Multiplicity of data in trial reports and the reliability of meta-analyses: empirical study

Ref ID 468
First Author B. Tendal
Journal BMJ
Year Of Publishing 2011
URL https://www.bmj.com/content/bmj/343/bmj.d4829.full.pdf
Keywords • General medical
• Cochrane
• Multiplicity
Problem(s) • Weaknesses identified in some Cochrane reviews
• Multiplicity of outcomes and lack of pre-specification for outcome reporting
Article Type Empirical
Article Subtype Meta-epidemiological analysis
First Author Country Denmark
Aim To assess the extent of multiplicity of data in trial reports and to assess the impact of multiplicity on meta-analysis results of a sample of Cochrane systematic reviews published 2006 to 2007. Reviews with protocols and meta-analyses that presented data with a standardised mean difference were included.
Level of Investigation Analytical
Summary of Findings Review protocols often lacked information about which data to choose. There was multiplicity of data in at least one trial report. 29% trials reported data for multiple intervention groups, 36% reported data for multiple time points, and 35% reported the index outcome measured on multiple scales. The impact of multiple data in trial reports regarding intervention groups, time points, or measurement scales on meta-analysis results varied greatly across meta-analyses, ranging from almost no effect (0.04 standard deviation units) to a substantial one (0.91 standard deviation) units, corresponding to a large treatment effect.
Number of systematic reviews included 19
Number of eligible systematic reviews assessed 32