使用R语言扩展中国疾病预防控制信息系统预测重庆市结核病疫情

Research on function expanding of China Information System of Disease Control and Prevention using R language to predict the tuberculosis prevalence

目的扩展《中国疾病预防控制信息系统》,实现结核病发病的在线预测功能,为将来的疾病预防控制工作的资源调配及优化起到辅助决策支撑作用。方法采用离线的《中国疾病预防控制信息系统》中重庆市2005年至2009年每个月的结核病发病登记数据,以Java平台为基础运行环境,采用R语言分析预测引擎,应用神经网络模型、灰色模型、时间序列模型对结核病发病率进行在线预测分析。结果灰色模型和时间序列模型对结核病预测的误差较大,而神经网络模型对于结核病发病的预测相对误差不超过5%。结论使用R语言扩展《中国疾病预防控制信息系统》具有可行性,可用于指导疾病预防控制工作中的资源调配及决策支撑。系统扩展后对于结核病发病可实现单机、网络的在线预测。

Objective To expand the function of China Information System of Disease Control and Prevention for realizing the online prediction of tuberculosis prevalence,in order to provide the support of decision making for allocation and optimization of resources in disease prevention and control.Methods The off-line registered data on tuberculosis morbidity in Chongqing originated from China Information System of Disease Control and Prevention during 2005-2009 were collected.Then these data were analyzed by computing engine programmed with R language,using different models such as neuro-network model,gray model and time series model to realize the online predicting of tuberculosis prevalence based on Java platform.Results Among these three different models,the relative error of neuro-network in predicting the TB prevalence was less than 5% and therefore was much better than the other two models in prediction.Conclusion The function expanding of China Information System of Disease Control and Prevention is feasible using R language.Moreover,the expanded system could make the prediction of TB prevalence available on personal computer or online.Therefore,it could be used to provide guidance and support in resources allocation and decision-making in disease control and prevention.