Heterogeneity can impair the results of Cochrane meta‐analyses despite accordance with statistical guidelines

Ref ID 173
First Author H.-J. Malling
Journal ALLERGY
Year Of Publishing 2008
URL https://onlinelibrary.wiley.com/doi/pdfdirect/10.1111/j.1398-9995.2008.01908.x?download=true
Keywords • Cochrane
• Immunology
• Heterogeneity
Problem(s) • Inadequate analysis of heterogeneity
• Weaknesses identified in some Cochrane reviews
Article Type Editorial
Article Subtype Comment/letter
First Author Country Denmark
Aim A letter which re-examines the meta-analysis of a Cochrane review on the effect of specific immunotherapy in patients with seasonal allergic rhinitis published in 2007.
Level of Investigation Descriptive
Summary of Findings The authors re-examined the meta-analysis and ultimate treatment effect of the Cochrane review on the effect of specific immunotherapy in patients with seasonal allergic rhinitis published in 2007. They found that excluding studies identified as extreme outliers from the original meta-analysis (as the authors state is generally advised within statistics) using influence statistics helped identify outliers and reasons for heterogeneity. As such, justifications for exclusion due to clinical or methodological heterogeneity were provided for exclusions of outlier studies in the meta-analysis.
Number of systematic reviews included 1
Number of eligible systematic reviews assessed 1
Treatment impacted Yes
Treatment impacted description The precision of the treatment effect was greater when outliers were excluded to reduce heterogeneity but the direction of effect was unchanged.
Interpretation impacted No
Interpretation impacted description