Application of moving epidemic method in early warning of influenza incidence intensity in winter-spring season in Ningbo city

ObjectiveTo explore the application of moving epidemic method (MEM) in early warning of influenza incidence intensity in winter-spring season in Ningbo city and to provide a reference for developing effective intervention measures. MethodsInfluenza surveillance data for the period of January 2013 – May 2022 were collected from 2 national influenza sentinel hospitals in Ningbo city, Zhejiang province and a part of the data on positive rate of influenza virus detection among the registered cases in winter-spring seasons from January 2013 to December 2019 were extracted to establish a MEM model. The constructed MEM model was adopted to analyze the beginning, the end and incidence intensity of seasonal influenza epidemic during 2019 – 2020 (40th week, 2019 – 20th week, 2020) in Ningbo city; the model analysis results were compared with actual situation of the epidemic. ResultsFor the virus-positive cases in the city during the 2013 – 2022, the proportions of isolated viral strains were 24.59% for influenza A (H1N1), 37.03% for influenza A (H3N2), and 38.38% for influenza B (mainly Victoria and Yamagata strain), respectively. The established MEM model showed a better goodness-of-fit, with the parameter δ of 2.7, the sensitivity of 87.97%, the specificity of 87.68%, and the Yoden index of 0.76. Based on the fitting results of the established MEM model, the thresholds of virus-positive rate were 22.76% and 25.05% for identifying the onset and the end of the influenza epidemic and the thresholds were 43.18%, 63.22%, and 74.83% for indicating a moderate-, high-, extremely high-intensity of the influenza epidemic during 2019 – 2020 influenza season. With the established MEM threshold estimations, the trajectory of the influenza could described as following: pre-epidemic stage from 40th week to 48th week of 2019, onset/low-intensity stage from 49th week of 2019, moderate-intensity stage from 51th week of 2019 to third week of 2020, subsequent low-intensity stage from 4th week of 2020, late stage from 7th week of 2020, and the end of the epidemic by the 10th week of 2020. Compared to the seasonal epidemics between 2013 – 2019, the 2019 – 2020 winter spring influenza epidemic occurred one week earlier but ended three weeks sooner. Conclusion MEM model could be adopted effectively in early identification and intensity warning of seasonal influenza epidemic.