Round 3

We have now completed Round 3.

In Round 3, scenarios broadly corresponded to a European Commission communication document, and were meant to cover a range of future COVID-19 burden scenarios (from “a diminished threat” all the way to “unmanageable winters/a new pandemic”).

With these scenarios, we were transitioning toward vaccination programme evaluations, envisioning a situation where COVID-19 vaccination has been integrated into routine vaccination programmes. Further, the uptake was chosen optimistically on purpose to demonstrate its potential effect, and it was assumed that new-generation vaccines were catching up with viral evolution. These types of scenario analysis are crucial for planning of such large-scale programmes that need strategic planning, security of funding and supplies etc.

• We request submissions by 16 September 2022
  • Scenario details
  • Submission information and key dates
  • Let us know if you need more time

Summary of changes for current modelling teams

In this round, we have made important modifications to what we ask from teams' model projection output. See full submission detail below.

• Projection horizon
  • We ask for projections over a period of at least two years and up to ten years
  • We encourage the longest possible projection horizon
• Age groups
  • We ask for submissions to include results for each of three age groups
  • If a breakdown by age is not possible, please continue to submit for all age groups combined
• Updated targets
  • New: vaccine doses administered
  • Unchanged: infections, hospitalisations, deaths
  • Removed: incident ICU; incident cases
• Updated fields
  • New: age_id
  • Modified: scenario_id
  • Removed: origin_date, target_end_date
• Submission
  • Please gzip your file (.csv.gz) before submitting it
  • If your file is larger than 100MB after gzipping it, please create one subfolder per target (contact us for more details or any trouble with this)

---

Round 3 scenarios

Notes * Vaccination as planned = Primary schedule + 2 booster doses ** Seasonal vaccination campaigns: Autumn = 15 September to 15 December; Spring = 15 March to 15 June	No further vaccination Vaccination as planned* No further vaccination	Annual vaccination Vaccination as planned* 2023 onwards: annual vaccination programme in autumn** Uptake higher in 60+ than <60 age groups	Bi-annual vaccination Vaccination as planned 2023 onwards: bi-annual (2 per year) vaccination in spring** and autumn** Uptake higher in 60+ than <60 age groups
Optimistic variant scenario 20% reduction in immunity against infection 0% reduction in immunity against severe outcome One new variant every 9 months First introduction: 1st October 2022	Scenario A	Scenario B	Scenario C
Pessimistic variant scenario 75% reduction in immunity against infection 20% reduction in immunity against severe outcomes One new variant every 9 months First introduction: 1st October 2022	Scenario D	Scenario E	Scenario F

Below we cover scenario details for:

• Vaccination campaigns
• New variants
• Additional assumptions

See elsewhere: data links for cross-European data sources.

---

Vaccination campaigns

We consider three scenarios for future vaccination campaigns.

• Baseline

  • Vaccination as planned, then:
  • No further vaccination

• Annual vaccination

  • Vaccination as planned, then:
  • 2023 onwards: annual vaccination programme:
    • Autumn: 15 September - 15 December

• Bi-annual vaccination

  • Vaccination as planned, then:
  • 2023 onwards: two vaccination programmes per year:
    • Spring: 15 March - 15 June
    • Autumn: 15 September - 15 December

Shared assumptions

We ask all modellers to include the following parameters:

• Vaccination as planned includes a full primary schedule and 2 booster doses

• Vaccination uptake is higher in 60+ than <60 age groups (for example, 75% / 15% respectively)

• There is sufficient vaccine supply

Assumptions left to modeller judgement

Modellers should use their own judgement and relevant literature if making assumptions about the following:

• Vaccination coverage

  • Current and planned levels of coverage
  • Differential vaccination coverage between 60+ and <60 age groups

• Vaccine effectiveness

  • Previous vaccinations' effectiveness against COVID-19 outcomes
  • See below for vaccine effectiveness against new variants

New variants

We look at optimistic and pessimistic scenarios for a new variant introduced every nine months.

• Optimistic

  • 20% reduction in immunity against infection
  • 0% reduction in immunity against severe outcome

• Pessimistic

  • 75% reduction in immunity against infection
  • 20% reduction in immunity against severe outcomes

Shared assumptions

We ask all modellers to include the following parameters:

• Variant introduction

  • One new variant every 9 months
  • First introduction: 1st October 2022
  • At start date, assume 0.1% of infections/cases due to new variant (at minimum 1 infection)
  • After that it should not drop below 0.1%, until the next variant arrives

• Variant characteristics

  • The basic reproduction number R0, or the inherent transmissibility of variants, is constant
  • The inherent severity of variants is constant

• Changes to immunity

  • Vaccine effectiveness against a variant that seeded >6 months age is assumed not to be affected by immune escape (from the variant that seeded >6 months ago). This resembles updated vaccine formulations
  • The same immunity of a variant induced against reinfection with same variant; variant immune escape is the same towards all other previous variants

Additional assumptions

Shared assumptions

• No new public health and social measures (non-pharmaceutical interventions / NPIs)
• No changes in demography
• No novel drugs that strongly impact burden

Assumptions left to modeller judgement

• The absolute value of vaccine effectiveness
• Waning immunity
• Contact rates in response to infection rates

---

Submission information

• Projections
• Key dates

Projections

Changes to formatting

• We ask that teams no longer include the following fields (columns):
  • origin_date (redundant with the file name)
  • target_end_date (redundant with the horizon field)
• Please gzip your file (.csv.gz) before submitting it
  • If your file is larger than 100MB even after gzipping it, please create one subfolder per target (contact us for more details or any trouble with this)

Targets

Teams can submit any of the below targets.

Target	target_variable	Notes
Incident infections	inc infection	Please only provide incident cases (inc case) if unable to model infections
Incident hospital admissions	inc hosp	We no longer ask for ICU (critical care) admissions
Incident deaths	inc death
Vaccine doses administered	inc dose	New target - essential to future calculations of vaccine cost-effectiveness between scenarios

Age groups

We ask for submissions to include results for each of three age groups, identified in a new column age_id. We suggest age brackets below.

Age group	age_id
Children (0-17 years)	child
Adult (18-59 years)	adult
Older population (60+ years)	older
Age breakdown not available	Do not include age_id column

• If your model is already age-structured and it is difficult to change the exact age-bands, please follow as closely as you can and let us know where your age-groups differ
• If a breakdown by age is not possible, please continue to submit for all age groups combined (and leave out the age_id column)

Scenario IDs

Please use the following codes to identify between scenarios in the scenario_id column, with one for each value (row) in your submission csv.

Scenario	scenario_id
Scenario A. No vaccination, optimistic variants	A
Scenario B. Annual vaccination, optimistic variants	B
Scenario C. Biannual vaccination, optimistic variants	C
Scenario D. No vaccination, pessimistic variants	D
Scenario E. Annual booster campaign, pessimistic variants	E
Scenario F. Biannual booster campaign, pessimistic variants	F

Model abstract

Please include an abstract for your model as part of your submission.

• Copy the Round 3 Abstract template
• Complete where you can - all comments are appreciated
• Save this to the model-abstracts/round-3 folder, adjusting the filename to MyTeam-MyModel.md

Key dates

The dates for Round 3 are:

	Round 3 date
Submissions due	September 16 2022
End date for fitting data (last day of epiweek)	September 10 2022
origin_date (first possible date of a daily simulation)	September 11 2022