Journal
STATISTICS IN MEDICINE
Volume 29, Issue 30, Pages 3088-3102Publisher
WILEY
DOI: 10.1002/sim.4055
Keywords
meta-analysis; diagnostic studies; posterior prediction; hierarchical models; Bayesian modeling; BUGS
Categories
Funding
- Swiss National Science Foundation [51A240-104890]
Ask authors/readers for more resources
In the last decades, the amount of published results on clinical diagnostic tests has expanded very rapidly. The counterpart to this development has been the formal evaluation and synthesis of diagnostic results. However, published results present substantial heterogeneity and they can be regarded as so far removed from the classical domain of meta-analysis, that they can provide a rather severe test of classical statistical methods. Recently, bivariate random effects meta-analytic methods, which model the pairs of sensitivities and specificities, have been presented from the classical point of view. In this work a bivariate Bayesian modeling approach is presented. This approach substantially extends the scope of classical bivariate methods by allowing the structural distribution of the random effects to depend on multiple sources of variability. Meta-analysis is summarized by the predictive posterior distributions for sensitivity and specificity. This new approach allows, also, to perform substantial model checking, model diagnostic and model selection. Statistical computations are implemented in the public domain statistical software (WinBUGS and R) and illustrated with real data examples. Copyright (C) 2010 John Wiley & Sons, Ltd.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available