サーベイランスにもとづく疾病流行の予測 : 福岡県における川崎病を例に

概要

論文の詳細を見る
The validity of disease surveillance is evaluated using the data about Kawasaki Disease from two different sources, one from the surveillance of infectious diseases of Fukuoka Prefecture, and the other from the nationwide survey of Kawasaki Disease conducted by the Japan Kawasaki Disease Research Committee (Chief: Dr. Tomisaku Kawasaki). The time trend of cases of Kawasaki Disease reported weekly in Fukuoka Prefecture from 1981 to 1984, both in the prefectural surveillance and in the nationwide survey, are compared and the validity and usefulness of the prefectural surveillance data in forecasting and quick recognition of epidemics are discussed.The results can be summarized as follows:1) The correlation coefficients between the numbers of cases reported weekly by the two sources were statistically significant.2) The number of epidemic weeks, which were defined as the weekly number of patients in the nationwide survey being more than 10 (or the mean number in the non-epidemic years, 1983 and 1984+2 standard deviations or more), was 45 in the 4 observation years (207 weeks).3) A higher correlation coefficient between the weekly numbers of cases in the nationwide survey and the prefectural surveillance was obtained by taking the data reported in the three previous weeks in the prefectural surveillance as well as the corresponding week. The weight of the respective data from the previous weeks was estimated through a multiple regression analysis.4) The efficacy of the estimation based on the formula thus obtained was evaluated. With the most effective screening criteria, the sensitivity rate or probability of classifying the epidemic week in the nationwide survey as epidemic was as high as 88.9%, and the specificity rate or probability of classifying the non-epidemic week as non-epidemic was 70.4%.5) From these observations, it is concluded that the number of patients observed in the three previous weeks, as well as in the corresponding week, seems to give information important for deciding the existence of epidemics when utilizing prefectural surveillance data.
日本衛生学会の論文

サーベイランスにもとづく疾病流行の予測 : 福岡県における川崎病を例に

スポンサーリンク

概要

著者

関連論文

スポンサーリンク