Prediction Of Cardiovascular And Cerebrovascular Disease Visits Based On Meteorological Factors And Multivariate Time Series Analysis

Objective:To explore the influence of meteorological factors on cardiovascular and cerebrovascular diseases in Changsha City,Hunan Province,this passage analyzed the differences in the number of cardiovascular and cerebrovascular disease visits by gender,disease type,and season,and established a model for predicting the number of disease visits,hoping to provide a scientific basis for the prevention of local cardiovascular and cerebrovascular disease susceptible populations and the scheduling of medical resources in public health departments.Methods:We collected daily outpatient and emergency attendance data for cardiovascular and cerebrovascular diseases at a tertiary care hospital in Changsha,Hunan Province,from 2014 to 2018 and day-by-day meteorological data（including weather data and air pollution data）for Changsha,Hunan Province,during the same period.We applied a generalized additive model to analyze the correlation of meteorological factors for cardiovascular and cerebrovascular disease visits.Based on the results of correlation analysis,we constructed a multifactorial ARIMAX model to fit and predict monthly cardiovascular and cerebrovascular disease visits.Results:Temperature,humidity,wind speed,barometric pressure,cloudiness,PM₁₀,NO₂,and O₃＿8h were positively correlated with cardiovascular and cerebrovascular disease visits;the rising temperature had a greater effect on women,and the effect of NO₂ was only for women and cardiovascular disease patients in the subgroup analysis;humidity,wind speed,barometric pressure,and air pollution factors were positively correlated with seasonal disease visits,while the rising temperature was negatively correlated with disease visits in spring and winter and positively correlated with summer.The ARIMAX model predicted better than the simple time series ARIMA model,and the average absolute percentage error of the prediction after the introduction of relevant meteorological factors was 22.52%,which was 17.6%less than the error of the ARIMA model.Conclusions:Hunan Province has a mild and humid climate,and the average indicators of weather factors are close to the human comfort zone,so its upward trend will cause an increase in the number of disease visits;while during the study period,there were not many days when air pollution exceeded the national good air quality standards,but its upward trend will still increase the number of cardiovascular and cerebrovascular disease visits.The effect of NO₂ is only evident in women and cardiovascular disease attending groups,which need to be regarded as susceptible groups for priority protection.Meteorological factors can have different effects in different seasons,which may be related to seasonal climatic characteristics,environmental self-purification capacity,and discharge intensity.