Cognitive impairment in psychiatric patients and length of hospital stay

The purpose of this study was to determine whether brief neuropsychologic screening at the time of admission to an acute-care psychiatric unit predicts the length of inpatient hospital stay (LOS). Over a 4-month period, 41 consecutively admitted patients on a general psychiatric inpatient unit were administered the Mini-Mental State Examination (MMSE), the Trails A and B subtests from the Halstead-Reitan Neuropsychological Battery, and the Visual Reproduction subtest (VR) of the Wechsler Memory Scale within 72 hours of admission. A setwise, hierarchic multiple regression model examined the contributions of demographic factors, clinical features, and the four neuropsychologic tests to LOS variance. VR and Trails A were the best predictors of LOS, responsible for 21.6% and 16.3% of outcome variance, respectively. The independent variables (IVs) comprising the best-fitting model include Trails A, VR, MMSE, and patient diagnosis. These variables collectively accounted for 48.8% of the variance in LOS, and predicted 70.4% of patients with LOS <21 days and 82.4% of patients with LOS >21 days. The results of this study support the hypothesis that LOS is related to cognitive impairment associated with psychiatric illness, and suggest that neuropsychologic testing upon admission to an inpatient unit can be useful in predicting LOS.