Designing efficient emergency departments: Discrete event simulation of internal-waiting areas and split flow sorting.

OBJECTIVE: Evaluate nine different models, the interaction of three flow models (ESI, intake attending physician, and no split flow) and three physical design typologies (zero, one, and two internal-waiting areas), on Emergency Department (ED) flow and patient-centered metrics. METHODS: Discrete Event Simulation (DES) was used to systematically manipulate flow and physical design. Three base models were developed and validated using ED and patient specific data. Subsequently, systematic manipulations of flow and internal-waiting areas were performed on other models. Five outcomes of interest were tracked - length of stay (LOS), bed utilization rate, door to provider time, left without being seen rate, and number of movements per patient. Models were compared for statistical significance and effect size using ANOVA, and linear and non-linear regression. RESULTS: The shortest LOS (mean 175.2 min) and highest bed utilization rate (5.02 patients/bed/day) were obtained with flow split by an intake attending physician with two internal-waiting areas. These represented improvements of 54 min and 1.48 patients/bed/day over the control model. Two-way ANOVA demonstrated that both physical design and flow type were statistically significant predictors of all outcomes of interest (p < .0001). Depending on flow type, adding one additional internal-waiting area resulted in decreased LOS (range 10.6-21.8 min), increased bed utilization (range 0.23-0.40 patients/bed/day), decreased D2P (range 1.3-4.8 min), and decreased LWBS (0.66%-2.0%). CONCLUSION: Based on a DES model with empirical data from a single institution, combining flow split by an intake attending physician and multiple internal-waiting areas resulted in improved ED operational and patient-centered metrics.