Treatment of heterogeneity in systematic reviews of diagnostic tests in cancer
Susan Mallett, Jon Deeks and Doug Altman
Systematic reviews of diagnostic tests frequently find results differ between studies. This heterogeneity of study results affects decisions on how to summarise results especially in meta-analysis. Understanding of the sources of this heterogeneity can increase the clinical and scientific understanding of the summary estimates resulting from systematic reviews.
This study describes the treatment of heterogeneity in meta-analyses within systematic reviews of tests used to diagnose cancer in symptomatic patients. Reviews were assessed for: discussion of heterogeneity in review abstract; display of study results used to show presence of heterogeneity; treatment of threshold heterogeneity; and the contribution to heterogeneity between study results from between study differences in study methods, participants and index test methods.
We found that only *% of meta-analysis abstracts acknowledge heterogeneity of study results and only 57% include a graphical display of heterogeneity between individual study results. Only 60% of meta-analyses included assessment of threshold effect, with the treatment of threshold effect considered as adequate in 51%. There was a high awareness of heterogeneity being due to differences in study methods, participants or index test differences.
This survey highlights deficiencies in assessment and treatment of heterogeneity between studies included in systematic reviews for diagnostic tests in cancer. A new Cochrane handbook for systematic reviews of diagnostic tests is being written and will help guide future reviewers in the process of testing and summarising results in the presence of study result heterogeneity.
