Date of Award
Open Access Thesis
Master of Public Health (MPH)
School of Public Health
Respiratory syncytial virus (RSV) is a common virus infecting the respiratory system and can cause severe disease in vulnerable populations. It has been shown that the RSV epidemic in the United States is seasonal, peaking in late fall in Florida and a few months later in the upper Midwest. Although a seasonal trend has been described, it is still unclear if there are any spatial trends between states. For this paper, RSV laboratory data from the continental United States was used to model the transmission dynamics of RSV in each state. We conducted an explanatory analysis to investigate the presence of spatial autocorrelation in parameters describing RSV transmission between states using a two-stage approach. In stage one we estimated the parameters using a dynamic mathematical model. In stage two we utilized Bayesian methods, where we considered two modeling options: spatial independence and spatial correlation. To model spatial correlation, we included a state-specific spatial parameter w(si), where w(si) is assigned the intrinsic conditional autoregressive (CAR) model. The two models were compared to determine if spatial correlation is present in the data. The seasonal offset and amplitude of seasonality in transmission rate parameters both showed spatial autocorrelation in preliminary analyses. Spatial modeling, using stage two, was implemented for these two parameters. The spatial model showed that spatial correlation was present in the data for the seasonal offset and amplitude of seasonality parameters, suggesting the need to account for spatial correlations in future work.
Emanuele, Samantha, "Bayesian Spatial Modeling Of Respiratory Syncytial Virus Transmission In The United States" (2015). Public Health Theses. 1077.