日均气温与呼吸系统疾病急诊人次相关性的时间序列分析

Relationship between daily mean temperature and emergency department visits for respiratory diseases:a time-series analysis

目的:定量分析和评价北京市日均气温对呼吸系统疾病急诊人次的影响。方法:收集北京市某三甲医院2004年1月至2009年6月急诊呼吸系统患者病例资料、同期北京市的气象因素及大气污染物数据,采用时间序列分析方法的广义相加模型,通过三次样条函数控制长期趋势、短期波动、星期几效应及其他可能的混杂因素,拟合日均气温与呼吸系统疾病急诊人次的暴露-反应关系模型,并探讨对不同特征人群的影响差异。结果:共收集呼吸系统急诊病例总计35073人次,其中男性14 707(41.93%,14 707/35073)人次,女性19 122(54.52%,19 122/35073)人次,性别缺失1244(3.55%,1244/35073)人次。日均气温与呼吸系统疾病急诊人次之间的关系基本呈"V"字型,最适日均温度在4℃左右,日均气温滞后0～3 d的效应明显。日均气温低于最适温度时,日均气温每降低1℃,当天因呼吸系统疾病而发生急诊治疗的超额危险度,总人群为3.75%(95%CI of RR:0.9383～0.9653),男性3.10%(95% CI of RR:0.9492～0.9891),女性4.09%(95%CI of RR:0.9407～0.9778);日均气温高于该最适温度时,日均气温每升高1℃,当天因呼吸系统疾病而发生急诊治疗的超额危险度,总人群为1.54%(95%CIof RR:1.0066～1.0243),男性为1.80%(95%CI of RR:1.0053～1.0309),女性为1.51(95%CI of RR:1.0032～1.0272)。低温的效应随年龄增大而增大,高温对45～59岁人群的效应最大。结论:北京市日均气温与人群呼吸系统疾病急诊人次的关系基本呈"V"字型,最适日均温度在4℃左右。日均气温在低于最适温度时降低以及在高于最适温度时升高,均造成发生呼吸系统疾病急诊的危险度增大,而且前者的效应比后者的大,不同年龄组、不同性别人群的气温效应有差异。

Objective： To quantitatively evaluate the influences of daily mean air temperature （DMT） on Emergency Department Visits （EDVs） for the respiratory diseases. Methods： The EDV data from medical records for respiratory diseases in Peking University Third Hospital between January 2004 and June 2009 were collected. The data of the air pollutants （SO2, NO2 and PM10） and meteorological factors at the same time periods were also collected from the local authorities of Beijing. Time-series analysis and generalized additive models （GAM） were used to explore the exposurrre-response relationship between DMT and EDVs for respiratory diseases. Results： A total of 35 073 patients [ males 14 707 （41.93%, 14 707/35 073）, females 19 122 （54.52%, 19 122/35 073） and gender missing 1 244 （3.55%, 1 244/ 35 073） ] EDVs for respiratory diseases were included. The relationship between DMT and EDVs for the respiratory diseases was mainly of ＂V＂ shape, the optimum temperature （OT） was about 4 ℃ and the effect of DMT was significant with a 0 - 3 day lag structure for most of the models. When DMT ≤ OT, each 1℃decrease in DMT corresponded to 3.75% （95% CI of RR： 0. 938 3 -0. 965 3） , 3. 10% （95% CI of RR：0.949 2-0.989 1） , 4.09% （95% CI of RR：0.940 7 -0.977 8） increase of EDVs for the overall, male, and female, respectively. When DMT 〉 OT, the value caused by each increase in 1℃ in DMTwas 1.54% （95% ClofRR：1.006 6-1.024 3）, 1.80% （95% ClofRR：l.005 3- 1. 030 9 ）, and 1.51 （95 % CI of RR ： l. 003 2 - 1. 027 2 ）, respectively. The effect was statistically significant within the 0 - 3 day lag. When DMT ≤ OT, the effect was stronger for the older people, while the effect was strongest for the 45 -59 years old people. Conclusion： The relationship between DMT and EDVs for respiratory diseases is mainly of ＂V＂ type, with an optimum temperature of 4 ℃. Both DMT de- crease when DMT ≤ OT and increase when DMT 〉 OT correspond to different increase of EDVs for respiratory diseases. Low DMT has stronger effect than high DMT. Different age group and gender have different effects.