Cluster Analysis of Emergency Medical Center Regional Use Patterns among Critically Ill Emergency Patients in Korea (2014-2018)

Abstract

The aim of this study was to analyse the inflow and outflow of critically ill emergency patients in Korea using the National Emergency Department Information System data of the last five years (2014-2018). Using the relevance index (RI) and the commitment index (CI) for analysis, the optimal number of clusters was determined. Moreover, K-means cluster analysis was performed using the determined number of clusters in cities, counties, and districts across the country. The difference between the RI and the CI clusters by year was analysed. The total data analysed included 5,551,616 critically ill emergency patients. The most appropriate number of clusters was two for the years 2014-2018. Cluster 1 was a patient outflow region with a l ow R I and high C I. C luster 2 w as a p atient i nflow region w ith a high RI and low C I. T here w ere no significant differences in the RI and CI each year based on the patient inflow of c ritically i ll emergency patients. In an annual comparison of the CI, significant differences were noted between 2014 and 2017. We expect to use such results to develop a plan for introducing an emergency medical system tailored to each region and for formulating an emergency medical policy.