Background and objective:
First COVID-19 case was detected in Portugal on March 2020. More than 25,000 COVID-19 deaths have been reported, but it may not reflect the burden of the pandemic, especially, when competing factors occur simultaneously. We aimed to estimate all-cause mortality attributable to COVID-19 by SARS-CoV2 variant of concern, influenza and extreme temperatures (ET) during pandemic period in Portugal.

Methods:
To estimate mortality attributable to COVID-19, influenza and extreme temperatures, we used an adaptation of the FLUMOMO model. Considering the weekly number of all-cause deaths as the dependent variable, an additive Poisson regression model adjusted for overdispersion, trend and seasonality, was fitted with Influenza activity (Goldstein index), new COVID-19 hospitalizations adjusted for concern variants (Pre-Delta; Delta; Omicron BA 1 and 2; and Omicron BA 4 and 5), and ET as independent variables. We estimated the mortality attributable fraction of each independent variable between 2020-W10 and 2022-W38. We performed the analysis for all population and age groups.

Results:
We estimated 43,766 deaths above the expected, 36,810 attributable to COVID-19, 544 to influenza and 6412 to ET. 60%, 15%, 11% and 15% of COVID-19 attributable mortality was due to, respectively, Pre-Delta, Delta, Omicron BA 1 and 2, and Omicron BA 4 and 5 variants. The mortality estimated as attributable to COVID-19 is higher than the reported officially as COVID-19 specific mortality, which was expected because we estimate mortality directly and indirectly attributable to COVID. The highest mortality attributable to COVID-19 was estimated in the age group ≥ 80 years. These results are provisional.

Conclusion:
This analytical approach allows a detailed assessment of the impact of the COVID-19 pandemic, direct and indirect factors, adjusting it for competing risks for excess deaths like influenza and ET. We were able to estimate the fraction of mortality attributable to SARS-CoV2 variant of concern.