"Risk of hospitalization of diagnosed COVID-19 cases during the pandemic: a time-series analsys to unveil short- and long-run dynamics"

Objectives: With the outbreak of the SARS-CoV-2 pandemic, the unprecedented rise in demand for hospital care brought health systems worldwide to the brink of collapse. The dynamics of the COVID-19 pandemic have alternated periods of high incidence with others of low incidence, making it difficult to separate short- and long-run relationship between the number of COVID-19 cases diagnosed and the demand for hospital beds. The aim of this study is to model the risk of hospitalization of diagnosed cases during the pandemic. Methods: Time series techniques are applied to evaluate the short- and long-run relationship between daily number of COVID-19 cases diagnosed and daily number hospital admissions. Drawing on daily Spanish data from 11 May 2020 to 20 March 2022, we propose an error correction model that introduces a short-run mechanism to adjust transitory disequilibrium in the long term. The impact of the Omicron variant and vaccination on the need for in-patient care are assessed. To examine changes during different life stages, the same analysis is performed by age group. Results: Dynamics between the number of positive cases and demand for hospital beds tends to the equilibrium in the long run, with 9% of any deviation being corrected after one period. Individuals aged between 50 and 69 have benefited most from the reduced severity of the Omicron variant, while vaccination had proved to be less effective for people aged over 80. Conclusions: Models discriminating between the short- and long-run dynamics provide health planners with a valuable demand forecasting tool which should be useful for developing both structural programs and emergency interventions.