Misleading meta-analysis

BMJ 1995; 311 doi: https://doi.org/10.1136/bmj.311.7007.753c (Published 16 September 1995) Cite this as: BMJ 1995;311:753
  1. Matthias Egger, Senior research fellow,
  2. George Davey Smith, Professor of clinical epidemiology
  1. Department of Social and Preventive Medicine, University of Berne, CH-3012 Berne, Switzerland
  2. Department of Social Medicine, University of Bristol, Bristol BS8 2PA

    EDITOR,--In their response1 to our editorial on misleading meta-analysis2 A Perry and R Persaud state that we ignored the degree of heterogeneity among different studies. We examined the case of intravenous magnesium in acute myocardial infarction, a treatment recently shown to be of no benefit in a trial.3 We argued that the asymmetrical funnel plot (a plot of the estimates of effect sizes against the sample size) should have alerted meta-analysts to the possible presence of bias. The erroneous conclusion …

    View Full Text

    Log in

    Log in through your institution


    * For online subscription