Analyzing risk factors, resource utilization, and health outcomes of hospital-acquired delirium In elderly emergency department patients

BACKGROUND: Delirium is a disorder that is characterized by an acute change in cognitive functioning including inattention, and disordered thinking. Delirium disproportionately affects the population over the age of 65, and is associated with increased costs, worse outcomes and longer lengths of stay. Although delirium is estimated to affect approximately 10% of elderly patients in the emergency department (ED) and 42% of elderly inpatients, it often goes unrecognized by the clinical staff. There is evidence that delirium can be prevented through non-pharmacologic prevention strategies, however it is less clear which patients should be targeted for these measures. OBJECTIVES: The objective of this study is to identify risk factors for development of hospital-acquired delirium during the most proximal aspect of a patient’s hospital course, namely the ED. Secondary objectives of this study are to analyze resource utilization and outcomes associated with the development of hospital-acquired delirium. METHODS: This study is a secondary analysis of a prospective observational study conducted over 3 years at a single urban university hospital. Patients over the age of 65, who could complete a structured cognitive assessment interview, were screened for delirium by a trained research assistant. Patients that were judged to be not delirious in the ED, and who were then admitted to an inpatient unit were included in the final cohort. A validated chart review method was used to determine if patients developed delirium during the course of their hospitalization. Potential predictors of hospital-acquired delirium, including demographics, laboratory values, comorbidities and outcomes, were also abstracted from the medical chart. We performed a univariate analysis of these predictors and included those covariates with a p values ≤0.2 in multivariate analysis. We allowed 1 predictor per 10 outcomes in the final model to avoid over-fitting and evaluated the discriminatory ability and calibration of the model using the c-statistic and Hosmer-Lemeshow goodness-of-fit test. RESULTS: Of the 520 patients included in our cohort, 77 developed delirium over the course of their inpatient visit. Multivariate analysis identified 7 risk factors to predict delirium in elderly emergency department patients admitted to the hospital. Patients were more likely to develop delirium during their stay if they were age 80 or older, had a history of dementia, had a history of stroke or transient ischemic attack, were hypoxic or hyponatremic in the ED, or had an ED admitting diagnosis of acute stroke/transient ischemic attack or fall. The model had a c-statistic of 0.73 and a non-significant p-value of 0.7 in the Hosmer-Lemeshow goodness-of-fit test. CONCLUSION: The predictive model that we created may help identify a population to target for delirium prevention strategies in elderly emergency department patients, thereby reducing delirium incidence in hospitalized patients, and the associated morbidity, mortality, and healthcare utilization