Risk Assessment Model And Early Warning Research Of Swine Fever, Porcine Reproductive And Respiratory Syndrome, And Pseudorabies Based On Seroepidemiology

Classical swine fever virus(CSFV)belongs to the genus Pestivirus of the family Flaviviridae.Pigs can be infected with clinical symptoms such as high fever,systemic hemorrhage,diarrhea,and death.Porcine reproductive and respiratory syndrome virus(PRRSV)is a single stranded RNA virus encapsulated in a capsule,belonging to the Arteriviridae family,Arteriviridae genus,characterized by severe reproductive failure in sows and respiratory confusion in young and growing and fattening pigs.Pseudorabies virus(PRV)is a double-stranded linear DNA virus that can infect various animals,including humans.From December 2017 to May 2021,blood samples were collected in 12 provinces and two cities in China in our study.The antibodies of the three viruses were detected by the method of Enzyme-linked immunosorbent assay(ELISA).Meanwhile,we collected the background information from samples and pig farms and investigated the serological prevalence of the three viruses in China in recent years.Bayesian latent class models were constructed to estimate the true prevalence of the three viruses in China.The risk factors related to the serological status of the three viruses were analyzed through logistic regression models.Furthermore,Sa TScan 9.6 software was used to predict the spatiotemporal distribution characteristics of CSFV,PRRSV,and PRV.An additive Bayesian network was used to construct a network structure model of farm-level variables to study the interaction among the three viruses.Autoregressive Moving Average(ARMA)method was used to analyze the time-series data of the antibody-positive rate of the three viruses.Moreover,using @RISK software(Version 7.0)to simulate and sample the time series models of the three diseases to predict future epidemiological trends.Finally,the simulation results are compared with the third-party testing company’s monitoring data to evaluate the model prediction’s accuracy.All three diseases have caused huge economic losses to pig farming and are the most important viral diseases in pig farms in China.A total of 40,489 pig serum samples were collected from 553 pig farms.The total positive rate of CSFV antibody was 60.40%(95%CI: 59.92%-60.88%),and the positive rate at the farm level was 100%.The highest rates of CSFV antibody in boars and multiparity sows were 92.57%(95%CI: 91.78%-93.31%)and 92.29%(95%CI: 91.84%-92.72%),respectively.14,134 blood samples were collected from 316PRRSV-unvaccinated pig farms.The overall PRRSV antibody-positive rate at the animal level was 61.84%(95%CI: 61.04%-62.65%),and the PRRSV antibody-positive rate at the pig farm level was 90.51%.The positive rates of PRRSV antibodies in gilts were highest at 74.59%(95%CI: 72.15%-76.91%).A total of 40,024 samples were collected from 545 pig farms in China.The positive rate of PRV g E antibody at the animal level was 25.04%(95%CI: 24.61%-25.46%),and the positive rate of PRV g E antibody at the farm level was 55.96%.The positive rates of PRV g E antibody in suckling piglets and weaning nursery pigs were 30.74%(95%CI: 29.31%-32.20%)and 30.33%(95%CI:29.27%-31.41%),respectively.Significant differences existed in the positive rates of the three virus antibodies in various provinces,municipalities,and pig herds.The true positive rates estimated by the hierarchical Bayesian model were shown as follows.The true positive rate of CSFV antibodies in each province is the highest in Sichuan Province at 90.88%(95%Cr I: 87.97%-93.25%).In contrast,the lowest in Henan Province is 81.88%(95%Cr I: 78.08%-84.84%).The true positive rate of PRRSV estimated showed that the PRRSV infection in Henan,Guangdong,Tianjin,and Liaoning provinces was the most serious at more than 75%.The PRV infection was the most serious in Henan Province,Shandong Province,and Tianjin City,which was 53.74%,55.54%,and 75.51%.Through multivariate logistic analysis,it was found that variables such as ’sampling season,’ ’geographical division of pig farms,’ ’the topography of pig farms,’ ’before and after the outbreak of ASF,’ and ’PRRSV prevention and control’ are risk factors that affect the CSFV serological status(negative/positive)of pigs at animal level.Variables such as farm size,ASF outbreak,and farm PRV purification were found to be risk factors affecting farm-level PRRSV infection.The farm’s geographical location,topography,the ASF outbreak,and the PRRSV control situation were the risk factors affecting the serological status of PRV at the farm level.Three significant regional clusters with high PRRSV-positive rates were detected for the first time in China.All three PRRSV clusters occurred before the outbreak of African swine fever.Five significant clusters of high PRV g E seroprevalence were also detected.The time range of the five PRV clusters occurred from 2017/12/1 to 2019/7/31.The possibility of construction of pig farms from the south in the plains was 0.508(95%CI: 0.0318-0.0782)times higher than that in the mountains.Then if the pig farm is located in a plain area,the possibility of the pig farm being a PRV-positive farm is 3.3003(95%CI: 2.2367-4.8695)times that of a PRV-negative farm.The possibility of PRV-positive pig farms also being PRRSV-positive farms was 12.2558(95%CI: 3.9988-54.5436)times that of PRV-negative farms.At the same time,PRV-positive farms were0.6035(95%CI: 0.4206-0.8668)times more likely to have high levels of CSFV antibodies than low levels.ASFV outbreaks and PRRSV serological status of pig farms also affected the CSFV antibody levels in pig farms,but the association was not significant.A time series model of the three diseases’ antibody positivity rates was established,simulated,and predicted using @RISK software.The results showed that the antibody-positive rate of swine fever and PRRS did not have apparent periodicity or trend with time.The averages predicted by the CSFV and PRRSV models remained unchanged,fluctuating around 80.68% and 55.64%,respectively.The time series model of the PRV antibody positive rate showed an apparent downward trend,and the mean value of the monthly average change rate was-0.826%.The probability of decreasing the positive rate of PRV antibody per month is 0.868,and the probability of increasing is 0.132.The epidemic trends of the three diseases predicted by @RISK software simulations are consistent with the monitoring data results of third-party testing companies,showing a positive correlation.The correlation coefficients were CSFV(0.631),PRRSV(0.354),and PRV(0.609).The established risk early warning models for the three diseases have the characteristics of high prediction accuracy and clinical application value.These research results provide data reference and a theoretical basis for understanding the epidemic situation of CSFV,PRRSV,and PRV in China in recent years,the risk factors of disease occurrence,time-varying law,temporal-spatial distribution characteristics,and interaction relationships.At the same time,it makes reasonable predictions and early warnings for the epidemic trends of the three viruses in China in the future,which will help decision-making departments prevent and control related diseases.