By using time-varying excess mortality rates and comparing to COVID-19 death rates, one may adjust for mortality displacement. Our method to adjust for mortality displacement is intuitive. Here the peak of the excess mortality comes simultaneously with the peak of the Covid-19 pandemic. However, the absolute differences between the two curves depends on the level of the dry tinder effect, the level of the mortality displacement and the frequency of testing for the virus. The latter you can estimate and adjust for, while there is no method to calculate and adjust for the dry tinder effect and mortality displacement. We have argued that both effects should be completely incorporated in the calculations. However, when comparing estimates in Norway and Sweden, one must have in mind that the estimated RR would probably have been different if the pandemic had arrived at another time.
Furthermore, this analysis does not rely on statical modelling involving multiple assumptions. Calculations are transparent.
Statistical modelling has so far focused on extrapolating the trends before 2020
[16],
[17],
[18]. We show that the change in expected mortality and excess mortality after 12–18 months with high mortality of COVID-19, primarily depends on what happened after 2020 (mortality displacement).
Mutations of virus during the follow-up may confound the analyses. It is generally assumed that new mutations were less fatal, but more contagious
[22],
[23]. Our MRR estimates relies on an intricate interaction between the number of individuals been infected and natural immunity to COVID-19 in the population. If for example the
delta virus infected less people and were more fatal in 2020, infection with
omicron in 2021 may affect more people, but the
case fatality rate was lower, and this may yield a higher excess mortality as well as lower.