Background: Increased emergency department (ED) length of stay (LOS) is associated with poor patient outcomes and worse satisfaction. Front-end operations in the emergency department (ED) impact LOS and are the processes that occur between patient arrival and a comprehensive provider evaluation. There is no consensus front-end model of what is optimal. We examined the effects of a front-end flow model designated the rapid assessment zone (RAZ) on multiple emergency department (ED) operational metrics.
Methods: This was a retrospective, pre-post study of consecutive patient visits at an urban, community ED. Six-month periods were compared before and after an intervention in 2017. The intervention replaced a waiting room with treatment spaces and created a split-flow model that emphasizes whether patients are likely to remain ambulatory (‘vertical flow’) instead of using vital sign or Emergency Severity Index (ESI) criteria. It also utilizes direct bedding, bedside registration, and advanced triage team methods. The outcomes were ED length of stay (LOS), door-to-provider time, and the rate of left without being seen (LWBS). Data were analyzed with multiple linear regression controlling for patient visit characteristics, ESI, and ED daily census volumes.
Results: We analyzed 43,847 visits in the pre- and 44,792 visits in the post-intervention periods. 4,502 (4.8%) of visits were excluded due to placement in ED observation status for behavioral health or rehabilitation placement purposes. The intervention was associated with the following changes: mean ED LOS from 3.82 to 3.26 hours (-14.7%, 95% CI -0.52 to -0.60 hours), mean door-to-provider time from 49.4 to 21.6 minutes (-56.5%, 95% CI -27.3 to -28.5 minutes), and LWBS from 3.1% to 0.5% (95% CI -2.4% to -2.8%). Regression analysis demonstrated that the reduced LOS after RAZ implementation persisted across ESI levels 2-5 and all ED daily census levels.
Conclusion: The RAZ was associated with improvement in operational metrics at a high-volume, urban ED. This site represents a cohort of centers with great operational needs and few published reports of front-end models, and this may represent a useful model for similar centers.