摘要
目的应用季节性指数平滑法对深圳市宝安区2015、2016年的涂阳肺结核病人数进行预测。方法收集2005~2014年深圳市宝安区每月确诊的涂阳结核人数,通过SPSS 20.0统计学软件进行指数平滑法预测,经过简单季节性模型、Winters加法模型和Winters乘法模型的指标分析、统计量分析、残差图分析对三种模型进行了全面评估,同时给出了2015、2016年的结核病发病估计人数。结果简单季节性模型、Winters加法模型和Winters乘法模型的R^2分别为0.69、0.69、0.67,Ljung-Box的概率为0.86、0.79、0.67,Alpha值为0.40、0.40、0.32,均显示模型拟合良好,可以做出良好的预测结果。结论涂阳肺结核结核病在深圳市宝安区进入下降通道,但幅度缓慢,应进一步加强结核病的防治工作。
Objective To predict the onset number of smear positive pulmonary tuberculosis in Bao＆#39;an District of Shenzhen City in the year 2015 and 2016 with seasonal exponential smoothing method. Methods The monthly number of smear positive tuberculosis from 2005 to 2014 in Bao＆#39;an District of Shenzhen City was collected to predict the re-sults with exponential smoothing method by SPSS 20.0. Three kinds of models were overall evaluated after index analy-sis, statistic analysis, residual plot analysis of simple seasonal model, Winters addition model and Winters multiplica-tion model, and the onset number of smear positive pulmonary tuberculosis in the year 2015 and 2016 was estimated. Results The R2 of simple seasonal model, Winters addition model and Winters multiplication model was 0.69, 0.69 and 0.67 respectively; P value of Ljung-Box was 0.86, 0.79 and 0.67; Alpha value was 0.40, 0.40 and 0.32 respectively. The model fitting and predictions was good. Conclusion The estimated onset number of smear positive pulmonary tu-berculosis will be decreased in Bao＆#39;an District of Shenzhen city, but the rate is slow, tuberculosis prevention should be further strengthened.
出处
《中国医药导报》
CAS
2015年第18期39-42,共4页
China Medical Herald
基金
广东省深圳市科技计划项目(201203266)
关键词
指数平滑法
结核病
预测
宝安区
Exponential smoothing method
Tuberculosis
Prediction
Bao＆#39
an District
作者简介
张磊（1981．2-），男，中南大学2011级流行病与卫生统计学专业在读博士研究生；研究方向：传染性疾病控制。
