Urinary incontinence (UI) fluctuates in prevalence in the developing world. Factors that influence the change in prevalence are a key area of interest, and knowledge of these would provide the opportunity for appropriate planning for preventive, primary and secondary care programs. The objective of this systematic review was to synthesize the best available evidence determining the prevalence of UI in adult women in a population setting.

All studies that included women with UI in a population setting were enrolled. A comprehensive search strategy was employed to find both published and unpublished studies that examined individual exposures. Databases searched included PubMed, Embase, Scopus, Web of Science and Google Scholar. We used the standardized critical appraisal instrument from the JBI Meta-Analysis of Statistics, Assessment, and Review Instrument (JBI-MAStARI) to appraise the included studies. We registered the title in http://joannabriggs.org/research/registered_titles.aspx.

54 studies were included in this systematic review. The total number of UI patients was 138,722. The mean age ranged from 10 to 90 years. Prevalence of UI ranged from 2.8 in Nigeria to 57.7 in Iran. In fixed method analysis prevalence was 29.4 (CI 95%, 29.1%-29.6%) but I-square was more than 50% which demonstrates the high heterogeneity of the studies. In the random effect model, the prevalence was reduced to 25.3 % (CI 95%, 22.1%-28.9%). Prevalence of different types of incontinence was 12.6% (CI 95%, 10.3%-15.4%), 5.4% (CI 95%, 3.5%-8.2%) and 9.1% (CI 95%, 7.0%-11.8%) for stress urinary incontinence, urgency urinary incontinence and mixed urinary incontinence respectively in random effect analysis.

To assess the publication bias of the extracted data, the funnel plot was established. It seems that the sample size of the included studies are almost appropriate but the pattern of the distribution is not symmetric completely which can be related to publication bias or methodological issues of the studies. However, we did not exclude any of these studies and we have done some sub-group analysis because only one in left side below is totally out of distribution. 
The different method used in the studies for assessment of UI was one of the difficulties for conducting the current study. Using different questionnaires either (validated or not validated) with different definitions of UI made the process of analysis difficult resulting significant differences in each method of inquiry. Geographical location and a wide variety of ethnicities and races made was also a problematic issue. Another major aspect was the lack of prevalence studies assessing UI in the developing countries leaving us with limited resources for an international estimate of UI. We did not find studies in most of the defined countries some countries had more than expected number of studies. Most of the included studies assessed the total prevalence of UI however, a fewer number of studies addressed the prevalence of subtypes of UI (Stress, Urge and Mixed). Another problem was studies that were in clinical or hospital settings. These studies were not included since the author included only population-based studies. Different age ranges made it difficult to reach a final conclusion in the definition of adulthood. Big sample size of some studies and the high study power resulted in an extra weight for these studies. Limited access to the gray literature stopped us from adding more relevant studies. The mentioned problems were expected due to the nature of prevalence studies.

Prevalence of UI is in a very wide spectrum from very low to very high. In addition, the background methodology and the definition were different. When facing a prevalence of 2.8 or 57.7% both can be meaningful.  However, surprisingly in the smaller regions, the prevalence is much fluctuated. A uniform multinational study is necessary for health policymaking in the developing world. Furthermore, excluding the elderly population did not affect the final results and the result might be due to the low weight studies.