Patterns in emergency-department arrivals and length of stay: Input for visualizations of crowding

Research output: Contribution to journalJournal articlepeer-review

Documents

  • EOJ2016

    Final published version, 927 KB, PDF document

  • Morten Hertzum
Crowding is common in emergency departments (EDs) and increases the risk of medical errors, patient dissatisfaction, and clinician stress. The aim of this study is to investigate patterns in patient visits and bottlenecks in ED work in order to discuss the prospects of visualizing such patterns to help manage crowding. We analyze two years of data from a Danish ED for patterns in the patient visits and interview six clinicians from the ED about bottlenecks in their work. The hour of the day explains 50% of the variance in the number of patient arrivals. In addition, there are weekly and yearly patterns in patient arrivals. With respect to the flow of patients through the ED, length of stay increases from low to medium triage levels and then decreases from medium to high triage levels. Also, length of stay increases with patient age. The bottlenecks in the work in the ED relate to patient input (mornings, boom days), patient throughput (staff work hours, linear workflows, manual data entry, overview of patient progress, personal competences), and patient output (no admissions at night, scheduling patient transfers, home transports). The patterns in patient arrivals and length of stay capture factors important to the evolving balance between the demand for ED services and the available resources. Visualization of the patterns, thus, appears a promising tool in managing ED crowding. However, visualizations presuppose reliable data and are expected by the clinicians to be accurate and prognostic. We propose three visualizations.
Original languageEnglish
JournalThe Ergonomics Open Journal
Volume9
Pages (from-to)1-14
Number of pages14
ISSN1875-9343
DOIs
Publication statusPublished - 2016

    Research areas

  • Faculty of Humanities - emergency department crowding, healthcare, Length of Stay, temporal patterns, visualization

Number of downloads are based on statistics from Google Scholar and www.ku.dk


No data available

ID: 161349706