Overview authors rarely defined systematic reviews that are included in their overviews

J Clin Epidemiol. 2019 May:109:70-79. doi: 10.1016/j.jclinepi.2019.01.004. Epub 2019 Jan 23.

Abstract

Objectives: To report systematic review definitions that are published in overviews of reviews and to propose a new classification of systematic reviews.

Study design and setting: In this review of overviews, we searched PubMed for systematic review definitions that were reported in overviews of reviews that were published in the medical literature between November 2017 and May 2018. Two independent authors extracted and descriptively reported the systematic review definitions from the overviews. The definitions were evaluated regarding whether the concepts of comprehensiveness and reproducibility were incorporated into them, as suggested by some published systematic review definitions.

Results: Initially, 138 documents were retrieved, and 111 overviews and protocols of overviews were included. Eight (8%) overviews explicitly reported a systematic review definition, whereas 25 (24%) overviews reported heterogeneous information about the criteria that were used to include systematic reviews in the overviews. Seventy-two (68%) overviews did not report any definition/criteria for including a systematic review. Two (2%) overviews reported a definition based on reproducibility, and none of the overviews reported the need to search for grey and unpublished literature for a review to be considered systematic.

Conclusion: Overview authors rarely define systematic reviews that are included in their overviews and the few that do include a definition that provides heterogeneous criteria.

Keywords: Bias; Meta-analysis; Methods; Overview; Research design; Review; Systematic review.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Biomedical Research*
  • Female
  • Humans
  • Male
  • Middle Aged
  • Publications / standards*
  • Publications / statistics & numerical data*
  • Reproducibility of Results
  • Research Personnel / statistics & numerical data*
  • Research Report / standards*
  • Selection Bias*
  • Systematic Reviews as Topic*