579 Communicating Crowding Status: Developing and Validating a Traffic Light Framework for Surge Scores

Document Type

Conference Proceeding

Publication Date

9-1-2025

Publication Title

Ann Emerg Med

Abstract

Study Objective: Quantifying levels of emergency department (ED) crowding that can be incorporated into healthcare system ‘surge’ description requires accounting for variation in hospital size and structure while still reflecting local levels of strain. The National Emergency Department Overcrowding Scale (NEDOCS) developed in 2004, is a validated communication tool used to quantify ED crowding. However, applying NEDOCS across diverse ED settings presents challenges related to generalizability. The objective of this project was to develop and validate a standardized approach for defining thresholds of ED crowding across variably sized EDs allowing more meaningful integration into a systemwide surge score. Methods: This study was conducted within a health system that includes an urban academic level 1 trauma center (D), four community and mixed academic hospital EDs (J, M, Wy, and WB), and five freestanding EDs (FSED) (B, C, F, P, S). Recognizing that the original NEDOCS variable, number of ventilated patients, underestimates the burden of critical care delivered in the today’s ED’s, we modified the score by substituting this variable with the number of critical care or ICU boarders. Initial validation began at the urban academic ED, which sees the highest patient volume in the system. ED crowding thresholds were defined using a traffic light-style color schema: Green, Yellow, and Red. Project stakeholders and hospital throughput leaders were mindful of the risk of desensitization if the Red threshold was set too low. An Excel-based modeling tool was used to create “what if” scoring scenarios across a wide range of variable combinations used in the modified NEDOCS (mNEDOCS) calculation. For modeling purposes, “ED Beds” and “Hospital Beds” were treated as fixed values. These “what if” scenarios were created to test our assumptions, identify leverage points, assess tactics and set thresholds. To validate thresholds across EDs in the system, each ED Medical Director selected one of two approaches: (1) prospective collection of mNEDOCS variables and perception of crowding at specific times of day to determine best-fit thresholds, or (2) use of the Excel-based “what if” tool to simulate score scenarios and identify Green, Yellow, and Red thresholds based on local conditions. Results: The urban academic level 1 trauma center was the first ED in the health system to validate the mNEDOCs model. Thresholds selected established a traffic light-based communication tool describing ED crowding, categorizing states as Green (not busy, busy, and extremely busy but not overcrowded), Yellow (overcrowded and severely overcrowded), and Red (dangerously overcrowded) (Figure 1). For system-wide expansion, each Medical Director selected the “what if” analysis tool for validation. Among the three highest-volume EDs (D, J, M), identical thresholds were set for all crowding categories. The next two mid-sized EDs (Wy, WB) established the same threshold for dangerously overcrowded, but this was 20 points lower than the larger EDs' threshold. Among the five freestanding EDs (FSEDs), there was significant variation in the dangerously overcrowded threshold, ranging from 106 to 180, a 75-point spread. The geometric mean threshold among FSEDs was 141.5 (SD = 23.8), with a median of 145. Conclusion: Development and validation of a standardized tool using a modified NEDOCS (mNEDOCS) score to describe ED crowding across a health care system is feasible. By establishing locally meaningful thresholds mapped to a traffic light schema, the tool enables consistent communication of crowding across diverse ED settings. In larger health systems with centralized coordination of transfers and bed placement, mNEDOCS has the potential to inform resource allocation, identify operational leverage points, andsupportsystemwide surge response. [Formula presented] No, authors do not have interests to disclose

Volume

86

Issue

3

First Page

S249

Find It @ Sladen

Share

COinS