Abstract
Bronchiolitis has a high morbidity in children under 2 years old. Respiratory syncytial virus (RSV) is the most common pathogen causing the disease. At present, there is only a costly humanized monoclonal RSV-specific antibody to prevent RSV. However, different immunization strategies are being developed. Hence, evaluation and comparison of their impact is important for policymakers. The analysis of the disease with a Bayesian stochastic compartmental model provided an improved and more natural description of its dynamics. However, the consideration of different age groups is still needed, since disease transmission greatly varies with age. In this work, we propose a multivariate age-structured stochastic model to understand bronchiolitis dynamics in children younger than 2 years of age considering high-quality data from the Valencia health system integrated database. Our modeling approach combines ideas from compartmental models and Bayesian hierarchical Poisson models in a novel way. Finally, we develop an extension of the model that simulates the effect of potential newborn immunization scenarios on the burden of disease. We provide an app tool that estimates the expected reduction in bronchiolitis episodes for a range of different values of uptake and effectiveness.
Highlights
Bronchiolitis is a common lower respiratory tract infection (LRTI) that mainly affects children under 2 years old, with the greatest burden occurring in infants younger than months [1]
Because around 90% of bronchiolitis cases in children less than 2 years of age are treated in primary care offices and very few cases have a Respiratory syncytial virus (RSV) microbiological confirmation, the analysis of laboratory-confirmed cases of RSV bronchiolitis is less useful for incidence estimation due to underreporting
It is important to emphasize that these parameters represent features of bronchiolitis dynamics and contact patterns that do not depend on the number of infected children in previous weeks
Summary
The most commonly used models in the analysis of infectious disease counts are compartmental models, which divide the population being studied into different compartments according to disease status and describe the evolution of infection through changes in the number of individuals in each compartment. By allowing the w w w =1 transmission rate exp(rt ) to vary over time, the stochastic model provided an improved and accurate description of the pattern of disease. Held et al [25] proposed a stochastic model for the analysis of disease counts based on a Poisson (or negative binomial) model with two components, which describe endemic seasonal patterns and localized epidemics. Multivariate extensions of that model can be found in [26] for the joint analysis of multiple time series of counts, where each component corresponds to a geographical region j or a certain age group. Paul et al [33] extended this multivariate model by allowing the autoregressive parameters to depend on the age-group; that is, λ j and φ j for each time series. In these multivariate models, the autoregressive parameters λ and φ are not allowed to vary over time
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
