Statistical heterogeneity in systematic reviews of clinical trials: a critical appraisal of guidelines and practice
By Julian P. T. Higgins, S. Thompson, J. Deeks, Douglas G. Altman. Journal of Health Services Research and Policy, 2002
751 citations recorded by the source corpus.
Subfields: Molecular Biology, Statistics and Probability
View Paper: https://www.semanticscholar.org/paper/b50d00efec24e886bcd15b4cb3be6b0dbecf294f
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Molecular Biology
In Statistics and Probability
- Julian P. T. Higgins
- J. Wetterslev
- C. Gluud
- S. Thompson
- J. Jakobsen
- David J. Spiegelhalter
- Georgia Salanti
- Paul T. von Hippel
Citations in the corpus
- Thresholds for statistical and clinical significance in systematic reviews with meta-analytic methods (2014, result): “If there is absence of statistical heterogeneity (the between trial variance of the estimated intervention effects is close to zero [23]), then the fixed-effect and the randomeffects models will show identical results [13].”
- Opioids in chronic non‐cancer pain: systematic review of efficacy and safety (2004, methodology): “Heterogeneity tests were not used (Gavaghan et al., 2000; Higgins et al., 2002) though homogeneity was examined visually (L’Abbé et al., 1987).”
- A re-evaluation of random-effects meta-analysis (2009, background): “Although sources of heterogeneity should ideally be investigated (Thompson, 1994), selection of covariates to explain heterogeneity can be difficult, both practically (Higgins et al., 2002) and theoretically (Thompson and Higgins, 2002).”
- How should meta‐regression analyses be undertaken and interpreted? (2002, background): “For example, in a study of 39 Cochrane systematic reviews, only one contained more that 15 studies with available data for the primary outcome.(26) Given such numbers of studies, the potential for robust conclusions from meta-regression analyses is clearly very limited.”
- Controlling the risk of spurious findings from meta‐regression (2004, methodology): “However, we have observed that in practice reviewers often fail to follow these guidelines and may be tempted to examine numerous covariates in an attempt to explain observed heterogeneity [28].”