Regularised B-splines projected Gaussian Process priors to estimate the age profile of COVID-19 deaths before and after vaccine roll-out


Abstract in English

The COVID-19 pandemic has caused severe public health consequences in the United States. The United States began a vaccination campaign at the end of 2020 targeting primarily elderly residents before extending access to younger individuals. With both COVID-19 infection fatality ratios and vaccine uptake being heterogeneous across ages, an important consideration is whether the age contribution to deaths shifted over time towards younger age groups. In this study, we use a Bayesian non-parametric spatial approach to estimate the age-specific contribution to COVID-19 attributable deaths over time. The proposed spatial approach is a low-rank Gaussian Process projected by regularised B-splines. Simulation analyses and benchmark results show that the spatial approach performs better than a standard B-splines approach and equivalently well as a standard Gaussian Process, for considerably lower runtimes. We find that COVID-19 has been especially deadly in the United States. The mortality rates among individuals aged 85+ ranged from 1% to 5% across the US states. Since the beginning of the vaccination campaign, the number of weekly deaths reduced in every US state with a faster decrease among individuals aged 75+ than individuals aged 0-74. Simultaneously to this reduction, the contribution of individuals age 75+ to deaths decreased, with important disparities in the timing and rapidity of this decrease across the country.

Download