Abstract
OBJECTIVES: This study was performed to develop and evaluate statistical and mathematical models for predicting the magnitude of the seasonal respiratory syncytial virus (RSV) epidemic during winter in the Republic of Korea. METHODS: Models, including the linear regression model, sigmoid function model, and Susceptible-Exposed-Infectious-Recovered (SEIR) model, were fitted to RSV epidemic data spanning 2015–2024 to explore whether key parameters from each model, including temperature, humidity, and number of cases, could predict epidemic trends in the early stages of each seasonal epidemic. RESULTS: The linear regression model performed well in predicting epidemic scale based on early stage information. The sigmoid function was suitable for forecasting epidemic peak timing. The SEIR model estimated the effective contact rate, but showed limited ability to predict this parameter using early stage data. CONCLUSIONS: This study demonstrates that simple statistical models can effectively predict RSV epidemic scales, providing valuable insights for winter healthcare preparation.