Hirotoshi Watase1, Yoshiko Nakanishi. 1. Fukagawa Health Consultation Bureau, 3-4-301, Shirakawa, Koto-ku, Tokyo 135-0021, Japan. h-watase05@city.koto.tokyo.jp
Abstract
PURPOSE: To assess geographic variations in the incidence of tuberculosis in Tokyo. METHODS: Using information on tuberculosis incidence, 2000 to 2002, from the annual notification report, patients were categorized into 8 groups by sex and age. We then calculated the Standardized Incidence Ratio (SIR) for each of the 23 wards in Tokyo. The SIR map was described by spatial interpolation and evaluated by cross validation. Spatial scan statistics were used to detect the significance of high-risk areas across the region. We compare this with the proportional distribution of those receiving public assistance and according to the SIR. RESULTS: The geographic variations of SIR did not show a uniform pattern for each group. Spatial scan statistics clearly identified locations, that were significantly high for male groups over 20 years old. Groups under 20 years old and all female groups did not produce high incidence cluster, which are likely to demonstrate spatial features of the proportion of those receiving public assistance. CONCLUSIONS: The geographic distribution of the proportion of those receiving public assistance should impact upon the geographic distribution of the high incidence clusters. However, considering the results of the young age group and each female group, we suggest that recent infection risk among local habitants was almost uniform, with a slightly higher tendency in urban locations of Tokyo.
PURPOSE: To assess geographic variations in the incidence of tuberculosis in Tokyo. METHODS: Using information on tuberculosis incidence, 2000 to 2002, from the annual notification report, patients were categorized into 8 groups by sex and age. We then calculated the Standardized Incidence Ratio (SIR) for each of the 23 wards in Tokyo. The SIR map was described by spatial interpolation and evaluated by cross validation. Spatial scan statistics were used to detect the significance of high-risk areas across the region. We compare this with the proportional distribution of those receiving public assistance and according to the SIR. RESULTS: The geographic variations of SIR did not show a uniform pattern for each group. Spatial scan statistics clearly identified locations, that were significantly high for male groups over 20 years old. Groups under 20 years old and all female groups did not produce high incidence cluster, which are likely to demonstrate spatial features of the proportion of those receiving public assistance. CONCLUSIONS: The geographic distribution of the proportion of those receiving public assistance should impact upon the geographic distribution of the high incidence clusters. However, considering the results of the young age group and each female group, we suggest that recent infection risk among local habitants was almost uniform, with a slightly higher tendency in urban locations of Tokyo.