Analysis of the excess mortality and factors associated with deaths from COVID-19 versus other causes in Central Tuscany (Italy) in 2020

OBJECTIVES: to assess the extent of the excess mortality from all causes in 2020 compared to 2015-2019 in Central Tuscany (Italy) as a proxy to estimate COVID-19-related excess mortality and to identify demographic and clinical differences between subjects who died from COVID-19 and those who died from other causes in 2020. DESIGN: descriptive analysis of the temporal trend of general mortality. SETTING AND PARTICIPANTS: the study population is represented by the 1.6 million residents living in the territory of the Central Tuscany Healthcare Authority in Central Italy, i.e., little less than half of the population of Tuscany, in an area of just over one fifth of the entire region, where the provinces of Florence, Pistoia, and Prato are comprised. MAIN OUTCOME MEASURES: using the Italian National Resident Population Registry (ANPR) as a source of mortality data, standardized mortality ratios with 95% confidence intervals were calculated to compare the number of deaths in 2020 with the number of deaths expected on the basis of mortality data from 2015 to 2019. Furthermore, after record linkage with data from the integrated surveillance of cases of SARS-CoV-2 virus infection and with the MaCro dataset of comorbidities, the characteristics of subjects who died from COVID-19 were compared with those of patients who died from other causes using a multivariate logistic regression model; odds ratios with 95% confidence intervals were calculated. RESULTS: a statistically significant excess mortality was observed during the first pandemic wave in March and April, and during the second wave in the fall; it ranged between +9% in March and +51% in November. On the contrary, in January, February, and May, all-cause mortality was significantly lower than in previous years. The male gender, dyslipidaemia, and dementia were positively associated with death from COVID-19 rather than from all other causes. On the contrary, heart failure and recent tumours were more represented among deaths from other causes. CONCLUSIONS: much of the over-mortality observed in spring is attributable to the harvesting effect COVID-19 exerted on a segment of population with serious underlying chronic conditions and who in the previous months had survived a mild winter and a flu season of medium intensity. In the second pandemic wave, in autumn, the impact of both direct and indirect effects of COVID-19 was substantially higher. Consistently with the available evidence, death from COVID-19 was related to the male gender and to clinical conditions such as dyslipidaemia and dementia.