Which indicators to include in a crowding scale in an emergency department? A national French Delphi study

Background Emergency department (ED) crowding is a serious international public health issue with a negative impact on quality of care. Despite two decades of research, there is no consensus on the indicators used to quantify crowding. The aim of our study was to select the most valid ED crowding indicators. Materials and methods The Delphi method was used. Selected indicators originated from a literature review and propositions from FEDORU (National Emergency Department Observatory Network) workgroup. Selected national experts were emergency physicians with a special interest in ED crowding. They had to assess each indicator in terms of validity out of a Likert scale from 1 to 9. Indicators withdrawal criteria after each round (consensus) were over 70% of answers of at least 7 with interquartile range less than 3 (positive consensus) or over 70% of answers of at least 4 and interquartile range less than 3 (negative consensus). The decision to stop the delphi procedure was based on the stability of answers between the rounds. Results 41 (89.13%) experts answered the first round and 37 (80.43%) answered the second round. Among the 57 indicators included, 15 reached consensus: four input indicators, six throughput and five output ones. For those three categories of at least 7 answers rates were, respectively, 80.9, 76.9 and 75.0%. Five indicators were deducible from the mandatory Emergency Department Discharge Summary. They obtained 80.2% of at least 7 answers. Conclusion Our study results enable the construction and validation of a crowding measuring tool from indicators approved by experts. It is necessary to further reflect about ED crowding as a concept and what is expected from a complex score.