北京灰霾事件期间空气污染与医院成人门急诊量的浓度-反应关系研究

Study on the concentration-response relation between air pollution and outpatient and emergency room visits during haze episode in Beijing

目的研究灰霾期间空气污染与医院门急诊量的浓度-反应关系。方法门急诊量数据来自北京市海淀区某三甲综合性医院,空气质量监测数据来自中国环境科学研究院超级环境监测站,气象数据来自天气网。研究时间包括2012年10月—2013年1月(灰霾期,T1序列)和2011年10月—2012年2月(对照期,T2序列)。采用广义相加模型确定浓度-反应关系,以多重结构断裂点检测方法确定浓度-反应关系曲线上的拐点,应用约束性分段线性函数估计相应的危险度。结果在污染最严重的2013年1月7日前后,呼吸内科和内科急诊的就诊量也达到最高,比对照期同期分别增加60.1%和64.2%。空气污染与医院门急诊量的浓度-反应关系为非线性曲线。T1序列分析显示,呼吸内科门诊与SO2、神经内科与NOx、内科急诊与NOx、SO2和PM10之间的相关关系有统计学意义,累积危险度分别为8.8%(avg03),5.7%(lag0),5.1%(avg06),3.3%(avg04)和5.6%(avg04);T2序列中仅神经内科门急诊量与NOx显著相关,累积危险度为18.7%(lag0);未见空气污染与心血管内科门诊量有显著的相关关系。研究中发现具有统计学意义的浓度-反应关系曲线多为"C"型,其常在污染物浓度较低时出现,而观察到少数的"S"和"J"型曲线分别在中等和高浓度时出现。结论空气污染与医院多个科室的门急诊量之间存在非线性相关关系,以"C"曲线最为多见,高污染范围时可以见到"S"、"J"型曲线,提示空气污染严重时成人门急诊量呈上升趋势。

Objectives To study and descript the concentration-response relationship between air pollution and number of outpatient visits and emergency room visits during 2013 haze episode in Beijing. Methods Daily number of outpatient visits and emergency room visits were collected in a comprehensive hospital in Beijing. Data of air pollutants were provided by Chinese Research Academy of Environmental Sciences and data of weather were gained from China Weather website. The observation period was from Oct. 2012 to Jan. 2012（T1） and the same period in previous year（T2）. Generalized additive models（GAM） were used to establish the curve of concentration-response relationship. Then structural-change-detection algorithm was employed to identify the change points dynamically. Finally constrained piecewise linear functions was used to estimate the relative risks of air pollutants. Results On 7 Jan. 2013, daily visiting number for respiratory department and internal medicine emergency reached their peaks value, which were increased 60.1% and 64.2%, respectively, comparing with the same day of control period. The curve of concentration-response relationship between air pollution and outpatient visits and emergency room visits was showed the non-liner shape. The results of T1 series analysis showed that there were significant correlation between SO2 and daily number of respiratory department visits, NOxand neurology medicine, NOx, SO2, PM10 and internal medicine emergency,the accumulated risks of which were 8.8%（avg03）, 5.7%（lag0）, 5.1%（avg06）, 3.3%（avg04） and 5.6%（avg04）,respectively. The results of T2 series analysis showed that there was significantly correlation between NOxand neurology medicine,the accumulated risk was 18.7%（lag0）. There was no significant relationship between air pollutants and cardiovascular medicine. The curves of concentration-response relationship were more likely to be C-shape, which usually was found at low level of pollutants, while less S- and J-shape curve was observed at high pollutant level. Conclusion The curves of concentration-response relationship between air pollutants and daily number of outpatient visits and emergency room visits are showed non-linear shape,which are more likely to be C-shape. S- and J-shape curves can be found at high pollutant level,which suggestes that outpatient and emergency room visits will increase during severe air pollution.