Advantages to transforming the receiver operating characteristic (ROC) curve into likelihood ratio co‐ordinates

Traditionally, the receiver operating characteristic (ROC) curve for a diagnostic test plots true positives (sensitivity) against false positives (one minus specificity). However, this representation brings with it several drawbacks. A transformation to positive and negative likelihood ratio co-ordinates, scaled by base-ten logarithms, offers several advantages. First we motivate the use of positive and negative likelihood ratios, emphasizing their relationship to modification of the odds ratio. Then we highlight properties of likelihood ratios using the traditional ROC axes. Finally, we demonstrate ROC curves and their properties after conversion to likelihood ratio co-ordinates. These graphs do not waste space for tests lacking diagnostic power, and offer a simple visual assessment of a test's impact on the odds ratio.