A novel model to predict age of respiratory syncytial virus infection from birth timing in relation to RSV circulation.

Respiratory syncytial virus (RSV) is a common respiratory virus that infects all children by age 2 to 3 years of age and causes the greatest morbidity at the extremes of life. Recent evidence suggests that early-life RSV infection, defined using active and passive surveillance with quantitative polymerase chain reaction- and serology-identified infection, is causal for childhood asthma. As such, identifying infants that are likely to be infected with RSV during this critical susceptibility window has important implications for determining who is most at risk for chronic respiratory sequelae like asthma. However, identifying the age of RSV infection is impractical in large populations, as not all infections are symptomatic, and measurement thus requires time- and cost-intensive surveillance. To address this, we developed the first probability model for age of first RSV infection. It uses an infant's birthdate, demographic covariates, and publicly available RSV circulation data to determine the probability they were first infected at any age from birth to one year. Our model is easy to interpret, provides an exceptional fit for the data, and generalizes across populations, where we use it to accurately predict age of first infection in two independent cohorts. Our work represents a major development in RSV research, as it facilitates, for the first time, reliable estimation of the age of infant RSV infection during the first year of life in populations without the need for active surveillance.