Correlation and causality of influenza incidence pattern between regions in the case of South Korea

Abstract

We investigate the interregional relationships of influenza incidence pattern for South Korea. The time series of the number of disease incidence by region was counted from the insurance claim documents. Similarity and influence flows between regions using Pearson correlation and Transfer Entropy (TE) respectively were measured and they are compared with volume of human traffic between regions. The similarity of the incidence pattern increases as the traffic volume increases between regions. It corresponds that the frequent human interactions cause rapid spread of the influenza. The magnitude of influence flow between regions where the traffic volume is neither too much nor too little is generally somewhat large. It can be seen that there is an influence of inducing influenza between regions where are neither too synchronized nor too independent. We realized that these results are unique feature of infectious disease comparing with the results of two chronic diseases.