Abstract:

Objective To analyze the prevalence of viral hepatitis B（HBV） in Guiping City from 2013 to 2024，construct a incidence prediction model of HBV，and provide a reference basis for relevant departments to carry out HBV prevention and control work. Methods Descriptive epidemiological methods were used to analyze the distribution characteristics by temporal，spatial，and demographic of HBV in Guiping City，and autoregressive integrated moving average（ARIMA） and Bayesian structured time series（BSTS） models were constructed to fit and predict the number of monthly cases of HBV in Guiping City，and the performance of the two models was compared. Results A total of 21 359 cases of HBV were reported in Guiping City from 2013 to 2024，with an average annual incidence rate of 88.42/100 000，and the incidence rate was generally on the rise（t=3.330，P<0.01）. There were no obvious seasonal characteristics of the incidence of disease，and the male-to-female sex ratio was 2.32 ∶ 1，with young adults aged 30-<45 years predominantly. In the distribution of occupations，farmers were the main group（16 868 cases，accounting for 78.97%）. The difference in the average annual reported incidence rate between different regions was statistically significant（χ²=637.303，P<0.01）. The root mean square error of the ARIMA model and the BSTS model was 29.69 and 31.57，respectively，and the mean absolute percentage error was 21.32% and 21.92%，respectively. The monthly incidence trend of HBV in Guiping City in 2025 predicted by ARIMA model was relatively stable. Conclusion The overall incidence of HBV in Guiping City shows an increasing trend，and prevention and control work should not be neglected. It is recommended to continue to strengthen comprehensive prevention and control measures focusing on vaccination to reduce the incidence of HBV. The ARIMA model developed for fitting and prediction was slightly better than the BSTS model in general.

Key words: Viral hepatitis B（HBV）, Epidemiological characteristics, Prediction, ARIMA model, BSTS model