146 Patients with Hospital Delirium Are Most Often Identified in the Emergency Department and Experience Longer Hospital Stays

Delirium is a common, costly, and devastating condition which affects up to 50% of hospitalized older patients, has an associated hospital mortality rate of up to 33%, and procures annual health care expenditures exceeding $152 billion. By identifying where delirium occurs in a patient’s hospital course, there is potential to improve how we manage their condition. Our goal was to evaluate where patients were detected with hospital delirium and whether or not they experienced longer hospital stays. We conducted a single-center, retrospective chart review of all patients 18 years or older who presented to the emergency department (ED) of an academic, quaternary-care, level one trauma center between January 1st - December 31st of 2018, who were subsequently admitted to the hospital. We defined delirium as any diagnosis of agitation, hallucinations, metabolic encephalopathy, or delirium in the electronic medical records in addition to surrogate measures of delirium management including: PRN orders for psychotropic medication such as Haloperidol, Quetiapine, Valproic Acid, or Lorazepam and/or orders for Enhanced Supervision, Constant Observation, Restraints, or the Confusion Assessment Method-ICU (CAM-ICU). Patients were excluded if they had insufficient hospital visit time data and/or presented to the ED with baseline altered cognition such as a chief complaint of intoxication, alcohol/drug withdrawal, altered mental status, suicidal ideations, delusions, psychiatric evaluations, dementia, or delirium. The control group included all other patients 18 years or older who did not meet criteria for delirium and were admitted via the ED with usable hospital visit time data. For both groups, we assessed whether hospital length of stay differed between patients detected with delirium and those who were not using a Mann-Whitney U Test. The effect of potential confounders such as Charlson Comorbidity Index (CCI), age, race, ethnicity, and sex were assessed using bivariate analyses. We analyzed a sub-sample of delirium patients with complete medication/order histories to determine where delirium was identified. Patients who did not meet any criteria for delirium were in the control group (n = 24552). Of the 2045 patients meeting the criteria for delirium, 1950 (95.4%) had complete medication/order histories and time data; 147 (7.6%) were identified in triage or while waiting for an ED bed, 1370 (70.2%) were identified in the ED core, and 433 (22.2%) were identified as an inpatient. Patients identified with delirium versus those who were not had a significantly longer median hospital length of stay (5.0 days vs. 4.3 days, p<0.001). Whereas age, sex, and ethnicity were not significantly different between both groups (p = 0.94, p = 0.98, p = 0.63), patients identified with delirium had significantly higher CCI scores (3.2 vs. 2.0, p<0.001). There was also a significant difference in the racial distribution between delirium and control groups (p<0.001). Patients who are treated for delirium have significantly longer hospital stays relative to all other admitted patients. These patients are most frequently detected with delirium during their ED stay. Further studies will be conducted to determine how we can improve the identification and management of delirium to improve patient outcomes.