結核疫学指標の標準化に関する研究

第1報 1970年都道府県別標準化死亡率

抄録

Because of the recent rapid development of Japanese economy, the interprefectural differences in age distribution of population have become very remarkable. Therefore, for accurate evaluation of certain epidemiological indices, such as death rate, by prefecture, the necessity of standardization according to the age distribution of the standard population, has been increasing especially for chronic diseases prevalent among the middle or old aged people.
This study was conducted to compare the tuberculosis death rates by prefecture on standardized basis, and to examine the usefulness of our new indicator on community health level (Uematsu's Indicator) for evaluating the interprefectural differences in tuberculosis mortality.
Standardized death rates of tuberculosis (all forms) according to the indirect method and the Uematsu's Indicators on tuberculosis death, were calculated for all prefectures and 7 large cities based on the data from the Population Census and the Vital Statistics in 1970.
The Uematsu's Indicator is calculated by the following formula.1/√ΣN_ai…Number of subjects in each age category of the observed population (each Prefecture)
P_ai…Death rate in each age category of the observed population
Psi…Death rate in each age category of the standard population (entire country)
The standard error of this indicator is given by _??_ Nai and the test of significance between the values for the observed population and that for the standard population can be easily done by t-test.
The results are as follows.
Comparing the standardized death rates for both sexes combined, Osaka (22.4 per 100, 000) shows the highest rate. Besides it, 10 prefectures and 3 large cities (most of them are located in Kyushu Island or in the Western Honshu, see Fig. 1) show remarkably significant high rate (p<0.001) from that for the entire country (15.4 per 100, 000).
Among them, 7 prefectures (Osaka, Kagoshima, Kumamoto, Nagasaki, Fukuoka, Hyogo and Oita) and 3 cities (Osaka, Kobe and Nagoya) are thought to be the prominently higher mortality areas, because remarkably significant high values are also shown by the Uematsu's Indicator.
Nagano shows the lowest rate (6.6 per 100, 000). The lower mortality prefectures are located in the Eastern Honshu (except the northern end) and in the Japan Sea coast districts of the Western Honshu.
From these results, the authors made the following conclusions.
1) If the death rates are evaluated without standardization, those in urban areas might be underestimated and those in rural areas might be overestimated.
2) The Uematsu's Indicator is considered to be a useful index to evaluate the interprefectural differences in tuberculosis mortality, since it reflects sensitively the difference in death rate between the observed and the standard population in each age category.