Missing binary data extraction challenges from Cochrane reviews in mental health and Campbell reviews with implications for empirical research

Ref ID 658
First Author L. M. Spineli
Journal RESEARCH SYNTHESIS METHODS
Year Of Publishing 2017
URL https://onlinelibrary.wiley.com/doi/pdfdirect/10.1002/jrsm.1268?download=true
Keywords • Cochrane
• Campbell
• General medical
• Missing data
• Error
Problem(s) • Failure to address missing outcome data in analyses
• Data extraction errors and double counting
Article Type Empirical
Article Subtype Meta-epidemiological analysis
First Author Country Germany
Aim Aims to examine the impact on the summary statistics from meta-analyses when missing outcome data from unclear or unaccceptable data extraction are imputed
Level of Investigation Analytical
Summary of Findings More than half of 113 eligible Cochrane meta‐analyses were evaluated as “unclearly” extracted (53%); 42 (37%) were “unacceptably” extracted and only 11 (10%) were “acceptably” extracted. The direct implication of “unclear” extraction is unnecessary inflation in uncertainty of estimated odds ratio irrespective of method used to address missingness.
Number of systematic reviews included 113
Number of eligible systematic reviews assessed 113
Treatment impacted Yes
Treatment impacted description
Interpretation impacted No
Interpretation impacted description