The HSROC, or hierarchical summary receiver operating characteristic, curve offers a more rigorous method of analyzing the threshold effect. The HSROC curve jointly analyzes sensitivity and specificity using both the within- and between-study heterogeneity measures.
Breaking down the HSROC curve
The HSROC curve also plots specificity on the x-axis and sensitivity on the y-axis. The x-axis is inverted such that the value of 1 is the origin while the value of 0 lies at the rightmost extreme of the plot. As in the SROC curve, each study's specificity-sensitivity pair is plotted in the ROC space using dots whose size corresponds with the size of the study. A line of best fit, which is the HSROC curve, passes through these dots. The region marked by the dotted lines represents the 95% confidence interval for the studies included.
# HSROC curve