Figure 21 presents comparisons of point estimates and confidence interval widths of summary logit-transformed sensitivity and specificity (univariate random effects vs. bivariate random effects inverse variance methods, both using a normal approximation for within-study variability). The scatter plots compare the estimated logit-transformed sensitivity, specificity and their corresponding confidence interval widths from univariate random effects meta-analyses over bivariate random effects inverse variance (DerSimonian-Laird and multivariate DerSimonian-Laird) meta-analysis (with an approximate normal likelihood to describe within-study variability). Point estimates and confidence intervals from the two methods are similar, both for sensitivity and specificity.

Figure 21Comparison of point estimates and confidence interval widths of summary sensitivity and specificity (logit scale, univariate random effects vs. bivariate random effects inverse variance methods, both using a normal approximation for within-study variability and a noniterative estimator for heterogeneity)

Note: Scatter plot of estimated logit-transformed sensitivity, specificity and their corresponding confidence interval widths from univariate random effects meta-analyses over bivariate random effects inverse variance (DerSimonian-Laird and multivariate DerSimonian-Laird) meta-analysis (with an approximate normal likelihood to describe within-study variability).

CI = confidence interval; DL = DerSimonian-Laird; mult. DL = multivariate 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.