Does It Work in the Long Term?—A Systematic Review on Pelvic Floor Muscle Training for Female Stress Urinary Incontinence

BØ K and Hilde G: Neurol and Urodynam 32:215-223 (2013)

Ann Dunbar PT, DPT, MS, WCS

April 9, 2014

Primary Aim:  To present long-term results of pelvic floor muscle training (PFMT) with or without biofeedback on stress urinary incontinence (SUI) and mixed urinary incontinence (MUI) with predominant SUI symptoms (including both RCTs and pre- and post-evaluation studies)

Study Design:  Systematic Review

Methods:

• Inclusion criteria: Pre-and post-test design, non-RCT and RCT’s using PFMT with or without biofeedback to treat SUI and SUI predominant MUI
• Authors completed a computerized search on PubMed
1. Pelvic floor AND (training OR exercise OR physical  activity)AND (urinary incontinence  OR stress urinary incontinence) AND (follow-up OR long-term)
2. Limits: humans, female, clinical trial, English, only adult subjects
• Authors also completed a computerized search on the PEDro database, abstracts from Intern. Continence Soc., and Internat. Assoc. of Urogynecology from 1990 to present
• Authors also did hand search of reference lists of studies eligible for inclusion in former systematic reviews and guidelines
• Long-term defined as time frame greater than or equal to 1 year after cessation of the original PFMT intervention
• Information was extracted by 2 researchers according to preset criteria (see topic headings on Table I
• Outcomes: Primary was surgery rate during the follow-up period; Secondary was report of cure/maintenance of improvement
• PRISMA statement (evidence-based minimum set of items for reporting in meta-analyses and systematic reviews ) was followed
• PEDRo scores for the studies were  used if available; if unavailable, the reviewing authors scored the study using the PEDro scoring system (10 point system)

Results

• PubMed search identified 44 studies with 17 of those being long term and fulfilling inclusion criteria;  hand search identified 2 additional studies
• The 19 studies included 1,141 women (see Table I)
• Three research groups completed a second, long-term follow up from their original study and both studies are reported (Cammu et al, BØ et al, Kondo et al)
• Five studies were excluded due to follow-up periods less than 1 year
• Nine studies were based on original pre- and post-study designs (noncontrolled)
• Nine studies were follow-up studies of original RCTs
• One follow-up study based on non-randomized design with a control group
• Mean PEDro score for RCTs was 5.1
• Variables differed between the studies: 8 compared different methods or intensities of PFMT;  1 RCT and 1 non-randomized study compared PFMT with untreated control groups; 1 RCT compared PFMT with surgery.
• Cross-overs: in 2 RCTs with untreated control groups, the control group crossed over to PFMT after short treatment period and analysis of long-term results couldn’t be completed.
• Follow up periods: ranged from 1 to 15 year.
• Incentives for continuing training in the follow-up period:  None offered in all but 2 studies
• Loss to follow up: reported in most;  range was 0 to 39%.
• Adherence: range between 10 and 70% of women doing PFMT;  6/17 studies did not report adherence to PFMT at follow-up or during the follow-up period.
• Outcomes: most used self-report questionnaires; 8 studies interviewed subjects and/or used various pad tests, tested PFM function or applied urodynamic assessments.
1. Outcome tools:  8 studies used instruments tested for validity and reliability
2. Surgery: 12 studies reported surgery rates occurring in the follow-up period and they ranged from 4.9% at 28 mo. to 58% after 4 to 8 years; in 2 studies with longest follow-up, surgery rates were 8% at 10 years and 50% at 15 years
• No meta-analysis was completed due to high heterogeneity in multiple variables
• Five of the 19 studies stated initial success rate was maintained at follow-up.
• Long-term outcome based on short-term success reported in 7 studies.
• All studies reported that effect was better maintained in responders than in non-responders to the original program.
• In Kondo et al, increase in m strength during original program was only reported parameter predicting positive long-term effect 8 years later.

Discussion

• Though 19 long term studies were found, not possible to make comparisons because original short-term studies too heterogeneous; length of follow-up period with long term studies, use of different outcome measures, co-interventions and losses to follow-up, introduce more variability “a double heterogeneity problem.”
• Since it is not possible to blind subjects and therapists during PFMT, authors state that highest PEDro score possible for these studies would be 8 therefore a  range of 4 to 6 would be considered moderate (authors consider this range would make a meaningful meta-analysis).
• Considering only 2 studies provided exercise classes during follow-up or gave specific advice on exercise, some of the studies demonstrated surprisingly good results assessed by surgery rates or self-report.
• Authors recommend for future that same outcome measures be used at both short and long term; use only measures that have been tested and found to be reliable, valid, responsive
• Authors chose surgery rate as outcome since PFMT for SUI is considered treatment to delay or avoid surgery
1. Short-term follow-up: surgery was clearly best short-term effect; however, short term effect of both surgery and PFMT were maintained after 4 to 8 years.
2. At 15 year follow-up, 50% of women in both the originally randomized groups had undergone surgery; more women in the less intensive training group had surgery after 5 years
3. At 15 year follow up, no differences in reported amount or frequency of leakage between operated vs. non-operated women and women who underwent surgery reported significantly more severe leakage and were more bothered by UI with ADLs
4. Authors conclude because of selection biases for surgery and variability between hospitals and countries as to when surgery is offered, this optiona difficult outcome to analyze and compare between studies; They suggest assessment of actual leakage (pad test and 3 day report of UI) and assessment of perceived QOL and impact on life for future studies
• Studies reported that the effect of intervention was better maintained with the short term responders than non-responders
• Go back to main question:  Can long-term outcome be expected after cessation of active PFMT intervention?
1. Consider in general, strength gain declines at slower rate than the rate with which it  increases (5 to 10% loss of m. strength/wk after training cessation)
2. Losses vary according to age as well and majority of loss is from weeks 12 to 31 after training cessation
3. To maintain strength gains, intensity needs to be maintained but volume and frequency of training can be reduced from training period (1 to 2 days/wk for persons already doing resistance training program)
• No studies have investigated the number of PFM contractions needed to maintain PFM strength after cessation of formal training.
1. One study  found women with MUI more likely to lose effect and women with SUI had best long-term effect but only 39% were exercising daily or ‘when needed’ (Lagro-Janssen at al 1998)
2. Other studies suggest conscious PFM pre-contraction before cough or increase in IAP may help maintain long-term effect 
• Authors discuss issue of exercise adherence
• Alewijnse et al 2002 found the following to predict adherence at 50%: 1- high short term adherence levels 2- positive intention to follow through  3—frequent episodes of UI before and after training period  4—positive self-efficacy expectations

Strengths and Weaknesses

Comprehensive review of literature found in extensive search

Low risk for publication bias due to published high quality SR of short-term effects

Limitations: quality of individual studies

Clinical Analysis

1-      Study findings suggest that benefits of exercise are better maintained in those who were responders. What do you feel is important to assure women have a good response with their therapy intervention?

2-      Will any of these findings change your practice?  What exercise prescription do you give at discharge for the long term?

3-      Authors suggest including PFMT in general fitness classes though state this will only reach the women highly motivated to exercise? Any other thoughts