Article type
Year
Abstract
Background: Systematic reviews that synthesize data about the prevalence of conditions or risk factors form an important part of the evidence base for many research questions. The number of systematic reviews of prevalence has increased substantially. However, several studies have shown diverse methods for conducting these reviews and the way in which they are reported.
Objectives: To assess the quality of reporting of systematic reviews of studies of prevalence in adult populations.
Methods: We searched for systematic reviews of prevalence from January 2010 to September 2020. We used the PRISMA 2009 checklist and assigned one point for each completed item. The standard 2009 checklist was designed initially for systematic reviews of interventions, and because some items were written explicitly for these reviews, we modified the scoring system. The maximum score was 26 points. We report the highest and lowest compliance items and the median (with interquartile range, IQR) score over time.
Results: We included 1,150 systematic reviews. The median PRISMA score was 20 points IQR (17-23); the median score in 2010 was 16 IQR (11-19) and 23 (19-25) in 2020. The items with highest compliance were related to the title, abstract, rationale and objectives, and the discussion. Items with poor compliance were mainly in the methods and results sections. For example, in the methods section, only 24% of reviews reported having published a protocol (item 5), at least one search strategy (item 5) was reported only by 48%, and 64% of authors reported the use of a risk of bias; however, only 54% of authors presented in the results section information on the risk of bias of each study in Figure 1.
Conclusions: Overall, the reporting of systematic reviews of prevalence has improved between 2010 and 2020. But the reporting of important methodological items did not improve. Specific guidance for reporting might be needed for systematic reviews of prevalence. In addition, further development of methodological guidance for the conduct of systematic reviews of prevalence could improve the quality of the evidence synthesis about prevalence.
Objectives: To assess the quality of reporting of systematic reviews of studies of prevalence in adult populations.
Methods: We searched for systematic reviews of prevalence from January 2010 to September 2020. We used the PRISMA 2009 checklist and assigned one point for each completed item. The standard 2009 checklist was designed initially for systematic reviews of interventions, and because some items were written explicitly for these reviews, we modified the scoring system. The maximum score was 26 points. We report the highest and lowest compliance items and the median (with interquartile range, IQR) score over time.
Results: We included 1,150 systematic reviews. The median PRISMA score was 20 points IQR (17-23); the median score in 2010 was 16 IQR (11-19) and 23 (19-25) in 2020. The items with highest compliance were related to the title, abstract, rationale and objectives, and the discussion. Items with poor compliance were mainly in the methods and results sections. For example, in the methods section, only 24% of reviews reported having published a protocol (item 5), at least one search strategy (item 5) was reported only by 48%, and 64% of authors reported the use of a risk of bias; however, only 54% of authors presented in the results section information on the risk of bias of each study in Figure 1.
Conclusions: Overall, the reporting of systematic reviews of prevalence has improved between 2010 and 2020. But the reporting of important methodological items did not improve. Specific guidance for reporting might be needed for systematic reviews of prevalence. In addition, further development of methodological guidance for the conduct of systematic reviews of prevalence could improve the quality of the evidence synthesis about prevalence.