Skip to main content
← allmeta
Bivariate DTA / HSROC
Meta-analysis of diagnostic test accuracy: bivariate random-effects model on logit(sensitivity) and logit(1-specificity), with a parametric HSROC curve. Approximates Reitsma/Harbord via method-of-moments + IGLS iteration. For rigorous estimates use R's mada.
Input
Studies — TP, FP, FN, TN, label (per line)
Fit
Per study (Clopper-Pearson 95% CI)
Study Se Sp logit(Se) logit(1-Sp)
Summary estimates
SROC plot
Model: bivariate random-effects on logit-Se, logit-FPR, correlation ρ — Reitsma JB et al. J Clin Epidemiol 2005;58:982–990; Rutter CM, Gatsonis CA. Stat Med 2001;20:2865–2884 (HSROC). Implemented here as alternating univariate DerSimonian-Laird pools on each margin plus an empirical residual correlation — this is a Harbord-Whiting-style approximation (Harbord RM et al. Biostatistics 2007;8:239–251) and is NOT iterated to convergence. For rigorous Reitsma/HSROC estimates use mada::reitsma (also available via WebR Studio). 0.5 continuity correction when any cell is zero. HSROC curve plotted as the locus of (FPR, Se) pairs consistent with the mean linear predictor.