Receiver operating characteristic curve analysis of clinical risk models

Receiver operating characteristic (ROC) curve analysis is a useful method t o measure the ability of a clinical risk model to discriminate between hosp ital deaths and survivors. Its use in medicine originated as a method for s ynthesizing the specificity and sensitivity of diagnostic tests across a sp ectrum of possible cut points. The area under the ROC curve can be interpre ted as a probability of correct classification or prediction. We illustrate its use in three steps: first, with a dichotomous variable to introduce sp ecificity and sensitivity; next, with a categorical risk factor (surgical u rgency) to produce a primitive ROC curve; and finally, with a continuous ri sk factor (age) to approximate the usual ROC curve used for clinical risk m odels. (C) 2001 by The Society of Thoracic Surgeons.