Applications of Estimating the Hospitalization Burden of Respiratory Syncytial Virus

Abstract

Annual epidemics of respiratory viruses such as influenza and RSV result in a major burden of associated severe outcomes. Some of those outcomes stem from the underlying health conditions that infected individuals carry, such as chronic lower respiratory disease (CLRD). Other outcomes are the result of the synergistic effect of respiratory tract infections from viruses and bacteria, such as S. pneumonia. These complications are of particular concern for both the infant and elderly population due to their impaired ability to fight off these illnesses. Understanding the scope of this burden presents a strong case for development of preventative tactics, such as vaccination and administration of antiviral medicines. For RSV, this information should shed light on the cost-effectiveness of developing a vaccine and employing it for different population categories. Quantifying the rates of hospitalization due to influenza and RSV is not easy, however, as influenza- and RSV- associated outcomes different population groups (save for RSV-associated bronchiolitis in young children) are rarely laboratory-confirmed. Consequently, a variety of mathematical and statistical models have been proposed to capture the true burden of these viruses and the risks for hospitalization outcomes. In this thesis, we employ a statistical framework that incorporates appropriate proxies for the incidence of the different respiratory viruses, a flexible model for the baseline of severe outcome not associated with these viruses, as well as other relevant constructs to ascertain both annual and average annual estimates of hospitalization rates associated with both influenza and RSV for each age group. These estimated hospitalization rates are used to predict the expected benefit of reducing the (high) costs of hospitalizations associated with influenza and RSV illnesses. This analysis builds on the existing literature, with the following extensions: (1) Application of an influenza- and RSV-specific model previously proposed by Goldstein et al. ([14]) to a novel dataset detailing hospitalization influenza data from eighteen states with a more granular age stratification. (2) Estimating the expected benefit and cost-effectiveness of developing an RSV vaccine. (3) Examining relative roles in driving RSV epidemics, with the aim of informing population-wide effects of RSV vaccines in different age groups.