1985 Volume 59 Issue 9 Pages 869-876
In order to investigate statistical relation between numbers of patients with streptococcal infectious disease and meteorological data, we carried out multiple regression analysis on mumbers (dependent variable) of the patients per a hospital per a decade of a month and meteorological data (independent variables) of mean temperature (°C), max.temperature (°C), min. temperature (°C), relative humidity (%), mean vapor pressure (mb), wind speed (m/sec), total radiation on a horizontal surface (MJ/m2), no. of hours with sunshine (h), and precipitation (mm) per a decade of a month during the period from September in 1976 through August in 1977 and from September in 1978 through September in 1984, respectively, using a personal computer of type Canon CX-1 and its program for multiple regression analysis.
Obtained multiple regression equation with high variance ratio (43.8971), multiple correlation coefficient (0.9219) and proportion (0.8499) was Y=2.7415 log TMAX+8.7521 log RH-3.4656 log MVP -4.4631 log TRH -12.6089, when all meteorological data were transformed to logarithmic values. Namely, selected independent variables in the equation were max. temperature, relative humdity, mean vapor pressure and total radiation on a horizontal surface, and one-sided probabilities of t-distribution of their partial regressin coefficients were from P=0.0000 to P=0.0055.As far as the above results are concerned, it seems to be able to estimate numbers of the patients from meteorological data by the multiple regression equation.
