急性一氧化碳中毒与气象因子大气污染物相关分析及预测方法研究  

Acute Carbon Monoxide Poisoning Meteorological Factors and Atmospheric Pollutants Correlation Analysis and the Forecast Method Research

作　　者：阮海林[1] 胡灼君 邓旺生[1] 刘华[1] 王瑶 陈剑兵 洪伟良 RUAN Hailin;HU Zhuojun;DENG Wangsheng;LIU Hua;WANG Yao;CHEN Jianbing;HONG Weiliang(The Fourth Affi liated Hospital of Guangxi Medical University/Liuzhou Workers Hospital,Liuzhou 545005,China;Liuzhou Meteorological Bureau,Liuzhou 545001,China;Liuzhou Environmental Monitoring Station,Liuzhou 545001,China)

机构地区：[1]广西医科大学第四附属医院/柳州市工人医院,广西柳州545005 [2]柳州市气象局,广西柳州545001 [3]柳州市环境保护监测站,广西柳州545001

出　　处：《中国医药指南》2020年第34期4-5,9,共3页Guide of China Medicine

基　　金：广西壮族自治区卫生厅立项科研课题(合同号Z2014395);广西急诊与医学救援人才小高地;广西高校急诊医学重点实验室(课题编号GXJZ201416)。

摘　　要：目的调查分析急性一氧化碳中毒(ACOP)与气象因子、大气污染物的相关性,建立区域性ACOP气象大气污染物预报模型,为ACOP防治提供科学依据。方法收集2015年1月至2017年12月广西柳州市四家定点医院收治的ACOP患者临床资料和同期气象要素、大气污染物相关资料,进行相关性分析,构建CO中毒风险等级的有序多分类Logistic回归模型,并进行效用验证。结果ACOP患者3年共4181人次,男1623例,占38.82%,女2558例,占61.18%;每年的12月至次年3月为发病高峰期;日时间分布以夜间21时到次日凌晨3时为明显高峰期;日平均CO中毒人数与平均温度,平均风速以及二氧化硫(SO2),臭氧(O3)的空气浓度呈显著负相关(r范围:-0.079^-0.152,所有P<0.01),但与24 h温度变化和空气中的一氧化碳(CO)浓度呈正相关(r分别为0.232,0.098;均P<0.01)。预测模型为Logit[P(y≤j)=βjo+β1X1+…+β6X6],j=1,2,…4。预测风险3及以上等级有效率达92%。结论ACOP的发生与温度、风速、24 h温差、SO2、CO、O3污染物的浓度密切相关,可利用气象因子和大气污染物建立区域CO中毒预报模型,加强预警预报服务。Objective To investigate and analyze the correlation between acute carbon monoxide poisoning(ACOP)and meteorological factors and atmospheric pollutants,and to establish regional forecast model of ACOP meteorological and atmospheric pollutants,so as to provide scientifi c basis for the prevention and treatment of ACOP.Methods Clinical data of ACOP patients admitted to four designated hospitals in Liuzhou city,Guangxi from January 2015 to December 2017 were collected.Meanwhile,meteorological elements and relevant data of atmospheric pollutants were collected for correlation analysis to build an ordered multi-classification Logistic regression model of CO poisoning risk level and verify its effectiveness.Results There were 4,181 patients with ACOP in 3 years,including 1623 males(38.82%)and 2558 females(61.18%).The peak of the disease was from December every year to March of the following year.The daily time distribution was from 21 o'clock at night to 3 o'clock in the morning of the next day.The daily average number of CO poisoning was signifi cantly negatively correlated with the average temperature,the average wind speed,the air concentration of sulfur dioxide(SO2)and ozone(O3)(r range:-0.079 to-0.152,all P<0.01),but positively correlated with the 24 h temperature change and the concentration of carbon monoxide(CO)in the air(r=0.232,0.098,respectively,P<0.01).The prediction model is Logit[P(y≤j)=βjo+β1X1+…+β6X6],j=1,2,…4.The eff ective rate of predicting risk 3 and above was 92%.Conclusions The occurrence of ACOP is closely related to temperature,wind speed,24 h temperature diff erence,concentration of SO2,CO and O3 pollutants.Meteorological factors and atmospheric pollutants can be used to establish regional CO poisoning forecast model and strengthen early warning and forecast services.

关 键 词：急性 一氧化碳中毒 气象因素 污染物 流行病学 预报 

分 类 号：R595.1[医药卫生—内科学]