2005-2023年湖北省武汉市急性乙型肝炎流行特征及发病预测分析

Epidemiological characteristics and incidence prediction of acute hepatitis B in Wuhan, Hubei, 2005−2023

  • 摘要:
    目的 分析2005—2023年湖北省武汉市急性乙型肝炎(乙肝)的流行特征,预测急性乙肝发病趋势,为急性乙肝防控策略提供参考依据。
    方法 采用描述性流行病学方法分析武汉市急性乙肝流行特征,以2005—2021年数据作为观测值建立自回归移动平均(ARIMA)模型和贝叶斯结构时间序列(BSTS)模型,2022年和2023年数据作为测试值比较两种模型预测的准确性。
    结果 2005—2023年武汉市共报告急性乙肝3 991例,年均报告发病率为2.10/10万;男性发病率高于女性;无明显季节性发病高峰;各区年均发病率为1.21/10万~3.37/10万。 病例职业主要为家务及待业(782例,19.59%)、农民(697例,17.46%)。 2005—2010年各年龄组发病率均高于平均水平,20~<25岁组最高,达9.01/10万,2011—2016年各年龄组发病率略低于平均水平,2017—2023年各年龄组发病率显著低于平均水平,0~<5岁组最低,为0.03/10万,病例主要集中在25~<65岁组(75.04%)。 BSTS模型、ARIMA模型预测的均方根误差分别为3.67、8.86,平均绝对百分比误差分别为49.63%、75.17%。 基于BSTS模型预测的武汉市2024年1—6月急性乙肝发病数与实际值比较,预测的均方根误差为3.32,平均绝对百分比误差为18.61%。
    结论 武汉市急性乙肝发病率总体呈下降趋势,各年龄组发病率均维持较低水平,尤其是0~<5岁组儿童,需加强重点人群的急性乙肝防控工作,建立的BSTS模型在预测急性乙肝的发病趋势方面优于ARIMA模型。

     

    Abstract:
    Objective To analyze the epidemiological characteristics and predict the incidence trend of acute hepatitis B in Wuhan, Hubei province, from 2005 to 2023, and provide reference for the prevention and control strategies of acute hepatitis B.
    Methods Descriptive epidemiological methods were used to analyze the epidemiological characteristics of acute hepatitis B in Wuhan, and autoregressive integrated moving average (ARIMA) model and Bayesian structural time series (BSTS) model were established with the incidence data of acute hepatitis B from 2005 to 2021 as observed values, and the incidence data from 2022 and 2023 as test values to compared the prediction accuracy of the two models.
    Results A total of 3 991 cases of acute hepatitis B were reported in Wuhan from 2005 to 2023, the average annual incidence rate was 2.10/100 000. The incidence was higher in men than in women. There was no obvious seasonal incidence peak. The average annual incidence rate ranged from 1.21/100 000 to 3.37/100 000 in different districts. The cases were mainly the unemployed (782 cases,19.59%) and farmers (697 cases,17.46%). The incidence rates in all age groups were higher than the average level during 2005−2010, and the incidence rate was highest in age group 20−<25 years ( 9.01/100 000). The incidence rates in all age groups were slightly lower than the average level during 2011−2016. The incidence rates in all age groups were significantly lower than the average level during 2017−2023. The incidence rate was lowest in age group 0−<5 years (0.03/100 000). Most cases occurred in age group 25−<65 years (75.04%). The root mean square error of ARIMA model and BSTS model were 3.67 and 8.86, respectively, and the mean absolute percentage error was 49.63% and 75.17 %, respectively. The predicted value of the incidence of acute hepatitis B from January to June 2024 based on BSTS model were compared with the actual value, the root mean square error was 3.32, and the mean absolute percentage error was 18.61%.
    Conclusion The incidence of acute hepatitis B showed a downward trend in Wuhan from 2005 to 2023. In recent years, the incidence remained at low levels in all age groups, especially in age group 0−<5 years. The prevention and control of acute hepatitis B in key groups should be strengthened. BSTS model has better fitting and prediction performance compared with ARIMA model.

     

/

返回文章
返回