Epidemiological Characteristics And Risk Factors Of Infectious Diseases In Shandong Province Based On Medical Big Data

BackgroundThe ability and effect of prevention and treatment of infectious diseases are the important indicators to evaluate a country’s social and economic development and the level of health care.In recent years,China has made remarkable achievements in the prevention and treatment of infectious diseases,but we are still facing the dual threat of old and new infectious diseases.The prevention and control of infectious diseases should not only be limited to notifiable infectious diseases,but also should pay attention to non-notifiable infectious diseases and the spectrum and dynamic characteristics of infectious disease.With the development of information technology and the promotion of national health information construction,the massive data of infectious disease in different medical and health databases,such as hospital electronic medical records,resident electronic health files and medical insurance data,provides a new way to further enrich and improve the surveillance of infectious diseases.The integration and utilization of infectious diseases data in various medical and health systems are of great significance to improve the national notifiable diseases reporting system and strengthen its integrity and accuracy.However,comprehensive and accurate studies about the information of infectious diseases which integrating multiple medical data sources are limited.At present,the relationship between air pollution and meteorological factors and human health has attracted increasing attention.Previous studies on the health effects of air pollution on infectious diseases mainly focused on specific areas or populations,and there was a lack of reports on the differences between urban and rural areas reflecting the impact of air pollution on infectious diseases.Under the background of global warming,a growing number of studies showed that temperature could affect the risk of respiratory diseases.However,the conclusions of these studies are inconsistent.There is an urgent need to reflect the differences in the effects of temperature on infectious diseases in different climate regions,and studies about the quantitative evaluation of the risk of infectious diseases caused by exposure to non-optimal temperatures and its influencing factors from the perspective of public health are sparse,and little is known about the modified effects of air pollution on the health effect of temperature.Based on the regional health information platform,this study integrates medical big data from multi-source medical and health systems,described the spectrum of infectious diseases and evaluated the incidence and dynamic trends of various infectious diseases.Pneumonia and influenza,two major respiratory infectious diseases with high incidence,were selected to reveal the short-term effects,lag effects and vulnerable populations of air pollutants on the incidence of pneumonia,to explore the regional differences in the impact of temperatures on the incidence of influenza,attributable risks and their influencing factors,the modification effects of air pollutants,and establish a high-resolution influenza incidence risk prediction model.The results will provide a scientific basis for improving the existing infectious disease surveillance system,and contribute to formulate infectious disease prevention and control measures and optimize the allocation of public health resources.Objectives1.To describe the spectrum of infectious diseases in Shandong Province from 2013 to 2017,and evaluate the incidence and distribution characteristics of various infectious diseases.2.To analyze the lag effects of air pollutants on the incidence of pneumonia in Qingdao and identify vulnerable groups.3.To evaluate the regional differences in the influence of temperature on the incidence of influenza in Shandong Province,and to analyze the attributable fraction of influenza caused by non-optimum temperatures and potential influencing factors,and the effect modifications of air pollutant PM2.5 on the relationship between influenza and temperature.4.To establish a practical and high-precision influenza risk prediction model incorporating natural and social factors based on the maximum entropy niche model.Methods1.Data collectionBased on Shandong Multi-Center Healthcare Big Data Platform,the hospital electronic medical records,residents’ medical insurance system,residents’ electronic health records and other medical data from multiple sources were integrated.According to the law of the People’s Republic of China on the prevention and treatment of infectious diseases and ICD-10,the incidence information of all infectious diseases in the sample population from January 2013 to June 2017 was collected.Daily air pollution data during the study period were collected from China’s National Environmental Monitoring Center,including PM2.5,SO2,NO2,CO and O3.The meteorological data were collected from National Meteorological Information Center of China,including average temperature,average relative humidity,accumulated rainfall,sunshine duration and wind speed.The surveillance data of influenza in Shandong Province during 2014-2016 were obtained from the National Notifiable Disease Surveillance System of Chinese Center for Disease Control and Prevention,and the raster data of population density,number of children under 5 years old,number of people over 65 years old and built-settlements in Shandong Province during the study period were obtained from the official website of Worldpop.The data of normalized difference vegetation index(NDVI)was obtained from National Aeronautics and Space Administration(NASA).2.Statistical analysis(1)Based on morbidity data of infectious diseases and demographic data,the annual incidence density of various infectious diseases during the study period in Shandong Province was calculated by using an accurate method to calculate the person years,and their distribution characteristics were described.Joinpoint regression was used to analyze the temporal trends of infectious diseases.(2)The exposure-response relationship between weekly air pollution exposure and pneumonia in Qingdao was estimated using distributed lag nonlinear model,and subgroup analysis was conducted by gender,region and age to identify vulnerable populations.(3)A two-stage model was used to analyze the regional differences in the impact of temperature on influenza in Shandong Province.In the first stage,a distributed lag nonlinear model was used to explore the impact of temperature on influenza in each city.In the second stage,multivariate meta-was used to combine the exposure response of each city to obtain the average effect of theanalysis province.Multivariate meta-regression model was used to explore the heterogeneity sources of regional differences.Based on the exposure-response relationship of each city obtained from the above model,the population attributable fraction caused by low temperatures and high temperatures in each city and the whole province were calculated,and the correlations between the social variables at the city level and the attributable fraction of cold and heat were analyzed using two separate meta-regression models.The effects of temperature on influenza transmission and air pollutants on the relationship between temperature and influenza were explored using distributed lag nonlinear model.(4)Ecological niche model was constructed to analyze the impact of natural and social factors on the incidence of influenza.The data of influenza in winter of 2014-2015 were used to establish an ecological niche model and the distribution of influenza in 2016 was predicted,and then the actual surveillance data in 2016 were used for external validation.Results1.During the study period,a total of 130 infectious diseases were reported in a sample population of 3987573 in Shandong Province,with a total case number of 106289 and an average annual incidence density of 694.86/100000 person-years.About two-thirds of the cases were 35 notifiable infectious diseases,while there were 95 non-notifiable infectious diseases,with a total of 32488 cases.The most common infectious diseases were respiratory diseases,with an average annual incidence density of 405.00/100000 person-years.The top three diseases were influenza,herpes zoster and pneumonia,accounting for 61.38%of the total cases.The incidence of 130 infectious diseases in urban areas was significantly higher than that in rural areas,with a relative risk(RR)of 1.25(95%CI:1.23-1.27).Among the five types of infectious diseases,the incidence of gastrointestinal diseases in rural areas was significantly higher than that in urban areas(RR=1.19,95%CI:1.15-1.24),and the incidence of the other four types of infectious diseases in urban areas was higher than that in rural areas(P<0.05).The incidence density of infectious diseases increased from 364.81/100000 person-years in 2013 to 1071.80/100000 person-years in 2017,and the increase rate of non-notifiable infectious diseases was higher than that of notifiable infectious diseases,with an average growth rate of 32.84%and 20.30%,respectively.The risk of infectious diseases in Jiaodong Peninsula was higher than that in inland areas(RR=4.66,95%CI:4.61-4.71).2.Between 2015 and 2017,the incidence density of pneumonia in the study population in Qingdao city was 54.33 per 100,000 person-years.There were statistically significant differences in the distribution of pneumonia cases in urban and rural areas and different age groups during the study period(P<0.01),but there were no significant differences in the distribution of pneumonia cases in different gender groups(P=0.1 5).The RR of pneumonia was 2.10(95%CI:1.06-4.13)for each 10μg/m3 increase in SO2 concentration after 4 weeks lag.Subgroup analysis showed that PM2.5 and SO2 had stronger impacts on the incidence of pneumonia in females than in males,and the RR of pneumonia in females were 2.38(95%CI:1.48-3.82)and 2.14(95%CI:1.12-4.10)for each 10μg/m3 increase of PM2.5 and SO2 concentration,respectively.The impact of NO2 and O3 on the incidence of pneumonia was more significant in urban areas than in rural areas,and the RR of pneumonia was 2.46(95%CI:1.165.22)and 4.29(95%CI:1.25-14.74)for each 10μg/m3 increase of O3 and NO2 concentration,respectively.The short-term health effects of different pollutants on different age groups were not significantly different.3.The combined effect at provincial level showed that the risk of influenza was higher at low temperatures,and the cumulative risk of influenza increased with the decrease of temperature,while the risk of influenza at high temperatures was not statistically significant.The heterogeneity of the relationship between temperature and influenza in different cities included annual mean temperature,longitude,meteorological region,proportion of retired people and proportion of kindergarten students.More attributable risk of influenza were due to low temperatures,with a attributable fraction of 36.66%.The attributable fraction of influenza caused by low temperatures was positively correlated with urbanization,and negatively correlated with NDVI,summer NDVI and winter NDVI.The transmission of influenza was relatively higher at low temperatures,and the elevated concentration of air pollutants at low temperatures may increase the risk of influenza.4.The internal verification results of the maximum entropy niche model showed that the mean AUC of the constructed model was 0.842,which indicating the model fitting effect was good.Population density and the number of children under 5 years old contributed the most to the model,with the contribution percentages of 44.7%and 41.8%,respectively.The results of the empirical study showed that 91%of the actual cases were distributed in the medium-risk and high-risk areas,reflecting the feasibility and reliability of the prediction model.Conclusion1.In recent years,the incidence of infectious diseases,especially non-notifiable infectious diseases,has been on the rise in Shandong Province,and the incidence of infectious diseases in urban areas is significantly higher than that in rural areas.The risk of infectious diseases in Jiaodong Peninsula is significantly higher than that in other areas.Children,the elderly and urban residents are vulnerable groups of infectious diseases,and the prevention and control of infectious diseases among these populations should be strengthened.PM2.5 and SO2 have a greater impact on female pneumonia,while urban population is more susceptible to NO2 and O3.Active mea2.Elevated concentrations of SO2 will increase the risk of pneumonia in the population.sures should be taken to further reduce the concentration of air pollutants and strengthen the prevention and control of pneumonia among the susceptible groups.3.Low temperatures could increase the risk and transmission ability of influenza.There is some heterogeneity in the effects of temperature on influenza in Shandong Province.The attributable fraction of influenza caused by low temperatures was positively correlated with urbanization level and negatively correlated with NDVI value.At low temperatures,higher concentrations of PM2.5 may increase the risk of influenza.4.The influence of population density,the number of susceptible children under 5 years old and the level of urbanization should be considered when constructing the prediction model of influenza.The maximum entropy ecological niche model is effective in predicting the risk of influenza in winter in Shandong Province and this model has a high accuracy,which can provide a reference for the prediction and early warning of infectious diseases.