Objective: Emergency department (ED) overcrowding has been quantified with a scale that reflects the degree of overcrowding (National ED Overcrowding Scale, or NEDOCS) in general academic EDs. However, validity of the 5-question NEDOCS scale has not been established for a pediatric ED. Our primary objectives were to validate the NEDOCS model in our institution's pediatric ED and explore the possibility of another pediatric ED overcrowding model that would be better than the NEDOCS model.
Methods: Objective data were determined by prospectively collecting 20 variables at 42 random site-sampling times in one pediatric ED. Data were obtained by counting patients, determining patient's times, and obtaining information from registration, triage, and ancillary services. The 5 questions needed for the NEDOCS scale were among the data collected. Expert consensus (EC) was obtained using a Likert scale completed by the charge nurse and ED physicians who rated the degree of overcrowding. National ED Overcrowding Scale scores were compared with EC score to determine predictive validity of a model for a pediatric ED. Spearman correlation and multivariable linear regression were used to evaluate individual variables.
Results: Overcrowding based on EC score was found in 18 (44%) of 41 times in the pediatric ED. In pediatric EDs, high correlations were found between EC score and NEDOCS (0.68), number of patients in the waiting room (0.74), full rooms (0.64), and total registered patients (0.65). In a multivariable analysis, a combination of patients in the waiting room and total registered patients had a high correlation (0.80) with EC score in the pediatric ED.
Conclusions: Overcrowding is quantifiable in a pediatric ED. Although the NEDOCS performed well in the pediatric ED, it was outperformed by other variables and other variable combinations. In this pediatric ED, a combination of 2 variables, total registered patients and patients in the waiting room, was a better model than the NEDOCS score for quantifying pediatric ED overcrowding.