After consulting the study by Legrand et al. and your paper again, I agree that the weighting of the transition rates is consistent and correct. The difference is that the study by Legrand defines a case fatality ratio delta that is then used to compute the scaling factors delta_1 and delta_2. In your manuscript, you call delta_1 and delta_2 case fatality rate and this might not be correct. From the ordinary differential equations, one can directly calculate the case fatality rate for the hospitalized as gamma_DH*delta_2/(gamma_DH*delta_2 + gamma_IH*(1-delta_2)). For the parameters from Sierra Leone, that would be 88% and not 75%.

If you model your system as shown in Figure 2 of the Appendix, delta_2 indeed corresponds to the case fatality rate in hospitals. In this case, you should change your ordinary differential equations to match the model scheme in Figure 2 of the Appendix.

]]>– Changed one symbol (iota) to theta_1 in order to be more consistent with Legrand et al and slightly changed the wording of this parameter’s meaning.

– Corrected a typo in the compartmental flow diagram.

– Added the appendix, which is a technical note on the model equations addressing some comments from readers of the article. ]]>