An observational study found large methodological heterogeneity in systematic reviews addressing prevalence and cumulative incidence.

The objective of this study was to assess reporting and methodological aspects of systematic reviews (SRs) on prevalence and cumulative incidence data. We searched PubMed up to 18 April, 2018, and drew a random sample of eligible SRs. The included 215 SRs were reported in 187 different journals. 58.1% were published between 2015 and 2018. Few SRs were registered with PROSPERO (5.6%). One-quarter considered articles without languages restrictions (25.1%). Regional restrictions of included studies were applied in 22.8%. A meta-analysis was carried out in 40.5% of the SRs. One hundred and six studies (49.3%) assessed risk of bias or study quality. A total of 41 different existing tools as well as 15 tools developed by the authors themselves were used. The most commonly applied tools were the Newcastle-Ottawa Scale (15.1%), STROBE (13.5%), and the Cochrane Collaboration's tool for assessing risk of bias (7.9%). We found large heterogeneity in characteristics, reporting, and methodological aspects of SRs on prevalence and cumulative incidence data, especially when compared with other types of SRs. Newly developed or revised guidance on how to conduct and report SRs as well as instruments for critical appraisal should consider the diversity of review types.

Informations de copyright

Copyright © 2019 Elsevier Inc. All rights reserved.