Delirium in Hospitalized Older Persons: Outcomes and Predictors

Abstract

The purpose of this study was fourfold; to determine the rate of delirium among hospitalized older persons, to contrast the clinical outcomes of patients with and without delirium, to identify clinical predictors of delirium, and to validate the predictive model in an independent sample of patients. Two prospective cohort studies Medical and surgical wards of 2 university teaching hospitals. In the derivation cohort, 432 patients were enrolled from the University of Chicago Hospitals. Patients 65 years of age or older admitted to 1 of 4 wards were eligible. Subjects were excluded if they were discharged within 48 hours of admission, unavailable to the research assistants during the first 2 days of hospitalization, or judged too impaired to participate in the daily interviews. In the test cohort, 323 patients 70 years of age or older admitted to Yale-New Haven Hospital were studied. Subjects were screened for delirium daily and referred to experienced clinician investigators if acute mental status changes were observed. The clinician investigators assessed the patient for delirium based on DSM-III-R criteria. Duration of hospitalization was adjusted for diagnosis-related groups (DRG) and mortality rates were determined at discharge and 90 days after discharge. Sociodemographic characteristics, cognitive and functional status, comorbidity, depression, and alcoholism were examined as predictors of delirium. The rate of delirium in the derivation cohort was 15%; subjects with delirium had longer hospital stays and an increased risk of in-hospital death. Cognitive impairment, burden of comorbidity, depression, and alcoholism were found to be independent predictors of delirium. The ability of the model to stratify patients as low, moderate, or high risk for developing delirium was validated in the test cohort in which the rate of delirium was 26%. This study confirms the high rate of delirium among hospitalized older persons and the associated adverse outcomes of prolonged hospital stays and increased risk of death. Patients can be stratified according to their risk for developing delirium using relatively few clinical characteristics which should be assessed, on all hospitalized older persons.