Pharmacologic and Nonpharmacologic Treatments for Urinary Incontinence in Women A Systematic Review and Network Meta-analysis of Clinical Outcomes

Background: Urinary incontinence (UI), a common malady in women, most often is classified as stress, urgency, or mixed. Purpose: To compare the effectiveness of pharmacologic and nonpharmacologic interventions to improve or cure stress, urgency, or mixed UI in nonpregnant women. Data Sources: MEDLINE, Cochrane Central Register of Controlled Trials (Wiley), Cochrane Database of Systematic Reviews (Wiley), EMBASE (Elsevier), CINAHL (EBSCO), and PsycINFO (American Psychological Association) from inception through 10 August 2018. Study Selection: 84 randomized trials that evaluated 14 categories of interventions and reported categorical cure or improvement outcomes. Data Extraction: 1 researcher extracted study characteristics, results, and study-level risk of bias, with verification by another independent researcher. The research team collaborated to assess strength of evidence (SoE) across studies. Data Synthesis: 84 studies reported cure or improvement outcomes (32 in stress UI, 16 in urgency UI, 4 in mixed UI, and 32 in any or unspecified UI type). The most commonly evaluated active intervention types included behavioral therapies, anticholinergics, and neuromodulation. Network meta-analysis showed that all interventions, except hormones and periurethral bulking agents (variable SoE), were more effective than no treatment in achieving at least 1 favorable UI outcome. Among treatments used specifically for stress UI, behavioral therapy was more effective than either alpha-agonists or hormones in achieving cure or improvement (moderate SoE); alpha-agonists were more effective than hormones in achieving improvement (moderate SoE); and neuromodulation was more effective than no treatment for cure, improvement, and satisfaction (high SoE). Among treatments used specifically for urgency UI, behavioral therapy was statistically significantly more effective than anticholinergics in achieving cure or improvement (high SoE), both neuromodulation and onabotulinum toxin A (BTX) were more effective than no treatment (high SoE), and BTX may have been more effective than neuromodulation in achieving cure (low SoE). Limitation: Scarce direct (head-to-head trial) evidence and population heterogeneity based on UI type, UI severity, and history of prior treatment. Conclusion: Most nonpharmacologic and pharmacologic interventions are more likely than no treatment to improve UI outcomes. Behavioral therapy, alone or in combination with other interventions, is generally more effective than pharmacologic therapies alone in treating both stress and urgency UI. Primary Funding Source: Patient-Centered Outcomes Research Institute. (PROSPERO: CRD42017069903)