2012—2021年内蒙古自治区开鲁县道路交通伤害与气象因素相关性研究  

Research on the correlation between road traffic injuries and meteorological factors in Kailu County,Inner Mongolia Autonomous Region from 2012 to 2021

作　　者：陈文婕[1] 王肖 钱永刚[1] 刘慧敏 田舒媛 王虎虎 CHEN Wenjie;WANG Xiao;QIAN Yonggang;LIU Huimin;TIAN Shuyuan;WANG Huhu(Inner Mongolia Autonomous Region Center for Disease Control and Prevention,Hohhot,Inner Mongolia Autonomous Region 010031,China;不详)

机构地区：[1]内蒙古自治区疾病预防控制中心,内蒙古自治区呼和浩特010031 [2]鄂尔多斯高新技术产业开发区科技人才局

出　　处：《医学动物防制》2024年第12期1244-1248,共5页Journal of Medical Pest Control

基　　金：内蒙古自治区卫生健康委医疗卫生科技计划项目(202201168)。

摘　　要：目的收集内蒙古自治区开鲁县伤害监测点2012—2021年道路交通伤害的门急诊数据,结合气象数据,初步探索内蒙古自治区不同气象因素与道路交通伤害的相关性。方法收集内蒙古自治区开鲁县伤害监测哨点医疗机构2012—2021年收集的道路交通伤害病例数据和同期开鲁县气象数据。采用计算方差膨胀因子的方法去除共线性影响后,再拟合负二项回归模型进行相关性研究。结果负二项回归模型显示,残差通过正态性检验(W=0.83,P<0.001)。月平均温度(X_1)和月雾天数(X_8)对月道路交通伤害发生数呈正相关性,月平均能见度(X_4)对月道路交通伤害发生数呈负相关性。利用负二项回归模型对2012—2020年数据作为训练集进行回归模型拟合[平均绝对误差(mean absolute error,MAE)为36.20,预测准确率(prediction accuracy,PA)为63.81%],2021年数据作为测试集进行预测效果评价,MAE为23.84,PA为67.90%,结果良好。结论月平均温度(X_1)、月平均能见度(X_4)、月雾天数(X_8)对道路交通伤害的影响显著,模型能通过残差检验,模型拟合以及预测效果良好,适合用于此研究。Objective To collect outpatient and emergency data on road traffic injuries at the injury monitoring point in Kailu County,Inner Mongolia Autonomous Region from 2012 to 2021,and to preliminarily explore the correlation between different meteorological factors and road traffic injuries in Inner Mongolia Autonomous Region by combining the data with meteorological information.Methods Data on road traffic injury cases from 2012 to 2021 were collected from the injury monitoring sentinel medical institutions in Kailu County,Inner Mongolia Autonomous Region,and meteorological data in Kailu County during the same period were collected.After removing the influence of collinearity by calculating the variance inflation factor,the negative binomial regression model was fitted for correlation research.Results Negative binomial regression model showed that the residual passed the normality test(W=0.83,P<0.001).The monthly average temperature(X_(1))and the monthly fog days(X8)were positively correlated with the number of monthly road traffic injuries,while the monthly average visibility(X_(4))was negatively correlated with the number of monthly road traffic injuries.The negative binomial regression model was used to fit the regression model with the data from 2012 to 2020 as the training set mean absolute error(MAE)was 36.20,prediction accuracy(PA)was 63.81%,and the data from 2021 were used as the test set to evaluate the prediction effect,MAE was 23.84,PA was 67.90%,the result was good.Conclusion The influence of monthly average temperature(X_(1)),monthly average visibility(X_(4))and monthly fog days(X8)on road traffic injury is significant.The model can pass the residual test and has good fitting and prediction effects,which is suitable for this study.

关 键 词：道路交通伤害 气象因素 负二项回归模型 相关性 研究 

分 类 号：R195[医药卫生—卫生统计学]