摘要
目的分析我国肺结核发病的流行病学特征,预测我国肺结核发病人数,为我国肺结核病的防治和疫情监测提供参考依据。方法 2005～2011年卫生部公开发布的每月肺结核发病数。采用时间序列分析方法建立预测模型,经过统计学检验和评价后再进行预测。结果拟合得到的最佳模型是没有常数项的SARIMA(0,1,1)(0,1,1)12和没有常数项的CE+SARIMA(1,1,0)(0,1,1)12,但后一个事件模型更优;春节所在月份报告的肺结核新发病人数偏少,平均每日少733人;根据拟合的模型估计得到了2012年我国各月份肺结核新发病人数。结论我国肺结核的发生具有明显周期性和季节性,一年一周期,冬春季节多发。
Objective To understand the characteristics of the tuberculosis epidemic in China, and to forecast the incidence cases of tuber- culosis in China. Methods Surveillance data of tuberculosis incidence cases in China between 2005 to 2011. Based on the tuberculosis incidence case surveillance data of China since 2005, we describe the temporal charac- teristics and trend of the tuberculosis epidemic in China,build models for the tuberculosis epidemic using Box-Jenkins methodology, find out the best model and make forecasts based on the models. Results Both SARIMA models and event models are built, to study the data. Results from both models suggest that the number of tuberculosis incidence cases in a month not only depends on the levels of tuberculosis before the month, but also as- sociates with its level at the end month of last year. The event model CE ＋ SARIMA（ 1,1,0） （0,1,1）12 with no constant gives the best result, which indicates that there will be less 733 incidence cases of tuberculosis reported daily in the month when the Chinese Spring Festival is coming, and there will be less incidence cases of tuberculosis in China in 2012. Conclusion The tuberculosis epidemic in China since 2005 exhibits distinct periodici- ty and a seasonal behavior：one year one cycle, and tuberculosis cases are reported more in winter and spring.
出处
《中国卫生统计》
CSCD
北大核心
2013年第2期158-161,共4页
Chinese Journal of Health Statistics
关键词
肺结核
新发病人数
时间序列分析
预测
Tuberculosis
Incidence cases
Time seriesanalysis
Forecast
