COMPARISON OF RECEIVER OPERATING CHARACTERISTIC CURVES ON THE BASIS OF OPTIMAL OPERATING POINTS

Rationale and Objectives. We developed a method of comparing receiver operating characteristic (ROC) curves on the basis of the utilities as sociated with their optimal operating points (OOPs). Methods. OOPs wer e computed for paired ROC curves on the basis of isocost lines in ROC space with slopes ranging from 0.1 to 3.0, For each pair of OOPs corre sponding to a single isocost slope, the difference in costs and the va riance of this difference was computed. A sensitivity analysis was thu s obtained for the difference between the two curves over a range of i socost slopes. Three published data sets were evaluated using this tec hnique, as well as by comparisons of areas under the curves and of tru e-positive fractions at fixed false-positive fractions. Results. The O OPs of paired ROC curves often occur at different false-positive fract ions. Comparisons of ROC curves on the basis of OOPs may provide resul ts that differ from comparisons of curves at a fixed false-positive fr action. Conclusion. ROC curves may be compared on the basis of utiliti es associated with their OOP. This comparison of the optimal performan ce of two diagnostic tests may differ from conventional statistical co mparisons.