Diagnostic accuracy is the ability of a measure/test to discriminate between the target condition/disorder and health. We used receiver operating characteristic (ROC) curves to graphically display tradeoff between sensitivity (true positive rate) and specificity (true negative rate) of FH measures for their ability to predict a dichotomous outcome, here DSM-5 AUD diagnosis. We also used area under the ROC curve (AUC) as a global measure to compare diagnostic accuracy across the FH measures. A measure/test with perfect diagnostic accuracy has an AUC of 1.0 and an uninformative measure/test is no better than chance where AUC=0.5 (50% probability/random chance) (Florkowski, 2008).
