Systematic reviews with individual patient data meta-analysis to evaluate diagnostic tests

Systematic literature reviews with meta-analysis of reported data in diagnostic research studies about the accuracy of tests assist clinicians in decision-making. However, there are limitations to this approach as the analysis of such data often does not allow reviewers to explore the diagnostic information gained from combinations of tests. In recent years meta-analysis of individual patient data (IPD) has been introduced as gold standard analytic approach in systematic reviews of randomised controlled trials. Application of IPD meta-analysis in systematic reviews of diagnostic tests can allow advanced analysis to decipher the real value of testing. In particular, the additional information provided by testing can be examined in light of the information already known from history and examination. To our knowledge there are no IPD meta-analyses in diagnostic research so far. This commentary aims to highlight the benefits of IPD meta-analysis in the diagnostic domain and demonstrates how strategies for diagnostic work-up of women with postmenopausal bleeding could be improved using this approach in systematic reviews of diagnostic research on accuracy of ultrasound and hysteroscopy. (C) 2003 Elsevier Science Ireland Ltd. All rights reserved.