Implementation of a Front-End Split-Flow Model to Promote Performance in an Urban Academic Emergency Department

In an urban academic emergency department (ED), a front-end split-flow model, which entailed deployment of an attending-physician intake model, implementation of a 16-bed clinic decision unit, expanded point-of-care (POC) testing, and dedicated ED transportation services, was created.

A retrospective, observational, pre-post intervention comparison study was conducted at a large academic urban hospital with 74,000 ED annual visits that serves as a Level 2 Trauma Center. The new flow model was implemented in April 2013, coincident with the opening of a new ED space.

During the six-month pre- (July 2012-December 2012) and postimplementation (July 2013-December 2013) periods, there were 17,307 and 27,443, respectively, walk-in encounters during the intake times. Despite this 59% increase and a 35% increase in overall ED patient census, implementation of the innovative novel process redesign resulted in a clinically meaningful reduction (median minutes pre vs. post and one-year post) in (1) overall length of stay (LOS) for all walk-ins (220 vs. 175 and 140), discharged (216 vs. 170 and 140), and inpatient admissions (249 vs. 217 and 181); (2) door-to-physician time (minutes) (54 vs. 15 and 12); and (3) left without being seen (LWBS) rates (5.5% vs. 0.5% and 0.0%). The left before visit complete (LBVC) rates were 0.8% vs. 1.1% and 0.6%. The average total relative value unit (RVU) per patient discharged from intake was 2.31. During the pre-post analysis periods, no significant increase in reported safety events were identified (10 vs. 9 per 1,000 patient encounters).

Implementation of a novel multifaceted process redesign including an attending physician-driven intake model had a clinically positive impact on ED flow. Validation of this model should be conducted in other practice settings.