A study evaluated predictors of adverse outcomes in geriatric patients admitted to the hospital with a hip fracture requiring surgery. The study compared an old multidimensional score (Model 1) with a new classification system (Model 2). The following results were reported: AUC = area under the curve
Which of the following is the best statement about the study results?

Question

A study evaluated predictors of adverse outcomes in geriatric patients admitted to the hospital with a hip fracture requiring surgery.  The study compared an old multidimensional score (Model 1) with a new classification system (Model 2).  The following results were reported:  AUC = area under the curve
Which of the following is the best statement about the study results?

Quizplus · Accepted Answer

Explanation: A receiver-operating characteristic (ROC) curve for a given diagnostic test plots sensitivity on the y-axis and (1 − specificity), or equivalently (100% − specificity), on the x-axis. It can help determine the best cutoff point to use depending on the optimal desired parameters for sensitivity and specificity. For example, for the multidimensional score models described in the scenario, the ROC curves can help determine which score to use as cutoff for a high score (eg, with higher risk of mortality, discharge to rehabilitation facility, or postoperative myocardial infarction or stroke) or a low score, if desired. In addition, the area under the curve (AUC) of a ROC curve is a reflection of diagnostic accuracy. A larger AUC means better discrimination and higher diagnostic accuracy. In general, an ideal test would have an AUC of 1.0 (100% sensitivity and specificity) and a nondiscriminating test would have an AUC of 0.5 (similar to a coin toss). Model 1 has a higher AUC than Model 2 in terms of predicting mortality, discharge to rehabilitation facility, and postoperative myocardial infarction or stroke; AUC differences are also statistically significant (P <0.05) in all 3 cases. Therefore, Model 1 outperforms Model 2 in all 3 areas. In general (Choice A) The sensitivity and specificity of the models depend on the cutoff points used. (Choice C) Model 1 outperforms Model 2 in predicting mortality, discharge to rehabilitation facility, and postoperative myocardial infarction or stroke. (Choice D) The AUC of Model 1 is higher for mortality than for postoperative myocardial infarction or stroke, so Model 1 predicts mortality more accurately than postoperative myocardial infarction or stroke. (Choice E) There may be instances where the Model 2 classification system is useful in clinical practice (eg, if its components are easier to obtain than those of Model 1). Educational objective: The area under the curve (AUC) of a receiver-operating characteristic (ROC) curve is a reflection of diagnostic accuracy. A larger AUC means better discrimination and higher diagnostic accuracy. References: Diagnostic accuracy measures (http://www.ncbi.nlm.nih.gov/pubmed/24135733)

A Study Evaluated Predictors of Adverse Outcomes in Geriatric Patients