Inclusion of nonrandomized studies in Cochrane systematic reviews was found to be in need of improvement

Ref ID 755
First Author S. Ijaz
Journal JOURNAL OF CLINICAL EPIDEMIOLOGY
Year Of Publishing 2014
URL https://www.ncbi.nlm.nih.gov/pubmed/24725644
Keywords • Cochrane
• Observational studies
• Risk of bias
Problem(s) • Weaknesses identified in some Cochrane reviews
• Flawed risk of bias undertaken
• Inclusion of observational / non-randomised studies
Article Type Empirical
Article Subtype Cross-sectional survey/Methodological systematic review
First Author Country Finland
Aim To assess the justification of inclusion of non-randomised studies from all Cochrane reviews that considered non-randomised studies published between 2000 to 2013.
Level of Investigation Descriptive
Summary of Findings Of the included 202 included Cochrane reviews that considered non-randomised studies published between 2000-2013 most reviews (n=114; 56%) did not justify including non-randomised studies. When they did, most were not in line with Cochrane recommendations. The reasons were divided into two major categories: non-randomised studies were included because randomized controlled trials are wanted (N 5 81, 92%) but not feasible, lacking, or insufficient alone or because RCTs are not needed (N 5 7, 8%). Risk of bias assessment varied across reviews and needs improvement.
Number of systematic reviews included 202
Number of eligible systematic reviews assessed 508
Treatment impacted No
Treatment impacted description
Interpretation impacted Not Applicable
Interpretation impacted description