广东省1986～2005年各初筛实验室汇总HIV检测阳性率时间序列分析  

TIME SERIES ANALYSIS OF POSITIVE RATE OF HIV TESTED IN ALL THE PRELIMINARY SCREENING LABS IN GUANGDONG FROM 1986 TO 2005

作　　者：徐英[1] 胡伟[2] 

机构地区：[1]广东药学院流行病与卫生统计系,广州510310 [2]广东省卫生统计信息中心

出　　处：《现代预防医学》2008年第16期3058-3059,3076,共3页Modern Preventive Medicine

摘　　要：[目的]利用时间序列分析广东省1986～2005年各初筛实验室汇总HIV检测阳性率变化趋势。[方法]利用SAS9.0软件对资料进行ARIMA模型的模拟建模,用所得的模型对HIV阳性率进行预测,并与实际资料进行比较。[结果]数据分析显示ARIMA(0,3,2)模型拟合效果较好,并预测2006年的HIV检测的阳性率为0.222%(0.095%～0.350%)。[结论]ARIMA(0,3,2)模型较好的拟合了HIV检测阳性率的变化趋势,并对未来的HIV检测阳性率进行预测,为艾滋病的防治工作提供了一定的依据。[ Objective ] To analyze the trend of positive rate of HIV tested in all the preliminary screening labs in Guangdong from 1986 to 2005 by time series analysis. [Methods] SAS9.0 software was used to construct ARIMA model, then the mode were performed to predict the positive rate of HIV and compared with the actual positive rate. [ Results] The analyzed da- ta indicated that fitting effect of ARIMA （0, 3, 2） model was favorable, and the positive detection rate of HIV in 2006 predicted was 0.222% （0.095%-0.350%）. [Conclusion] ARIMA （0, 3, 2） better fit the trend and positive detection rate of HIV, and could predict the positive detection rate of HIV so that to provide basis for the preventing and controlling HIV.

关 键 词：HIV检测阳性率 时间序列分析 ARIMA模型 预测 

分 类 号：R512.91[医药卫生—内科学]