Simulation model of pandemic influenza in the whole of Japan

Abstract

An influenza pandemic-an outbreak of a novel influenza virus that spreads around the world-is a real threat. To predict the spread of an influenza pandemic across the whole of Japan, this study applies a real individual-based model (ribm) to the whole of Japan and simulates how a single case of influenza could spark pandemic influenza throughout Japan. We used Person-Trip (PT) data for nine regions (the Tokyo metropolitan area, Kansai, Chukyo, Fukuoka, Sendai, Sapporo, Miyazaki, Okinawa, and northern Ibaragi). PT data were collected from randomly chosen persons and contain information on their locations and their uses of all transportation modes including trains, cars, buses, bicycles, and walking. In total, the nine regions have a population of about 72 million, of whom more than 2.20 million individuals participated in the PT surveys. Estimates on the probability of movement among the nine PT regions are based on the results of the Third National Survey for Movement in 2003. Disease transmission in each region or on a train is assumed to occur within a 1-m radius. The approximate numbers of new cases arising on day 14 after the arrival of the first infected person are estimated to be 322,000 in Tokyo, 25,000 in Kansai, 4,800 in Chukyo, 3,600 in Sapporo, 2,600 in Fukuoka, 600 in Sendai, 17 in Okinawa, and 300 in Miyazaki. The disease seems to spread more slowly in smaller cities such as Miyazaki and Okinawa than in big cities such as Tokyo and Kansai. Area quarantines may help to contain outbreaks in smaller cities, but probably not in larger cities. Because this study demonstrates that ribm is useful for simulating the spread of pandemic influenza in Japan, it is hoped that this modeling technique will be used in the creation of preparedness plans.