Date of Award
Spring 2022
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
Department
Public Health
First Advisor
Weinberger, Daniel
Abstract
Respiratory syncytial virus (RSV) is the leading cause of respiratory infection and hospitalization for infants, children, and the elderly [1]. In this dissertation, I present five quantitative studies aimed at improving our understanding of the disease burden, epidemic timing, geographic spread, and prevention of RSV. Specifically, I applied statistical and mathematical models to (1) estimate respiratory hospitalizations attributable to RSV across age and socioeconomic groups; (2) identify drivers that impact RSV epidemic spatiotemporal patterns in a typical winter season; (3) estimate the timing and intensity of RSV reemergence following the COVID-19 pandemic; (4) compare the geographic spread of RSV epidemics in summer 2021 to a typical winter season; and (5) predict the potential effectiveness of three RSV prevention strategies. Together, these five studies shed light on the mechanisms of RSV transmission and offer recommendations for future RSV prevention strategies. In the first study, I estimated the incidence of respiratory hospitalizations attributable to RSV by age and by socioeconomic status using regression models. I found that the estimated annual incidence of respiratory hospitalizations due to RSV was highest amongst infants <1 year of age coming from low socioeconomic neighborhoods. RSV incidence was also considerable in older adults ≥65 years of age across all socioeconomic statuses. The incidence of hospitalizations recorded as being due to RSV represented a significant undercount, particularly in older adults. These results suggest that RSV causes a considerable burden of hospitalization in young children and older adult populations in the U.S, with variations in burden that depended on socioeconomic status. Incidence of hospitalizations due to RSV in older adults based on the recorded diagnoses likely represents an underestimate. The estimated incidence of hospitalizations is roughly 40 times higher than indicated by recorded diagnoses in older adults. In the second study, I identified factors associated with local epidemic timing using a two-stage Bayesian meta-regression model. I found that earlier epidemics were associated with larger household size and greater population density. Nearby localities had similar epidemic timing. My findings suggest that RSV epidemics spread faster in areas with more local contact opportunities, and that epidemic spread follows a spatial diffusion process based on geographic proximity. Our findings can help inform the timing of delivery of RSV extended half-life prophylaxis and maternal vaccines and guide future studies on the transmission dynamics of RSV. In the third study, I examined the association of different factors, including mitigation strategies, duration of maternal-derived immunity, and importation of external infections, with the dynamics of reemergent RSV epidemics following the COVID-19 pandemic. I applied mathematical models to reproduce RSV epidemics that occurred before the COVID-19 pandemic and evaluate the impact of different factors on the timing and intensity of reemergent RSV epidemics. I found that virus introduction from external sources was associated with the emergence of the spring and summer epidemic in 2021. There was a tradeoff between the intensity of the spring and summer epidemic in 2021 and the intensity of the epidemic in the subsequent winter. Reemergent RSV epidemics in 2021 and 2022 were predicted to be more intense and affect patients in a broader age range than in typical RSV seasons. In the fourth study, I compared the relative onset timing of RSV epidemics in 2021 with those that occurred in 2016–2019 at the state level. I used a hierarchical Bayesian regression model to assess the similarity between timing of RSV epidemic onset in a typical year and the timing observed in 2021 across all states. Despite the unusual out-of-season summer timing, the relative timing of RSV epidemics between states in 2021 shared a similar spatial pattern with typical winter RSV seasons. Our results suggest that the onset of RSV epidemics in Florida can serve as a baseline to adjust the initiation of prophylaxis administration and clinical trials in other states regardless of the seasonality of RSV epidemics. In the fifth study, I predicted the potential effectiveness of maternal immunization, live-attenuated pediatric vaccines, and extended half-life monoclonal antibodies (mAbs) against RSV hospitalizations. I fitted age-structured transmission models within a Bayesian framework to RSV hospital discharge data and added distinct immunized compartments to evaluate the impact. I discovered that live-attenuated vaccines were predicted to be the most effective immunization strategy in children under 5 years of age. Maternal immunization and long-lasting mAbs were predicted to protect the most vulnerable newborn infants but failed to provide substantial indirect effects on the overall pediatric population. A seasonal vaccination program at the country level would provide only a slight efficiency advantage over a year-round vaccination program. In conclusion, RSV causes substantial hospitalizations in young children and older adults, especially those residing in low socioeconomic areas. Prevention strategies can avert more than half of the total RSV hospitalizations in infants under 6 months of age. The onset of RSV epidemics follows a notable spatial pattern. Thus, initiation of RSV prophylaxis and vaccine administration can be based on local epidemic timing to efficiently prevent RSV hospitalizations in young children and older adults.
Recommended Citation
Zheng, Zhe, "Synthesizing Epidemiological Evidence of Respiratory Syncytial Virus to Predict the Potential Effectiveness of Three RSV Prevention Strategies" (2022). Yale Graduate School of Arts and Sciences Dissertations. 1079.
https://elischolar.library.yale.edu/gsas_dissertations/1079
