Nonsurgical Management of Urinary Incontinence in Women: A Clinical Practice Guideline From the American College of Physicians Clinical Guideline

Description: The American College of Physicians (ACP) developed this guideline to present the evidence and provide clinical recommendations on the nonsurgical management of urinary incontinence (UI) in women. Methods: This guideline is based on published English-language literature on nonsurgical management of UI in women from 1990 through December 2013 that was identified using MEDLINE, the Cochrane Library, Scirus, and Google Scholar. The outcomes evaluated for this guideline include continence, improvement in UI, quality of life, adverse effects, and discontinuation due to adverse effects. It grades the evidence and recommendations by using ACP’s guideline grading system. The target audience is all clinicians, and the target patient population is all women with UI.

U rinary incontinence (UI), the involuntary loss of urine, has a prevalence of approximately 25% in young women (aged 14 to 21 years) (1), 44% to 57% in middle-aged and postmenopausal women (aged 40 to 60 years) (2), and 75% in elderly women (aged Ն75 years) (3). However, these statistics may be underestimated because one study showed that at least half of incontinent women do not report the issue to their physicians (4). Risk factors for UI include pregnancy, pelvic floor trauma after vaginal delivery, menopause, hysterectomy, obesity, urinary tract infection, functional and/or cognitive impairment, chronic cough, and constipation (5). The effects of UI range from slightly bothersome to debilitating. Urinary incontinence also contributes to high medical spendingapproximately $19.5 billion was spent in the United States in 2004 -and it accounts for 6% of nursing home admissions for elderly women, costing approximately $3 billion (6).
The 2 types of UI are based on the dysfunctional mechanism: stress and urgency. However, the distinction is not always clear, particularly for older women. Stress UI is related to urethral sphincter failure associated with intraabdominal pressure and results in the inability to retain urine when laughing, coughing, or sneezing (7). Urgency UI is the involuntary loss of urine associated with a sudden and compelling urge to void (7).
Mixed UI is a combination of stress and urgency UI. Overactive bladder is a constellation of symptoms that includes urinary urgency (with or without UI), usually accompanied by frequency, and nocturia (5).
The primary goal of treatment is to achieve or improve continence (8,9). Clinically successful treatment has been defined as that which reduces the frequency of UI episodes by at least 50% (10). Treatments addressed in this guideline include lifestyle changes, pelvic floor muscle training (PFMT), and various approved drugs ( Table 1) (8). Surgical treatments, available for women in whom conserva-tive therapy has failed or who have anatomical abnormalities, are not addressed in this guideline.
This guideline from the American College of Physicians (ACP) presents the available evidence on the nonsurgical (pharmacologic and nonpharmacologic) treatment of UI in women in the primary care setting. It does not fully evaluate nonsurgical treatments, such as botulinum toxin or percutaneous nerve, magnetic, or electrical stimulation, because they are not typically used by or available to primary care physicians. The target audience includes all clinicians, and the target patient population is all women with UI. This guideline is based on a systematic evidence review sponsored by the Agency for Healthcare Research and Quality (11) and an updated literature search (Supplement, available at www.annals.org).

METHODS
This guideline is based on a systematic evidence review (11) that addressed the following key questions related to the diagnosis and nonsurgical management of UI: 1. How effective is the nonpharmacologic treatment of UI in women?
1a. How do nonpharmacologic treatments affect incontinence, severity and frequency of UI, and quality of life compared with no active treatment?
1b. How do combined methods of nonpharmacologic treatments with drugs affect incontinence, severity and frequency of UI, and quality of life compared with no active treatment or monotherapy?
1c. What is the comparative effectiveness of different nonpharmacologic treatments?
1d. What are the harms of nonpharmacologic treatments compared with no active treatment?
1e. What are the comparative harms of different nonpharmacologic treatments?
1f. Which patient characteristics, including age, type and severity of UI, baseline disease that affects UI, adherence to treatment recommendations, and comorbid conditions, can modify the effects of nonpharmacologic treatments on patient outcomes, such as continence, quality of life, and harms? 2. How effective is the pharmacologic treatment of UI in women?
2a. How do pharmacologic treatments affect continence, severity and frequency of UI, and quality of life compared with no active treatment or combined treatment methods?
2b. What is the effectiveness of pharmacologic treatments compared with each other or with nonpharmacologic treatments of UI?
2c. What are the harms of pharmacologic treatments compared with no treatment?
2d. What are the harms of pharmacologic treatments of UI compared with each other or with nonpharmacologic treatments?
2e. Which patient characteristics, including age, type and severity of UI, baseline disease that affects UI, adherence to treatment recommendations, and comorbid conditions, can modify the effects of pharmacologic treatments on patient outcomes, such as continence, quality of life, and harms?
The systematic evidence review was done by the Minnesota Evidence-based Practice Center. The literature search included English-language studies published between 1990 and December 2011 identified using MEDLINE, the Cochrane Library, Scirus, and Google Scholar as well as manual searches of reference lists from systematic reviews. Literature was updated through December 2013, focusing on treatments most relevant to primary care (see the Supplement for details). Data were extracted using a standardized form, and study quality was assessed according to the Methods Guide for Effectiveness and Comparative Effectiveness Reviews (12). This guideline focuses on treatments most relevant to primary care clinicians; the full report (11) and published article (13) contain more details.
This guideline rates the evidence and recommendations by using ACP's guideline grading system ( Table 2). Details of the guideline development process can be found in the methods paper (14).

DIAGNOSIS
Because most women with UI do not report it to their physicians (4), physicians should proactively ask female patients about bothersome UI symptoms as part of a routine review of systems. Clinicians should take a focused history and ask specific questions, such as the time of onset, symptoms, and frequency (4). Clinicians should also do a focused physical examination and evaluate neurologic symptoms. Asking such questions as "Do you have a problem with urinary incontinence (of your bladder) that is bothersome enough that you would like to know more about how it could be treated?" as part of a quality improvement intervention has been shown to increase appropriate care by 15% in patients aged 75 years or older (15).

TREATMENT
Complete continence, a clinically important improvement in UI (defined as reducing UI frequency by Ն50%), and quality of life were the primary outcomes assessed in the systematic review to evaluate the effectiveness of nonpharmacologic and pharmacologic treatments.

Nonpharmacologic Treatment
Appendix Other Treatments. Evidence was insufficient to determine the effectiveness of behavioral modification programs, a soy-enriched diet, or acupuncture for improving UI in women with mixed UI (11).

Comparative Effectiveness of Nonpharmacologic Treatments
No evidence showed that one nonpharmacologic treatment was superior to another in the various comparisons assessed for stress, urgency, or mixed UI. Further details are available in the full systematic review (11) and the Supplement.

Stress UI: Pharmacologic Treatment
Nonsystemic Estrogen Therapy Versus Placebo. Overall evidence was insufficient to determine the effectiveness of topical estrogen therapies at improving UI. Evidence showed increased continence and improved UI with vaginal estrogen formulations, but transdermal patches were associated with worsened UI. Studies used a range of estrogen applications.
Duloxetine Versus Placebo. Low-quality evidence showed that continence was reduced less with duloxetine than placebo (44, 45). High-quality evidence showed that duloxetine did not statistically significantly improve UI compared with placebo (NNT B , 13 [CI, 7 to 143]) (44, 46 -49). Low-quality evidence showed that duloxetine improved quality of life (45, 49, 50). However, quality of life did not improve in women with severe stress UI or overactive bladder (46, 51).

Urgency UI: Pharmacologic Treatment With Antimuscarinics
Darifenacin Versus Placebo. High-quality evidence showed that darifenacin improved UI compared with placebo (NNT B , 9 [CI, 6 to 18]) (52-54). Achieving complete continence was not studied as an outcome with darifenacin treatment. High-quality evidence also showed that darifenacin improved quality of life (11).
Solabegron Versus Placebo. Evidence was insufficient to determine the effect of solabegron on continence or improving UI, but low-quality evidence showed that it decreased the frequency of UI episodes in a dose-dependent manner (98).

Urgency UI: Other Pharmacologic Treatments
Evidence was insufficient to determine the clinical effectiveness of resiniferatoxin or nimodipine compared with placebo for treatment of UI (11).
Oxybutynin Versus Tolterodine. Low-quality evidence showed no difference between oxybutynin and tolterodine for achieving continence (100). Moderate-quality evidence showed no difference for improving UI (66,68,100,101).
Tolterodine Versus Trospium. Low-quality evidence from 1 study showed that tolterodine and trospium were similarly effective at treating urgency UI (100).
Solifenacin Versus Tolterodine. Evidence was insufficient to compare solifenacin with tolterodine for effects on continence or improvement of UI (11).
Trospium Versus Oxybutynin. Low-quality evidence showed no differences between trospium and oxybutynin for effects on continence or improvement of UI (100).
Other Comparisons. Evidence was insufficient to determine the comparative effectiveness on continence or improvement of UI for darifenacin, propiverine, solifenacin, or flavoxate versus oxybutynin; solifenacin versus darifenacin; or tolterodine or solifenacin versus propiverine (11).

Comparative Effectiveness of Pharmacologic Versus Nonpharmacologic Treatments
Low-quality evidence from 1 study showed that PFMT plus bladder training improved UI more than tolterodine alone (102).

Role of Patient Characteristics on Outcomes of Pharmacologic Treatments
Age. Moderate-quality evidence showed that age did not modify clinical outcomes associated with pharmacologic treatment (11). High-quality evidence showed that trospium, oxybutynin, and darifenacin effectively improved UI and quality of life in older women (52, 71,92). High-quality evidence also showed that solifenacin achieved continence more often than placebo, regardless of age (77).
Race. Evidence was inconclusive about differences among various racial groups.
Baseline Frequency of UI. Low-quality evidence showed that the baseline frequency of UI was not associated with statistically significantly different clinical outcomes for any drugs examined (11). However, women with more frequent UI episodes had slightly greater benefits with active pharmacologic treatment than placebo (103,104).
Prior Treatment Response. High-quality evidence showed that solifenacin achieved continence more than placebo regardless of the response to previous treatments; a larger dose did not improve an initially poor response (77).
Concomitant Treatments. Moderate-quality evidence indicated that trospium reduced the number of urgency UI episodes regardless of whether the patient was receiving other drugs. Patients receiving 7 or more concomitant medications had more adverse effects than those receiving fewer than 7 (105).
Obesity. Evidence did not show any difference in effectiveness of trospium in achieving continence in obese or nonobese patients (106).

Adverse Effects Nonpharmacologic Treatments
The risk for adverse effects associated with nonpharmacologic treatments was low.

Pharmacologic Treatments
Appendix Table 2 summarizes the adverse effects associated with pharmacologic treatments, which were similar within drug classes. The most commonly reported adverse effects associated with antimuscarinics included dry mouth, constipation, and blurred vision. Evidence showed that fesoterodine (high-quality; number needed to treat for harm [ (99). Dry mouth and insomnia were more frequently reported for oxybutynin than for tolterodine (100).
Dizziness was more frequently reported for trospium, and dry mouth and insomnia were more frequently reported for oxybutynin (100). Tolterodine has also been associated with increased risk for hallucinations (118). Nasopharyngitis and gastrointestinal disorders were more frequent with mirabegron than placebo (96,97).

SUMMARY
Nonpharmacologic therapies were effective at managing UI, had a large magnitude of benefit for increasing continence rates, and were associated with a low risk for adverse effects. Pelvic floor muscle training alone and in combination with bladder training or biofeedback and weight loss with exercise for obese women were effective at achieving continence and improving UI. Evidence was insufficient to compare nonpharmacologic therapies with one another or with pharmacologic therapies; head-to-head comparisons would be useful.
Pharmacologic therapies were effective and equally efficacious at managing urgency UI and had a moderate magnitude of benefit in achieving continence rates but were associated with adverse effects. In addition, evidence showed that some patients were likely to discontinue pharmacologic treatment because of adverse effects. Solifenacin was associated with the lowest risk for discontinuation due to adverse effects, whereas oxybutynin was associated with the highest risk.
Only darifenacin and tolterodine had risks for discontinuation due to adverse effects similar to placebo. Evi- dence was insufficient to compare most drugs with one another for safety and efficacy. Tolterodine and oxybutynin resulted in the same benefits, but tolterodine caused fewer harms. Of note, many studies did not fully specify details of the patient populations studied, including whether they received prior treatment for UI, which could potentially influence treatment response. The NNT B and NNT H should be interpreted with care because of inherent limitations with statistics expressing absolute benefits or harms. The patient population, disease severity, and treatment duration are factors that influence the NNT B and NNT H . In addition, because these statistics are derived from the risk difference, they are ultimately an expression of a specific treatment versus a specific control (active or placebo) and should not be used to indirectly compare NNT B and NNT H across various treatments. See the Figure for    UI is a common and important health care problem in women that is underreported and underdiagnosed. Clinicians should take a detailed history and ask specific questions about UI, such as the time of onset, symptoms, and frequency. Clinicians should use nonpharmacologic management for UI, such as PFMT for stress UI, bladder training for urgency UI, and PFMT with bladder training for mixed UI, because they are effective, have few adverse effects, and are cheaper than pharmacologic therapies. Although pharmacologic therapy can improve UI and provide complete continence, many patients discontinue medication because of adverse effects. Vulnerable populations include women aged >65 y, nursing home residents, and women receiving Medicare home care services. At least one half of women with UI do not report the issue to their physician. Pharmacologic treatment should be based on harms, because most drugs are similarly efficacious. Identifying and managing conditions that may cause UI, such as urinary tract infections; metabolic disorder; excess fluid intake; and impaired mental conditions, such as delirium, are important. Clinicians should identify whether patients are receiving medications that may cause or worsen UI.

SUMMARY OF THE AMERICAN COLLEGE OF PHYSICIANS GUIDELINE ON
PFMT ϭ pelvic floor muscle training; UI ϭ urinary incontinence.

Clinical Guideline Nonsurgical Management of Urinary Incontinence in Women
Pelvic floor muscle training increased continence rates and improved UI and quality of life in women with stress UI. Nonpharmacologic therapy with PFMT should be first-line treatment for women with UI.

Recommendation 2: ACP recommends bladder training in women with urgency UI. (Grade: weak recommendation, low-quality evidence)
Bladder training improved UI for women with urgency UI. The addition of PFMT to bladder training did not improve continence compared with bladder training alone for urgency UI.

Recommendation 3: ACP recommends PFMT with bladder training in women with mixed UI. (Grade: strong recommendation, high-quality evidence)
Pelvic floor muscle training combined with bladder training improved continence and UI in women with mixed UI.

Recommendation 4: ACP recommends against treatment with systemic pharmacologic therapy for stress UI. (Grade: strong recommendation, low-quality evidence)
Treatment of stress UI with standard pharmacologic therapies used for urgency UI has not been shown to be effective. Vaginal estrogen formulations improved continence and stress UI, but transdermal estrogen patches worsened UI.

Recommendation 5: ACP recommends pharmacologic treatment in women with urgency UI if bladder training was unsuccessful. Clinicians should base the choice of pharmacologic agents on tolerability, adverse effect profile, ease of use, and cost of medication. (Grade: strong recommendation, highquality evidence)
Pharmacologic therapies were effective and equally efficacious at managing urgency UI and had a moderate magnitude of benefit in achieving continence rates. However, they were associated with adverse effects and evidence showed that some patients were likely to discontinue pharmacologic treatment because of these effects. For urgency UI, oxybutynin, tolterodine, darifenacin, solifenacin, fesoterodine, and trospium increased continence rates and improved UI.
Evidence was insufficient to evaluate the comparative effectiveness of different drugs and to determine the longterm safety of pharmacologic treatments for UI. Patient characteristics, such as age, race, comorbid conditions, or baseline UI, did not affect the outcomes of the various pharmacologic medications. However, adherence to pharmacologic treatments for UI was poor.
Adverse effects were a major reason for treatment discontinuation. Clinicians and their patients should compare the risk for pharmacologic adverse effects with the severity and bothersomeness of the patient's symptoms. Appendix Table 2 shows the quality of evidence for outcomes of continence and improvement of UI as well as the adverse effects for the various drugs.
Evidence was insufficient to evaluate the comparative effectiveness of nonpharmacologic versus pharmacologic treatments for UI, and nonpharmacologic treatment should be considered first-line therapy. Evidence showed that nonpharmacologic treatments were better than no treatment in achieving continence and improving UI with a large magnitude of effect and are associated with a low risk for adverse effects. Pharmacologic treatments are associated with adverse effects that may be intolerable and lead to discontinuation of treatment. Clinicians and patients should keep in mind the costs of treatment, especially long-term costs, when choosing treatment.

Recommendation 6: ACP recommends weight loss and exercise for obese women with UI. (Grade: strong recommendation, moderate-quality evidence)
Weight loss and exercise improved UI in obese women with no evident harms. In addition, the benefits of weight loss in obese women extend beyond improvement of UI.

ACP HIGH-VALUE CARE
Urinary incontinence is a common and important health care problem in women that is underreported and underdiagnosed. Clinicians should take a detailed history and ask specific questions, such as the time of onset, symptoms, and frequency. Clinicians should use nonpharmacologic management for UI, such as PFMT for stress UI, bladder training for urgency UI, and PFMT with bladder training for mixed UI, because these therapies are effective, have few adverse effects, and are cheaper than pharmacologic therapies. Although pharmacologic therapy can improve UI and provide complete continence, many patients discontinue medication because of adverse effects. Cooke reports support for travel to meetings for the study or other purposes from the American College of Physicians; board membership on the National Board of Medical Examiners; consultancy for the University of Texas; employment at the University of California, San Francisco; and travel/accommodations/meeting expenses unrelated to activities listed from the American Board of Internal Medicine (ABIM) and the Accreditation Council for Graduate Medical Education. Dr. Fitterman is a member of the ABIM Examination Committee. To protect the integrity of Board Certification, the ABIM enforces the confidentiality and its ownership of ABIM exam content, and Dr. Fitterman has agreed to keep ABIM exam content confidential. No ABIM exam content is shared or otherwise disclosed in this article. Dr. Schwartz reports other support from the National Heart, Lung, and Blood Institute, National Institutes of Health, during the conduct of the study; personal fees from Allergan, Bayer, the Blue Cross and Blue Shield Association, General Electric, UBC, and Genentech; and grants from Pfizer. Authors not named here have disclosed no conflicts of interest. Authors followed the policy regarding conflicts of interest described at www.annals.org/article .aspx?articleidϭ745942. Disclosures can also be viewed at www.acponline .org/authors/icmje/ConflictOfInterestForms.do?msNumϭM13-2410. A record of conflicts of interest is kept for each Clinical Guidelines Committee meeting and conference call and can be viewed atwww.acponline .org/clinical_information/guidelines/guidelines/conflicts_cgc.htm.

ANNALS PERSONAE PHOTOGRAPHS FOR ACP'S CENTENNIAL
In recognition of the American College of Physicians' 100th anniversary, Annals of Internal Medicine is seeking photographs of internal medicine physicians to feature on each issue of the journal during 2015, the ACP's centennial year. In choosing from among submitted photographs, we will seek photos that capture personality and celebrate the diversity of individuals who devote their professional lives to the practice of internal medicine. Readers and others are encouraged to submit photographs of internal medicine physicians for consideration. In an effort to bring people to the pages of the Annals of Internal Medicine, the editors began publishing photographs of people in 1999. Annals published photographs in a section of the journal called "Personae" from 1999 to 2000, and photographs have appeared on the cover since 2000.
Written permission to publish the photograph from the subject (or subjects) of the photograph or the subject's guardian or next of kin must accompany submissions. The subject must understand that, if selected for publication, the photograph will not only appear on the cover of the journal but also in digital versions of the journal and associated publications. Photographs can be published without the subject's permission only under the following circumstances: 1) the subject is unidentifiable in the photograph or 2) the photograph was taken in a public venue, is not potentially damaging to the subject, and is accompanied by a written statement from the photographer vouching that the photograph was taken in a public venue with the subject's consent. A cover letter ensuring no prior publication of the photograph and providing permission from the photographer for Annals to publish the image should accompany all submissions. In addition, the letter should indicate the name and specialty of the internist depicted and the photographer's name, academic degrees, institutional affiliation, mailing address, telephone number, and e-mail address. Photographers must relinquish copyright to the American College of Physicians before publication. Pictures from photographers unwilling to do so will not be considered. Clinically important improvement in UI was defined as a Ն50% reduction in UI frequency. † Adapted from reference 11. Relative risk/absolute risk where noted. "Improve" signifies that the treatment provided benefit vs. the comparator.