Modeling geographic vaccination strategies for COVID-19 in Norway

Vaccination was a key intervention in controlling the COVID-19 pandemic globally. In early 2021, Norway faced significant regional variations in COVID-19 incidence and prevalence, with large differences in population density, necessitating efficient vaccine allocation to reduce infections and severe outcomes. This study explored alternative vaccination strategies to minimize health outcomes (infections, hospitalizations, ICU admissions, deaths) by varying regions prioritized, extra doses prioritized, and implementation start time. Using two models (individual-based and meta-population), we simulated COVID-19 transmission during the primary vaccination period in Norway, covering the first 7 months of 2021. We investigated alternative strategies to allocate more vaccine doses to regions with a higher force of infection. We also examined the robustness of our results and highlighted potential structural differences between the two models. Our findings suggest that early vaccine prioritization could reduce COVID-19 related health outcomes by 8% to 20% compared to a baseline strategy without geographic prioritization. For minimizing infections, hospitalizations, or ICU admissions, the best strategy was to initially allocate all available vaccine doses to fewer high-risk municipalities, comprising approximately one-fourth of the population. For minimizing deaths, a moderate level of geographic prioritization, with approximately one-third of the population receiving doubled doses, gave the best outcomes by balancing the trade-off between vaccinating younger people in high-risk areas and older people in low-risk areas. The actual strategy implemented in Norway was a two-step moderate level aimed at maintaining the balance and ensuring ethical considerations and public trust. However, it did not offer significant advantages over the baseline strategy without geographic prioritization. Earlier implementation of geographic prioritization could have more effectively addressed the main wave of infections, substantially reducing the national burden of the pandemic.

• We have added a sentence to define "community" while it is first mentioned in Section 4.2.1.

2.2)
In paragraph between lines 287 and 291, please cite source and give a brief account of the methodology that led to the proportions of infections in each of the settings explored.Moreover, the Community social setting is not listed here and you only mention "other settings"; are these "other settings" the definition for the Community setting that was missing as it was first mentioned?
• We have added a citation showing that the data were collected by Norwegian Surveillance System for Communicable Diseases (MSIS).
• We have replaced the "other settings" by "community" to be consistent throughout the entire article.
3) The authors mention that contacts are modeled based on an unpublished social contact studies (105-106).
3.1) Later on, they explain that contact are uniformly distributed for agents within household, schools, universities and workplaces, which leaves only the Community setting (which was already vague) to follow any heterogeneous contact structure.I therefore find the statements in lines (105-106) misleading as it implies data integration of observed contact patterns to a greater degree than actually performed.Again, in lines 506 and 507, the authors mention that heterogeneous contact structure was employed in all four contact settings; this clashes with their explanation that contacts were uniformly distributed in the This needs to be addressed!
• We have added the explanation of contacts within households, schools, universities, and workplaces in Section 4.2.1.

3.
2) Contact structured is a fundamental element of ABM, and therefore cannot be obscured.Please, address -if there are valid reasons for this data to remain unpublished; -if it will be published soon and give enough information for us and the readers to be able to find it later on; -or include reasonable summary of these studies and their results in the SI (if it was already done and I missed it, please cite the appropriate section of the SI in your original lines).4) The approach to contact modeling was quite limiting, mainly because 4.1.1)if does not account for heterogeneities in the contact venues employed (different types of workplaces can have very different contact structures with terrible consequences for disease propagation like in accounts of meat processing plants in various countries); • We did not model heterogeneous contact patterns in "workplaces" in the ABM.This was due to the lack of available data that could inform such a level of granularity.We acknowledge this limitation, recognizing that our model assumes the same pattern for all types of interactions among colleagues.
4.1.2) it does not account for multiplicity of roles of a contact venue; for example: where does a restaurant or supermarket place in your venue classification?It is certainly a very relevant social setting and their workers can be effective super spreaders, but at the same time this is not a "workplace" for their customers, so are they not meeting with the workers in your contact model?Then, how could the model hope to represent super-spreading events in these and any other setting?And If so, fundamental mechanism of social interaction which are very important for disease propagation and play a large role in indirect protection conferred by good vaccination policies are not being captured at all and the true potential of ABMs is not being realized.
• In the ABM, interactions in "workplaces" pertain to transmissions between colleagues, constituting a smaller fraction, about one-tenth of all transmissions.• On the other hand, interactions such as those between customers and workers are categorized as "community", as transmissions occur during activities like grocery shopping.More than one-third of all transmissions occur in these community interactions.• We assumed that the number of contacts in the "community" follows a negative binomial distribution, representing potential super-spreading events.We have added this in Section 4.2.1.
4.1.3)the assignment of roles to agents (e.g.front line workers) in the ABM model is not described in its appropriated place (section 4.2.1) and only assumed in the Results section.
How are these roles assigned, and are any other types of roles assigned to the agents?How do the roles affect contacts?
• In the ABM, we do not assign higher contact rates to healthcare workers; rather, we assume that individuals in these roles share the same contact patterns as others.Our model only prioritizes vaccination as a primary focus.We have added our assumption in Section 4.2.1.
4.2) I do not hope for a rework of the models to improve contact mechanics, and I believe that your primary intent of documenting the efforts for policy design during that pandemic needs to be respected.However, I would expect improvement in the discussion of the limitations of the contact model employed.These limitations and others you already mentioned are shared by most models and this is a point that the community needs to be better informed.