A systematic survey shows that reporting and handling of missing outcome data in networks of interventions is poor

Ref ID 436
First Author L. M. Spineli
Journal BMC MEDICAL RESEARCH METHODOLOGY
Year Of Publishing 2018
URL https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6201503/pdf/12874_2018_Article_576.pdf
Keywords • General medical
• Protocols
• Missing data
• Network meta-analysis
Problem(s) • Failure to address missing outcome data in analyses
• No registered or published protocol
Article Type Empirical
Article Subtype Cross-sectional survey/Methodological systematic review
First Author Country Germany
Aim To assess the prevalence, reporting and handling of missing outcome data in systematic reviews with network meta-analysis and acknowledgement of their impact on the conclusions. Systematic reviews comparing at least three interventions and published from January 2009 until March 2017 were eligible.
Level of Investigation Descriptive
Summary of Findings 72% did not mention whether a protocol was available; 17% mentioned that a protocol was developed but they did not make the protocol available. Only 16% reviews explicitly defined (either in their protocol or publication) missing outcome data however 70% of reviews explicitly reported that there was missing outcome data in the included trials. 56% applied at least one strategy to address missing outcome data in their analysis. Overall, there was insufficient information in most systematic reviews with network meta-analysis to judge the actual strategy applied to handle missing outcome data.
Number of systematic reviews included 387
Number of eligible systematic reviews assessed 893
Treatment impacted No
Treatment impacted description
Interpretation impacted Not Applicable
Interpretation impacted description