Systematic reviews of diagnostic tests in cancer: Assessment of methodology and reporting quality.
Susan Mallett, Jon Deeks, Victoria Cornelius, and Doug Altman with Steve Halligan (Northwick Park) and Sally Hopewell (UK Cochrane Centre)
Systematic reviews of evaluations of diagnostic tests are complex and methods of conducting and reporting these studies are not well established. We have undertaken a descriptive study of the current practice of methodology and reporting of systematic reviews of cancer diagnosis, with a view to establishing key areas where improvements need to be made for reviews to be of value to healthcare workers in cancer.
Our quality assessment of reviews included reporting on: literature search; patient spectrum; adequate description of the diagnostic test and gold standard reference test; study design; reporting of individual study results; study bias; heterogeneity; and meta analysis methods.
Preliminary results from this study have already informed our own systematic review on CT colonoscopy, as well as contributing to the new Cochrane Collaboration Handbook currently being written to provide guidelines for systematic reviews of diagnostic studies.
