A revised classification algorithm for assessing emergency department visit severity of populations.

Area of Science:

• Health Services Research
• Public Health
• Emergency Medicine

Background:

• Accurate analysis of emergency department (ED) utilization relies on visit classification algorithms using administrative data.
• Existing algorithms, like the New York University Emergency Department Algorithm (NYU-EDA), require refinement for comprehensive ED use characterization.

Purpose of the Study:

• To present an expanded and revised algorithm, the NYU/JHU-EDA, for classifying ED visits.
• To characterize patterns of ED use and injury severity across US hospitals and health plan enrollees using the new algorithm.

Main Methods:

• The Johns Hopkins University (JHU) team enhanced the NYU-EDA by categorizing uncoded diagnoses and adding injury severity levels, creating the NYU/JHU-EDA.
• The study utilized the National Hospital Ambulatory Medical Care Survey (NHAMCS) and health plan billing data to compare ED visit classifications across different algorithm versions.

Main Results:

• The NYU/JHU-EDA successfully classified 99% of ED visits, identifying a substantial proportion of non-emergent visits.
• Approximately 86% of injury-related ED visits were classified as non-severe, providing new insights into injury management.
• The analysis indicated that a majority of clinically emergent ED visits may not require ED-level care or could be prevented through improved primary care access.

Conclusions:

• The updated NYU/JHU-EDA taxonomy provides a more accurate and detailed understanding of population-level emergency department utilization.
• The findings highlight opportunities for optimizing ED resource allocation and improving primary care interventions to reduce unnecessary ED visits.