For the purpose of forecasting the date of prevalenc of Japanese encephalitis in Japan, we estimated the median date of this epidemic time curve of the prevalence from the regression equation of one variable and multiple regression equation from the above factors using an electronic computer. And the following results were obtained.
1) To estimate median date (y) and mode (z) of the epidemic curve of the prevalence, we can use the next equation:
The regression equation to estimate y or z from T. 6, 7 (x
1) is as follows.
y=3.75x
1+144.47 σ
0=12.4……(1)
z=-3.80x
1+157.26 σ
0=14.9 ……(1)'
The regression equation from D. pos. swine (x
2) is as follows.
y=0.68x
2+31, 82 σ
0=9.7 ……(2)
z=0.76x
2+40.71 σ
0=12.0……(2)'
The multiple regression equation from T 6, 7 and D. pos. swine is as follows.
y=1.07x
1+0.62x
2+59.37 σ
0=9.7……(3)
z=-0.79x
1+0.71x
2+61.02 σ
0=12.0……(3)'
The multiple regression equation from T 6, 7 D. pos. swine, latitude and longitude is as follows.
y=1.01x
1+0.58x
2-0.26x
3+0.37x
4+18.50 σ
0=9.8……(4)
z=-0.32x
1+0.52x
2+2.05x
3+0.54x
4-87.81 σ
0=11.8……(4)'
These factors (x
1, x
2)were obtained as the results from 1965 to 1967.
2) We got the estimated value of median and mode in 18 prefectures in 1968 and 14 prefecture, in 1969 in Kyushu, Chugoku, Shikoku, Kinki and Kanto provinces from (1) or (2) or (3) or (4) equation. Ten prefectures out of 17 prefectures by equation (1), 12 out of 18 by (2), 13 out of 18 by (3), and 13 out of 18 by (4) could be estimated with about 10 days (σ
0) error or less in 1968. And in 1969, 9 out of 14 by (3). 7 out of 13 by (4) could be estimated. In the case of mode, 10 out of 18 by equation(1)', 6 out of 17 by (2)', 8 out of 17 by (3)', 12 out of 18 by (4)' could be estimated with about 12 days (σ
0) error or less in 1968, and 9 out of 14 by equation (3)' and 9 out of 13 by equation (4)', in 1969.
3) For the estimation of median and mode of prevalence of Japanese encephalitis two new factors; the date when the numbers of caught mosquitoes (Q. t.) were found to be maximum, and the date when hazardous mosquitoes were found at first, were added to above four factors (T. 6, 7 and D. pos. swine, latitude and longitude) by the data on multiple regression equation. But accuracy of estimation did not increase significantly.
View full abstract