Figure 1 is a scatter plot presenting a comparison of point estimates and standard errors of summary logit-transformed sensitivity and specificity (univariate DerSimonian-Laird random effects vs. fixed effect inverse variance). Point estimates from the two methods are often similar; however, confidence interval widths are substantially larger with the DerSimonian-Laird method.

Figure 1Comparison of point estimates and standard errors of summary sensitivity and specificity (logit scale; univariate DL random effects vs. fixed effect inverse variance)

Scatter plot of estimated logit-transformed sensitivity, specificity and their corresponding confidence interval widths from univariate random and fixed effects meta-analyses of sensitivity and specificity. CI = confidence interval; DL = DerSimonian-Laird.

From: Results

Cover of An Empirical Assessment of Bivariate Methods for Meta-Analysis of Test Accuracy
An Empirical Assessment of Bivariate Methods for Meta-Analysis of Test Accuracy [Internet].
Dahabreh IJ, Trikalinos TA, Lau J, et al.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.