2014-2023年云南省乙肝流行特征分析及发病预测

Epidemiological characteristics and incidence prediction of hepatitis B in Yunnan, 2014−2023

  • 摘要:
    目的 分析2014-2023年云南省乙型病毒性肝炎(简称乙肝)的流行特征并预测发病趋势,为制定精准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云南省乙肝流行特征,并构建季节性自回归差分移动平均模型(SARIMA)与支持向量机回归模型(SVR)组合模型进行预测。
    结果 2014-2023年云南省累计报告乙肝174 305例,年均发病率36.70/10万,以慢性乙肝为主(140 582例,80.65%),报告发病率总体呈显著下降趋势(Z=−73.86,P<0.001),时间分布以冬春季为主;地区分布显示部分市(自治州)在2016-2023年呈现不同程度上升趋势。乙肝报告发病病例年龄集中在35岁以上(143 254例,82.18%),男性发病率(47.91/10万)高于女性(24.66/10万),差异有统计学意义(χ2=17 537.054,,P<0.001),职业以农民(占77.21%)为主。SARIMA-SVR组合模型预测的均方根误差(RMSE)为98.34,平均绝对百分比误差(MAPE)为7.88,均优于单一SARIMA模型(RMSE=126.86,MAPE=10.79),表明组合模型在乙肝报告病例数预测中具有更高的精度和稳定性。模型预测2024年云南省乙肝报告病例数为12 762例,较2023年下降3.77%。
    结论 2014-2023年云南省乙肝疫苗接种及母婴阻断措施对低龄人群乙肝发病控制效果显著,但乙肝总体发病水平仍处于一定水平;男性、35岁及以上人群和农民群体仍是乙肝防控的重点人群;建立的SARIMA-SVR组合模型可较好地预测云南省乙肝发病趋势。

     

    Abstract:
    Objective To analyze the epidemiological characteristics of hepatitis B in Yunnan province from 2014 to 2023, predict future incidence trend and provide reference for the development of targeted hepatitis B prevention and control strategies.
    Methods Descriptive epidemiological methods were used to analyze the epidemiological characteristics of hepatitis B in Yunnan. A hybrid forecasting model combining Seasonal Auto-regressive Integrated Moving Average (SARIMA) model and Support Vector Regression (SVR) model was constructed to predict the incidence of hepatitis B.
    Results A total of 174 305 hepatitis B cases were reported in Yunnan from 2014 to 2023, with an average annual incidence rate of 36.70/100 000. In the reported hepatitis B cases, most were chronic ones (140582 cases, 80.65%). The reported incidence of hepatitis B showed a significant downward trend during the study period (Z=−73.86, P<0.001), with a seasonal peak occurring in winter and spring. Several prefectures showed varying degrees of increase of reported hepatitis B incidence during 2016−2023. The reported hepatitis B cases were mainly individuals aged ≥35 years (143254 cases, 82.18%). The incidence rate in men (47.91/100 000) was significantly higher than that in women (24.66/100,000), and the difference was significant (χ2=17537.054, P<0.001). Farmers were mostly affected occupational group (77.21%). The SARIMA-SVR hybrid model outperformed the single SARIMA model, with a lower root mean square error (RMSE: 98.3448 vs. 126.8590) and mean absolute percentage error (MAPE:7.8797% vs. 10.7856%), indicating better predictive accuracy and stability. The model predicted that 12 762 hepatitis B cases would be reported in Yunnan in 2024, a 3.77% decrease compared with that reported in 2023 (13 283 cases).
    Conclusion The reported incidence of hepatitis B in Yunnan showed an overall declining trend in younger age groups from 2014 to 2023, which could be attributed to the hepatitis B vaccination program and the blocking of mother to infant transmission; however, the overall incidence of hepatitis B remained at certain level. Men, individuals aged 35-<60 years, and farmers were identified as key populations for hepatitis B prevention and control. The SARIMA–SVR hybrid model demonstrated good performance in the prediction of hepatitis B incidence trends in Yunnan .

     

/

返回文章
返回