Abstract
Background: Declining vaccine confidence post-pandemic necessitates transparent estimates of vaccination impact to regain public trust. Mathematical modelling offers tools to estimate both direct and indirect impacts on mortality and healthcare burden, thereby informing vaccination strategies. This study retrospectively assesses the impact of COVID-19 vaccination on hospitalizations, intensive care units (ICU) admissions, and deaths in Slovenia in 2021, employing an extended Susceptible-Exposed-Infected-Recovered model (SEIR).
Methods: We utilized Slovenian COVID-19 national surveillance data from January to December 2021, encompassing cases, vaccination coverage, hospitalizations, ICU admissions, and deaths, to fit an age-stratified extended SEIR model. This model incorporates compartments for both vaccinated and unvaccinated populations and comprises five models, linked by a mixing matrix, collectively modelling the epidemic dynamics of 5 age groups: 0-24, 25-44, 45-64, 65-74, and over 75. This allows for exploration of vaccine impact across scenarios with varying vaccination coverage by age groups.
Results: With 56% overall coverage in 2021, vaccination alone led to a 70% decrease in hospitalizations (2,507 vs. 752 per 100,000 population), a 74% decrease in ICU admissions (538 vs. 138 per 100,000 population) and a 70% decrease in COVID-19-related deaths (461 vs. 140 per 100,000 population). A population-wide vaccination strategy led to an additional 29% decrease in deaths in the over 65 years old group compared to a scenario vaccinating this high-risk group only.
Conclusions: The COVID-19 vaccination program in Slovenia significantly reduced the burden of the pandemic in terms of preventable deaths. Support for national and EU surveillance systems is needed to improve modelling capacities, which are essential to guide public health interventions. When relevant, modelling the indirect effects of vaccination on the unvaccinated population regarding disease transmission is relevant for prioritizing age groups for vaccination. Further research on social mixing patterns is warranted to improve such estimates.