호남지역 응급실 환자의 유출입 유형 분석 (Inflow and outflow type analysis of emergency department patients of the Honam region)

Objective: This study examined the inflow and outflow patterns of emergency department patients with si-gun-gu in the Gwangju, Jeonbuk, and Jeonnam areas.
Methods: Data from the Gwangju, Jeonbuk, and Jeonnam were extracted from the National Emergency Department Information System in 2016. The extracted data (on 42 areas in Gwangju, Jeonbuk, and Jeonnam) using the variables of the patient’s address (zip code) and the emergency medical institution code (emergency medical institution address) were used to calculate the relevance index and commitment index. The calculated indices were classified into the regional types by applying NbClust and cluster analysis (K-means) of the R package.
Results: The relevance indices ranged from 12.5% to 90.4%, and the commitment indices ranged from 9.2% to 90.3%.
The results of cluster analysis with the relevance indices and commitment indices revealed three types for 39 areas. In cluster 1, the relevance indices ranged from 43.5% to 61.6%, and the commitment indices ranged from 9.2% to 49.5%.
Three out of the thirty-nine areas were classified as the inflow type. In cluster 2, the relevance indices ranged from 12.5% to 56.0% and the commitment indices ranged from 62.5% to 90.3%; 12 areas were classified as the outflow type. The areas in cluster 3 were classified as the self-sufficient type, with relevance indices ranging from 60.1% to 90.4% and commitment indices ranging from 59.0% to 89.7% for 24 areas.
Conclusion: Three area types and 11 out of 12 areas classified as outflow types were found to be emergency medical vulnerable areas. The results of this study can be used to establish local emergency medical policies.

Objective: This study examined the inflow and outflow patterns of emergency department patients with si-gun-gu in the Gwangju, Jeonbuk, and Jeonnam areas.
Methods: Data from the Gwangju, Jeonbuk, and Jeonnam were extracted from the National Emergency Department Information System in 2016. The extracted data (on 42 areas in Gwangju, Jeonbuk, and Jeonnam) using the variables of the patient’s address (zip code) and the emergency medical institution code (emergency medical institution address) were used to calculate the relevance index and commitment index. The calculated indices were classified into the regional types by applying NbClust and cluster analysis (K-means) of the R package.
Results: The relevance indices ranged from 12.5% to 90.4%, and the commitment indices ranged from 9.2% to 90.3%.
The results of cluster analysis with the relevance indices and commitment indices revealed three types for 39 areas. In cluster 1, the relevance indices ranged from 43.5% to 61.6%, and the commitment indices ranged from 9.2% to 49.5%.
Three out of the thirty-nine areas were classified as the inflow type. In cluster 2, the relevance indices ranged from 12.5% to 56.0% and the commitment indices ranged from 62.5% to 90.3%; 12 areas were classified as the outflow type. The areas in cluster 3 were classified as the self-sufficient type, with relevance indices ranging from 60.1% to 90.4% and commitment indices ranging from 59.0% to 89.7% for 24 areas.
Conclusion: Three area types and 11 out of 12 areas classified as outflow types were found to be emergency medical vulnerable areas. The results of this study can be used to establish local emergency medical policies.