Applicability of the modified Emergency Department Work Index (mEDWIN) at a Dutch emergency department

Steffie H. A. Brouns*, Klara C. H. van der Schuit, Patricia M. Stassen, Suze L. E. Lambooij, Jeanne Dieleman, Irene T. P. Vanderfeesten, Harm R. Haak

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

Background

Emergency department (ED) crowding leads to prolonged emergency department length of stay (ED-LOS) and adverse patient outcomes. No uniform definition of ED crowding exists. Several scores have been developed to quantify ED crowding; the best known is the Emergency Department Work Index (EDWIN). Research on the EDWIN is often applied to limited settings and conducted over a short period of time.

Objectives

To explore whether the EDWIN as a measure can track occupancy at a Dutch ED over the course of one year and to identify fluctuations in ED occupancy per hour, day, and month. Secondary objective is to investigate the discriminatory value of the EDWIN in detecting crowding, as compared with the occupancy rate and prolonged ED-LOS.

Methods

A retrospective cohort study of all ED visits during the period from September 2010 to August 2011 was performed in one hospital in the Netherlands. The EDWIN incorporates the number of patients per triage level, physicians, treatment beds and admitted patients to quantify ED crowding. The EDWIN was adjusted to emergency care in the Netherlands: modified EDWIN (mEDWIN). ED crowding was defined as the 75 th percentile of mEDWIN per hour, which was >= 0.28.

Results

In total, 28,220 ED visits were included in the analysis. The median mEDWIN per hour was 0.15 (Interquartile range (IQR) 0.05-0.28); median mEDWIN per patient was 0.25 (IQR 0.15-0.39). The EDWIN was higher on Wednesday (0.16) than on other days (0.14-0.16, p<0.001), and a peak in both mEDWIN (0.30-0.33) and ED crowding (52.9-63.4%) was found between 13: 00-18: 00 h. A comparison of the mEDWIN with the occupancy rate revealed an area under the curve (AUC) of 0.86 (95% CI 0.85-0.87). The AUC of mEDWIN compared with a prolonged ED-LOS (>= 4 hours) was 0.50 (95% CI 0.40-0.60).

Conclusion

The mEDWIN was applicable at a Dutch ED. The mEDWIN was able to identify fluctuations in ED occupancy. In addition, the mEDWIN had high discriminatory power for identification of a busy ED, when compared with the occupancy rate.

Original languageEnglish
Article numbere0173387
Number of pages14
JournalPLOS ONE
Volume12
Issue number3
DOIs
Publication statusPublished - 10 Mar 2017

Keywords

  • OUTCOMES
  • NEEDS
  • STAY
  • CARE

Cite this