Characteristics and completeness of reporting of systematic reviews of prevalence studies in adult populations: a meta-research study.

Abstract

Objective: The Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) statement, first published in 2009, has been widely endorsed and compliance is high in systematic reviews of intervention studies. Systematic reviews of prevalence studies are increasing in frequency, but their characteristics and reporting quality have not been examined in large studies. Our objectives were to describe the characteristics of systematic reviews of prevalence studies in adults, evaluate the completeness of reporting and explore study-level characteristics associated with the completeness of reporting. Study design and setting: We did a meta-epidemiological study. We searched 5 databases from January 2010 to December 2020 to identify systematic reviews of prevalence studies in adult populations. We used the PRISMA 2009 checklist to assess completeness of reporting and recorded additional characteristics. We conducted a descriptive analysis of review characteristics and linear regression to assess the relationship between compliance with PRISMA and publication characteristics. Results: We included 1172 systematic reviews of prevalence studies. The number of reviews increased from 25 in 2010 to 273 in 2020. The median PRISMA score for systematic reviews without meta-analysis was 17.5 out of a maximum of 23 and, for systematic reviews with meta-analysis, 22 out of a maximum of 25. Completeness of reporting, particularly for key items in the methods section was suboptimal. Systematic reviews that included a meta-analysis or reported using a reporting or conduct guideline were the factors most strongly associated with increased compliance with PRISMA 2009. Conclusion: Reporting of systematic reviews of prevalence was adequate for many PRISMA items. Nonetheless, this study highlights aspects for which special attention is needed. Development of a specific tool to assess the risk of bias in prevalence studies and an extension to the PRISMA statement could improve the conduct and reporting of systematic reviews of prevalence studies.