B104-01 Association of Emergency Department Crowding and Disposition Decisions in Sepsis

Abstract Rationale Emergency Department (ED) crowding is associated with delays in diagnosis and treatment, increased rates of hospital admission, and worse patient outcomes. For patients with sepsis in the ED, while ED crowding has not been found to influence mortality, crowding is associated with longer door-to-antibiotic time. It is unclear whether changes in ED disposition during times of ED crowding may influence treatment and outcomes in sepsis. We sought to measure the association between ED crowding and ED disposition and hypothesized that crowding would be associated with increased likelihood of hospital admission. Methods We performed a retrospective cohort study of adults presenting to four Utah EDs from 2013 to 2017 who met Sepsis-3 criteria in the ED. We excluded patients whose ED disposition was not discretionary (e.g., death in the ED, discharge to hospice, left against medical advice, ED intubation, ED receipt of vasopressors). The primary outcome was ED discharge (vs. admission); the secondary outcome was ICU vs. ward admission among admitted patients. The primary exposure, ED crowding, was measured by ED occupancy rate (ratio of patients registered in the ED to licensed ED beds at patient’s ED arrival). We used multivariable mixed-effects logistic regression adjusting for site, demographics, comorbidities, and illness severity on ED arrival. Secondary analyses evaluated alternative measures of ED crowding including overcrowding status (occupancy rate ≥1), patient-to-nurse ratio, patient-to-physician ratio, ED Work Index score, and concurrent arrival volume (number of patients arriving to the ED 30 minutes before and after the index patient). Results There were 12,111 ED patients with median age 63 years, 57% male, and median ED SOFA score 3. Of these, 17% were discharged from the ED, 62% were admitted to a ward, and 21% to an ICU. The median ED occupancy rate was 0.66 and 13% of encounters occurred during overcrowding. After adjustment, ED occupancy rate was not associated with discharge (OR 1.04, 95% CI 0.85-1.29) or ICU vs. ward admission (OR 1.06, 95% CI 0.79-1.42). Results were consistent across alternative measures of ED crowding (Table 1). Conclusions In this large cohort of ED patients with sepsis, ED crowding metrics were not associated with ED disposition. There may be multiple compensatory processes that explain how crowding leads to delays in treatment (e.g., longer median door-to-antibiotic time) but does not alter disposition. Alternatively, more granular clinician-level metrics of decision-making under high workload may reveal subtler effects not revealed in this study. This abstract is funded by: NIGMS (R35GM151147)