A different crowd, a different crowding level? The predefined thresholds of crowding scales may not be optimal for all emergency departments

Introduction Previous studies indicate that crowding scales may not perform well in low-volume emergency departments (EDs). In this study, face-validity of the Modified National ED OverCrowding Score (mNEDOCS) was assessed in a high-volume ED as well as in a low-volume ED. Methods A prospective observational cohort study was performed in the Netherlands. The correlation of the mNEDOCS with ED staff perceptions of crowding were assessed, using weighted Kappa (κ) and Pearson correlation. Subsequently, ED process measures (elapsed target times to triage, elapsed target times to treatment and patients’ LOS) were described under different levels of ED crowding. Results Correlation between the categorized crowding scores was low (weighted κ 0.34 resp. 0.26). However, good correlations of 0.73 and 0.82 were found between the uncategorized mNEDOCS and ED staff’s perception of crowding. Percentages of patients with elapsed target times to treatment increased simultaneously with increasingly busy periods when measured with mNEDOCS. Conclusions The uncategorized mNEDOCS correlates well with perceived crowding, even at a low-volume ED. Determining a cut-off level at which a specific ED can be identified as crowded is important, because the predefined mNEDOCS categories may not be optimal for all EDs