Excess mortality in Italy in 2020 by sex and age groups accounting for demographic changes and temporal trends in mortality

Abstract

Background: Differences between total deaths registered during the COVID-19 pandemic and those registered in a previous reference period are the most frequently used measures of the pandemic effect. However, these measures do not consider demographic changes and temporal trends in mortality. In this study we estimated the excess mortality in 2020 in Italy considering demographic changes and temporal trends in mortality.

Methods: We used daily mortality and population data for the 2011-2019 period to estimate the expected deaths in 2020. Expected deaths were estimated, separately by sex, through an overdispersed Poisson regression model including calendar year and age group as covariates, a smooth function of the year's week, and the logarithm of the population as offset. The difference between observed and expected deaths was considered a measure of excess mortality.

Results: In 2020, 746,146 deaths occurred in Italy. We estimated an excess mortality of 90,725 deaths (95% CI: 86,503-94,914), which became 99,289 deaths after excluding January and February, when mortality was lower than expected. The excess was higher among men (49,422 deaths) than women (41,303 deaths) and it was mostly detected at ages ≥80 (60,224 deaths) and ages 65-79 (25,791 deaths), while among the population aged 25-49 and 50-64 we estimated an excess of 281 and 4764 deaths, respectively.

Conclusions: After considering demographic changes and temporal improvement in mortality the excess deaths in 2020 still remains above 90,000 deaths. More important, considering these factors, the excess at ages <80 years is revised upwards, while the excess at older ages is revised downwards.