Cross-validation of the delirium rating scale in older patients

OBJECTIVE: To cross-validate the Delirium Rating Scale (DRS). DESIGN: Cross-sectional. SETTING: Geriatric medicine and geriatric psychiatry assessment units and consultation services. PARTICIPANTS: A total of 104 older patients on the above services. MEASUREMENTS: Mini-Mental State Examination (MMSE) score, Barthel Index score, DRS score, Blessed Dementia Scale score, clinical diagnoses using DSM-III-R criteria. MAIN RESULTS: The mean DRS score was highest in the delirium group. Cronbach's alpha was .90, and inter-rater reliability of total scores was .91 (intra-class correlation). Receiver Operating Characteristic curve analysis showed that the area under the curve for the DRS was significantly higher than the MMSE as a test for delirium. At its published cutpoint of 10, the sensitivity of the DRS is .82 and the specificity is 94. The value at which the sensitivity of the DRS is .90 is 8, at which specificity is .82. CONCLUSIONS: The DRS appears to a feasible instrument. In a sample with a high proportion of delirious patients, it has acceptable measurement properties when used by expert observers.