A time varying parameter regression model to investigate the relationship between intensive care occupancies and confirmed COVID-19 deaths in European NUTS-2 Regions

The impact of the COVID-19 pandemic varied significantly across different countries, with important consequences in terms of population health status and medical resources allocation. In this paper, to investigate the relation between the occupancy of intensive care units by COVID-19 patients and the number of confirmed deaths through time and space, we apply a Bayesian approach for multivariate time series. The model provides a flexible framework for the analysis of time series data, allowing the analysis of different features of the series, such as spatial correlations, time varying parameters and clustering. We evaluate the effect of intensive care units occupancy on the death counts recorded at regional level for several European countries in the period from March 2020 to April 2021.