基于ARIMA模型的医院流感样病例特性分析与预测研究

Characteristic Analysis and Prediction of Hospital Influenza-Like Illness Based on ARIMA Model

目的基于自回归移动平均模型分析预测医院流感样病例就诊特点,探讨疫情防控应急管理策略。方法采用回顾性研究收集某三甲医院2017-2022年流感样病例,分析不同季度、科室流感样病例流行病学特征,采用时间序列拟合模型分析流感样病例流行趋势及变化。结果2017—2022年医院就诊的5048123例门急诊病例中监测到118781例流感样病例,其中2017—2019年12月至次年3月呈现出流感样病例发病高峰,2020—2022年由于新冠疫情防控措施使得流感样病例就诊未呈现典型的季节性波动。采用时间序列拟合模型分析后,预测2023—2024年流感样病例及门急诊病例数,预测季节性高峰将于2023年12月、2024年3月以及2024年12月出现。结论利用ARIMA模型对流感样病例进行分析预测有助于医院统筹医疗资源,加强应急管理,降低流感传播风险。

Objective To analyze and predict the outpatient characteristics of influenza-like illness in hospitals based on autoregressive integrated moving average(ARIMA)model,and to explore the emergency management strategy for epidemic prevention and control.Methods A retrospective study was conducted to collect the influenza-like illness cases in a tertiary hospital from 2017 to 2022.The epidemiological characteristics of influenza-like illness cases in different seasons and departments were analyzed,and the time series fitting model was used to analyze the epidemic trend and change of influenzalike illness.Results From 2017 to 2022,118781 influenza-like illness cases were detected among 5048123 outpatient and emergency cases treated in hospitals,with a peak of influenza-like illness occurring from December to March next year from 2017 to 2019,and no typical seasonal fluctuation of influenza-like illness from 2020 to 2022 due to the prevention and control measures of the COVID-19 epidemic.After using the time series fitting model,the number of influenza-like illness cases and outpatient and emergency cases from 2023 to 2024 was predicted,and the seasonal peak was predicted to occur in December 2023,March 2024 and December 2024.Conclusion The use of ARIMA model to analyze and predict influenzalike illness can help hospitals to coordinate medical resources,strengthen emergency management,and reduce the risk of influenza transmission.