|
Abstract:
|
A standard logistic meta-analysis estimates the effect of a dichotomous covariate on the probability of an adverse outcome through the log(odds ratio). When it comes to physician decision-making, diagnostic statistics such has sensitivity, specificity, positive/negative predictive value, and positive/negative likelihood ratio are of interest. We present an extension of the standard Bayesian logistic meta-analysis model that provides inference about these diagnostic statistics. We illustrate the model in a series of meta-analyses using data from studies which look at adverse events after syncopal episodes.
|
