The impact of study size on meta-analyses: examination of underpowered studies in Cochrane reviews

Ref ID 545
First Author R. M. Turner
Journal PLOS ONE
Year Of Publishing 2013
URL https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3609745/pdf/pone.0059202.pdf
Keywords • General medical
• Cochrane
• Power
Problem(s) • Weaknesses identified in some Cochrane reviews
• Perpetuates citation of poor quality primary study data
Article Type Empirical
Article Subtype Meta-epidemiological analysis
First Author Country United Kingdom
Aim To assess the distribution of power available in studies within meta-analyses published in Cochrane reviews, and investigate the impact of underpowered studies on meta-analysis results.
Level of Investigation Analytical
Summary of Findings In 70% of CDSR meta-analyses, all studies had less than 50% power to detect a 30% relative risk reduction, and only 17% of meta-analyses included at least two studies with at least 50% power to detect a relative risk reduction of 30%.
Number of systematic reviews included 14886
Number of eligible systematic reviews assessed 14886
Treatment impacted Yes
Treatment impacted description The results of excluding underpowered studies from meta-analyses correspond to a slight shift towards the null value when underpowered studies were removed, more so under the random-effects model in which small studies have greater influence.
Interpretation impacted No
Interpretation impacted description