PURPOSES: The authors derived two forecasting models which can be used as objective tools for detecting epidemics and predicting the future frequencies of communicable diseases. METHODS: In this study, regression analysis using trigonometric functions, Box and Jenkins's seasonal ARIMA model were applied to the monthly accumulated data of five nationally notifiable communicable diseases from January 1987 to December 1998 in Korea. RESULTS: Between two forecasting models, seasonal ARIMA model gives more precise predicted frequencies than regression model in the neighborhood of the current time points and future time, but the regression model is better in overall agreement between the predicted and observed frequencies during 7 years(1992-1998). CONCLUSIONS: These forecasting models can be usefully applied in deciding and carrying out a national policy in preventing epidemics in the future, and graphic program is much helpful to understand the present status of disease occurrence.
Communicable Diseases* ; Forecasting* ; Korea* ; Moclobemide ; Residence Characteristics ; Seasons