2025 年 12 巻 1 号 p. 93
Introduction and Objectives: Disaster exercises provide an opportunity to evaluate response capabilities and identify problems in disaster preparedness. However, debriefing and questionnaires are often used for evaluation, which are insufficient for continuous improvement or comparative evaluation with other hospitals. The objective of this study was to collect various data during exercise and perform a quantitative analysis to evaluate response performance. This study also aimed to develop a computational model of hospital disaster response based on our analysis.
Approaches: We collected data during the annual exercise at a major Japanese hospital designed to respond to a mass casualty incident involving approximately 50 dummy patients and more than 100 staff. We tracked the flow of patients and documents by using NFC tags and smartphone applications. We analyzed the length of stay of each patient, the number of patients in each area, the accuracy of the triage, and compared the exercise performance over two years. We also developed a simulation model based on the analysis to simulate both normative and descriptive responses.
Results: We tracked the patient flow with our method; however, we could not track the document flow well. The analysis showed that while the triage accuracy was not significantly different, the length of stay and crowding in each area varied over the two years. We also performed test simulations and confirmed that the simulations replicated the exercise results.
Discussion: The performance measures used in this study are objective and quantitative; therefore, we can compare the results across years and hospitals. If many hospitals conduct the same analysis and share the results, disaster preparedness can be effectively improved, not only for individual hospitals, but also for regional communities. We also expect that computer simulations of normative responses can provide a benchmark for such cross-hospital evaluations.
