COVID-19 prediction models are characterized by uncertainties due to fluctuating parameters, such as changes in infection or recovery rates. While deterministic models often predict epidemic peaks too early, incorporating these fluctuations into the SIR model can provide a more accurate representation of peak timing. Predicting R0, the basic reproduction number, remains a major challenge with significant implications for government policy and strategy. In this study, we propose a tool for policy makers to show the effects of possible fluctuations in policy strategies on different R0 levels. Results show that epidemic peaks in the United States occur at varying dates, up to 50, 87, and 82 days from the beginning of the second, third, and fourth waves. Our findings suggest that inaccurate predictions and public health policies may result from underestimating fluctuations in infection or recovery rates. Therefore, incorporating fluctuations into SIR models should be considered when predicting epidemic peak times to inform appropriate public health responses.
© 2023. The Author(s).