Emergency Department Boarding Duration and Incident Delirium: Evidence From the Epic Cosmos Database

Emergency department (ED) boarding is a pervasive consequence of hospital crowding, yet its impact on acute cognitive outcomes remains poorly characterized. Delirium is a common, preventable complication, but prior studies linking it to boarding have been limited in scale. Our objective was to evaluate whether prolonged ED boarding is independently associated with incident delirium using a national dataset.

We conducted a retrospective cohort study of elderly ED patients (aged 65+) admitted between January 1, 2024, and 2025 Dec 1 using the Epic Cosmos database. We excluded patients with pre-existing delirium, primary psychiatric admissions, or missing boarding data. ED boarding was categorized as: <4 hours (referent), 4–8 hours, and >8 hours. The primary outcome was a composite of incident delirium occurring within a clinically reasoned post-admission window (e.g., first 24 hours) identified via standardized diagnostic code and parenteral sedative administration. We used multivariable logistic regression to estimate adjusted odds ratios (aOR), controlling for patient demographics, comorbidities (Charlson index), and pre-admission medication exposure.

The final cohort included 51,419 patients (mean age 78.04; 41.9% female). Delirium incidence was 2.7% overall. The risk of delirium increased progressively with boarding duration, with an absolute increase of 0.023 percentage points per additional hour of boarding (p < 0.05). Compared to those boarding <4 hours, patients boarding 4–8 hours had an aOR of 1.10 (95% CI: 1.08-1.11), and those boarding >8 hours had an aOR of 1.20 (95% CI: 1.17-1.24) of experiencing delirium. These associations remained statistically significant across clinically relevant subgroups defined by age, sex, race, pre-existing dementia status, baseline comorbidity burden (Charlson score), polypharmacy status, prior psychiatric medication use, and overnight boarding status.

Prolonged ED boarding is associated with an increased risk of incident delirium. These findings identify boarding as a modifiable operational risk factor with significant implications for patient safety. Future research will utilize hierarchical modeling to account for hospital-level clustering and evaluate whether boarding-reduction interventions can decrease delirium rates