Impact of COVID-19 on total excess mortality and geographic disparities in Europe, 2020–2023: a spatio-temporal analysis

Background COVID-19 dramatically reshaped mortality across Europe. This study aimed to assess its impact on total mortality in European countries taking into consideration the relationship with selected country-level socioeconomic indicators, non-pharmaceutical interventions, and vaccine uptake. Methods We obtained weekly mortality data from 2010 to 2023 from the Short-term Mortality Fluctuations data series, the annual population data from the United Nations archives, selected sociodemographic and economic indicators from the World Bank’s database, the stringency index and the percentage of the population fully vaccinated from Our World in Data. A quasi-Poisson regression model trained on pre-pandemic years was used to estimate expected number of deaths in 2020–2023 in 29 European countries. Excess mortality was estimated using three different metrics: excess deaths (number), relative excess mortality (% different from expected deaths) and age-standardized excess death rate per 10,000 population. The relationship between socioeconomic indicators and excess mortality was evaluated using linear regression models, which included both linear and quadratic terms for the predictors to account for possible non-linear relationships. Findings We estimated 1,642,586 excess deaths (95% confidence interval, CI: 1,607,161–1,678,010) across all countries over the four years (+8.0% compared to the expected number of deaths). Excess mortality was mainly concentrated in 2020–2022 (0.52 million excess deaths in 2020, 0.57 million in 2021 and 0.44 million in 2022), with no substantial excess (0.11 million) estimated for 2023. Over the period 2020–23, the highest number of excess deaths was estimated for Italy (227,736 deaths, +8.7%), Poland (223,735 deaths, +13.7%), and Germany (218,111 deaths, +5.6%), while the highest excesses in relative terms were in Bulgaria (72,328 deaths, +17.2%), Lithuania (23,813 deaths, +16.1%), and Slovakia (31,984 deaths, +14.9%). The age-standardised death rates ranged from 1.8 per 10,000 population in Sweden to 24.7 in Bulgaria. The percentage of the population living below the poverty line and the Gini index were significantly associated with an increased excess death rate, with p-values for the linear and quadratic terms being 0.003 and 0.003 for the Gini index, and 0.024 and 0.017 for the population living below the poverty line. Conversely, gross domestic product per capita (p-values for the linear and quadratic terms: <0.001, 0.003), health expenditure (0.001, 0.273) and the percentage of people fully vaccinated by the end of 2021 (<0.001, 0.989) or 2022 (0.001, 0.890) were inversely associated with excess death rate. No significant association was observed with population density and stringency index. Interpretation The observed geographic disparities in total mortality excess across Europe can be related to differences in socioeconomic contexts, as well as to suboptimal vaccine uptakes in some countries.