摘要
目的利用自回归滑动平均混合模型（autoregressive integrated moving average,ARIMA）模型开展长沙市肾综合征出血热（haemorrhagic fever with renal syndrome,HFRS）发病率预测,为相关政策的改进提供科学依据。方法收集长沙市1971-2010年HFRS发病率资料,用Excel 2003软件建立数据库,采用SPSS13.0统计分析软件实现模型构建,经过模型识别、参数估计、诊断检验后,利用建立好的模型对长沙市2011-2015年HFRS发病率进行预测。结果根据模型原理及与实际值的比较,ARIMA（1,0,0）模型适合模拟长沙市HFRS的发病规律,模型拟合值与实际值非常接近,模型回归系数有统计学意义。2011-2015年长沙市HFRS年发病率预测值分别为1.69/10万（95%CI：0.00~9.67/10万）、2.23/10万（95%CI：0.00~12.51/10万）、2.66/10万（95%CI：0.00~14.22/10万）、3.00/10万（95%CI：0.00~15.35/10万）、3.27/10万（95%CI：0.00~16.13/10万）。结论 2011-2015年长沙市HFRS发病有增加趋势,需进一步加大防控策略。
Objective To forecast the morbidity of haemorrhagic fever with renal syndrome（HFRS） in Changsha by autoregressive integrated moving average（ARIMA） model,so as to provide scientific evidence for the improvement of corresponding prevention and control policies.Methods The incidence data of HFRS in Changsha from 1971 to 2010 were collected and the database was established with Excel 2003 software.SPSS 13.0 software was employed to construct an ARIMA model after model identification,parameter estimation and diagnosis test.The morbidity of HFRS from 2011 to 2015 in Changsha was forecasted with this model.Results According to the theory of the model and comparison with the actual data,ARIMA（1,0,0） model was appropriate for simulating the morbidity pattern of HFRS in Changsha.The fitted values and actual values had a high fitting precision,and the regression coefficient of the model had statistical significance.The predicted morbidity of HFRS in Changsha from 2011 to 2015 was 1.69/100,000（95%CI：0.00-9.67/100,000）,2.23/100,000（95%CI：0.00-12.51/100,000）,2.66/100,000（95%CI：0.00-14.22/100,000）,3.00/100,000（95%CI：0.00-15.35/100,000） and 3.27/100,000（95%CI：0.00-16.13/100,000）,respectively.Conclusions Forecast by ARIMA model suggests an increase tendency for the morbidity of HFRS in Changsha from 2011 to 2015.Prevention and control strategies should be further strengthened.
出处
《实用预防医学》
CAS
2012年第12期1782-1785,共4页
Practical Preventive Medicine
基金
湖南省科技厅项目(2010SK3007)
关键词
时间序列
肾综合征出血热
自回归滑动平均混合模型
预测
Time series
Haemorrhagic fever with renal syndrome（HFRS）
Autoregressive integrated moving average（ARIMA） model
Forecast
