新陈代谢GM(1.1)模型和Verhulst模型预测心脏病死亡

Metabolism GM(1.1) model and Grey Verhulst model predict death from heart disease

目的研究心脏病的流行趋势和预测,为心脏病预防工作提供科学依据。方法以香港特别行政区居民2008-2017年心脏病死亡资料建立不同维数的常规GM(1.1)模型,针对常规GM(1.1)模型存在的不足,在此基础上建立不同维数的新陈代谢GM(1.1)模型,用相邻做差的方式建立2008-2017年每年新增病死人数Verhulst模型。结果 5维新陈代谢GM(1.1)模型是最合适的预测模型,心脏病死亡率5维新陈代谢GM(1.1)模型精度检验结果显示Q=0.0055,C=0.2568,P=1.00,发展系数-a=0.0209≤0.3,模型判为优,可用于预测。预测2018-2022年香港心脏病死亡率分别为80.8/10万、79.1/10万、77.5/10万、75.9/10万和74.3/10万。每年新增病死人数灰色Verhulst模型回代结果平均相对误差为97.2%,模型判为差,预测2018-2022年心脏病新增死亡人数分别为-155、-167、-216、-253和-302。结论分析结果说明在未来几年内香港特别行政区居民心脏病死亡率有下降的趋势,灰色Verhulst模型分析结果显示每年新增病死人数已达到饱和状态。

Objective To study the prevalence and prediction of heart disease and provide scientific basis for the future heart disease prevention work.Methods A general GM(1.1) model with different dimensions was established based on the cardiac death data of Hong Kong SAR residents from 2008 to 2017.In view of the shortcomings of the general GM(1.1) model,a metabolic GM(1.1) model with different dimensions was established on this basis,and the Verhulst model with adjacent difference was used to establish the annual number of new deaths from 2008 to 2017.Results The five-dimensional metabolic GM(1.1) model was the most suitable prediction model.The accuracy test results of the five-dimensional metabolic GM(1.1) model showed that Q=0.0055,C=0.2568,P=1.00,and the development coefficient-a=0.0209≤0.3.The model was judged as excellent and could be used for prediction.The predicted mortality rates of heart disease in Hong Kong from 2018 to 2022 were 80.8/100 000,79.1/100 000,77.5/100 000,75.9/100 000,74.3/100 000.The average relative error of the gray Verhulst model of the new deaths number per year was 97.2%,and the model was poor.It predicted that the number of new deaths from heart disease from 2018 to 2022 would be-155,-167,-216,-253 and-302.Conclusion The analysis results indicate that the mortality rate of heart disease in Hong Kong special administrative region residents will decrease in the next few years.The results of the grey Verhulst model show that the number of new deaths per year has reached the saturation state.