Differences in patient population and length of stay between freestanding and hospital-based emergency departments

Differences in patient population and length of stay between freestanding and hospital-based emergency departments

Simon Erin L; Shakya Sunita; Muir McKinsey; Fertel Baruch S

The American Journal of Emergency Medicine

2019

2019-09

BACKGROUND: Freestanding emergency departments (FEDs) represent over 10% of emergency departments (EDs) in the United States. Little is known about differences in encounter characteristics. We compared ED length of stay (LOS) clinical demographics, method of arrival, acuity level, and patient disposition for encounters to FEDs vs. hospital-based EDs (HBEDs). METHODS: A multi-center retrospective analysis was performed. Study sites included 6 FEDs and 13 HBEDs from 10/1/2017 to 9/30/2018. Data was abstracted from ED records and discharge summary within the electronic health record. Descriptive statistics were reported with prevalence (95% Confidence Interval [CI]) for categorical variables and mean (standard deviation [SD]) for continuous variables. Multivariable linear regression assessed the relationship between ED facility (FEDs vs. HBEDs) and ED length of stay (LOS). RESULTS: 1,263,297 encounters were analyzed. Mean ED LOS was shorter at FEDs (146.62 min (±97.04)) vs. HBED (249.70 min (+287.50)). Nine percent of FED encounters arrived via EMS vs. 21% at the HBEDs. FEDs saw 5.47% emergency severity index (ESI) level 2 vs. 13.76% at the HBEDs. Medicaid and Medicare patients were more prevalent in HBEDs (64.2%) than in FEDs (50.6%). FEDs admitted 13% of patients and HBEDs 27%. All results were significant (p < 0.001). After adjusting for potential confounding variables, patients utilizing FEDs had 16.2% shorter ED LOS vs. HBEDs (β = -0.18 [95% CI: -0.18 to -0.17]). CONCLUSION: Overall ED LOS was significantly less for FED vs. HBED patients. Acuity level, insurance status, method of arrival, and patient disposition were significantly different at FEDs vs. HBEDs.

Article information provided for research and reference use only. All rights are retained by the journal listed under publisher and/or the creator(s).

1738-1742

9

37