Essay |
Department of Social Medicine, University of Bristol, Bristol, UK;
1 Department of Social and Preventive Medicine, University of Bern, Switzerland
Correspondence to: Penny.whiting{at}bristol.ac.uk
Making an accurate diagnosis is essential to ensure that a patient receives appropriate treatment and correct information regarding their prognosis. Characteristics of diagnostic tests are quantified in test accuracy studies, but many such studies have methodological flaws. The HSRC evidence-based diagnosis programme has focused on methods for systematic reviews of test accuracy studies, and the wider context in which tests are ordered and interpreted. We carried out a range of projects relating to literature searching, quality assessment, meta-analysis, presentation of results, and interactions between doctors and patients during the diagnostic process. We have shown that systematic reviews of test accuracy studies should search a range of databases and that current diagnostic filters do not have sufficient accuracy to be used in test accuracy reviews. Summary quality scores should not be used in test accuracy reviews; the Quality Assessment of Studies of Diagnostic Accuracy included in Systematic Reviews (QUADAS) tool for assessing test accuracy studies is acceptable for quality assessment. We have shown that the hierarchical summary receiver operating characteristic (HSROC) and bivariate models for meta-analysis of test accuracy are statistically equivalent in many circumstances, and have developed an add-on module for the statistical software package Stata that enables these statistically rigorous models to be fitted by those without expert statistical knowledge. Three areas that would benefit from further research are literature searching, synthesis of results from individual patient data and presentation of results.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
