1988-2010年全国疟疾发病率的灰色预测模型研究  

A study on the trend of malaria incidence in China in the recent 20 years with GM(1,1)

作　　者：郭海强[1] 丁海龙[1] 曲波[1] 孙高[1] 

机构地区：[1]中国医科大学公共卫生学院卫生统计教研室,辽宁沈阳110001

出　　处：《热带医学杂志》2011年第6期639-640,643,共3页Journal of Tropical Medicine

基　　金：国家自然科学基金资助项目(30700690)

摘　　要：目的描述分析我国近20年疟疾的发病率变化趋势,并进行预测,为疟疾的科学防控提供参考。方法根据全国1988-2008年疟疾的报告发病率数据,拟合GM(1,1)灰色预测模型,并进行回代预测,外推预测2009和2010年的发病率。结果 1988-2009年全国疟疾报告发病率总体呈下降趋势,由12.56/10万降至1.06/10万。利用1988-2008年数据建立的GM(1,1)预测模型为:y(t+1)=-97.0998e-0.113t+109.5398,模型拟合精度较高(后验差比值C为0.4162,小误差概率P为0.95)。模型预测2009、2010年全国疟疾发病率分别为1.08/10万和0.97/10万。结论灰色模型GM(1,1)计算简便,预测结果准确,是一种较为理想的预测疟疾发展趋势的方法。Objective To describe and analyze the trends of malaria incidence in China during the recent 20 years, and to forecast the rate in the short future. It aims to provide scientific base for prevention and control of malaria. Methods The reported incidence rates of malaria from 1988 to 2008 were collected to establish GM （ 1,1 ） model for prediction. Those rates of 2009 and 2010 were predicted with the established model. Results The actual rates declined from 12.56 per lOs to 1.06 per 105 in the period of 1988 to 2009.The established GM （ 1,1 ）model was the following ： ~ （t ＋1 ）=-97.0998e-Zt^3t＋lO9.5398.The prognosticating precision of GM （ 1,1 ）was desirable in accordance with the criterion of grey model prognostication assessment. The predictive incidences in 2009 and 2010 were 1.08 per 105 and 0.97 per 105. Conclusion Since the calculation of GM （ 1, 1 ） is simole and its 13redictive result is orecise, it is suitable to fit and oredict the trends of malaria incidence.

关 键 词：疟疾 发病率 灰色预测模型 

分 类 号：R195.4[医药卫生—卫生统计学]