, Italy. ON Canada. Neurourology and urodynamics. Neurourology and urodynamics. 2010;29(6):1093-4. 2011;30(SUPPL 1):11-2. 331. Glazener C, Boachie C, Buckley B, Cochran C, 324. Morihoro N, Masatsugu I, Shinji K, Kenichi T, Dorey G, Grant A, et al. Conservative treatment Kazumasa M, Shiro B. Effectiveness of sacral for urinary incontinence in Men After Prostate

187 Surgery (MAPS): two parallel randomised con- technique for curing detrusor overactivity: trolled trials. Health technology assessment Stoller afferent neurostimulation. Scandinavian (Winchester, England). 2011;15(24):1-296. journal of urology and nephrology. 2005;39(3):230. 332. Voorham-van der Zalm PJ, Stoetman A, Putter H, Bevers R, Pelger R. Effect of preoperative 341. Peters KM, MacDiarmid SA, Wooldridge LS, pelvic 1226 floor physiotherapy versus Leong FC, Shobeiri SA, Rovner ES, et al. standard care on incontinence in men Randomised trial of percutaneous tibial nerve undergoing radical laparoscopic stimulation versus extended release tolteridine: prostatectomy: an ongoing study. (Abstract Results from the overactive bladder innovative number 590). Proceedings of the Joint Meeting thereapy trial. Journal of Urology. of the International Continence Society (ICS) 2009;182(3):1055-61. and the International Urogynecological As- sociation; August 23-27, 2010; Toronto, 342. Finazzi Agrò E, Campagna A, Sciobica F, Petta Canada 2010. F, Germani S, Zuccalà A, et al. Posterior tibial nerve stimulation: is the once-a-week protocol 333. Breyer BN, Phelan S, Hogan PE, Rosen RC, the best option? Minerva urologica e Kitabchi AE, Wing RR, et al. Intensive Lifestyle nefrologica = The Italian journal of urology and Intervention Reduces Urinary Incontinence in nephrology. 2005;57(2):119. Overweight/Obese Men with Type 2 Diabetes: Results from the Look AHEAD Trial. Journal of 343. Seth J, Gonzales G, Haslam C, Ochulor J, Urology. 2014;192(1):144-9. Elneil S, Vashisht A, editors. Single centre randomised pilot study of two regimens 334. Sønksen J, Ohl DA, Wedemeyer G. Sphincteric (30mins daily or 30 mins weekly for 12 weeks) events during penile vibratory ejaculation and of Transcutaneous Tibial Nerve Stimulation electroejaculation in men with spinal cord using an adhesive skin patch for the treatment injuries. Journal of Urology. 2001;165:426-9. of Overactive Bladder (OAB) Symptoms Proceedings of the 44th Annual Meeting of the 335. Laessøe L, Sønksen J, Bagi P, Biering- International Continence Society (ICS), 2014 Sørensen F, Ohl DA, McGuire EJ, et al. Effects Oct 20-24; 2014. of ejaculation by penile vibratory stimulation on bladder capacity in men with spinal cord 344. Monteiro ES, Coin de Carvalho LB, Fukujima lesions. Journal of Urology. 2003;169:2216-9. MM, Lora MI, Fernandes do Prado F. Electrical Stimulation of the Posterior Tibialis Nerve 336. Sønksen J, Ohl DA, Bonde B, Laessøe L, Improves Symptoms of Poststroke Neurogenic McGuire EJ. Transcutaneous mechanical Overactive Bladder in Men: A Randomized nerve stimulation using perineal vibration: A Controlled Trial. Urology. 2014;84:509-14. novel method for the treatment of female stress urinary incontinence. Journal of Urology. 345. Peters KM, Carrico DJ, Wooldridge LS, Miller 2007;178(5):2025-8. CJ, MacDiarmid SA. Percutaneous tibial nerve stimulation for the long-term treatment of 337. Peters KM, Carrico DJ, Perez-Marrero R, Khan overactive bladder: 3-year results of the STEP AU, Wooldridge LS, Davis GL, et al. study. The Journal of urology. Randomized trial of percutaneous tibial nerve 2013;189(6):2194. stimulation versus Sham efficacy in the treatment of overactive bladder syndrome: 346. MacDiarmid SA, Peters KM, Shobeiri SA, results from the SUmiT trial. The Journal of Wooldridge LS, Rovner ES, Leong FC, et al. urology. 2010;183(4):1438. Long-term durability of percutaneous tibial nerve stimulation for the treatment of 338. Booth J, Hagen S, McClurg D, Norton C, overactive bladder. The Journal of urology. MacInnes C, Collins B, et al. A feasibility study 2010;183(1):234. of transcutaneous posterior tibial nerve stimulation for bladder and bowel dysfunction in 347. van Balken MR, Vergunst H, Bemelmans BLH. elderly adults in residential care. Journal Of Prognostic Factors for Successful The American Medical Directors Association. Percutaneous Tibial Nerve Stimulation. 2013;14(4):270. European Urology. 2006;49:360-5. 339. Vohra AK, Britchford A, Neale E, Husain I, Waterfall N, editors. The efficacy of stoller afferent nerve stimulation in frequency/urgency syndrome: A randomised control trial. Proceedings of the International Continence Society (ICS) 32nd Annual Meeting, Aug 28-30; 2002. 340. Karademir K, Baykal K, Sen B, Senkul T, Iseri C, Erden D. A peripheric neuromodulation

188