A simple and robust method for multivariate meta-analysis of diagnostic test accuracy

Meta-analysis of diagnostic test accuracy often involves mixture of case–control and cohort studies. The existing bivariate random-effects models, which jointly model bivariate accuracy indices (e.g., sensitivity and specificity), do not differentiate cohort studies from case–control studies and thus do not utilize the prevalence information contained in the cohort studies. The recently proposed trivariate generalized linear mixed-effects models are only applicable to cohort studies, and more importantly, they assume a common correlation structure across studies and trivariate normality on disease prevalence, test sensitivity, and specificity after transformation by some pre-specified link functions. In practice, very few studies provide justifications of these assumptions, and sometimes these assumptions are violated. In this paper, we evaluate the performance of the commonly used random-effects model under violations of these assumptions and propose a simple and robust method to fully utilize the information contained in case–control and cohort studies. The proposed method avoids making the aforementioned assumptions and can provide valid joint inferences for any functions of overall summary measures of diagnostic accuracy. Through simulation studies, we find that the proposed method is more robust to model misspecifications than the existing methods. We apply the proposed method to a meta-analysis of diagnostic test accuracy for the detection of recurrent ovarian carcinoma.