The Application Of Generalized Additive Model In The Study Of Correlation Between Meteorological And The Number Of Ambulance Visits In Beijing

Objective: To study the relationship between number of ambulance visits in Beijing from March 2016 to February 2019 and the change of meteorological factors in the same period,and explore the application of the generalized additive model in the study of the correlation between the weather and ambulance visits number,so as to provide a certain reference for the allocation of related medical emergency resources.Methods: Weekly emergency visits quantity from March 2016 to February 2019 were collected from Beijing Municipal Health Commission(http://wjw.beijing.gov.cn/)and Beijing Emergency Medical Center(http://www.beijing120.com/),.Daily meteorological data were obtained from China Meteorological Data Service Center(http://data.cma.cn/)that corresponds to the March 2016-February 2019,which statistical index mainly including average air pressure,average maximum air pressure,average minimum air pressure,average air temperature,average maximum air temperature,average minimum air temperature,average daily air temperature range,average relative humidity,average wind speed,average sunshine hours,average land surface temperature,average maximum land surface temperature,average minimum land surface temperature.SPSS20.0 statistical software and Microsoft Excel 2016 were used for data merge,descriptive analysis,time series diagram,curve fitting.Spearman’s correlation analysis was used to study the relation between meteorological factors and ambulance visits number.Generalized Additive Model was applied with R 3.5.2 statistical software to explore the correlation effects between meteorological factors and emergency visits number.Results :(1)The overall trend of the number of Beijing emergency visits was on the rise from March 2016 to February 2019.Beijing Emergency Medical Center sent out1,141,627 ambulances total,and 626,593 for five kinds of diseases(trauma,cardiovascular disease,cerebrovascular disease,respiratory disease and digestive disease),accounting for 54.89% of the total.The proportion of five kinds of major diseases trauma,cardiovascular disease,cerebrovascular disease,respiratory disease and digestive disease were 27.99%,24.37%,20.86%,13.40%,13.38% respectively.The average air pressure about 1000 h Pa,the highest and lowest average air temperature or land surface temperature were about 35℃ and 0℃,respectively;the average daily air temperature range basically stable between 5℃ and 15℃;the average relative humidity almost 50%;the average wind speed basically stable at about 2m/s;the average sunshine time about 7 hours.(2)From March 2016 to February 2019,the number of emergency visits in the trauma,cardiovascular disease,cerebrovascular disease were higher than other disease,trauma and digestive system disease were positively correlated with average air temperature and average land surface temperature and relatively humidity.However,the number of first aid visits of cardiovascular,cerebrovascular and respiratory system were negatively correlated with mean air temperature,mean land surface temperature and relatively humidity.(3)Meteorological factors have different lag effects on various diseases in different periods.The average air temperature in one week lag lower than 5℃ and higher than 25℃,respectively,the number of ambulance visits had upward trend;the RR of the average temperature with 1-week lag for the total number of ambulance visits was 0.20%(95%CI: 0.03%-0.37%);when the average relative humidity was lower than 40% and higher than 60%,the number of ambulance visits had increased trend when the average relative humidity was lag one week,the RR of the total number of ambulance visits was increased by 0.11%(95%CI: 0.04%-0.18%).Mean air pressure above 1020 h Pa had more effect on cardiovascular disease and was positively correlated.With one-lag week,the average temperature was below 5℃ and above 25℃ and as the temperature decreased and increased,respectively,the number of ambulances with cardiovascular diseases both had upward trend;the RR of the average temperature with a lag of 1-week with cardiovascular diseases increased by 0.66%(95%CI: 0.31%-1.00%).Conclusion:(1)There was a close correlation between meteorological factors and the number of ambulances in Beijing,especially air pressure,air temperature and relative humidity(2)The number of ambulance visits for cardiovascular,cerebrovascular diseases and trauma were higher than other diseases from March 2016 to February 2019.(3)The relative risk of meteorological factors and emergency visits number were also various in delay 0-week,1-week and 2 weeks according to different diseases based on generalized additive model,which means the model has certain application value in the study of correlation between meteorological and the number of ambulance visits.