论文部分内容阅读
目的探讨应用Hockey Stick回归和时间序列分析中的SARI MA模型进行细菌性痢疾疫情预测的可行性。方法收集辽宁省葫芦岛市1990-2006年的逐月及逐年细菌性痢疾疫情资料和当地气象数据。首先,利用描述统计分析细菌性痢疾季节性发病规律,使用Spearman等级相关分析疫情同气象因素的关系,根据Hockey Stick回归确定发病阈值。其次,进行扩充迪基富勒的平稳性单位根检验。再次,根据自相关函数图和偏自相关函数图识别逐月疫情间的相关性。应用Eviews3.1、Stata8.2和SPSS12.0软件对1990-2005年逐月发病率进行上述统计分析。最后,利用所得到的模型对2006年各月发病率进行预测,并与实际发病率进行比较。结果最低气温、平均气温和最高气温所确定的阈值分别为11.42℃、17.17℃和22.98℃;(1,0,0)×(0,1,1)12模型为最优SARI MA模型,此模型对2006年各月发病率的预测值符合实际发病率变动趋势。结论Hockey Stick回归和SARI MA模型可较好地模拟细菌性痢疾疫情在时间序列上的变动趋势,并对未来的发病率进行一定预测,能够为传染病防制工作提供一定决策支持。
Objective To explore the feasibility of using SARI MA model in Hockey Stick regression and time series analysis to predict the epidemic situation of bacillary dysentery. Methods The monthly and yearly bacterial dysentery epidemic data and local meteorological data of Huludao City in Liaoning Province from 1990 to 2006 were collected. First, the descriptive statistics were used to analyze the seasonal incidence of bacterial dysentery. Spearman rank correlation analysis was used to analyze the relationship between epidemic situation and meteorological factors. Hickey Stick regression was used to determine the incidence threshold. Second, conduct a root-mean-square test to augment Dickinson’s. Thirdly, the correlation between monthly epidemics was identified based on the autocorrelation function and partial autocorrelation function. Eviews3.1, Stata8.2 and SPSS12.0 software were used to carry out the above statistical analysis on the monthly incidence of 1990-2005. Finally, the resulting model was used to predict morbidity rates by month of 2006 and to compare them with actual morbidity. Results The thresholds of minimum temperature, average temperature and maximum air temperature were 11.42 ℃, 17.17 ℃ and 22.98 ℃, respectively. The model of (1,0,0) × (0,1,1) 12 was the optimal SARI MA model. The predictive value of morbidity in each month in 2006 accords with the trend of actual morbidity. Conclusions Hockey Stick regression and SARI MA model can better simulate the trend of time-series changes of bacterial dysentery epidemic, and make some predictions on the future incidence, which can provide some decision support for the prevention and control of infectious diseases.