The SROC, or summary receiver operating characteristic, curve relates the between-study heterogeneity and the sensitivity and specificity of diagnostic test studies, where specificity is measured on the x-axis while sensitivity is measured on the y-axis. A regular ROC curve plots the relationship between the false positive rate, or when specificity is 1, and the true positive rate, when sensitivity is 1.
Breaking down the SROC curve
The SROC curve plots the specificity and sensitivity from each study in the ROC space to observe the threshold effect in diagnostic studies. The specificity-sensitivity pairs from each study are represented by dots, whose size depends on the size of the study. The line passing through these points, or the SROC curve, is the line of best fit while the dotted 45 degree line is an indicator for the case of complete randomness. The more the dots gravitate toward the top left corner, the better is the sensitivity or true positive rate.