Empirical Bayes estimates generated in a hierarchical summary ROC analysis agreed closely with those of a full Bayesian analysis

摘要: Abstract Background and objective A range of fixed-effect random-effects meta-analytic methods are available to obtain summary estimates measures diagnostic test accuracy. The hierarchical receiver operating characteristic (HSROC) model proposed by Rutter Gatsonis in 2001 represents a general framework for the meta-analysis studies that allows different parameters be defined as fixed effect or random effects within same model. Bayesian method used fitting is complex, however, not widely used. this report show how may fitted using SAS procedure NLMIXED compare results fully analysis an example. Methods HSROC model, its assumptions, interpretation described. advantages over usual ROC (SROC) regression outlined. complex example estimated SROC curves, expected points, confidence intervals alternative approaches Results empirical Bayes obtained agree closely with those analysis. Conclusion This more straightforward makes accessible meta-analysts.