ObjectiveTo calculate the influenza epidemic threshold, intensity thresholds and predict peak, and provides a theoretical basis for the surveillance for influenza activity levels and early warning.

MethodsThe positive rates of influenza virus in influenza seasons from 2011 to 2018 in Yangzhou (from week 40 to week 20 of the next year) was used as the historical baseline data. Moving epidemic method (MEM) was used to analyze the epidemic threshold and three intensity thresholds, and the average number of weeks from the day reaching epidemic threshold to the day reaching peak was calculated to predict peak of 2018 —2019 influenza season in Yangzhou.

ResultsThe data of influenza seasons from 2011 to 2018 were used to establish the model, the optimal parameter δ value was 2.10. The average sensitivity, specificity, positive predictive value, negative predictive value and Yoden index were 0.81, 0.91, 0.87, 0.87 and 0.72, respectively. In 2018 —2019 influenza season, the epidemic period began in the third week of 2019. The predicted peak was consistent with the observed peak (week 8). The intensity level was high.

ConclusionThe MEM using influenza virology data is an effective tool for assessment of the intensity of influenza epidemic and prediction of the peak of influenza season in Yangzhou.