The abstract is read at the publisher; the summary is JClub's.
Including school and public holiday effects plus an annually periodic transmission component improves 28-day forecasts of RSV, influenza, and SARS-CoV-2 case incidence across five Australian jurisdictions, beating a random-walk baseline.
- Why it matters: Forecasts are used for public health decisions, but predictive accuracy can be limited when models ignore calendar-driven and seasonal changes in transmission and reporting.
- What they did: A Bayesian renewal model decomposes R(t) into log-additive components for school and public holidays and an annually periodic Gaussian process, while allowing reporting rates to vary with holidays; it forecasts three pathogens over 28 days in five Australian jurisdictions.
- The result: For the 2025 winter season, forecasts outperform the random-walk baseline across all settings, and holiday and seasonal transmission effects improve performance; school holidays are linked to population-wide reductions in R(t) of 5-15% and generally not to reporting changes.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.