Table 6.

Classification of Medical Evaluation Board or Return to Duty Status from Logistic Regression Models with One of the Predictors Dropped from the Full Model^*

Model predictors	Sensitivity	Specificity	AUC (95% CI)
RBANS, ADC, 3+ TBI	0.82	0.67	0.77 (0.66, 0.86)
RBANS, 3+TBI	0.78	0.70	0.76 (0.66, 0.98)
RBANS, 3+TBI	0.78	0.70	0.76 (0.66, 0.98)
RBANS, ADC	0.80	0.63	0.74 (0.62, 0.85)
ADC, 3+ TBI	0.70	0.59	0.67 (0.54, 0.79)

RBANS, Repeatable Battery for Neuropsychological Status; ADC, apparent diffusion coefficient; TBI, traumatic brain injury.

^*The classification threshold is chosen to maximize the sum of sensitivity and specificity. The area under the receiver operating characteristic curve (area under the curve [AUC]) is higher for a better classifier. The AUC is lowest when RBANS is left out of the model.