Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Bias in meta-analysis detected by a simple, graphical test

M Egger1, G Davey Smith, M Schneider

  • 1Department of Social Medicine, University of Bristol. m.egger@bristol.ac.uk

BMJ (Clinical Research Ed.)
|October 6, 1997
PubMed
Summary

Funnel plot asymmetry is a useful indicator of bias in meta-analyses. However, caution is advised when interpreting results from meta-analyses with few small trials, as bias detection may be limited.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Novel integrase mutations linked to genotypic DTG resistance in non-B HIV-1 strains from African participants: The DTG RESIST study.

medRxiv : the preprint server for health sciences·2025
Same author

Diagnosis of postpartum depression and associated factors in South Africa: a cohort study of 47,697 women.

Epidemiology and psychiatric sciences·2025
Same author

[Agonist-antagonist myoneural interface (AMI) : Innovative treatment option for lower limb amputees?]

Unfallchirurgie (Heidelberg, Germany)·2025
Same author

Biodegradation of polyethylene by the marine fungus Parengyodontium album.

The Science of the total environment·2024
Same author

Genetic risk for schizophrenia is associated with increased proportion of indirect connections in brain networks revealed by a semi-metric analysis: evidence from population sample stratified for polygenic risk.

Cerebral cortex (New York, N.Y. : 1991)·2022
Same author

Screening for hepatocellular carcinoma among adults with HIV/HBV co-infection in Zambia: a pilot study.

International journal of infectious diseases : IJID : official publication of the International Society for Infectious Diseases·2021

Area of Science:

  • Biostatistics
  • Evidence-based medicine
  • Meta-analysis

Background:

  • Funnel plots, which graphically represent effect estimates against sample size, are utilized to detect bias in meta-analyses.
  • Previous research suggests funnel plots may identify biases that are later contradicted by large-scale trials.

Purpose of the Study:

  • To evaluate if funnel plot asymmetry predicts discordant results between meta-analyses and large trials.
  • To assess the prevalence of bias in published meta-analyses.

Main Methods:

  • A Medline search identified meta-analysis and large trial pairs.
  • Funnel plots were analyzed for asymmetry using the intercept from regression of standard normal deviates against precision.
  • Data included 37 meta-analyses from general medicine journals and 38 from the Cochrane Database of Systematic Reviews.

Related Experiment Videos

Main Results:

  • Of eight identified meta-analysis and large trial pairs, four were discordant, with meta-analyses showing larger effects.
  • Funnel plot asymmetry was present in three of four discordant pairs but absent in concordant pairs.
  • Bias, indicated by funnel plot asymmetry, was found in 38% of journal meta-analyses and 13% of Cochrane reviews.

Conclusions:

  • Simple funnel plot asymmetry analysis is a valuable test for detecting potential bias in meta-analyses.
  • Results from funnel plot analyses should be interpreted cautiously, especially when meta-analyses involve a limited number of small trials, due to limited bias detection capacity.