OBJECTIVE: To explore the lung cancer incidence rates from 1972 to 2001 and utilize varieties of models in forecasting trend up to 2010 in the city of Qidong, Jiangsu in order to provide baseline data for its control and prevention. METHODS: Using data from the cancer registry office in Qidong, we tried to reveal the trends of lung cancer incidence by analyzing the time-series on trends extrapolation, exponent smoothness, Box-Jenkins model etc. We also compared the prognostication precision, endow differ power, and established assembled forecast model. RESULTS: Data showed that there had been a rising trend of lung cancer from 1972 to 2001 and would still probably be on the increase in the future. The rate of male and female attained to 70 per 100 000 and 20 per 100 000, predicting that there would be a respective 33 percent and 10 percent increase in 2010. CONCLUSIONS: According to analysis of forecast models, it was right to prognosticate lung cancer incidence from time-series and improve forecast precision through developing combination models. The result of combination seemed close to ARIMA models which suggested that it could serve as a chief way to forecast the incidence of lung cancer.
OBJECTIVE: To explore the lung cancer incidence rates from 1972 to 2001 and utilize varieties of models in forecasting trend up to 2010 in the city of Qidong, Jiangsu in order to provide baseline data for its control and prevention. METHODS: Using data from the cancer registry office in Qidong, we tried to reveal the trends of lung cancer incidence by analyzing the time-series on trends extrapolation, exponent smoothness, Box-Jenkins model etc. We also compared the prognostication precision, endow differ power, and established assembled forecast model. RESULTS: Data showed that there had been a rising trend of lung cancer from 1972 to 2001 and would still probably be on the increase in the future. The rate of male and female attained to 70 per 100 000 and 20 per 100 000, predicting that there would be a respective 33 percent and 10 percent increase in 2010. CONCLUSIONS: According to analysis of forecast models, it was right to prognosticate lung cancer incidence from time-series and improve forecast precision through developing combination models. The result of combination seemed close to ARIMA models which suggested that it could serve as a chief way to forecast the incidence of lung cancer.