Defining and Validating a Metric for Emergency Department Crowding

Abstract

Aim: Emergency Department (ED) crowding intuitively impedes quality of care. Despite an extensive literature base, there is no agreement on which measure of crowding best reflects quality of care. This thesis aimed to determine which ED crowding metric(s) should be used as quality of care and performance indicator(s) to drive improvements in care quality. Methods: The thesis explains why ED is an important place to measure healthcare quality and describes the development of the Quality Indicator Critical Appraisal (QICA) tool to appraise indicators. A survey was conducted to evaluate the measurement of ED crowding in Australasia and a systematic review undertaken to determine the strength, direction and quality of evidence for associations between metrics and healthcare quality domains. Finally, the association between the metrics and mortality was tested in two million unique ED presentations in 25 hospitals in New Zealand from 2006-2012. Results: Sixteen metrics were identified in the survey. The review identified 198 studies investigating the association between these metrics and quality of care. After appraisal of each metric using the QICA tool, there was a strong recommendation that ED Length of Stay (LOS) should be used, with a conditional recommendation for using total ED Occupancy and Hospital Occupancy. There were also conditional recommendations that Time to Assessment, Boarding Time and Boarder Occupancy may be used. Conversely, there was a conditional recommendation that ED Treatment Time should not be used and strong recommendations that other metrics were not suitable. In New Zealand EDs after adjustment for confounders, 7- day mortality for new patients was highest when more than 10% of current patients awaiting admission had EDLOS greater than eight hours: HR=1.10 (1.05, 1.17) p<0.001 and when fewer than 90% of all ED patients had LOS less than four hours: HR=1.07 (1.01, 1.12) p=0.011. For ED Occupancy optimised between 85% and 99% mortality was lower: HR=0.91 (0.86, 0.97) p=0.005. Conclusion: Total EDLOS and ED Occupancy are both associated with mortality for new patients arriving at ED in New Zealand. Total EDLOS is the metric associated with the most quality domains and best satisfies the requirements of a good quality indicator.