【摘要】 目的分析湖南省2014—2020年水痘发病的流行特征，建立自回归滑动平均混合（autoregressive integrated moving average,ARIMA）模型对2021年和2022年水痘月发病率进行预测，评价模型对水痘发病的预测效果，为疫情的防控提供新的思路。方法 收集湖南省2014—2021年水痘月发病率数据，分析2014—2020年湖南省水痘的流行特征，使用2014—2020年的数据建立ARIMA模型对2021年、2022年水痘月发病率进行预测，以2021年的预测结果与实际发病数据进行对比评价模型的预测准确性。结果湖南省水痘高发年份为2017—2019年，5—6月以及10月至次年1月为水痘报告的高发月份。ARIMA最优预测模型为（2,0,0）（1,0,1）₁₂，其对2021年水痘月发病率预测的平均相对误差为15.20%。利用最优模型预测的水痘2022年发病趋势与2021年基本一致，6月、10月至次年1月仍是水痘高发月份。结论ARIMA模型对水痘的预测效果较为理想，可以为湖南省水痘的预测提供参考依据。更多还原

【Abstract】 Objective To understand the epidemiological characteristics of varicella in Hunan province from 2014 to 2020, and to build an autoregressive integrated moving average（ARIMA） model in order to predict the monthly incidence of varicella in 2021 and 2022, and to assess the prediction accuracy of the model.Methods The monthly incidence data of varicella in Hunan Province from 2014 to 2021 was collected, and the data from 2014 to 2020 was used to establish an ARIMA model to predict the monthly incidence of varicella in 2021 and 2022, the prediction accuracy of the model was evaluated by comparing the prediction results in 2021with the actual incidence data. Results The incidences of varicella in Hunan province were higher from 2017 to 2019, and majority of cases were reported from May to June and October to January. The optimal prediction model was ARIMA（2, 0, 0）（1, 0, 1）₁₂, and its average relative error in predicting the monthly incidence of varicella in 2021was 15. 20%. The incidence trend of varicella in 2022 predicted by ARIMA（2,0,0）（1,0,1）12model was the same as in 2021. Higher incidences were also in June and October to January. Conclusions The ARIMA model is able to accurately predict the incidence of varicella, which may be widely used for the prediction of varicella in Hunan province.更多还原