灰色模型GM(1,1)在全国甲乙类传染病发病率定量宏观评估中的可行性研究

Feasibility on the Application of GM(1,1)Model in the Macroscopic Quantitative Incidence Evaluation of Notifiable Communicable Diseases

目的探索灰色模型GM(1,1)在全国甲乙类传染病发病率定量宏观评估中的合理性与应用条件。方法收集1990~2011年中国大陆地区22年甲乙类法定传染病发病率资料。首先,利用1990~2007年甲、乙类法定传染病发病率建立灰色GM(1,1)模型,然后对2008~2011年甲、乙类法定传染病发病率进行外推预测。结果灰色模型GM(1,1)应用较好的全国甲乙类传染病包括:伤寒和副伤寒、流行性脑脊髓膜炎和流行性乙型脑炎。应用基本适合的包括:梅毒、百日咳、白喉、钩端螺旋体病、布鲁氏菌病和炭疽。这些传染病发病率均具有呈指数上升或下降的趋势。结论 GM(1,1)模型本质上是指数模型,当数据服从指数分布时,利用GM(1,1)效果较好,与指数函数回归结果基本一致。

Objective To explore the feasibility of GM（1,1） model in the macroscopic quantitative incidence evalua- tion of notifiable communicable diseases. Methods The national incidence of Class A and B notifiable communicable dis- eases during 1990-2011 were collected,Gray model and exponential regression were adopted for fitting and prediction. Results The precision of GM（1,1） could reach degree 1 when being appl.ied in typhoid and tSaratyphoid fever,epidemic cerebrospinal meningitis and epidemic encephalitis b. The precision of tl^e model could reach degree 2 when beigg applie（t in syphilis, pertussis, diphtheria, leptospirosis, brucellosis and anthraX. Mo＇d^is＇got good resuhs when incidence showed expo nential tendency. Conclusion Gray model GM（I, 1） can simdlate the incid：ence trbnd of the communicable diseases only when the data follow the exponential distribution and Grey Model can have the same power as exponential regression.