Cross validation of the MacNeill-Lichtenberg decision tree: Triaging mental health problems in geriatric rehabilitation patients

Objective: Cross validation of the MacNeill-Lichtenberg Decision Tree (MLDT ), a tool for triaging mental health problems in older medical patients. De sign: Sensitivity, specificity, and predictive power were calculated for pr ediction of cognition and depressive symptoms. Correlational analyses were performed to examine the relationship between demographic variables and MLD T measures. Correlational and logistic regression analyses were conducted t o examine the utility of the cognitive component of the MLDT. Setting: A fr eestanding urban, university affiliated rehabilitation hospital in the Midw est. Participants: One hundred eighty-seven elderly, predominantly African American medical rehabilitation patients who lived alone before hospitaliza tion. Main Outcome Measure: The MLDT was compared with the Mattis Dementia Rating Scale and the 30-item Geriatric Depression Scale. Results: The cogni tive component of the MLDT demonstrated high positive and negative predicti ve value; the affective component evidenced excellent negative predictive v alue. Conclusions: These findings are consistent with previous research and provide further evidence of the utility of the MLDT in (a) quickly and acc urately identifying geriatric rehabilitation patients with cognitive impair ment, (b) guiding the nature of the referral question, and (c) reducing the number of unnecessary referrals for both cognitive deficits and depression .