Dynamic SIR/SEIR-like models comprising a time-dependent transmission rate: Hamiltonian Monte Carlo approach with applications to COVID-19
Hristo Inouzhe,María Xosé Rodríguez-Álvarez,Lorenzo Nagar,Elena Akhmatskaya
DOI: https://doi.org/10.48550/arXiv.2301.06385
2023-01-16
Abstract:A study of changes in the transmission of a disease, in particular, a new disease like COVID-19, requires very flexible models which can capture, among others, the effects of non-pharmacological and pharmacological measures, changes in population behaviour and random events. In this work, we give priority to data-driven approaches and choose to avoid a priori and ad-hoc methods. We introduce a generalised family of epidemiologically informed mechanistic models, guided by Ordinary Differential Equations and embedded in a probabilistic model. The mechanistic models SIKR and SEMIKR with K Infectious and M Exposed sub-compartments (resulting in non-exponential infectious and exposed periods) are enriched with a time-dependent transmission rate, parametrized using Bayesian P-splines. This enables an extensive flexibility in the transmission dynamics, with no ad-hoc intervention, while maintaining good differentiability properties. Our probabilistic model relies on the solutions of the mechanistic model and benefits from access to the information about under-reporting of new infected cases, a crucial property when studying diseases with a large fraction of asymptomatic infections. Such a model can be efficiently differentiated, which facilitates the use of Hamiltonian Monte Carlo for sampling from the posterior distribution of the model parameters. The features and advantages of the proposed approach are demonstrated through comparison with state-of-the-art methods using a synthetic dataset. Furthermore, we successfully apply our methodology to the study of the transmission dynamics of COVID-19 in the Basque Country (Spain) for almost a year, from mid February 2020 to the end of January 2021.
Methodology,Applications,Computation