向量自回归模型在慢性阻塞性肺疾病患病率预测中的应用研究  

Application of Vector Autoregressive Models in Predicting the Prevalence of Chronic Obstructive Pulmonary Disease

作　　者：窦一峰 王富彪 贾新亮 赵一犇 吕劲松[1] DOU Yi-feng;WANG Fu-biao;JIA Xin-liang(Tianjin Baodi District People’s Hospital,Tianjin,301800,China)

机构地区：[1]天津市宝坻区人民医院,天津301800

出　　处：《黑龙江医学》2022年第23期2872-2875,共4页Heilongjiang Medical Journal

摘　　要：目的:探索向量自回归模型(vector autoregression model,VAR)在慢性阻塞性肺疾病(chronic obstructive pulmonary disease,COPD)患病率与气象因素之间的关系,为防控提供科学依据。方法:以天津市2000—2016年COPD患病率数据和同期气象数据为例建立VAR模型,通过脉冲响应函数和方差分解对模型进行定量分析,运用多指标为所建立的模型对COPD患病率的预测效果进行评价分析。结果:给出了COPD患病率作为因变量的VAR模型结果表达式,模型总的拟合优度为0.999 8,调整后的拟合优度值为0.999 6,利用该模型对未来5年COPD患病率的预测结果分别为6.354 3%、6.471 5%、6.586 6%、6.701 7%和6.816 9%,均方振误差(root mean square ewor,RMSE)、平均绝对误差(mean absolute ewor,MAE)、平均绝对百分比误差(mean absolute porcentage ewor,MAPE)分别为0.020 4、0.017 8和0.323 7%,希尔不等式系数(thiel ineguality coefficient,TIC)为0.001 9,远小于1。结论:向量自回归模型在天津市慢性阻塞性肺疾病患病率预测分析中具有较好的应用价值,能够为疾病防控提供参考。Objective:To explore the relationship between vector autoregressive models(VAR) in the prevalence of chronic obstructive pulmonary disease(COPD) and meteorological factors to provide a scientific basis for prevention and control.Methods:A VAR model was established with COPD prevalence data and contemporaneous meteorological data in Tianjin from 2000-2016, and the model was quantitatively analyzed by impulse response function and variance decomposition. The predictive effect of the developed model for COPD prevalence was evaluated and analyzed using multiple indicators.Results:The expression of the results of the VAR model with COPD prevalence as the dependent variable showed an overall model goodness-of-fit of 0.999 8 and an adjusted goodness-of-fit value of 0.999 6. The predictions of COPD prevalence for the next 5 years using this model were 6.354 3%, 6.4715%, 6.586 6%, 6.701 7%, and 6.816 9% for RMSE, MAE, and MAPE of 0.020 4, 0.017 8, and 0.323 7%, respectively. The coefficient of Hill’s inequality was 0.001 9, which was much smaller than 1.Conclusion:The vector autoregressive model has good application value in the prediction analysis of the prevalence of chronic obstructive pulmonary disease in Tianjin, and can provide a reference for disease prevention and control.

关 键 词：慢性阻塞性肺疾病 气象因素 向量自回归模型 预测 评价指标 

分 类 号：R657.4[医药卫生—外科学]