摘要
目的探讨三种模型进行甲肝发病率的预测,选择最优方法,为甲肝防治提供科学依据。方法对陕西省某地区2004-2011年甲型肝炎发病率分别采用动态数列法、年估计百分比变化(EAPC)及直线回归模型三种方法进行计算,预测该地区2012年甲肝发病率,并与实际值进行比较,选出最优方法。结果年估计百分比变化(EAPC)法预测数值与真实值最接近,且该方法将计算的前提假设(变化趋势稳定)与统计学检验结合起来,从理论上更加完善,故EAPC计算最佳。结论动态数列法适用于数据波动稳定的短期预测,直线回归法适用于观察值随时间变化呈线性关系的数据预测,EAPC法适用于对数据波动未知,长期变化数据趋势分析及预测。通过对甲肝发病率的预测,为指导该地区甲型肝炎的防治提供了帮助。
Objective To choose the best way from the three models to forecast the incidence of hepatitis A, so as to provide basis for the prevention and treatment of hepatitis A. Methods Using the dynamic series method, the estimated annual percentage change method and linear regression model to calculate the data from 2004 to 2011 in some areas of Shaanxi Province, to forecast the incidence of hepatitis A in 2012, and compared with the actual value to choose the best method. Results The result of EAPC was the closest forecast value in the three methods, and it was calculated on the assumptions which combined with statistical tests, theoretically more perfect, therefore, the EAPC calculation was the best one. Conclusions The dynamic series method was suitable for short-term and stable-data prediction, while linear regression method was suitable for the observation over time. EAPC applies to the prediction for the long-term, fluctuated, and uncertain data. Through forecasting the incidence of hepatitis A, it is helpful for the prevention and treatment of hepatitis A in the districts.
出处
《公共卫生与预防医学》
2014年第2期18-20,共3页
Journal of Public Health and Preventive Medicine
