BACKGROUND: Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic has rapidly spread globally. Due to different testing strategies, under-detection of positive subjects and COVID-19-related-deaths remains common. Aim of this analysis was to assess the real impact of COVID-19 through the analysis of 2020 Italian all-cause mortality data compared to historical series.
METHODS: We performed a retrospective analysis of 2020 and 2015-2019 all-cause mortality data released by the Italian National Institute for Statistics (ISTAT) for the time period 'January 1 - March 21'. This preliminary sample included 1,084 Italian municipalities showing at least 10 deaths during the above-mentioned timeframe and an increase in mortality of more than 20% as compared to the previous five years (2015-2019), with a resulting coverage of 21% of Italian population. The difference between 2020 observed and expected deaths (mean of weekly deaths in 2015-2019) was computed, together with mortality rate ratio (MRR) for each of the four weeks following detection of the first autochthonous COVID-19 case in Italy (23 February, 2020 - 21 March, 2020), as well as for this entire timeframe. Subgroup analysis by age groups was also performed.
RESULTS: Overall MRR was 1.79 [1.75-1.84], with an observed excess mortality of 8,750 individuals in the investigated sample, which in itself outweighs Italian Civil Protection report of only 4,825 COVID-19-related deaths across Italy, as of March 21. Subgroup analysis did not show any difference in mortality rate in '0-14 years' age group, while MRRs were significantly increased in older age groups, in particular in patients >75 years (MRR 1.84 [1.79-1.89]). In addition, week-by-week analysis showed a progressive increase in MRR during this period, peaking in the last week (15 March, 2020 - 21 March, 2020) with an estimated value of 2.65 [2.53-2.78].
CONCLUSIONS: The analysis of all-cause mortality data in Italy indicates that reported COVID-19-related deaths are an underestimate of the actual death toll. All-cause death should be seen as the epidemiological indicator of choice to assess the real mortality impact exerted by SARS-CoV-2, given that it also best reflects the toll on frail patient subsets (eg the elderly or those with cardiovascular disease).