Study On The Epidemic Characteristics Of Rash And Fever Syndrome And Prediction Of The Main Pathogen Among Under-five Years Children In Gansu Province

Objective:Through the analysis and research of the monitoring data of fever with rash syndrome in children under 5 years old in Gansu Province from 2011 to 2020,to observe the distribution characteristics of pathogens,reveal the epidemic pattern of pathogenic infection,and identify the main pathogen types;and to construct the main pathogens according to the distribution pattern.The positive rate prediction model was constructed to estimate the epidemic trend of the main pathogens of fever with rash syndrome in children under 5 years old in Gansu Province,and to provide applicable methods and technical references for the corresponding disease prevention and control.Methods:An observational study method was used to analyze the distribution and pathogen composition of fever with rash syndrome in children under 5 years old in Gansu Province from2011 to 2020.Theχ~2test was used to reveal differences in the gender distribution and age composition of the pathogens,to find the main pathogen type and reveal the epidemic regular of pathogenic infection.Based on the epidemic pattern of main pathogens,the SARIMA model was constructed with R software.On this basis,the combined model of SARIMA-ERNN and SARIMA-GRNN was established by Matlab software.MAE,RMSE and MAPE were used to evaluate the prediction effect of each model,and the optimal model was selected.Results:1.From 2011 to 2020,a total of 3136 cases of children under 5 years of age with fever with rash syndrome were monitored,and 1735（59.30%）were positive cases,of which cases under 3 years old accounted for 59.71%of the total.There was no statistical difference in the positive rate of different genders（P=0.144）,but different among each age groups（P<0.05）.The positive cases showed a seasonal"three peaks"epidemic,and the positive rate in the eastern region was higher than the Hexi region and the central region（P<0.001）.2.The pathogens were mainly viruses（99.83%）,and the top three pathogens were enteroviruses（74.87%）,measles virus（18.27%）and varicella-zoster virus（5.88%）.The pathogen composition was different in different years.The composition ratio and positive rate of measles viruses were the highest in 2015 and 2017,and enteroviruses was the highest in the rest years.3.Except for varicella-zoster virus,the positive rate of other pathogens were not significantly difference between genders（P>0.05）;the pathogen composition of the 0-month group and the6-month group was mainly measles virus,with a composition ratio respectively were 70.13%and57.33%,the pathogen composition of the remaining age groups is dominated by enterovirus,the composition ratios were 81.74%,87.50%,87.70%and 78.55%,respectively.The positive rates of enterovirus,measles virus and varicella-zoster virus in different seasons were statistically different（P<0.05）;a pairwise comparison corrected by Bonferroni method found that the positive rate of summer enterovirus is higher than other seasons（χ~2=72.233,P<0.001）,the positive rate of measles virus in spring was higher than other seasons（χ~2=129.071,P<0.001）,and the positive rate of varicella-zoster virus in winter was higher than other seasons（χ~2=31.262,P<0.001）.The"three peaks"of enteroviruses were prevalent,with the peak of the pandemic from May-August,the second peak of epidemic in October,and the peak of small epidemic from January to February of the following year.The epidemic season of measles virus was from January to May.The peak of herpes virus prevalence was from November to February of the following year.4.Enterovirus was the main pathogen of fever with rash syndrome in children in Gansu Province.Its pathogen subtypes are diverse.Among them,EV71 positive cases are the most（435cases）,accounting for 41.31%of the pathogen type;the second is CVA16（337 cases）and other types（239 cases）,accounting for 32.00%and 22.70%,respectively.The 6-month group was mainly other types（43.33%）and CVA16（26.67%）;the 12-month group pathogenic subtypes were mainly EV71（36.81%）and other types（31.60%）;other age groups were mainly EV71 type and CVA16 type.EV71 showed an epidemic peak every 3-4 years,and EV71 accounted for a higher proportion from November to May of the following year;while CVA16 had an epidemic peak every 2-4 years,and CVA16 accounted for a higher proportion from June to October.5.Based on the monthly positive rate of enterovirus,the overall prediction error MAE of the SARIMA（1,1,1）（0,1,1）₁₂model,SARIMA-ERNN combined model and SARIMA-GRNN combined model was 0.222,0.194,and 0.085,respectively.The RMSE was 0.301,0.256,and0.228,respectively,and the MAPE was 38.080%,32.265%,and 31.016%,respectively.It can be seen that the prediction accuracy of the SARIMA-GRNN combined model was better than other models.Using this model to predict the monthly positive rate of enteroviruses from January to December 2021,it was found that its epidemic trend was consistent with the past.The positive rate from August to September was higher than the historical level in the same period,and CVA16was the main type.Conclusion:1.The positive cases of fever with rash syndrome in children under 5 years old in Gansu Province were concentrated in the age group under 3 years old.2.The pathogen composition of fever with rash syndromes of children under 5 years old in Gansu Province was mainly viral infections.The main pathogens were enterovirus,measles virus and varicella-zoster virus.Enterovirus and measles virus infection were common,the main cause of enterovirus infection was over 12 months old,and the main cause of measles virus infection was under 12 months of age.3.Enteroviruses showed an obvious seasonal trend.Types EV71 and CVA16 were the main pathogenic subtypes of enterovirus infections in children under 5 years old in Gansu Province.EV71 was more common in winter and spring,and CVA16 was more common in summer and autumn.4.The SARIMA model,SARIMA-ERNN and SARIMA-GRNN combined model can be used to predict the positive rate of enterovirus in children under 5 years old in Gansu Province.The prediction effect of the combined model was better than that of the single model,and the SARIMA-GRNN combined model has a better prediction effect.It can be used as a predictive model for the monthly positive rate of enterovirus in children under 5 years old in Gansu Province.