Bridging individual and population dynamics of acute respiratory viruses to optimize antiviral interventions in nursing homes

Respiratory virus outbreaks impose a recurrent burden in nursing homes, where elderly residents face high risks of infection and severe complications. Although antiviral treatments exist, their use remains limited due to insufficient evidence on efficacy against virus transmission. In this context, outbreak control often relies on contact precaution and isolating symptomatic residents and/or their contacts. To evaluate pharmacological interventions, we developed a multiscale model integrating within-host viral kinetics, contact patterns, and transmission dynamics in nursing homes. The framework was applied to outbreaks caused by SARS-CoV-2, influenza virus (IAV), and respiratory syncytial virus (RSV) in nursing home populations. Calibrated to reflect typical nursing home individual interactions and the activity of antivirals reducing residents severity by 50%, the model shows that treating symptomatic individuals together with their close contacts could markedly reduce infection burden. In such scenarios, incidence of severe cases and overall cases decreased by up to 80% and 45-70%, respectively. Notably, such strategies were predicted to be at least as effective as isolation- and contact-reduction-based measures. In contrast, treating only symptomatic residents alone had limited impact on transmission, because of the strong role of pre- and asymptomatic transmission, especially for SARS-CoV-2. Antivirals with modest efficacy were unlikely to substantially mitigate SARS-CoV-2 outbreaks but could still prove effective against IAV and RSV, where transmission dynamics are more favorable. This modeling approach combines within- and between-host processes to provide a comprehensive understanding of respiratory virus outbreaks in nursing homes. It can guide the design of targeted interventions, optimize antiviral deployment, and reduce reliance on isolation, thereby helping to mitigate both outbreak size and severity.