摘要
目的对广州市5岁以下儿童死亡率(U5MR)进行分析和预测,为健全儿童保健工作提供科学依据。方法使用SPSS 19.0对2001-2017年广州市U5MR构成的时间序列建立整合移动平均自回归(ARIMA)模型。得到相对最优拟合模型后对广州市U5MR进行短期预测。结果 ARIMA(1,2,0)模型对广州市U5MR拟合效果较好,用该模型预测广州市2018年至2020年的数据符合U5MR的变动趋势,实际值均落在拟合模型的置信区间之中。结论 ARIMA模型适用于拟合U5MR的时间变化趋势,对短期的广州市U5MR预测具有较强的实用价值。
Objective To analyze and predict the mortality rate of children under five(U5MR)in Guangzhou,and to provide scientific basis for the improvement of children′s health care.Methods SPSS 19.0 was used to establish an integrated mobile average autoregression(ARIMA)model for the time series composed of U5MR in Guangzhou from 2001 to 2017.After obtaining the relative optimal fitting model,the short-term prediction of U5MR in Guangzhou was carried out.Results The ARIMA(1,2,0)model has a good effect on U5MR fitting in Guangzhou.The prediction with the mode lshows that the data of Guangzhou from 2018 to 2020 conforms to the U5MR trend,and the actual values fall into the confidence interval of the fitting model.Conclusion The ARIMA model is suitable for fitting the time change trend of U5MR and has strong practical value for short-term U5MR prediction in Guangzhou.
作者
刘涛
Liu Tao(Department of Medical Records and Statistics,the First Affiliated Hospital of Guangdong Pharmaceutical University,Guangzhou 510080,China)
出处
《中国医院统计》
2020年第3期259-261,共3页
Chinese Journal of Hospital Statistics
