Table B.2.

Parameters of the model. The parameters of the model with their description and unit of measure. For each country and wave, the following assumptions have been made: ${\tau }_8={(1/{\tau }_1+1/{\tau }_2)}^{-1},p_9=p_7,{\tau }_9={\tau }_7$. This is motivated by the fact that these parameters are relative to the undetected branch, and thus cannot be inferred from data. Thus, we assumed that undetected cases would evolve with the same parameter values as detected cases not entering hospitals/ICUs. $\mathcal{N}$ represents the number of times that the social interaction parameter changes: $\mathcal{N}=13$ for Luxembourg and Sweden and $\mathcal{N}=16$ for Austria. The values of the manual fit of all the rates and probabilities for each country and wave are summarized in Table H.6. The values and dates of change for the social interaction parameter ${\rho }_n$ are reported in Table F.3 for Luxembourg, in Table F.4 for Austria and in Table F.5 for Sweden.

Parameter	Description	Units
${\rho }_n$	social interaction $(n=0,1,\dots ,\mathcal{N})$	adim.
β	average contact rate	days−1
α	(mean incubation period)−1	days−1
${\tau }_1$	(mean time in I)−1	days−1
${\tau }_2$	(mean time in Q)−1	days−1
${\tau }_3$	(mean time in H)−1	days−1
${\tau }_4$	(mean time in ICU)−1	days−1
${\tau }_5$	(mean time in Hl)−1	days−1
${\tau }_6$	(mean time in AICU)−1	days−1
${\tau }_7$	(mean time in NI)−1	days−1
${\tau }_8$	(mean time in A)−1	days−1
${\tau }_9$	(mean time in NII)−1	days−1
$p_1$	probability of $E\to I$	adim.
$p_2$	probability of $Q\to H$	adim.
$p_3$	probability of $H\to H_l$	adim.
$p_4$	probability of $\mathit{ICU}\to \mathit{AICU}$	adim.
$p_5$	probability of $\mathit{Hl}\to R_{\mathrm{II}}$	adim.
$p_6$	probability of $\mathit{AICU}\to R_{\mathrm{II}}$	adim.
$p_7$	probability of $N_I\to R_{\mathrm{II}}$	adim.
$p_9$	probability of $\mathit{NII}\to R_A$	adim.