• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.

Hartmann KE, McPheeters ML, Biller DH, et al. Treatment of Overactive Bladder in Women. Rockville (MD): Agency for Healthcare Research and Quality (US); 2009 Aug. (Evidence Reports/Technology Assessments, No. 187.)

Chapter 3Results

KQ1. Prevalence and Incidence of Overactive Bladder

To understand the epidemiology of OAB and associated symptoms (frequency, urgency, and urge incontinence), we sought publications that provided estimates of prevalence, which is the proportion of the population with the condition, or those that examined the proportion with onset of new symptoms over time, which is termed incidence. The strongest estimates come from a well-specified population-base that allows sampling to generate a representative group.

We applied operational criteria to select those studies best suited to estimate prevalence and incidence. The key criterion for inclusion was that the study authors specified a population base a priori from which a sample of individuals was drawn as a representative selection of individuals to estimate the true proportion in the larger population. The population was required to be clearly defined and capable of being enumerated such as the population of a region, the participants in a health plan or individuals on voter registration roles. This “roster” defines the sampling frame from which stratified samples are drawn. Often the research team has an explicit statistical goal of randomly selecting participants with aggregate characteristics (range of age, race/ethnicity, income level, etc.) that approximate the entire population of a city, region, or country. Common approaches include random digit dialing within a specific geographical region, or mailings of questionnaires to a sample of individuals who are in a pool of registered voters, registered residents, or on the patient panels of a national health systems or large healthcare organizations. Outside the United States, household canvassing and administered interviews were also common.

The characteristic of being able to define a sampling frame from which participants were subsequently drawn was a required characteristic for inclusion in the prevalence and incidence summary. Other approaches to the study of large groups, such as enrolling individuals who are presenting for clinical care, or who have just had a health event like childbirth or surgery, can be informative, however, such studies do not by virtue of their design, generate participant pools who are representative of a larger population. They represent those with access to care, who have sought care for a problem visit, chronic condition, or preventive care, or who have a medical condition or experience in common. Such designs may also reflect characteristics of the site of care at which the cohort was recruited, such as a high proportion of indigent patients or specialty referral patients. For this reason, cohort studies and clinical samples, without a population-based sampling frame were excluded in order to emphasize the broadest picture of the epidemiology and natural history of OAB is as possible.

Prevalence of Overactive Bladder

Using this approach, we identified 75 publications,2–4, 10–81 from 60 distinct study populations. Detailed summaries for all studies are included in Appendix C and a summary is provided in Table 6.

Table 6. Study characteristics of the prevalence and incidence literature.

Table 6

Study characteristics of the prevalence and incidence literature.

Fifteen studies were conducted in United States populations; 24 in European populations; 13, Asian; and 8 other countries. Thirty-eight of these studies appeared in print after the 2002 International Continence Society revised definitions and 37 percent of these reported specifically about incidence or prevalence of OAB. One study prior to the consensus definitions used fully comparable definitions and the term “overactive bladder”.

A total of 15 studies provided information about OAB prevalence (Table 7). These studies included a total of 64,528 women in 16 distinct populations. The majority of respondents completed questionnaires returned by mail. The highest estimates of OAB prevalence was 31.3 percent in a telephone questionnaire study conducted in Korea in which the average age of participants was 59 years; the lowest was 7.7 percent in a mailed questionnaire conducted in the United Kingdom among women of nearly identical average age. Across all studies the weighted average prevalence of OAB was 13.7 percent. Excluding the highest and lowest estimates, an estimated 14.8 percent of women meet criteria for OAB, with 8.0 percent of those surveyed having OAB with a component of urge incontinence. Combined estimates from the two United States populations are similar: 14.7 percent with OAB and 11.3 percent with a component of urge incontinence.

Table 7. Prevalence of OAB.

Table 7

Prevalence of OAB.

No clear pattern of higher or lower estimates for prevalence of OAB was associated with survey response rates. The direction of bias is therefore difficult to estimate: researchers have noted both that those with symptoms may be more likely to be interested in the topic and to respond and that the social stigma or embarrassment associated with bladder control symptoms may prompt under-response. In each case in which the authors addressed non-response, they report the demographic characteristics of those who did not respond were similar to those who did with the exception of several authors who noted modest under-representation of the very oldest residents.

Age and prevalence of OAB. The relationship between OAB and age was fairly uniform across studies with a trend to increase with age.2, 3, 12–14, 38, 42 As discussed below, this increase appears to be more notable for OAB than UUI. Because OAB criteria can be met with urgency combined with incontinence, frequency, or nocturia, the “amplification” of age effects for OAB risk may in part be related to well-documented increases in stress urinary incontinence and therefore also mixed stress and urge incontinence with age. Prevalence of OAB for women in their 20s was in the range of 4.6 to 5.9 percent 2, 3, 11, 13 while uniformly double digits, 11.7 to 19 percent 2, 3, 11, 13 for women older than 60. Several researchers noted a threshold effect such that prevalence was not statistically different until an inflection in the 60s or 70s.12, 14, 38, 42 Few data were provided, however discussion materials often noted that all component symptoms: urgency, frequency, nocturia, and urge and mixed incontinence, increased with age therefore contributing to the rise in OAB.

Prevalence of Urge Urinary Incontinence

A larger number of studies (n=48) examined urge urinary incontinence as the primary prevalence estimate of interest. Twenty-six distinct study populations, with 36 publications, appeared after the ICS 2002 consensus definitions.2, 3, 10–14, 17–19, 21, 25, 27, 29, 31–38, 70, 72, 75, 80 These are presented here in greater detail because the definitions used are more similar to those that define “OAB-wet” and therefore are more likely to have comparability in the measures. Across all these populations, prevalence of urge incontinence ranged from a low of 1.5 percent in a large European study of all adult women from ages 18 to 70, 18 up to a high estimate of 22.0 percent in a United States mall-based consumer survey36 and 26.4 percent among a household sample of Jordanian women age 50 to 65.35 Each of the latter had somewhat ambiguous survey items. Average urge incontinence prevalence by region was:

United States: 9.7 percent among 37,596 respondents from nine populations 2–4, 10, 11, 13, 17, 19, 25, 27, 32, 36, 40, 42, 72

Europe: 10.6 percent among 68,051 respondents from seven studies including nine countries 18, 37, 43, 45, 48, 75, 81

Asia: 9.6 percent among 14,537 respondents from five populations in three countries 14, 21, 29, 31, 33, 34, 41, 47, 80

Other: 12.5 percent among 6,219 respondents in five populations from five countries 12, 35, 38, 49, 70

Though more than half of studies did not report the frequency of urge incontinence episodes required to meet the case definition (e.g., weekly, monthly), or used nonspecific terms such as “mostly” and “sometimes” in definition, these estimates are concordant with the range of estimates for the prevalence of OAB with urge incontinence features, suggesting that the measurement instruments captured similar features. These estimates are consistent with the AHRQ report on the Prevention of Fecal and Urinary Incontinence in Adults, which reported prevalence of UUI in community dwelling adults as increasing “from 5 percent in younger women to 10 percent in women 45–64 (32 studies), and to 12 percent in women older than 65 years (28 studies)”.196

In the nine studies in which the interval of occurrence required to meet the definition of urge incontinence was specified, there was not a clear pattern of relationship between length of the interval and proportion of women classified as having urge incontinence.10, 17, 27, 31, 34, 37, 38, 72, 80 Overall the average prevalence of UUI across these nine studies was 8.9 percent. The two studies that required weekly symptoms reported a prevalence of 7.0 and 6.5 percent (in populations which had an average age of 61, and an age range from 35 to 75 without a mean reported, respectively).21, 31, 38 The average across six studies that reflect a wide age range and required at least monthly episodes was 9.4 percent, and the sole study reporting any urge incontinence within the year was 7.9 percent with an average respondent age of 51.72 This lack of a pattern associated with the operational definition of frequency of urge incontinence episodes may result from variation in the methods of measurement or characteristics of the respondents that obscures any trend, or, more likely given large population-based samples, similarity in estimates may reflect that most women who are affected have fairly frequent symptoms and are detected and properly classified regardless of interval definition required.

Other Measures

Urgency symptoms. Individually, the symptoms of urgency and frequency are common. The range of prevalence for urgency symptoms (in post ICS studies) was from 8.0 percent in a young population of pre-menopausal Indian women,50 to the highest estimate of 45.4 percent among an even younger population of European women (mean age 34).48 Despite this wide range, which may reflect varied operational definitions used by researchers, nine of ten estimates for the prevalence of urgency are 10 percent or above.12, 14, 18, 34, 38, 47, 48, 50, 73

Frequency. Thirteen studies with fourteen groups reported on frequency. Most defined frequency as more than eight voids per day (or more). The range of estimates for women with daily frequency was 5.2 percent among Thai women ages 20 to 59,20 to 34 percent among Danish women ages 20 to 45,48 and 49 percent among Japanese women with an average age of 61.31 The average proportion of women with frequent voiding across study groups was 16.0 percent, with five estimates between 5 and 15 percent,12, 18, 20, 45, 50 and four estimates between 15 and 25 percent.14, 34, 38, 47

Age as a predictor of UUI. The relationship between urge urinary incontinence and increasing age was less consistent across studies than that for OAB and age, with some authors reporting nearly identical prevalence across all age brackets and no statistical trend15, 53 and others noting increases parallel to OAB,11, 12, 14, 42 more modest increases64 or threshold ages as in OAB which are inflection points at which prevalence was meaningfully increased, typically among the oldest age strata: >60,11 >65,42 >70,55 and >75.38 In aggregate, the trend was less pronounced or not apparent for a strong and continuous influence of age on prevalence of UUI. Thresholds in older age seem more likely with risk being similar across wide ranges of younger ages. We must also note that such effects may not be results of age per se but of comorbidities and medication use that change with age.

Other predictors. Other factors noted across studies were the influence of race and ethnicity in United States populations. Three studies (one of high quality and two of fair quality), several with more than one related publication, documented statistical association showing black women were more likely than white to have urge urinary incontinence (while noting higher risk of mixed and stress among white women compared to black).11, 17, 25, 197 Another high quality United States study found no difference with adjusted estimates of prevalence of 3.5 percent among black women and 3.6 percent among white women.10

Incidence of Overactive Bladder and Urge Urinary Incontinence

Ten studies provide incidence data. Three investigated the occurrence of new OAB symptoms over varied periods of time.22, 26, 39 Seven investigated the onset of urge urinary incontinence.15, 17, 49, 64, 74, 77, 81 Two provide information about resolution of symptoms among those with symptoms at baseline.64, 77

Estimates for annual incidence of OAB ranged from 2.6 to 143 cases per thousand. The lowest estimates came from a population-based estimate using the national health services database of the United Kingdom and including all adults; the highest is from the top age bracket ( 80 year) in another UK study.39 In that study, by decade beginning at 40, the annual incidence was 78, 65, 100, 117 and 143 per 1,000; the remaining report had participants with a mean of 59 years and an estimate of 54 new cases per 1,000 women per year.39

The largest study to address urge urinary incontinence was conducted in the United States within the Nurses Health Study. They report followup from 64,650 respondents with two-year data. In each five year age bracket from 35 to 55 years of age, the annual incidence among those without symptoms at baseline was 4 per 1,000 with the exception of the 46 to 50 category at 5 per 1,000.15 A separate analysis within the Nurses Health Study, using supplemental data collection to gather additional details, reported two-year incidence of urge incontinence of 14 per 1,000, which annualized to approximately 7 new cases per 1,000 women per year.69, 74 A six state urban sample that recruited 16,065 women and followed a subset of 3,032 for five years, reported 5-year incidence of urge incontinence of 15.9 percent or 3 cases per thousand per year.17 Work in Finland in a much smaller population sample of adults who were in their 50s and 60s at baseline and followed into 60s and 70s report the equivalent of 1.1 case per 1,000 per year, and if mortality is taken into account, 1.7 cases per year. A single study in Southern Australia, reports estimates that are meaningfully higher than these. Defining urge incontinence as that which occurs at least occasionally, and without providing a specific definition of how urge was queried, they report annual incidence of 226 cases per thousand. Of note the entire study population was 70 or older and no information about adjustment for competing morbidity is provided.49

A study of 2,025 women older than 65 who lived in rural Iowa provides additional information. This interview-based study was conducted prior to ICS consensus definitions, the research team used the index item “How often do you have difficulty holding your urine until you can get to a toilet?”, classifying those who answered “never” and “hardly ever” as free of urge incontinence. Estimated 3-year incidence of urge incontinence, over two rounds of followup was 20.4 and 24 percent, which annualizes to 6.8 to 8.0 per 1,000 among older women.64

The authors also report on remission: among those with UUI, 31.7 to 34.9 percent had resolution of their symptoms over the 3-year followup windows.64 Other research in a single United States county, among women 60 and older, found annual incidence of 17 per 1000, with 22.7 percent of women with urge incontinence symptoms reporting resolution within a year.77

KQ2: Outcomes of Treatment of OAB

This section presents results of our literature search and findings about outcomes of pharmacologic treatments for OAB. We review the basic mechanism of action of the family of medications classified as antimuscarinic agents. Six specific agents are available to United States practitioners: oxybutynin, tolterodine, fesoterodine, solifenacin, darifenacin, and trospium. Because a number of the studies are dose ranging and safety studies, we have included doses and preparations that may not be available. We also summarize studies of estrogen treatment for OAB. Each pharmacologic agent is presented individually with a thorough description of the content of the literature followed by the findings from trials and information from cohorts and case series, when those studies provided additional information beyond that provided by trials. Side effect and harms of treatment are reviewed together at the end of this section.

Pharmacologic Treatments

Pharmacologic treatments for OAB include antimuscarinic agents which have differing affinities for multiple subtypes of muscarinic receptors found both in the bladder as well as throughout the body. These agents generally bind to muscarinic receptors on the bladder muscle blocking the input required for contraction of the muscle. In short, such drugs prevent or decrease the intensity of involuntary bladder contractions.

Because muscarinic receptors are present throughout the body, nonselective agents affect other processes explaining the occurrence of side effects such as dry mouth (reduced action on salivation), constipation (slowing gut contractions), dry eyes (affecting tear ducts), and altered cognition (affecting central nervous system). In an attempt to decrease the effect on other organs and improve tolerability, the development of agents purported to be more specific in targeting subtypes of muscarinic receptors found in the bladder has been undertaken.

Medication is often initiated with the lowest dose of an agent and adjusted to the desired clinical effect while minimizing adverse effects. The two most frequently prescribed pharmacologic treatments for OAB in the United States are oxybutynin and tolterodine. Oxybutynin is available in an oral immediate-release (IR) and an extended-release (ER) form, a transdermal patch, and a topical gel, approved by the FDA in January 2009. Tolterodine is also available in both immediate and extended-release pills. Fesoterodine was approved by the FDA in October 2008. Fesoterodine is a first order metabolite of tolterodine with similar selectivity. Newer agents available in the United States include solifenacin, darifenacin, and trospium, which is available in both immediate and extended-release formulations.

We reviewed 110 studies,82–96, 98–109, 140–172, 174, 198–255 of which four were good quality, 75 fair and 31 poor. This section of the report presents summary data for each pharmacologic treatment, including varied doses, intervals, and delivery mechanism. Direct comparisons across or among types of treatment (including pharmacologic, behavioral and procedural, and others) are presented in KQ3. Additional information is also provided in that section about comparisons across extended release versus immediate release. Detailed information on all studies relating to pharmacologic treatment of OAB can be found in evidence tables in Appendix C.

The summary tables included here compile unique arms of randomized trials that did not allow dose adjustment within the arm. Each drug and dose combination, as well as the related placebo arms, is presented here if data for the outcome in the summary table was provided in the publication. When necessary, weekly or monthly measures were converted to daily in order to allow ready assessment of the baseline similarity, outcomes, and effect size across studies. Each of these pharmacologic agents has been shown to be statistically superior to placebo for some facet of OAB symptoms. Complete details of measurement approach and statistical comparisons are available in Appendix C; comparisons among doses of the same drug are discussed in this section, and direct comparisons between or among drugs are presented with KQ3.

Treatment with oxybutynin.

Content of the literature. We identified 13 randomized controlled studies of oxybutynin for treatment of OAB. A total of 22 study arms compared oxybutynin at various doses and intervals, and included five placebo arms (Tables 8 and 9). Most participants were recruited from non-primary care populations with seven studies performed in the United States,82–88 three in Europe,89–91 and one each in Japan,92 Taiwan,93 and South Korea.94 These studies included a total of 2,575 women in treatment arms, and 383 women in the placebo arms. Participants had an average age of 59.3 and 60.9, in the treatment and placebo groups, respectively.

Table 8. RCT arms for oxybutynin chloride effect on urge incontinence.

Table 8

RCT arms for oxybutynin chloride effect on urge incontinence.

Table 9. RCT arms for oxybutynin chloride effect on voids per day.

Table 9

RCT arms for oxybutynin chloride effect on voids per day.

Outcomes of treatment. Baseline numbers of incontinence episodes per day in the oxybutynin treatment groups ranged from 1.0 to 7.2, and in the placebo arms from 0.0 to 3.0. Because many trials were drug to drug comparisons, this difference in range does not reflect marked differences between placebo and comparison groups within studies, rather the higher severity of symptoms in the drug to drug comparison studies which were more likely to include more individuals with multiple urge incontinence episodes per day. Study participants using oral oxybutynin achieved a reduction in mean episodes of urge urinary incontinence of between 0.0 and 4.6 over the course of placebo-controlled studies that ranged in followup from 2 to 52 weeks. Of the formulations studied, 10 mg per day appeared to achieve the greatest reduction with a range of 2.9 to 7.2

(Tables 8 and 9). However, all estimates in this section of the report should be assessed with reference to the placebo arms, in which participants also reported reductions in incontinence episodes. In oxybutynin trials, in the placebo arms, reduction in UUI episodes ranged from 0 to 1.4.

Oxybutynin was also associated with a decrease in voids per day in both oral and transdermal formulations. Baseline voids per day in drug and placebo arms were similar (7.2 to 11.7 overall). Voids per day were lowered by 1.8 to 2.3 in the transdermal arms, 0.7 to 4.2 in oral oxybutynin IR, 3.4 to 4.1 in oral oxybutynin ER and 0.8 to 1.7 in the placebo arms.

Meta-analysis was possible for immediate release but not extended release. Immediate release reduced UUI by 1.49 episodes per day (95 percent CI: 1.18, 1.80); and voids by 2.18 episodes per day (95 percent CI: 1.75, 2.61).

Quality of life outcomes in these studies were measured with several validated tools, including the Kings Health Questionnaire, Incontinence Inventory Questionnaire (IIQ), and Urinary Distress Inventory (UDI) (Table 18). In all studies, statistically significant improvements were observed relative to baseline, and in drug groups compared to placebo.85, 86, 88, 92, 93 The MATRIX study assessed HRQoL in patients treated with transdermal oxybutynin.245, 246 This patient population included 2,508 women with an average age of 62.5 years, and followed patients for a minimum of six months. Sand and colleagues used the KHQ and the PPBC to assess general quality of life improvement, and reported improvement in nine of ten domains versus baseline (p<0.001).245 A second paper by Sand and colleagues employed the Beck-Depression Inventory (BDI-II) and the KHQ, and demonstrated improvement in embarrassment scores, effect on sex life, and relationships with partners, all p <0.001. The BDI-II also showed improvement in interest in sex from baseline to study end (p<0.001).246

Table 18. Effect on quality of life and satisfaction for pharmacologic treatment.

Table 18

Effect on quality of life and satisfaction for pharmacologic treatment.

Treatment with tolterodine.

Content of the literature. We included 20 RCTs83, 84, 86, 88, 89, 91, 94, 96, 97, 99, 140, 141, 154, 162, 168, 209, 223, 233, 240, 241 and 11 prospective cohort studies157, 160, 166, 169, 172, 212, 224, 227, 232, 239, 243 of tolterodine. Among the RCTs were 30 tolterodine arms and 14 placebo arms (Tables 10 and 11). Most were multinational studies conducted at centers in Europe, the United States, Australia, and Asia. A total of 6,746 women were in the treatment arms, with 3,298 women in the placebo arms. The average ages were 58.1 and 59.9, respectively.

Table 10. RCT arms for tolterodine tartrate effect on urge incontinence.

Table 10

RCT arms for tolterodine tartrate effect on urge incontinence.

Table 11. RCT arms for tolterodine tartrate effect on voids per day.

Table 11

RCT arms for tolterodine tartrate effect on voids per day.

Outcomes. At baseline, women reported between 1.6 and 5.0 episodes of urge urinary incontinence per day. Treatment with 4 mg of drug once a day reduced episodes of urge in continence an average of 0.9 to 3.7 episodes each day, compared to a reduction of 1.3 to 2.4 for 2 mg taken twice a day. These doses and intervals are common in clinical practice; other doses and intervals also are included in Tables 10 and 11. Overall, reductions in urge urinary incontinence episodes on active therapy ranged from 0.9 to 3.7, compared to reductions in the placebo arms of 0.6 to 2.1.

With respect to frequency of voiding, tolterodine was associated with 0.9 to 3.6 fewer voids per day in study populations that had baseline voiding frequencies of 7.2 to 13.7. The range of response for placebo was 0.4 to 2.2 fewer voids per day, in study populations that had baseline voiding frequencies of 10.3 to 12.3 per day. Reductions differed by treatment dosage and timing, with a reduction of 1.7 to 3.6 with 4 mg once daily, 0.9 to 1.1 for 4 mg twice daily, and 1.4 to 3.3 for 2 mg twice daily.

Estimates were developed in the meta-analysis for decreases in UUI episodes and voids per day in tolterodine immediate release and tolterodine extended release formulations. Tolterodine immediate release reduced UUI by 1.45 episodes per day (95 percent CI: 1.24, 1.66) and voids per day by 2.19 (95 percent CI: 1.76, 2.61). Tolterodine extended release reduced UUI by 1.75 episodes per day (95 percent CI: 1.65, 1.85) and voids per day by 2.48 (95 percent CI: 1.94, 3.02). The aggregate effect of placebo across all available study arms was a decrease in UUI episodes per day of 1.08 (95 percent CI: 0.86, 1.30) and voids by 1.48 (95 percent CI: 1.19, 1.71) per day.

Two RCTs included urodynamic parameters to assess intermediate effects of tolterodine treatment.221, 238 In a dose-finding study, tolterodine 2 mg twice a day demonstrated statistically significant increase in the volume at first contraction (p=0.03), and an in increase in maximal cystometric capacity following four weeks of treatment compared to baseline urodynamic parameters.221 In contrast, mean volume at first contraction was significantly increased after two weeks of treatment compared to baseline (p=0.046), but no improved outcome effect was seen for maximum cystometric capacity, bladder compliance, or number of detrusor contractions.238

Findings from cohort studies and case series were compatible with these results. Twelve such studies demonstrated improvement in urinary urgency, frequency, and urge incontinence with tolterodine treatment.142, 157, 160, 166, 169, 172, 212, 224, 227, 232, 239, 243, 247

Twelve studies including RCTs, prospective cohorts and extension studies evaluated the effect of tolterodine on patient reported outcomes, including quality of life in at least one arm.86, 88, 91, 99, 142, 158, 160, 174, 223, 240, 241, 247 Participants in tolterodine arms consistently reported greater changes in quality of life, including on various domains of the Kings Health Questionnaire, IIQ, and UDI-6 when compared to placebo that were similar to those experienced by participants of oxybutynin arms in the studies.

One study evaluated the impact of treatment with tolterodine ER 4 mg once daily on emotional and sexual health. Sexual Quality of Life Questionnaire- Female (SQoL-F), Pelvic Organ Prolapse/Urinary Incontinence Sexual Questionnaire (PISQ), and the Hospital Anxiety and Depression scale (HAD) were used to assess patient reported outcomes. Treatment with tolterodine resulted in improved scores versus placebo in SQoL-F; PISQ- total score, and HAD anxiety (p=0.004, p=0.009, p=0.03 respectively).241

One study stratified outcomes by treatment in patients with prolapse and found no significant difference in changes in quality of life between those with anterior vaginal wall prolapse and those without.243 Another study stratified patients by those with urge incontinence and those with OAB syndrome without incontinence, and found similar bother and severity scores in each group in an open label study of tolterodine ER.159

Most outcomes were equivalent in studies comparing tolterodine IR versus ER; however, in two RCTs, extended release tolterodine resulted in a greater reduction in incontinence episodes per day, p=0.036 and p=0.05, respectively.140, 141 Other outcome parameters including voids per day and pads per day showed a similar reduction compared to placebo with no statistical difference seen between the two formulations.

Treatment with fesoterodine.

Content of the literature. Two randomized controlled trials compared fesoterodine at 4 and 8 mg to placebo for reducing symptoms of OAB in this literature that met criteria for inclusion in the systematic review (Table 12).95, 96 These studies included a total of 1,017 women with a mean age of 57.7 in the treatment arms, and a total of 518 women with a mean age of 57.5 in the placebo arms. Both studies reported outcome data at 12 weeks of treatment. One study included a tolterodine 4 mg arm, discussed in KQ3.96 There is one paper included which is a post hoc analysis of this study population96 assessing treatment effect on health-related quality of life.142

Table 12. RCT arms for fesoterodine fumarate effect on urge incontinence and voids.

Table 12

RCT arms for fesoterodine fumarate effect on urge incontinence and voids.

Outcomes. At baseline, women in the treatment arms reported between 2.2 and 3.2 episodes of urge urinary incontinence per day. Treatment with drug resulted in reductions of 1.95 and 3.2 episodes per day. In the placebo arms at baseline, participants had an average of 3.7 episodes of urge incontinence per day. UUI in the placebo group was reduced by 1.14 to 1.48 episodes per day after 12 weeks of treatment.95, 96

At baseline, women in the treatment arms reported between 11.6 and 12.9 voids per day. Treatment with drug resulted in reductions of 1.4 to 1.9 voids per day. Women in the placebo arms reported between 12.0 to 12.2 voids per day at baseline, and reported a reduction of voids per day ranging from 0.7 to 0.95.95, 96

In the meta-analysis, fesoterodine decreased UUI per day by 2.03 (95 percent CI: 1.74, 2.31) episodes and voids per day by 1.84 (95 percent CI: 1.64, 2.03) episodes. placebo reduces UUI episodes by 1.08 (95 percent CI: 0.86, 1.30), and voids by 1.48 (95 percent CI: 1.19, 1.71) per day. The aggregate effect of placebo across all available study arms was a decrease in UUI episodes per day of 1.08 (95 percent CI: 0.86, 1.30) and voids by 1.48 (95 percent CI: 1.19, 1.71) per day.

The post hoc analysis of fesoterodine arms showed significant improvement in HRQoL as assessed by KHQ and ICIQ-SF in both fesoterodine 8 mg arm and the tolterodine ER 4 mg arm versus placebo, but there was no significant difference between arms. The fesoterodine arm showed significant improvement in eight of nine domains (non significant in General Health domain). A subset of patients who reported incontinence at baseline showed similar improvements versus baseline; however, no significant difference was seen between arms as well as between subsets of those reporting incontinence at baseline versus those without.142

Treatment with solifenacin.

Content of the literature. Three RCTs investigated solifenacin compared to placebo for reducing symptoms of OAB (Table 13).97–99 These studies included at total of 1,541 women, with a mean age of 58.2 years in the solifenacin treatment arms, and a total of 638 women with a mean age of 59.3 in the placebo arms. Two were conducted at multiple centers across Europe, and the other in Japan at academic health centers. One of these trials compared solifenacin to tolterodine, without a placebo arm, and is described in KQ3.99, 211 A newly published RCT performed at multiple centers in the United States comparing solifenacin at various doses to placebo over twelve weeks focused on outcomes of urgency and “warning time” showing significant decrease in urgency episodes per day (p<0.001), and an increase of 31.5 seconds in median warning time (p=0.008).222 Voids per day and UUI episodes per day were significantly reduced compared to placebo (both p<0.001).222

Table 13. RCT arms for solifenacin succinate effect on urge incontinence and voids.

Table 13

RCT arms for solifenacin succinate effect on urge incontinence and voids.

Outcomes. The first study compared solifenacin 5 mg and 10 mg daily to placebo and propiverine 10 mg.98 Women in the two solifenacin arms (713 women with a mean age of 60.2) had an average decrease in the number of UUI episodes per of 1.45 and 1.52 at 12 weeks compared to a reduction in the placebo arm of 0.89 (p<0.001 for comparisons of both dosages to placebo).98 Participants had a significant decrease in the number of voids per day of 1.9 to 2.9 (p<0.001 for both doses)

The second study also studied 5 mg and 10 mg doses, compared to placebo. Women in the two solifenacin arms, 5 mg and 10 mg, (531 women with a mean age of 57.1) experienced reductions in number of UUI episodes per day of 1.4 and 0.9, respectively at 12 weeks. The placebo arm (267 women with a mean age of 57.7) demonstrated a reduction in urge urinary incontinence episodes per day of 0.6.97

Solifenacin was associated with a decrease of 1.46 episodes of UUI per day (95 percent CI: 1.32, 1.59) and of 2.19 voids per day (95 percent CI: 1.94, 3.02) in the meta-analysis.

Quality of life was assessed in four studies of solifenacin.98, 99, 218, 225 Quality of life measures improved significantly in the solifenacin arms218 including when compared to placebo.98, 225 Significant improvement in quality of life was maintained upon sub-analysis for African Americans and Hispanics.205, 229 Perception of bladder condition was better in one solifenacin arm that was compared to tolterodine99 (p=0.006).

Two prospective case series with multiple publications evaluated solifenacin in the treatment of OAB.205, 218, 220, 229 One study included 1,280 women with a mean age of 56.4 years.220 Participants had a 66 percent reduction in incontinence episodes per day (from 2.66 to 0.93), a 23 percent reduction in voids per day (12.16 to 9.18), a 63 percent reduction in urgency episodes per day (5.76 to 2.28), and a 32 percent reduction in nocturia episodes (1.95 to 1.25) as assessed by urinary diary at one-year followup.

A dose ranging study evaluating treatment with solifenacin 5 and 10 mg daily after washout from tolterodine 4 mg ER, darifenacin or trospium found significant improvement in UUI episodes per day and voids per day over a 12 week treatment period after 14 day washout period (p<0.001). Patient reported outcomes assessed by the PPBC and OAB-q questionnaires indicated significant subjective improvement (p<0.001 in all domains).208

Another prospective randomized trial compared Solifenacin 5 and 10 mg daily to placebo. This study used the PPIUS (Patient Perception of Intensity of Urgency Scale) to assess treatment effect on urgency, and demonstrated that Solifenacin at the two doses demonstrated significant improvement in severe urgency episodes over placebo (p<0.001). In addition, patient reported outcomes of urgency bother, patient PBC score, and treatment satisfaction were significantly reduced compared to placebo (p<0.001).206

Treatment with darifenacin.

Content of the literature. Five RCTs provided data on the effectiveness of darifenacin.82, 100–102, 210 Three compared 15 and 30 mg daily relative to placebo (Tables 14 and 15).82, 100, 101 These three studies included a total of 690 women, with a mean age of 57.6 years in the darifenacin treatment arms and a total of 304 women with a mean age of 57.6 in the placebo arms. Two of the three were conducted in the United States in a non-primary care population.82, 101 The other was performed at multiple centers in Europe including community-based ambulatory populations, and also included a 7.5 mg arm.100 The fourth study compared only darifenacin 7.5 mg to placebo and included 269 women with mean age of 57.5 in the treatment arm, and 129 women with an average of 58.5 in the placebo arm. This was a dose change study without report of outcomes by dosage.102 A fifth RCT also compared darifenacin 7.5 mg to placebo in a multinational, multicenter trial. This was a dose changing study without reporting of outcome by dosage. The treatment arm included 206 women with an average age of 72 years and a placebo arm including 100 women with an average age of 73 years.210 One of these studies82 also compared darifenacin to oxybutynin; the darifenacin arm is presented in Tables 14 and 15, but the comparison is described in the chapter on KQ3.

Table 14. RCT arms for darifenacin effect on urge incontinence.

Table 14

RCT arms for darifenacin effect on urge incontinence.

Table 15. RCT arms for darifenacin effect on voids per day.

Table 15

RCT arms for darifenacin effect on voids per day.

Outcomes. Baseline urge urinary incontinence episodes per day in the darifenacin arms ranged from 2.0 to 2.9. Urge urinary incontinence episodes per day on treatment ranged from 0.8 to 1.6, for a decrease in episodes of UUI per day ranging from 1.2 to 1.8. This is compared to an average reduction in the placebo groups of 0.8 to 1.0, with similar baseline measures of 2.3 to 3.0.

At baseline, participants reported an average of 10.3 to 11.0 voids per day. The treatment groups reported reductions ranging from 1.1 to 2.2 per day, compared to 0.8 to 1.8 in the placebo group. Study followup ranged from 2 to 12 weeks. Daily episodes of the symptom of urinary urgency also were reported to be reduced after treatment with darifenacin (range of reduction: 1.4 to 3.0 compared to 0.6 to 1.2 in placebo). The one study that reported reduction in urgency on a weekly basis, however, found no significant difference between treatment and placebo.101

Because of a lack of uniform information provided in the studies on variance measures, we could not use meta-analysis techniques to estimate effects of darifenacin on UUI or voids per day.

In the first dose-adjustment trial, urge urinary incontinence episodes were not reported, but voiding frequency was significantly reduced in the treatment arm versus placebo, p=0.001.102 In the second dose changing trial changes in urge urinary incontinence episodes per week were not statistically different between the two arms (p=0.328), but reductions in voids per day were statistically greater in the treatment arm versus placebo (p=0.006). This study also assessed urgency episodes per day, but found no statistically significant difference in effect (p=0.174).210

Three studies assessed quality of life (QoL) with validated questionnaires using Visual Analog Scale (VAS), OAB-q, King’s Health Questionnaire (KHQ), and ICIQ. The arms using 15 and 30 mg showed statistically significant improvement versus placebo, p=0.045 and p=0.011, respectively.100 Zinner evaluated 15 mg versus placebo and showed significant improvement in quality of life questionnaires, OAB-q (p<0.001), ICIQ (p<0.001) and KHQ (p<0.05).101 Chapple demonstrated improvement in all domains of the OAB-q versus placebo (p<0.001).210 A single prospective cohort study evaluated the efficacy and tolerability of darifenacin.219 This prospective, non-comparative, open-label extension study was conducted in multiple countries in Europe, Australia, and North America.102 The authors stratified results by age. All outcome parameters were similar between groups showing similar reductions in voids per day, urgency episodes per day, and severity of urgency in those over and under 65 years old.219 A second paper describing the open labeled extension study of those patients219 demonstrated significant improvement from baseline in eight of nine domains (p<0.001) in health related quality of life outcomes as measured by the KHQ at 24 months. Significant reductions were confirmed in the subgroup of subjects 65 years old and greater. The only domain that did not show improvement was the General Health Perceptions.217

Treatment with trospium chloride.

Content of the literature. Five RCTs evaluated trospium for reduction of symptoms of OAB. Four trials compared trospium to placebo, and one compared trospium to oxybutynin (5 mg twice daily) (Table 16). Four were conducted in the United States,103, 104, 105, 106 and the fifth at multiple centers in Europe and Asia (Tables 16 and 17).90 These studies included a total of 1,309 women, with a mean age of 59.8 years in the trospium treatment arms, and a total of 1,130 women with a mean age of 60.1 in the placebo arms.103–106 One trial included a comparison to oxybutynin (5 mg twice a day).90 The trospium arm is included in Tables 16 and 17; and the drug to drug comparison in noted in KQ3.

Table 16. RCT arms for trospium chloride effect on urge incontinence.

Table 16

RCT arms for trospium chloride effect on urge incontinence.

Table 17. RCT arms for trospium chloride effect on voids per day.

Table 17

RCT arms for trospium chloride effect on voids per day.

Outcomes. Reductions in numbers of urge urinary incontinence episodes per day at 12 weeks ranged from 1.8 to 2.5 relative to reductions in the placebo arms of 1.3 to 1.9. Reduction in voids per day on treatment ranged from 2.4 to 3.5 at 12 weeks, compared to placebo at 1.3 to 2.1

Meta-analysis could be used to estimate effects of trospium extended release only. This formulation reduced UUI episodes per day by 2.45 (95 percent CI: 2.19, 2.70) and voids per day by 2.68 (95 percent CI: 2.38, 2.98) episodes. The aggregate effect of placebo across all available study arms was a decrease in UUI episodes per day of 1.08 (95 percent CI: 0.86, 1.30) and voids by 1.48 (95 percent CI: 1.19, 1.71) per day.

Three studies evaluated urgency severity using the Indevus Urgency Severity Scale (IUSS). All three studies found statistically significant improvements in severity of urgency at trial end in the trospium arms versus placebo: p=0.0004, p 0.001, and p<0.0001 respectively.103–105

Quality of life was addressed in three studies using the OAB-SCS validated questionnaire103, 104 and the IIQ in one.106 Only one of the studies using the OAB-SCS reported statistically significant improvement in quality of life compared to placebo, p<0.05.103

Treatment with estrogen.

Content of the literature. Three studies assessed the role of hormonal therapy in different doses and formulations in the alleviation of OAB symptoms.107–109 All studies were performed in Europe and Scandinavia. They included a total of 508 women with a mean age of 62.4 years. Three were RCTs and one was a case series.

Outcomes. One RCT compared oral estriol 3 mg per day to placebo at academic centers in England. They found that estriol was not significantly superior to placebo at improving symptoms or objective measures for all patients at one or three months.107

The second RCT compared an estradiol releasing vaginal ring to an estriol vaginal pessary in a community population in Denmark, and found both equal in improvement of UUI episodes, urgency, and nocturia. In each arm, 58 percent of participants had decreased UUI at 24 weeks; 51 and 56 percent had decreased urgency, respectively.108 This is difficult to interpret in the absence of a placebo comparison.

A recently published RCT compared tolterodine 2 mg twice daily with and without CEE (conjugated equine estrogens) vaginal cream 0.625 mg twice weekly. Voids per day decreased significantly in the CEE arm (p=0.001), but did not show significant difference in UUI events per day. Subjective assessment with UDI-6 and IIQ-7 showed significant improvement in the tolterodine with CEE arm (p=0.001).247

One case series found oral HRT significantly improved frequency at six-month followup where dosage was not associated with outcome. Hormone replacement therapy in the form of estriol significantly decreased frequency in the OAB arm at six weeks.109 Outcomes were assessed with a five-point unvalidated rating system and no helpful comparison group.

Harms of Pharmacologic Treatments

Proportions of individuals reporting harms in RCTs of pharmacologic treatments ranged from 9.7 to 63.6 percent of study participants; however, harms were generally mild in nature, and withdrawals due to adverse events did not exceed 17 percent in any study (Table 19). The risk of occurrence of harms reported in treatment arms often overlapped those observed with placebo.

Table 19. Side effects and harms of pharmacologic treatment.

Table 19

Side effects and harms of pharmacologic treatment.

Dry mouth was the most commonly reported harm, ranging from 5.9 percent to 88 percent in studies of oxybutynin IR, compared to 1.6 to 21 percent in placebo arms. Studies of transdermal oxybutynin had the lowest reported estimates of dry mouth (2.6 to 9.6 percent). Impaired urination, not defined by the authors, was reported in studies of oxybutynin and tolterodine. It was highest in two studies of oxybutynin IR (14 to 29 percent), compared to 3.2 to 4.0 percent in two studies of oxybutynin ER, no events to 9.0 percent in six studies of tolterodine IR and 1.0 percent in one study of tolterodine ER. Urinary tract infections were reported by up to 11 percent of participants in eight placebo arms. Immediate release formulations of oxybutynin and tolterodine both had reports of up to about 18 percent of participants experiencing a UTI, compared to up to 12 percent in studies of oxybutynin ER and 4.1 percent in studies of tolterodine ER. Between 0 and 32 percent of participants in treatment arms reported constipation; again, the highest rate was reported in an oxybutynin IR arm. Darifenacin had the second highest proportion of participants reporting constipation (18.5 to 27.8 percent). Up to seven percent of participants in placebo arms also reported constipation.

Cardiac events, including new abnormalities on EKG, were very rare and reported in only a few studies. The highest reported rate of cardiac events was five percent in tolterodine ER. Generally, however, less than one percent of participants experienced any cardiac event, mostly tachycardia and arrhythmias in treatment arms, with events also occurring in placebo arms (0 to 0.9 percent).

Procedural and Surgical Treatments of OAB

We reviewed 18 studies, of which 11were fair quality and 7 poor. This section presents the results of our literature search and findings about outcomes of procedural and surgical treatments for OAB. These treatments include sacral neuromodulation, peripheral neuromodulation, electromagnetic nerve stimulation, injection or instillation of drugs into the bladder, bladder distention and bladder transection. No studies regarding augmentation cystoplasty or detrusor myomectomy met our search criteria. Detailed information on all studies relating to surgical management of OAB can be found in evidence tables in Appendix C.

Sacral neuromodulation. Stimulation of the sacral nerve roots is a technique in which an electrical stimulus directly stimulates the S3 sacral nerve root.256 The technique has evolved over time, but typically it is performed as a staged procedure. The first stage involves a “test” stimulation using a percutaneous needle to stimulate the S3 nerve root. If there is a favorable response during the trial period, then long-term stimulation can be provided by implanting an implantable pulse generator surgically. The implantable pulse generator is usually placed in the fatty tissues overlying the buttocks, although abdominal placement was used with some of the earlier studies. Recent evolutions in this technique now permit a permanent lead to be used for the test stimulation. If the test is unsuccessful, the lead can be removed, but if it is successful, this lead is attached to the permanent implantable pulse generator. This has the advantage of ensuring that stimulation is provided in the exact location as during the test period. (Previously, a new lead was placed at the time of the implantable pulse generator placement.) The mechanism by which neuromodulation acts to improve symptoms is not completely understood. The technique is used for urinary urgency, frequency, and urge incontinence refractory to other treatment modalities.256 It is also used for urinary retention. Given these seemingly contradictory applications, it is thought that the electrical stimulation affects the afferent nerves (which perceive bladder sensation), thus allowing them to appropriately transmit bladder sensations.

Peripheral neuromodulation. Other techniques for neuromodulation involve stimulating the S3 nerve fibers more peripherally, at the posterior tibial nerve or cutaneous stimulation of the pudendal nerve via an anal or vaginal probe.257 For the posterior tibial nerve stimulation, a needle is placed percutaneously near the ankle and is attached to an external electrical device. Instead of implanting an implantable pulse generator, the patient returns for periodic sessions, often weekly for a series of treatments. Small case series suggest that posterior tibial nerve stimulation may improve OAB symptoms.258–260 There were no studies involving this technique which met our search criteria. One study evaluating neuromodulation of the pudendal nerve with anal and/or vaginal probes met our search criteria. Similarly, this is performed on an outpatient basis with weekly treatment sessions.

For the purposes of this report, sacral neuromodulation will refer to techniques that directly stimulate the S3 nerve root. Peripheral neuromodulation will refer to nerve stimulation peripherally, such as the use of an anal or vaginal probe to stimulate the pudendal nerve. These approaches are reviewed in the same section of this text.

Electromagnetic sacral nerve stimulation. Electromagnetic stimulation is yet another modality to modulate the neurologic control of the bladder. Most treatments involve large, powerful magnets which require a dedicated facility as the magnets are not portable.261 The study included in this review evaluated the use of a smaller, portable electromagnetic system.

Bladder instillation/injection of a drug. Another approach to treating urinary frequency, urgency, and urge incontinence is to instill or inject a drug into the bladder. Numerous drugs have been administered using this approach. Two categories of intravesical drugs are included in this review, antimuscarinic agents and neurotoxins. By instilling or injecting the drug directly into the bladder, systemic adverse effects are theoretically avoided.

Two neurotoxins discussed in this review are resiniferatoxin and botulinum toxin. Botulinum toxin is a neuromuscular blocking agent which prevents nerve conduction. Typically botulinum toxin-A is used and can be injected directly into the wall of the bladder under cystoscopic guidance as a treatment for refractory OAB.262–264 It is not FDA approved for this indication at the time of the writing of this document. Concerns with this approach are the risk of urinary tract infections and urinary retention and that the ideal dosing has not yet been determined. Additionally, the effects of botulinum toxin are temporary and multiple courses of treatment would be anticipated. An RCT comparing botulinum toxin A to placebo for women with refractory urge incontinence found that incontinence episodes decreased from over 20 episodes per three day diary to less than 5 episodes per three day diary among those receiving botulinum toxin A, while those taking placebo had no difference in the number of incontinence episodes. Using a Patient Global Impression of Improvement score, approximately 60% of the women who received botulinum toxin A had a clinical response, and this response lasted six times longer than that achieved with placebo. The study was stopped early due to increased postvoid residuals in 43% of women receiving botulinum toxin A and a high rate of urinary tract infections among those with elevated postvoid residuals.265 Resiniferatoxin is a neurotoxin in the same category as capsaicin; these do not have FDA approval for the treatment of OAB. These agents block transmission along the C-fibers, nerve fibers involved in transmitting noxious stimuli.127 It has been hypothesized that inhibition of these fibers may be a treatment for overactive bladder.

A review of RCTs evaluating the use of intravesical botulinum toxin for OAB was published by the Cochrane Collaboration in 2007.128 Eight studies met their search criteria: five were published abstracts and three were full papers. Only one of the eight studies exclusively dealt with idiopathic OAB, the definition we used for our literature search. The remaining seven studies in the Cochrane review included subjects with neurogenic OAB.

The findings from the Cochrane review were that botulinum toxin injections were more effective than placebo, with fewer incontinence (unspecified type) episodes per day at 2 to 24 weeks and fewer incontinence episodes compared to baseline. Improvements following treatment were also seen in incontinence specific and overall quality of life. Urodynamic changes were also seen, including decreased pressure during a detrusor contraction and increased bladder capacity following treatment with botulinum toxin. The postvoid residual, the amount of urine left in the bladder after voiding, was also elevated. The numerical data were not provided by the author. Adverse events included cases of urinary retention requiring intermittent self catheterization following treatment. Of the patients requiring catheterization, 25 percent had a lower urinary tract infection.

In the Cochrane Collaboration, the limited number of studies, their small size and heterogeneous population highlight the need for more research regarding treatment with botulinum toxin.128 One study compared botulinum toxin to bladder instillations with resiniferatoxin and found lower rates of incontinence, increased bladder capacity and lower detrusor pressure during uninhibited bladder contractions at 6 to 18 months with the botulinum toxin treatment. The optimal dose or long term effects of elevated postvoid residuals has not been determined.

Bladder distention and bladder transection. Two treatments that are no longer in common practice, prolonged bladder distention and bladder transection, are also included in this review. The bladder distention study describes distending the bladder to maximum capacity at a pressure equal to systolic blood pressure for four hours.118 Bladder transection involves cutting the bladder wall and detrusor muscle. Both have significant morbidity and have been abandoned by most practitioners.

Outcomes of procedural and surgical treatments.

Content of the literature. We identified 18 studies reporting on surgical treatments and procedures for OAB (clinical trials reported in Table 20).110–127

Table 20. Outcomes of clinical trials of procedures.

Table 20

Outcomes of clinical trials of procedures.

Eleven were of sacral neuromodulation, one of peripheral neuromodulation and one of electromagnetic sacral nerve stimulation. Three studied bladder instillation or injection of drugs, one was of bladder distention and one was of bladder transection.

Six of the 11 studies on sacral neuromodulation come from a family of papers financially supported by one company.110, 111, 114, 115, 123, 124 The patient population for this family of studies included an RCT to evaluate sacral neuromodulation versus medical therapy for six months.124 The other studies involved similar inclusion and exclusion criteria, but varied in regards to the number of centers involved in recruitment (between 12 and 17 centers) and timing for enrollment. Given this, it is not possible to determine the degree of subject duplication.

This literature included 13 case series studies, which we operationally defined as descriptive analyses of a sequence of participants having the same type of procedure without a comparison to another type of surgery or treatment. Three of these studies are retrospective case series of a particular surgical treatment: two report on sacral neuromodulation119, 120 and one on bladder transection.121 Nine studies are prospective case series: six report on sacral neuromodulation,110–115 one on peripheral nerve stimulation with anal and/or vaginal probes,116 one on botulinum-A toxin injections117 and one on prolonged bladder distention.118 One case series had both a retrospective arm as well as a prospective arm; this study looked at sacral neuromodulation.122

One study is a prospective cohort that compared outcomes among subjects receiving sacral neuromodulation and subjects receiving a control surgery for sacral neuromodulation, but without active electrical stimulation.123

Four studies were RCTs: one looked at sacral neuromodulation versus medical therapy,124 one evaluated transcutaneous electromagnetic stimulation versus sham,125 and two evaluated instillation of a drug into the bladder versus placebo – one using oxybutynin126 and one using resiniferatoxin.127

The majority of the studies were conducted in Europe, six of which included involvement from Canada and the United States (there is subject duplication among these six studies). One study was conducted in the United States only. Two studies were performed in Australia and one in Brazil. Two of the studies involved national registries, one from Switzerland and one from Italy. Five studies specified that they occurred in an academic setting, three in a specialty setting and the remainder did not specify the clinical setting for the study.

The quality of the studies varied widely. Of the four randomized controlled trials, three had an appropriate control group, the fourth compared sacral neuromodulation to continued medical treatment among subjects who had already failed medical management for refractory OAB.124 This does not adequately control for the placebo effect, which is quite prominent among these treatments. It is challenging to create an appropriate control group when evaluating sacral neuromodulation. This is illustrated by a cohort study in which surgical controls received sacral neuromodulation, but no electrical stimulation was applied. Given that subjects feel the electrical stimulation, they would be aware that this has been de-activated following the test stimulation.123 Of the 11 studies on sacral neuromodulation, seven did not restrict the study population to OAB. Although sacral neuromodulation may be used to treat urinary retention, grouping these studies in such a fashion severely limited our ability to analyze the results. We restricted our discussion to the OAB findings in these papers. Case series studies are limited by the lack of a control group.

Outcomes assessed. Half of the studies reported on urge urinary incontinence outcomes (usually incontinence episodes per day as measured by a bladder diary, but also pad counts, pad weight, severity of incontinence or other measures). Nearly two-thirds of the studies reported some measure of urinary frequency, such as voids per day. Nearly all of the studies reported on a subjective measure of symptoms (such as percent cured, perceived severity, QoL). A third of the studies reported on urodynamic outcomes, such as the bladder volume at which there was a normal desire to void, and bladder capacity.

The majority (78 percent) of the studies, reported data on adverse effects or harms. Six studies reported on problems with constipation or gastrointestinal symptoms,111, 114, 116, 119, 120, 126 eight reported the presence of pain, six reported on infection and nearly half of the sacral neuromodulation studies reported the presence of lead migration. Other adverse effects were reported in 72 percent of studies.

Outcomes of sacral neuromodulation and peripheral neuromodulation.

The one RCT comparing sacral neuromodulation to medical therapy found a reduction in daily incontinence episodes from 9.7 to 2.6 in the intervention group, compared to an increase of 9.3 to 11.3 in the medical management group at six months (p<0.01).124 Of note, all subjects receiving medical therapy had already failed medical management; no benefit from continued medical therapy would be expected. The remaining six case series that reported on change in UUI had decreases in mean incontinence episodes per day of 51 percent to 80 percent111, 114, 115, 119, 122 and from a median of five down to zero incontinence episodes a day.112 Length of followup in these studies ranged for six months to five years.

Pad use per day also decreased with sacral neuromodulation. Most studies started with a baseline of five or more pads used daily. One RCT reported an 82 percent decrease in pad use from 6.2 to 1.1 pads daily, six months following initiation of sacral neuromodulation.124 Three case series evaluating sacral neuromodulation also found significant decreases in pad use ranging from 49 to 84 percent fewer mean pads111, 114, 115 and a 75 percent decrease in median pad use.112 Length of followup on these studies ranged from six months to five years.

Some of the studies tried to characterize the severity of incontinence episodes. One RCT and two case series found a 64 percent to 92 percent decrease in the number of moderate to heavy urge urinary incontinence episodes at six months to five years of followup.111, 115, 124 The RCT reported the highest rate of decreased heavy incontinence episodes with a mean baseline in the neuromodulation group of 3.4 per day, reduced to 0.3 per day six months after treatment. In comparison, those with refractory OAB receiving usual therapy experienced an increase in mean heavy episodes per day from 2.6 to 3.9.124

Improvement in episodes of urinary urgency without incontinence is difficult to measure as it is a more elusive symptom. Nonetheless, on a 3 point scale, (1=mild, 2=moderate, 3=severe), 69 percent of participants in one study reported improvement,114 with a second study showing no change in experience of urgency.115

Reduction in urinary frequency of between 31 and 45 percent is seen consistently across studies of sacral neuromodulation, regardless of study design. The majority of studies of sacral neuromodulation reported urinary frequency as the mean number of voids in 24 hours as recorded by a bladder diary. Six studies reported mean voids per day, one of which was a prospective cohort study comparing subjects with sacral neuromodulation and controls who had sacral neuromodulation placed, but did not receive active electrical stimulation.123 At six months, those receiving sacral neuromodulation had a 45 percent decrease in the number of voids per day; no change was seen in the controls. Similar results were seen in the other studies, which found 31 to 40 percent fewer voids per day with sacral neuromodulation, regardless of whether they were prospective or retrospective case series studies.114, 115, 119 One study reported its results as a 40 percent decrease in the median number of voids per day.112 Another study reported the results from an Italian national registry; they did not have baseline information regarding the number of voids in the registry prior to treatment, but 42 percent of subjects had fewer than 8 voids daily after treatment.122 The 31 to 45 percent decrease in mean (and median) voids per day seen across the studies was present at six months and up to two years following initiation of the sacral neuromodulation.112, 114, 115, 119, 123 The longest followup data was available from a prospective case series which found a 33 percent decrease in mean voids per day at one year which was reduced to a 23 percent decrease in mean voids per day at five years.115

Some studies also looked at the mean voided volume as a measure of treating urinary frequency. One cohort study and two case series found that sacral neuromodulation increased the mean voided volume between 1.7 to 1.9 fold, an increase of 78 mL to 108 mL per void.114, 115, 123

Peripheral neuromodulation and electromagnetic stimulation were clinically ineffective in changing voiding frequency. One prospective case series found a 12 percent decrease in the mean voids per 24 hours was seen six weeks following 12 sessions of peripheral neuromodulation with an anal and/or vaginal probe.116 No decrease in mean voids per day was seen with electromagnetic stimulation of the sacral nerves.125

Neither sacral neuromodulation nor peripheral neuromodulation with an anal and/or vaginal probe had a clinically relevant impact on nocturia rates, which were very low in these studies at baseline. Both treatment approaches reduced already low rates by approximately 30 percent.116, 119

Several of the studies comment on clinical “cures” or “improvements”, but lack definitions for these criteria. Moreover, several of the case series on sacral neuromodulation include diverse patient populations which may include urinary retention or pelvic pain in addition to patients with symptoms of overactive bladder. The clinical endpoints for improvement in urinary retention are very different from those for urinary urgency and frequency.

Studies evaluating improvements following sacral neuromodulation describe cure rates of 26 to 65 percent (cure was defined as “completely dry” at six to twelve months).111, 122, 124 One study looked at improvements following the initial test stimulation and found 39 percent had greater than a 90 percent improvement in urinary frequency and/or urgency.113 As would be expected, higher rates are seen for classifications of “success” or “improvement” as compared to “cure.”

Compared to these rates, peripheral neuromodulation only had an 8.1 percent cure rate, with 31 percent reporting no change.116

Several studies evaluated QoL, either with the KHQ, a VAS or other validated QoL questionnaire (Table 21). Two studies evaluated the impact of sacral neuromodulation on QoL and found the treatment beneficial.122, 123 There was no improvement in QoL with electromagnetic stimulation.

Table 21. Effect on quality of life and satisfaction of procedural treatments.

Table 21

Effect on quality of life and satisfaction of procedural treatments.

The evaluation of overactive bladder with urodynamics provides objective data, but the clinical relevance of this information is not always clear. Urodynamic parameters may indicate changes in bladder function, but unless symptoms were also evaluated, they do not necessarily carry over into improvements in clinical outcomes.

During bladder filling, known as cystometry, the first sensation of bladder filling can be measured. This is known to be one of the more variable urodynamic measurements.268 The bladder volume at first sensation increased by 80 mL after sacral neuromodulation in one case series 110 and by 36 mL after peripheral neuromodulation with an anal/vaginal probe.116 Both of these studies also looked for changes in bladder capacity and found this increased 1.4 fold for sacral neuromodulation and found no difference for peripheral neuromodulation.110, 116

By their very nature, surgical and procedural treatments are likely to have a higher incidence of adverse events than conservative and medical treatments for OAB. In the early studies of sacral neuromodulation, there was an average of 1.1 to 1.7 adverse events per participant.115, 124 Advances in technology, such as the use of tined leads, have decrease this rate and the more recent studies report 0.1 to 0.5 events per participant.112, 119 None of the studies exclusively used the newer technology. Of the adverse events, pain, lead migration or problems with the lead, infection and explantation of the device were the most common adverse events. From the test stimulation phase, pain at the needle site was 0.5 percent in a study employing newer technology112 and 7 percent in a study with older technology.113 When the implantation phase is included, pain rates and uncomfortable stimulation responses were seen in 3.9 to 43 percent of subjects, with studies employing new techniques at the lower end of this spectrum.112, 114, 115, 120, 122, 124 The one RCT found a pain rate of 19.1 percent, but this was compared to medical management, not a sham procedure.124 Pain at the implantable pulse generator implantation site was typically reported separately and occurred 15.4 to 27 percent of the time.114, 115, 120, 124 Problems with the lead or lead migration occurred between 3.3 and 11 percent of the time.114, 115, 120, 122, 124 This may be less common with the new tined leads, as one study had lead migration only with the older non-tined leads112 and another found a lower loss of efficacy, 12.3 percent, with the tined leads compared to the non-tined leads, 31.7 percent.119 Infection occurred in 1.9 to 6.1 percent of participants, sometimes requiring hospitalization for intravenous antibiotics or removal of the device.114, 119, 120, 122, 124 Two papers noted a 0.5 to 1.7 percent risk of neuropraxia or nerve injury.114, 120 There was a high rate of needing surgical revision, with most studies showing 33 to 48.3 percent of subjects required a return to the operating room.114, 120, 124 Lower rates were documented in a study using newer technology, 7 percent,112 and in a national registry, 9.7 percent.122 One study looked at return to the operating room rates at five years and found that that there was a 67 percent risk of return (numbers based on the population enrolled).115 However, at five years, many of the subjects were returning for new implantable pulse generator batteries, an expected development over time, thus this is not truly an adverse event. Unfortunately, the number of surgeries which were to replace implantable pulse generator batteries was not reported. Of the surgical adverse events, a significant proportion included explantations of the device. This occurred in 3.9 percent of subjects in the national registry;122 higher rates were seen in the remaining studies, 9.8 to 14 percent.111, 115, 120

Adverse events following peripheral neuromodulation with an anal and/or vaginal probe led to a 19 percent dropout rate, with 29 percent of the dropouts attributed to pain. The other most common complaint was bowel irritation.116 Currently, techniques of peripheral neuromodulation with an anal/vaginal probe are not routinely used.

Outcomes of bladder instillation/injection of drugs. An RCT of resiniferatoxin bladder instillations versus placebo found no improvement in the number of urge urinary incontinence episodes per day with either arm.127 However, bladder instillations with oxybutynin did decrease mean voids per day. In an RCT evaluating bladder instillations with oxybutynin versus a placebo with sterile water, those receiving oxybutynin had a 47 percent decrease compared to a 16 percent decrease in mean voids per day with the placebo at two weeks following completion of treatment.126 This decrease is similar to that seen with sacral neuromodulation. In a prospective case series evaluating the use of botulinum-A toxin injected into the detrusor muscle of the bladder wall, 74 percent of the subjects had 8 or fewer voids per day four weeks following treatment. All subjects had greater than 8 voids per day at the start of the study.117 In an RCT of bladder instillations with resiniferatoxin versus placebo, there was no difference in the mean voids per day four weeks following treatment.127 Bladder instillations with oxybutynin decreased nocturia 65 percent from 5 to 1.8 episodes per night.126 Following injection of botulinum-A toxin into the bladder, 66 percent had no urgency at 12 weeks and 80 percent reported no urge incontinence.117 No difference was seen compared to baseline or placebo in an RCT evaluating resiniferatoxin bladder instillations.127

Two studies looked at what volume the normal desire to void occurred, finding this increased 1.6 fold (55 mL increase) for instillation of oxybutynin into the bladder126 and 1.7 fold (83 mL) for botulinum-A toxin bladder injections.117 Theoretically these increases would result in less urinary frequency, and possibly less urgency. Similar changes were also seen in terms of bladder capacity after treatment, with bladder instillations of oxybutynin increasing capacity 1.5 fold (105 mL)126 and 1.6 fold (135 mL) with botulinum-A toxin injections.117

A potential adverse effect of intravesical oxybutynin or botulinum-A toxin is impairment of the bladders ability to empty.265 The goal of treatment with these agents is to provide a dose large enough to decrease the symptoms of OAB, without impairing the ability to void when physiologically necessary. An objective measurement that can serve as a proxy for this parameter is a postvoid residual, the volume of urine that remains in the bladder after voiding; however the clinical importance of an increased postvoid residual in the absence of symptomatic urinary retention is not clear. The mean postvoid residual following treatment with intravesical oxybutynin increased twofold to a mean of 40 mL (range 10 to 50 mL) and for botulinum-A toxin, increased fourfold to a mean of 75 ± 10 mL.117, 126 Both treatments seem to have similar effects on bladder capacity, but the risk of urinary retention may be higher with use of botulinum-A toxin.117, 126 Currently, oxybutynin and resiniferatoxin instillations are not routinely used for the treatment of OAB.

For instilled/injected bladder drugs, there were small benefits seen in the emotions, sleep/energy and symptoms severity subscales of the KHQ for subjects receiving the treatment compared to placebo. Incontinence impact scores improved a similar amount for both resiniferatoxin and placebo (p<0.05). Emotions, sleep/energy and symptom severity scores decreased more for resiniferatoxin following the intervention (p<0.05).127 Ninety percent of subjects receiving botulinum A toxin had an improvement in QoL scores at 3 months; this effect was waning at 9 months.117

Outcomes of bladder distention and transection. Bladder distention and transection are treatment methods no longer in routine use. Following bladder distention, 18 percent were symptom free at 18 months with 52 percent unchanged.118 The study on bladder transection reports 65 percent were cured at two to five years and 16 percent were unchanged, but provides no information about the criteria for these categories.121 Prolonged bladder distention had a 4 percent rate of bladder rupture.118 Bladder transection was associated with a 14 percent rate of vesicoureteral reflux on urodynamics, the clinical significance of which was not determined, and 1 percent rate of a persistent urine leak requiring reoperation. The authors also noted “minor chest and urinary tract infections” but the rates of these adverse events were not reported.121

Behavioral Treatments

This section presents the results of our literature search and findings about outcomes of behavioral techniques to reduce overactive bladder in women. Behavioral treatment options for OAB have been used for managing urinary incontinence for more than 50 years, although large-scale and well-designed studies on them are fairly uncommon. We reviewed 27 studies, of which 14 were fair quality and 13 poor.

Behavioral techniques include the use of bladder training, pelvic floor muscle exercises (PME), biofeedback, dietary changes, and multicomponent approaches that combine bladder training with PME and/or biofeedback. Detailed information on all studies related to behavioral techniques for OAB can be found in evidence tables in Appendix C.

Bladder training. Bladder training was introduced in the 1960s (Jeffcoate and Francis), and modified by Frewen in the 1970s. It involves education, a strict schedule of daytime voiding with progressive increases in time between voids, urgency suppression techniques, and positive reinforcement. Frewen recommended that women be treated initially on an inpatient basis, and the training was often combined with antimuscarinic medication or sedatives to manage extreme urgency. Inpatient bladder training is no longer standard practice, and the technique has been modified to be administered on an outpatient basis, with the use of patient education and bladder diaries. Although the underlying mechanism for bladder training in OAB is not well understood, it is thought to reverse dysfunctional habits, increase bladder capacity and provide techniques for handling feelings of urgency.

Pelvic muscle exercises. Training the pelvic floor muscles were originally suggested for patients with stress incontinence, the idea being that patients could learn to contract the periurethral muscles to occlude the urethra during activities that caused leakage. However, it may also be useful in inhibiting detrusor contractions. PME can be implemented alone or with additional techniques such as biofeedback to help patients identify and contract the appropriate muscles.

Multicomponent approaches. Although both bladder training and PME can and are administered alone, they may also be combined with or without biofeedback for a multicomponent approach to reducing incontinence.

Tools for behavioral training. Behavioral training can be administered with written materials, verbal feedback, coaching in person or on the phone, in groups, or using other strategies, such as cognitive approaches or biofeedback for increasing the potential for success. Behavioral approaches can also be combined with medications such as antimuscarinics. Various combinations of approaches are examined in the literature and are described in this review.

For the purposes of this report, we will use the term “bladder training” to refer to bladder training alone (i.e., without PME without biofeedback). “Behavioral training” will refer to a multicomponent approach that includes bladder training. We indicate when biofeedback is used in conjunction with other behavioral techniques.

Behavioral treatments.

Content of the literature. We identified 29 papers from 27 studies that included arms with outcomes of behavioral interventions. We have divided the studies into three primary categories: those that compare only behavioral approaches, those that compare behavioral approaches to pharmaceutical ones directly and those that measure the effect of adding a behavioral approach to a pharmaceutical one (combination approaches). The first category is presented here and the second two are covered in KQ3.

Nine studies met criteria and included only behavioral arms. Among these, two studies focused on comparing delivery mechanisms or approaches and provide insight into techniques for teaching and encouraging participants in behavioral management.132, 135

Four additional studies (represented in six papers) with multiple arms, including pharmacologic ones, provided data that allowed the comparison of behavioral management to placebo93, 143–145 or to another behavioral intervention.148, 201 These are counted in the direct comparisons section, and therefore are not represented in the counts below, but relevant outcomes are described here as well as the direct comparisons section, and they appear in tables in both sections. All three of these are RCTs.

The literature base of studies that had only behavioral arms included three retrospective case series, which we have operationally defined as a sequence of participants having the same intervention without a comparison to another type of intervention. One examined bladder training alone,129 one examined pelvic muscle exercises130 and the third reported on a series of participants who were provided either bladder training or biofeedback, but presented results only for the two groups combined.131 All three were conducted in community-based clinical settings.

One was a prospective cohort study comparing three bladder training approaches: self-administered, coaching and cognitive strategies.132

Five studies were randomized controlled trials. One compared bladder training to a “control” condition.133 One compared bladder training to pelvic muscle exercises.134 One included the following three arms: pelvic floor muscle training, pelvic floor muscle training assisted with biofeedback, and electrical stimulation.93 One compared three different approaches to multicomponent behavioral training: biofeedback, verbal feedback and self-administered using an instruction booklet.135 One compared bladder training to bladder training with an additional caffeine reduction component.136

Three of the studies were conducted in the United States, four in Europe, one in Australia and one in Taiwan. Four were conducted at academic medical centers; five were in community settings, of which two included an inpatient component.

Outcomes measured. For each type of intervention, we combed the publications for the outcomes and complications summarized in the analytic framework presented in Chapter 1.

Five studies reported a change in numbers of episodes of incontinence,93, 134–136, 143 although the time period varied, and three of these calculated a percent reduction in incontinence episodes.135, 136, 143 Three of the studies measured incontinence episodes over the course of a week,134, 135, 143 two did so over 24 hours.93, 136 One study reported that a significant decrease in frequency of voiding was observed, but did not provide data that could be included in this table.132

Five studies presented the outcome of cure or improvement in incontinence (using various definitions)129–131, 133, 269 with three defining cure as complete resolution of incontinence.129, 131, 133

One study presented data on episodes of urgency separate from incontinence.136

Changes in frequency were reported in three studies.131, 136, 145 Two specified voids per day as an outcome.136, 145 In addition, Jarvis and colleagues133 reported on the numbers of women indicating that they had nocturnal and diurnal frequency before and after treatment. One study recorded frequency over a three-day period132 and reported a statistically significant effect, but did not report numbers of episodes, and one reported time between voids.131 One reported that cure included achieving a minimum of three to four hours between voids129 and was therefore related to frequency but not summarized as effect on frequency.

Other patient-reported outcomes in this literature included reports of resolution or improvement and changes in severity, quality of life, and satisfaction.

Five of the studies provided some patient-reported outcome, with four reporting on perceived improvement.130, 131, 134, 135 Three provided some report on severity;130, 135, 143 four measured bother;132, 134, 135, 143 two measured impact or interference with daily activity;134, 135 one assessed changes in quality of life overall;134 and three reported on patient satisfaction.134, 135, 143

Outcomes of behavioral treatment.

UUI episodes. Lack of consistency in study design, interventions or comparison groups makes it impossible to provide consistent summary results across studies. Although each intervention was associated with reductions in incontinence episodes, no behavioral approach performed better than any other in any study in this category on incontinence outcomes over any time period greater than 12 weeks. Trials are summarized in Table 22.

Table 22. Outcomes of behavioral treatment trials.

Table 22

Outcomes of behavioral treatment trials.

One study evaluated the ability of biofeedback to improve outcomes associated with multicomponent behavioral therapy,135 relative to providing the training with verbal feedback. The intervention included multicomponent behavioral training with either biofeedback or verbal feedback, compared to a self-help booklet. Multicomponent behavioral training combined bladder training techniques (e.g., relaxation approaches in the presence of urge, extending time between voids) with pelvic muscle exercises. The interventions took place over an eight-week period with four visits, followed by a two-week bladder diary. Reductions (mean: 58.6 to 69.4 percent) in episodes of incontinence were seen in all groups, with median percentage reductions ranging from 70.4 (IQR: −29.4, 100) to 82.8 (IQR: 0, 100). The wide IQR makes this finding somewhat difficult to interpret, and overall there were no by-group differences (p=0.23), Nonetheless, patients’ perceptions of treatment benefit differed (see section below on patient-reported outcomes) with patients in the self-administered training significantly less satisfied (p=0.001).

Burgio et al 135 also compared their multicomponent behavioral approach to oxybutynin and placebo (study described in KQ3) and observed reductions in incontinence episodes of 81 percent among those in the behavioral group relative to 39 percent in the placebo arm.

Wyman and colleagues attempted to separate the role of pelvic muscle strengthening from bladder drill, and assess the potential for a combined impact. Combining the approaches provided the greatest reduction in incontinence episodes immediately after the 12 week intervention (p=0.050), but the difference did not persist at three months (p=0.587).134

Only one study in this body of literature included vaginal electrical stimulation.269 In this study of pelvic floor muscle training with or without biofeedback compared to electrical stimulation, about half of women reported subjective improvement or cure in OAB when treated with electrical stimulation or biofeedback assisted pelvic floor muscle exercises, compared to 38 percent of women instructed in pelvic floor muscle exercises and told to perform them at home.269

Frequency outcomes. Number of voids per day was a common, objective measure of treatment effectiveness for behavioral interventions – particularly as a key element of the training is generally encouraging patients to extend time between voiding progressively, with a goal of reaching three to four hours between voids. Bryant found greater reduction in frequency when a caffeine reduction component was added to bladder training (reduction of 4.3 versus 3.3; p=0.037).136 Women who simultaneously reduced caffeine intake had a 35 percent decrease in number of voids per day compared to 25 percent for those using bladder training alone. In the study described above of multicomponent behavioral intervention compared to placebo (and oxybutynin), frequency of micturition was reduced significantly in the behavioral arm, but not the placebo arm with a reduction of 1.8 micturitions per day.

Urodynamic outcomes. Three studies examined urodynamic measures pre- and post- treatment.131, 133, 135 In all of the studies, increased bladder capacity was seen along with changes in incontinence and frequency measures.

Pelvic muscle strength. Wang and colleagues269 found that biofeedback associated pelvic floor muscle exercises resulted in greater change in muscle strength than electrical stimulation, but the clinical significance of the change was not examined.

Patient reported outcomes. The nature of OAB is such that while specific morbidities and mortality are not primary outcomes, the interference that OAB creates in patients’ lives, along with embarrassment and stigma, is often the concern that results in treatment seeking. Therefore, characteristics that make up and affect quality of life are of concern for studies of women with OAB, and are the focus of the results presented below. Four studies presented data on patient-reported outcomes other than urgency and frequency or cure/improvement.130, 132, 134, 135

Wyman134 examined the potential impact of combining pelvic muscle exercises with bladder training in an RCT of 204 women, of whom 59 had detrusor instability (results for the 145 women with SUI are not reported here as they are not relevant to this report). Although both impact (measured by the IIQ-R) and quality of life (measured by the UDI) were most improved in the combination group, relative to pelvic muscle exercises alone immediately after treatment, those effects were not sustained three months later. The combination group also had greater perceived improvement immediately after treatment, but again, no differences by group were sustained.

In the Burgio study135 comparing biofeedback to verbal feedback and a pamphlet to teach multicomponent behavioral training, both biofeedback and verbal feedback performed equally well on all measures of patient perceptions of improvement, with verbal feedback better than the self-help booklet on five measures (accidents are smaller, p<0.006; comfortable with treatment, p=0.01; description of progress, p<0.001; satisfaction with progress, p<0.001; and restriction of activities, p=0.002). The biofeedback group performed better on three measures (description of progress, p<0.001; satisfaction with progress, p=0.03 and restriction of activities, p=0.002). The biofeedback and verbal feedback groups saw no significant differences. All three groups had significantly improved quality of life scores (per Hopkins Symptom Checklist and Incontinence Impact Questionnaire), with no by-group differences. Similarly, Wang and colleagues269 found that biofeedback produced greater change in the overall Kings Health Questionnaire score among the group with biofeedback assisted pelvic floor muscle training relative to those who received PFMT without biofeedback or those who received electrical stimulation. Quality of life and satisfaction outcomes are summarized in Table 23.

Table 23. Effects on quality of life and satisfaction of behavioral treatments.

Table 23

Effects on quality of life and satisfaction of behavioral treatments.

Complementary and Alternative Therapies

Complementary and alternative therapies span a broad range of which only a small subset of modalities were used in studies of OAB. Acupuncture is an ancient Chinese medical system based on the balance of subtle energy flows (chi) in which imbalance of energy flows can result in disease. Acupuncture therapy aims to manipulate these energies through the insertion of fine needles at key, highly specific points related to chi flow to specific organs for varying periods of time. Foot reflexology is a variation of acupressure that postulates all body organs have corresponding external “reflex points on the foot” and the manipulation of these points can enhance the flow of energy to the reference organ. Specific areas of the sole of the foot are treated for specific medical conditions or symptoms. Hypnotherapy involves direct suggestion of symptom removal through therapeutic relaxation. Treatment of OAB is aimed at reduction of the component symptoms of urgency, frequency, and UUI.

We identified three publications that used complementary and alternative medicine therapies to treat OAB: a fair quality trial of acupuncture,137 a fair quality trial of foot reflexology,138 and a poor quality prospective case series of hypnotherapy.139 See complete evidence tables in Appendix C.

Acupuncture. The acupuncture trial was conducted in Oregon at an academic teaching center and was notable for incorporating sham acupuncture treatment as the comparison group.137 Among 85 women randomized, 74 (87 percent) completed all four weekly treatment sessions and had complete outcome data at two to four weeks after treatment. Outcomes included comparison of baseline and post-treatment three-day voiding diaries as well as cystometrics, measurements of post-void residuals, the Urinary Distress Inventory, and the Incontinence Impact Questionnaire.

Episodes of urge urinary incontinence were statistically equivalent across groups at completion of four weeks of treatment. Number of voids per day were reduced 14 percent in the acupuncture group compared to 4 percent in the sham treatment group (p=0.03). This equated to a reduction of 1.4 voids per day among those receiving acupuncture. The experience of symptoms of urgency was reduced 30 versus 3 percent, with those treated having 1.6 fewer distinct episodes of awareness of urgency per day (p<0.02).

Some measures, including functional bladder capacity and cystometric maximum capacity, were modestly improved in the treatment group (p<0.05); others, including volume at urge to void and detrusor contractions during cystometry, were comparable across groups with no apparent trends. Scores improved meaningfully on both validated instruments that evaluate distress and impact on quality of life, with statistical significance.137 No subjects withdrew for adverse events and treatment was well tolerated.

Reflexology. The study of foot reflexology was conducted in an academic center in Hong Kong. They incorporated sham reflexology in the form of a nonspecific foot massage without deep pressure. Among 120 women randomized, 97 (81 percent) completed all treatments and the assessment at three weeks. The reflexology group had three of 60 drop out; the sham group had six of 60 drop out. Losses were related to various competing demands, including four individuals who reported fear of SARS which was a threat during the study period in Hong Kong, with five losses to followup for other personal or medical reasons. There were no withdrawals because of discomfort or complications of treatment. Outcomes included comparisons of baseline and followup 24-hour voiding diaries and the King’s Health Questionnaire.

At completion, number of urge incontinence episodes, urgency episodes, and nocturnal voids were equivalent across groups. Daytime voids were reduced by 1.9 voids in the reflexology treatment group compared to 0.55 in the sham massage group (p=0.03). Quality of life measures did not differ. The authors note that their participants may have been unmasked by their familiarity with what to expect from reflexology treatments: as 88.9 percent of those in the reflexology group and 67.4 percent of those in the sham massage group believed that they had received “true” reflexology. This difference in unblinding could bias the findings.

Hypnotherapy. The hypnotherapy study was conducted in a UK academic setting and followed 63 women who were prospectively enrolled, had urodynamics, received 12 weeks of weekly hypnotherapy, and had followup urodynamic evaluation. Descriptive information is provided with little statistical analysis.139

Ten of 63 participants discontinued hypnotherapy before completing all sessions. Of those completing all sessions, 29 were reported to be entirely free of OAB symptoms with 14 “considerably improved”. Of 44 women who had repeated urodynamics, 22 initially classified as having unstable bladder “converted to stability”, other improvements in cystometrics were also reported. This case series lacks masking of assessors and does not provide key patient reported outcomes.

In summary, a well-conducted small trial of acupuncture has intriguing results related to decreased frequency of voiding and reduced symptoms of urgency which are associated with changes in cystometrics related to improved bladder capacity that are logical intermediates of the improvement in symptoms. Women in the study felt they were improved as measured by scales that capture bother and quality of life. This evidence is insufficient to support definitive choice of acupuncture but offers preliminary information that promises modest improvements that are similar to those reported in many pharmacologic trials.

Reflexology is represented by a small trial with unmasking of participants that could have biased the results. There is not evidence to support choice of this modality. Likewise, hypnotherapy is not supported by the scant information provided by one case series with little detail, patient reported outcomes, or statistical assessment. Given the scope of placebo effects demonstrated in other well-conducted studies of OAB treatment, it is difficult to know whether to attribute any effect to hypnotherapy.

KQ3: Comparisons of Treatments

Direct comparisons of treatments are made between entire approaches to management of OAB (e.g., pharmacologic to behavioral) or within approaches (e.g., drug to drug). In this section we present the results of any direct comparisons of either type, beginning with comparisons within the pharmacologic approach, followed by procedures compared to medical therapy or one another, and finally any comparison involving behavioral approaches. This third group comprises three distinct subgroups: behavioral compared directly to pharmacologic; combination pharmacologic plus behavioral compared to pharmacologic alone, and combination behavioral plus pharmacologic compared to behavioral alone.

Comparisons between pharmacologic treatments

All trial arms for RCTs of pharmacologic treatment for OAB are presented in evidence tables in Appendix C. Specific comparisons have been made in the literature for the following pairs of drugs that describe differences in reduction in urge urinary incontinence or voids per day:

These studies are generally powered simplistically only to assess non-inferiority; studies would need to be much larger for full assessment of comparability at robust power for small differences between pharmacologic agents. Nonetheless, in the majority of comparisons, neither drug was reported more effective at reducing either urge urinary incontinence episodes or voids per day with a few exceptions (Tables 24 and 25).

Table 24. Direct comparisons of pharmaceutical treatments on urge incontinence.

Table 24

Direct comparisons of pharmaceutical treatments on urge incontinence.

Table 25. Direct comparisons of pharmaceutical treatments on voids per day.

Table 25

Direct comparisons of pharmaceutical treatments on voids per day.

Both oxybutynin and tolterodine in their extended release forms demonstrated superiority in reducing incontinence episodes over tolterodine immediate release.83, 140 In the OBJECT trial, oxybutynin 10 mg ER was compared to tolterodine 2 mg IR twice a day.83 At the end of 12 weeks, women taking oxybutynin reduced their episodes per week of urge urinary incontinence from 25.2 to 6.2, compared to a change from 25.1 to 8.5 in the tolterodine arm. The difference of 2.4 episodes per week between groups was statistically significant. However, upon stratifying by age group, the difference was maintained only among those age 64 and younger. Two studies compared the effectiveness of tolterodine 4 mg once per day to tolterodine 2 mg, taken twice per day, and found that the extended release formulation resulted in significantly greater reductions in incontinence episodes.140, 141

Sand and colleagues found that voids per week diminished from 91.7 to 68.0 in the oxybutynin 10 mg extended release arm, compared to 91.6 to 71.2 in the tolterodine 2 mg immediate release twice a day arm (p=0.024).83 However, as with the difference observed for incontinence episodes, upon stratifying by age, the difference was maintained only among those 64 years and younger. Diokno and colleagues (2003)84 observed greater reductions in voids per week (p=0.05) in women taking oxybutynin 10 mg ER compared to those taking tolterodine 4 mg ER in the OPERA trial. Both studies provided treatment for 12 weeks, and data were obtained via bladder diaries. Harms were rare in both studies, although Diokno and colleagues report significantly higher rates of dry mouth with oxybutynin (p=0.02).

No other comparisons yielded statistically significant differences in terms of our primary outcomes.

Comparisons between procedural and pharmacologic treatments

The only procedure to be compared to another treatment modality was sacral neuromodulation, which was compared to medical therapy in one RCT. In this study, 98 participants refractory to medical therapy were randomized to immediate sacral nerve stimulation or delayed sacral nerve stimulation. The delay group continued unspecified medical management for a six month period before having the procedure. The study found a reduction in daily urge urinary incontinence episodes from 9.7 to 2.6 in the sacral neuromodulation group, compared to an increase from 9.3 to 11.3 in the medical management group at six months (p<0.01).124 At 18 months, 76 percent of patients reported that they were completely dry or had experienced a reduction in symptoms of 50 percent or greater. It is important to note that those receiving medical therapy knew they were awaiting treatment with a modality that they were invested in believing was superior to their current level of symptom management. The differences in risk between sacral neuromodulation and medical management are important. Six patients had permanent explantation: three for pain, two for infection, and three for change in bowel function.

Comparisons between behavioral and pharmacologic treatments

Nine studies, with 11 publications, included behavioral and pharmaceutical arms in direct comparison to one another.93, 143–150, 201, 254

The literature base included one prospective cohort study,147 and eight RCTs.93, 133, 143–146, 148, 150

Three of the studies were conducted in Europe, two were in the United States, two were in Asia (Taiwan and Korea), one in Brazil, and one in New Zealand. Seven were conducted at academic medical centers; two were in community settings.

The behavioral approaches examined included bladder training,146–150, 254 multicomponent behavioral approaches,143–145 pelvic floor training,201, and electrical stimulation (Table 26).93

Table 26. Direct comparisons between pharmacologic and behavioral interventions.

Table 26

Direct comparisons between pharmacologic and behavioral interventions.

Episodes of Incontinence. Two RCTs considered urge urinary incontinence episodes as a primary outcome.93, 143 One compared multicomponent behavioral interventions to pharmacologic interventions143 and reported on changes in incontinence episodes. This study was a three-arm study in which multicomponent behavioral treatment was compared to pharmacologic treatment and to placebo. The behavioral arm had a significantly higher percent reduction in episodes of incontinence at the ten-week followup (p<0.001) (80.7 percent compared to 68.5 percent for pharmacologic and 39.4 percent for placebo). In a pilot study comparing oxybutynin to bladder retraining, no difference was observed in effectiveness by group.254 The authors calculated that to observe a difference between these arms in a full-scale study would require 165 women in each arm, rather than the approximately 20 in the pilot. This study also had a combination therapy arm. Two studies examined the effects of electrical stimulation with different comparison groups.93, 201 One93 found no difference in reduction of incontinence comparing electrical stimulation to oxybutynin or to placebo either within or by groups. The other201 showed no difference in effectiveness between three groups: electrical stimulation, oxybutynin (5 mg b.i.d.) or pelvic floor exercises.

Cure. Three studies reported on numbers of patients who achieved “cure” or resolution of UUI, without further definition.146, 147, 149 The results were inconsistent, the studies were of poor quality, and all were conducted prior to the ICS definition of OAB. The behavioral intervention in Jarvis, 1981149, was inpatient bladder drill, which is not a current treatment approach.149 Diokno (1995) is a report on a series of patients who chose their own treatment modality.147 Only Colombo (1995) in this series was an RCT.146 Bladder training was provided over a six-week period on an outpatient basis; although cure rates at the end of treatment were essentially the same, after six months, those who had received bladder training maintained a higher cure rate (96 percent of those initially cured in the bladder training group versus 57 percent in the oxybutynin group). However, the numbers reported in this study were quite small (only 53 in all groups at six months followup).

Frequency. Of the six studies that provided data on voids per day, all found that both pharmacologic and behavioral approaches could reduce frequency, but that there was no difference between the two approaches.93, 145, 148, 150, 201, 254 Goode and colleagues conducted a secondary analysis of the original Burgio and colleagues 1998 study. They used structural equation modeling to determine that changes in voiding frequency were not mediating factors associated with decreases in incontinence.145

Similarly, both bladder training approaches (either multicomponent or bladder drill) and pharmacologic treatment were associated with urodynamic changes, including increased maximum cystometric capacity overall and at first and strong desire to void. These changes were statistically significant, and large enough to be clinically relevant as well. However, Goode and colleagues145 examined the potential role of urodynamic changes to mediate the perceived effects of behavioral and pharmacologic treatment on incontinence and found, once again, that they did not seem to be associated with observed reductions in incontinence.145

Patient reported outcomes. Burgio and colleagues used the Hopkins Symptom checklist to consider psychological changes potentially associated with improvement in continence.144 Behavioral management and individually titrated oxybutynin were both associated with improvement in psychological status overall (including the placebo condition) and on a range of subscales, but psychological changes measured on the Hopkins Symptom checklist did not correlate with changes in rates of incontinence episodes. Satisfaction was highest among patients who received behavioral management (77.6 versus 54.7 percent with oxybutynin); interestingly, satisfaction on placebo was not substantially different than that on drug (43.1 percent). A very high proportion of the women receiving multicomponent behavioral training felt that they were “comfortable enough” to continue with the approach (96.5 percent) relative to those on drug (54.7 percent) or placebo (43.1 percent).

Comparisons of combined behavioral and pharmacologic treatment to pharmacologic treatment alone

Seven studies examined the effect of adding a behavioral intervention to drug compared to drug alone, one of which was a feasibility study.150–155, 254 In all but two studies, the drug was tolterodine. The literature included six RCTs 143, 150, 151, 154, 155 and two randomized open-label trial (Table 27).153, 254

Table 27. Comparisons of pharmaceuticals with and without behavioral interventions.

Table 27

Comparisons of pharmaceuticals with and without behavioral interventions.

Two of the studies were conducted in the United States, one was in Canada, one was in multiple Scandinavian countries, one in Korea, one in New Zealand, and one did not specify, but indicated that it took place internationally in multiple sites. The two United States studies, the Korean one, and the one in New Zealand were conducted at academic medical centers.

Burgio and colleagues (2008) provided multicomponent behavioral treatment as an adjunct to pharmacologic treatment compared to pharmacologic treatment alone152 to examine the potential for behavioral management to aid patients in ceasing medication use and staying off. The primary endpoint of interest was a combined effect of a 70 percent reduction in incontinence plus no medication use or other therapy for incontinence. This study examined outcomes at 10 weeks, immediately after treatment, and at 8 months after a period of no treatment to examine persistent effects. During the post-treatment period, participants could request a return to medication. The participants in the behavioral group reported a greater reduction in episodes of incontinence than those in the oxybutynin group at 10 weeks; however the difference did not persist to 8 months (ability to discontinue drugs was 41 percent in both groups at 8 months). The two groups also experienced similar percentage reductions in episodes of urge incontinence (20.4 percent in the behavioral group versus 18.5 percent in tolterodine alone). The behavioral group did report statistically significantly greater satisfaction or quality of life.

Three additional studies compared the effectiveness of combinations of drug and behavioral approaches to drug alone in changing episodes of incontinence as well as episodes of urgency per day.151, 153, 154 Differences between the groups were small and non-significant, although there were significant decreases within all groups. Episodes of urgency decreased in all groups as well, with decreases ranging from approximately 1.9 to 2.7 episodes per day, but again there were no differences between the study groups.

Two studies found that adding behavioral training to tolterodine was associated with further reductions in frequency compared to tolterodine alone;151, 152 two found no significant effect of adding behavioral training.150, 155 However, the type of the training provided differed dramatically – for example, the intervention provided by Herschorn155 was purely informational, while Burgio152 provided a multicomponent system that included biofeedback and pelvic floor muscle exercises. Per one study,154 addition of pelvic floor exercises alone in addition to tolterodine immediate release (2 mg b.i.d.) did not confer added reductions in frequency.

In those studies that measured quality of life and participant satisfaction, improvements were significantly greater among those patients receiving combination therapy compared to those receiving pharmacologic therapy alone (Table 28).

Table 28. Effect on quality of life and satisfaction of combination treatment.

Table 28

Effect on quality of life and satisfaction of combination treatment.

Comparisons of combined behavioral and pharmacologic treatment to behavioral treatment alone

Three studies measured the effect of adding a pharmaceutical approach to a behavioral one. All three of these studies used bladder training as the behavioral technique.

Ghei and colleagues253 describe a series of cases in which patients chose management approaches that could include either bladder training alone, or bladder training in addition to an antimuscarinic agent. Only 52 of 708 patients chose the bladder training alone, and although they experienced greater reduction in frequency (p<0.0001), those in the combination group had greater reductions in incontinence episodes (p=0.024).

A retrospective chart review of 92 patients treated with bladder retraining drill, among whom 36 also received antimuscarinics, was reported by Fantl et al.252 The outcome of cure, in this clinical populations with 6 months to 6 years of followup, was defined as no further episodes of incontinence and voiding every 3 to 5 hours with no associated symptoms. Cure was achieved in 83.3 percent of patients using bladder training with antimuscarinics and 78.6 percent of patients using bladder training alone; this difference was not significant (p>0.6).

In a double-blind, placebo-controlled RCT, patients were randomized to receive placebo or oxybutynin 2.5 mg twice a day in addition to bladder training.255 The patients taking oxybutynin had a greater reduction in daytime frequency when compared to the patients taking placebo (p<0.05). There were no differences between groups in the change in incontinence episodes.

Finally, Burgio and colleagues251 provided the opportunity to patients in their trial of biofeedback-assisted therapy versus oxybutynin143 whose treatment was not completely successful from the patient perspective to receive combined pharmacologic and behavioral management. Of the 35 individuals who met criteria and agreed to move onto combined treatment, eight crossed from behavioral alone to combination, and 27 went from pharmacologic alone to combined. Both groups experienced significant reductions in incontinence over the effect of the initial treatment. The behavioral to combined group went from 58 percent reduction at the end of single therapy to 89 percent reduction after combined therapy for an additional eight weeks. The pharmacologic to combined group also improved from 73 percent reduction to 84 percent reduction in incontinence episodes.

KQ4. Modifiers of Treatment Outcomes

This section includes information related to how individual characteristics may influence likelihood of responding to treatment and outcomes of treatment. We included publications that explicitly presented stratification by a baseline characteristic that can be determined in a clinical setting and that presented statistical analysis related to interpreting the influence of the characteristic on treatment effects as they relate to outcomes. Detailed analyses of personality characteristics or psychometrics that require specialized expertise or survey instruments unlikely to be used by those providing care for women with OAB are not reviewed here.

Identified Modifiers

Age. Eight publications examined the relationship of age to response to pharmacologic treatment.83, 102, 156–161

Tolterodine was the focus of four of these studies;156–158, 160 and one compared oxybutynin to tolterodine.83 The largest study of symptom-related outcomes was an open label clinical cohort with 2,250 patients from 462 urology practices in Germany. The mean dose received was 3.8 ± 1.2 mg, with a median of 2 mg. Average age of those treated was 61 ± 14 years with range not provided. Increasing age was associated with being more likely to have incontinence episodes which were not strictly required to be urge incontinence. In multivariable regression models, with age as a continuous variable, each year of increasing age was associated with small absolute reductions in global efficacy (OR for global efficacy 0.986; 95 percent CI: 0.98, 0.99 per year of age). Global efficacy was defined as eight or fewer voids a day, fewer than two urge episodes, and no incontinence episodes per day.157

The largest RCT evaluated age effects among 1,015 participants and found that tolterodine drug was superior to placebo among patients aged younger than 65 compared to those 65 and older. They reported that treatment effects on incontinence episodes per week, voids per day, and subjective reports of experience of urgency and the ability to hold urine, were comparable across age groups at 12 weeks. A side-light of interest was that placebo effect was more pronounced for reduction of voids per day in the younger group (p<0.045); however change from baseline in the treatment arm was comparable regardless of age.158

The IMPACT trial enrolled 896 individuals with urge incontinence from primary care settings and conducted an open-label evaluation of tolterodine extended release 4 mg once daily.160, 239 All parameters, including UUI, urgency, frequency, and nocturnal frequency, were improved at 12 weeks for both those younger than 65 and those older. However, decreases in frequency were less pronounced among the older group who experienced on average a 22.2 percent reduction in daytime frequency (95 percent CI: −26.7, −15.2) and 28.6 percent decrease in nighttime frequency (95 percent CI: −35.7, −20.0); while younger participants had a 33 percent decrease in daytime frequency (95 percent CI: −36.0, −30.3) and a 50 percent decrease in nighttime frequency (95 percent CI: −53.8, 40.0). The most common treatment-related adverse events among those younger than 65 were dry mouth (11.4 percent), constipation (2.7 percent), and dry eyes (1.0 percent); among those who were older dry mouth (6.6 percent), constipation (4.4 percent) and headache (2.2 percent) were most common. Retention (<1 percent) occurred only among those older than 75 and none of four individuals with this complication required catheterization. Statistical comparisons for harms by age were not provided.160

A prescription-event monitoring study, conducted in the United Kingdom to assess population impact of tolterodine (range 1 to 4 mg) entering the prescription drug market, monitored more than 14,500 patients who filled prescriptions over a minimum of six months.156 Average age was 63 ± 16. Analysis of risks of rare adverse events (fewer than 40 events in the full population) found the upper quartile of age, those over 74, had greater risk of rare events including hallucination, heart palpitations, and tachycardia. Those under 50 had the lowest risk of cardiac events. No note was made of whether more common side effects varied with age.

A small comparative study of oxybutynin (10 mg once daily) and tolterodine (2 mg twice daily) (n=315) found an advantage for oxybutynin extended release over tolterodine twice daily for decreasing urge incontinence, urgency, and frequency, among those 64 and younger. This effect was not apparent in older age groups in which both were comparably effective.83

Other pharmacologic treatments. A single placebo controlled RCT of trospium (20 mg twice a day) evaluated whether CNS adverse effects, specifically daytime drowsiness, varied with age. Using the Stanford Sleepiness Scale, they found fewer than 1.5 percent of those on trospium (and 2.5 percent of those on placebo) experienced a clinically relevant three or more point increase. Using continuous scores, neither age grouping as <65 and older or <75 and older revealed meaningful differences. Average changes in scores across groups were most often improvements of less than half a point, generally less than a quarter point.161 Darifenacin has been studied among those age 65 and older in a two-year, open-label extension of 716 participants, that documented comparable effectiveness among older and younger participants with respect to sustained or improved treatment response over time as defined by a global response score, and individual measures that included incontinence episodes, voiding frequency, urgency, and OAB-related nocturnal waking.159 All statistical testing for these outcomes across time points through 24 months had p<0.088 for the comparisons with baseline status.159 A trial that allowed dose adjustment of darifenacin over the course of the study found the mean age of those requiring dose adjustments “for additional efficacy” was equal to those who did not change dose.102

Though some studies reported reduced efficacy for specific endpoints among older participants in their study populations, none reported complete lack of benefit among older populations.

Prior treatment. Seven publications investigated whether prior treatment with antimuscarinics predicted treatment response.85, 86, 102, 162–165 In three placebo-controlled trials of oxybutynin patch,85, and tolterodine IR,162 participants who had previously been on antimuscarinics had comparable outcomes to those who were treatment naïve. The tolterodine study specifically commented on prior treatment failures, noting improvements in those who had failed prior treatments that were above placebo but not statistically significant; few participants were in this category.162

In two drug-to-drug comparison trials, outcomes, including UUI, total incontinence episodes, and “perceived improvement of bladder condition,” were likewise reported to be comparable for treatment naïve and previously treated participants. These studies investigated oxybutynin ER (5 mg and 10 mg) and tolterodine ER (2 mg and 4 mg);86 and oxybutynin ER (10 mg) and tolterodine ER (4 mg).164 A nine-month open label study of tolterodine 2 mg twice daily found that 89 percent of individuals previously unable to tolerate oxybutynin, tolerated tolterodine well.163 One study included participants who had not been on medications as well as those switching from oxybutynin immediate release to extended release. In a subanalysis of those switching from immediate to extended release, the total proportions continent at 12 weeks is similar to that presented for the whole study population; no statistical test related to this comparison is provided. Some who had been on IR dosing regimens had worsening of symptoms on comparable total doses of ER.165 This study was small (n=256 spread across 16 centers without placebo comparisons) which hinders interpretation. In a trial that allowed dose adjustments, prior treatment was associated with higher rates of dose increase.102

Baseline severity

Presence, type, and severity of incontinence. Two studies contrasted those with UUI at baseline to participants without UUI. In an open-label study of tolterodine 3,824 participants with nine months of treatment, urge, frequency, nocturia, and OAB scales were similarly improved regardless of UUI baseline status.166 The VOLT study of solifenacin was an open label study of flexible dosing.167 Participants who had UUI at baseline and reported it was their most bothersome symptom, reported improvements in urgency, UUI, frequency, and nocturia, from “moderate to severe” to “very minor to some minor.” With the exception of nocturia, the point estimates for improvement in individual measures were better in the group with UUI at baseline, although not always statistically significantly better.167 Severity of UUI was not significantly associated with improvement in UUI. Subjects with severe UUI (defined as >20 episodes per week) had 67.6 percent decrease in UUI episodes, compared with 71.4 percent in subjects with less severe UUI.168

The response of those with UUI (n=552) was compared to those with urge-predominant mixed urinary incontinence (n=171) in a 16-week single blind trial of tolterodine that allowed dose adjustments from 1 mg twice daily to 2 mg twice daily. The authors reported cure rates were comparable: 44 percent of those with UUI achieved “dryness” and 39 percent of those with mixed urinary incontinence; 24.0 and 23.5 percent respectively had normalized voiding frequency (<8 voids per 24 hours).169 Other studies included participants with urge-dominant mixed urinary incontinence, or failed to specifically address inclusion or exclusion of those with mixed urinary incontinence, and did not report assessing trial data for differential treatment effects.

Specific symptoms and outcomes. A single study of transdermal oxybutynin and tolterodine ER 4 mg reported on ability to improve frequency in the lowest quartile of frequency, which was those with <10 voids a day at baseline, and found that neither drug had a significant effect in that subgroup compared to placebo.88 The IMPACT trial, an open label trial of tolterodine ER in primary care practices, documented improvements among all groups of participants whether the most bothersome symptom was daytime frequency, nocturnal frequency, urge urinary incontinence, or urgency. Statistical testing was not provided to determine if specific groups were more likely to benefit; the data presented suggests similar improvements in self-reported improvement, symptom bother, coping, health-related quality of life, and AUA symptom index scores across all groups.239

Other authors used severity scales or single questionnaire items to group participants and to conduct sub-analyses by severity. In the German open-label, observational study of tolterodine 2 mg twice daily, patients with greater baseline symptom severity experienced greater magnitude of improvements. This trend was for global efficacy; individual component symptoms varied across groups and were not statistically significant. Overall, those with severe baseline symptoms were less likely than others be symptom free at 12 weeks.157 In the ACET study among those whose self-reported symptom scores were graded as moderate to severe, those on tolterodine ER achieved greater magnitude of improvements on a global improvement scale than those on oxybutynin ER; comparison across moderate-to-severe to less affected participants was not provided.86

Urodynamic findings. Four publications related baseline urodynamic findings to outcomes of treatment.89, 170, 171, 173 Three of these studies did not identify urodynamic findings that predicted poor response or non-response to treatment. The finding of detrusor overactivity compared to its absence was not associated with outcomes in a case series of 365 women treated with bladder retraining and oxybutynin 2.5 mg twice a day. Both women with and without detrusor instability had comparable benefits from treatment, as measured by voids per day and incontinence episodes.171 Likewise in an RCT of duloxetine (an SSRI) versus placebo which showed treatment benefits for OAB with and without UUI, the classification of participants as having detrusor overactivity or sensory urgency based on urodynamic findings was not statistically significant as an effect modifier or predictor of treatment outcomes.170 The third study to group participants by urodynamic finding had group sizes (n = 6, 25, 36, 40) too small to make definitive assessments but suggested in the two larger groups that those with low volume and high pressure profiles had comparable results to those with low volume and low pressure profiles. The last case series, reporting on 111 women with OAB based on symptoms and urodynamic diagnosis of detrusor overactivity, found that those women who had involuntary detrusor contractions with provocative maneuvers like coughing, washing hands in cold water, and the sound of running water, were less likely to respond to treatment, which was not operationally defined by the authors, with tolterodine 4 mg a day than those who did not respond to provocation (p=0.0008).172

A last study conducted urodynamics among 1,133 women of whom 132 met the criteria for detrusor overactivity “conforming” to the definitions of the ICS; the researchers then compared treatment outcomes among women with and without coital incontinence. Among women with DO treated with tolterodine ER 4 mg daily for 12 weeks, those women with coital incontinence at orgasm were more likely to be non-responders to treatment (“unimproved by self-report”) for their OAB (41.2 percent) than those without coital incontinence (17.0 percent), (p=0.23).173

Other candidate modifiers

Race and ethnicity. The VOLT study, an open label trial of solifenacin has resulted in two publications addressing race and ethnicity. The authors report these analyses were motivated by lack of literature that explicitly addresses whether minority groups achieve comparable outcomes to the overall study group; the publications focused on Hispanic and black participants in VOLT. They provide detailed symptom and outcome profiles and report that outcomes were similar to the larger cohort. However direct statistical comparison is not provided by race/ethnicity.205, 229

A single three-arm trial of tolterodine ER 4 mg, tolterodine IR 2 mg twice daily, and placebo (n=1,235) reported that women who were above the mean for BMI (> 27kg/m2) were more likely to have UUI at baseline but achieve comparable reductions in number of incontinence episodes, to those with lower BMI after 12 weeks of treatment.140

Anterior vaginal wall prolapse. Anterior vaginal wall prolapse was the sole pelvic organ prolapse measure that we identified in the literature about treatment outcomes. Anterior vaginal wall prolapse was defined as descent to at least 1 cm proximal to the hymen (POP-Q measurement stage IIa). Women with such prolapse in this clinical case series of women treated with 4 mg of tolterodine extended release formula once daily were more likely to show no improvement at 12 weeks (39.2 percent) compared to those without prolapse (14.1 percent) using a three-point scale of no change, improved, and cured (p=0.0002).243

Gender. Authors frequently reported that men, especially older men, fared less well in resolution of symptoms of OAB.85, 106, 156, 157, 174 This evidence review was focused on outcomes of treatment among women. However, in order to retain landmark studies we included a number of studies that enrolled men as long as the proportion of women in the study was 75 percent or more. This means that treatment effects may be attenuated when men are included.

KQ5. Costs of Overactive Bladder Treatment

This section presents the results of our literature search and findings about financial costs associated with treatment for overactive bladder. Direct medical costs, indirect medical costs, and lost productivity for individuals with OAB are significant issues, in part because these women may be at greater risk for comorbidities such as falls and fractures, urinary tract infections, depression, and skin conditions.188, 189 OAB symptoms can interfere substantially with work and other activities,250 affecting productivity. Nonetheless, the proportion of individuals with OAB who seek medical treatment for OAB has been estimated to be extremely low with possibly as few as 4 percent of individuals with symptoms seeking treatment.24 Therefore, estimates of costs of care for OAB are likely a reflection of care provided to only a small proportion of a group that may be able to benefit from effective treatment.

Although several studies have estimated the cost of illness of urinary incontinence, few studies in the United States have estimated the cost of illness of OAB, and even fewer have focused on treatment costs, as specified for this review. In particular it is difficult to estimate costs related to OAB because it is a symptom-based syndrome with significant inter- and intra-individual variation in etiology and in symptoms. To answer the question of costs related to the treatments reviewed in this systematic review, we restricted our analysis to studies that included direct costs of treatment for OAB in the United States.

Detailed information on all studies related to financial costs associated with management of OAB can be found in Appendix C. Studies analyzed costs for at least one of the following cost categories:

  • Direct costs. Depending on the bundle of treatment for OAB, costs may include PCP services, specialist care, skilled nursing facilities, home health care, and/or surgery. Direct costs include diagnostic costs, radiology costs, laboratory costs, treatment costs, and costs related to the consequences of OAB.
  • Indirect Costs. Indirect medical costs include lost productivity and an informal caregiver’s time.

Financial Costs

Content of the literature. We identified five studies on financial costs related to OAB that met criteria for inclusion.175–179 All studies included assessment of direct medical costs related to OAB, and two included costs due to lost productivity. One study additionally assessed financial implications for pain and suffering. Two additional studies do not meet criteria for measuring direct costs of treatment, but provide context related to health care utilization.24, 250

The literature base included three analyses of administrative-claims databases: one using Medicare data; one using a large, private health plan affiliated with Ingenix; one using seven plans administered by UnitedHealth Group. These studies analyzed claims for ICD-9 codes they selected as indicative of OAB; some ICD-9 codes were unique to specific studies.

One study was a national telephone survey of community-dwelling adults179 that was designed to measure prevalence in the community. A followup survey was sent to a selected set of age and sex-matched cases (with OAB) and controls to estimate treatment use, medication, routine care, OAB related consequences and work productivity.

One focused on community-dwelling adults, one on persons younger than 65 years of age who filled prescriptions for drug treatment of OAB, one on persons of any age with prescription-drug benefit who filled prescriptions for drug treatment of OAB, and one on persons who had failed tolterodine ER and sought alternative drug therapy. Each of the studies used a different definition of OAB (Table 29). Because Jensen’s study177 represents the Medicare population in 1994 to 1995 and is therefore likely substantially out of date, we do not summarize the data here. Data are available in the evidence table (Appendix C).

Table 29. Study definitions used for cost determination.

Table 29

Study definitions used for cost determination.

Costs of treatment. Only Varadharajan and colleagues provided expenditures on medications for OAB, comparing sets of matched patients who used either tolterodine or oxybutynin in their extended release form, or who used tolterodine ER compared to oxybutynin in its immediate release formulation. They did not provide estimates by gender. Costs of drug ranged from an average low of $56 over 12 months for oxybutynin immediate release to a high of $360 for extended release tolterodine. The extended release formulation of oxybutynin had intermediate costs at $317. Tolterodine extended release was significantly more expensive over the course of treatment than both oxybutynin ER and oxybutynin IR (p<0.001). These differences in part reflect higher number of prescriptions filled in the extended release groups over the study period compared to the immediate release group (p<0.0001). Nonetheless, total healthcare costs, including those specifically related to OAB, were highest for users of oxybutynin immediate release compared to the extended release formulations (p<0.0001).

Three studies examined total medical costs (not just treatment costs) for persons who filled prescriptions for OAB drug treatment (Table 30). These studies calculated the total healthcare costs for persons with OAB, per person per year by drug (PPPY). No studies did so by surgical or behavioral approach and only Varadharajan presented data for women only.175

Table 30. Total cost differences in annual medical care among persons filling prescriptions for OAB drug treatment.

Table 30

Total cost differences in annual medical care among persons filling prescriptions for OAB drug treatment.

In their followup survey of individuals with OAB, Hu and colleagues179 collected data on a range of health care utilization measures, including pharmacologic and surgical treatment for OAB. They used several sources of cost data, including the Red Book, to assign the annual costs of treatment for OAB in the United States in 2000. Pharmacologic treatment costs for women in 2000 were estimated at approximately $1.2 billion for women overall, with approximately equal costs across age groups. OAB surgical costs among women in 2000 were estimated at approximately $550 million.179 Overall health care costs, including lost productivity for the nation were $7.4 billion for women (approximately $3.1 billion for those under 65).

This body of literature is particularly challenged by the varying methodology for identifying and defining patients with OAB. In part, this variability is a reflection of the overall literature on OAB, which also uses varying definitions. Most problematic for definitional purposes is whether and when authors included incontinence in any of its forms.

None of the cost papers qualifying for inclusion used data beyond 2002, when the ICS definition of OAB was put into place.

Cover of Treatment of Overactive Bladder in Women
Treatment of Overactive Bladder in Women.
Evidence Reports/Technology Assessments, No. 187.
Hartmann KE, McPheeters ML, Biller DH, et al.

PubMed Health Blog...

read all...

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...