scenarios.Rmd

---
title: "Scenarios"
output:
  html_document:
    toc: false
---

```{r setup, include = FALSE}
knitr::opts_chunk$set(echo = FALSE, 
                      message = FALSE, warning = FALSE)
source("code/load/scenarios.R")
```

Information about past and present rounds of scenarios is below. For modellers participating in each round,  please refer to the [Wiki](https://github.com/covid19-forecast-hub-europe/covid19-scenario-hub-europe/wiki/Scenario-rounds) for the most up to date details. For more general information on our approach to setting scenarios, read more in the [About](about.html) section.

# {.tabset}

## Round 5 {.tabset .tabset-pills}

### Scenarios

Round 5 projections begin from `r scenarios[["round-5"]][["origin_date"]]`. 

```{r scenario-round-5, results = 'asis'}
cat(paste0("_", scenarios[["round-5"]][["scenario_caption"]], "_"))
cat("\n\n")
cat(scenarios[["round-5"]][["table"]])
```

### Assumptions

See our [Github Wiki](https://github.com/covid19-forecast-hub-europe/covid19-scenario-hub-europe/wiki/Round-5) for detailed guidance on the assumptions that modellers used in creating scenario projections.

### Findings

See the [full report](https://covid19scenariohub.eu/report5.html) for our findings from Round 5.

## Round 4 {.tabset .tabset-pills}

### Scenarios

Round 4 projections begin from `r scenarios[["round-4"]][["origin_date"]]`. 

```{r scenario-round-4, results = 'asis'}
cat(paste0("_", scenarios[["round-4"]][["scenario_caption"]], "_"))
cat("\n\n")
cat(scenarios[["round-4"]][["table"]])
```

### Assumptions

See our [Github Wiki](https://github.com/covid19-forecast-hub-europe/covid19-scenario-hub-europe/wiki/Round-4) for detailed guidance on the assumptions that modellers used in creating scenario projections.

### Findings

See the [full report](https://covid19scenariohub.eu/report4.html) for our findings from Round 4.

## Round 3 {.tabset .tabset-pills}

### Scenarios

Round 3 projections begin from `r scenarios[["round-3"]][["origin_date"]]`. 

```{r scenario-round-3, results = 'asis'}
cat(paste0("_", scenarios[["round-3"]][["scenario_caption"]], "_"))
cat("\n\n")
cat(scenarios[["round-3"]][["table"]])
```

### Assumptions

See our [Github Wiki](https://github.com/covid19-forecast-hub-europe/covid19-scenario-hub-europe/wiki/Round-3) for detailed guidance on the assumptions that modellers used in creating scenario projections.

### Findings

See the [full report](https://covid19scenariohub.eu/report3.html) for our findings from Round 3.

## Round 2 {.tabset .tabset-pills}

### Scenarios

Round 2 projections begin from `r scenarios[["round-2"]][["origin_date"]]`. 

```{r scenario-round-2, results = 'asis'}
cat(paste0("_", scenarios[["round-2"]][["scenario_caption"]], "_"))
cat("\n\n")
cat(scenarios[["round-2"]][["table"]])
```

### Assumptions

See our [Github Wiki](https://github.com/covid19-forecast-hub-europe/covid19-scenario-hub-europe/wiki/Round-2) for detailed guidance on the assumptions that modellers used in creating scenario projections.

### Findings

See the [full report](https://covid19scenariohub.eu/report2.html) for our findings from Round 2.

## Round 1 {.tabset .tabset-pills}

### Scenarios

Round 1 projections begin from `r scenarios[["round-1"]][["origin_date"]]`. 

```{r scenario-round-1, results = 'asis'}
cat(paste0("_", scenarios[["round-1"]][["scenario_caption"]], "_"))
cat("\n\n")
cat(scenarios[["round-1"]][["table"]])
```

### Assumptions {.tabset}

#### Booster campaign

We consider scenarios for a second booster vaccination campaign among the **population aged 60+** in each country. This vaccination would typically be a fourth dose, following a completed course of two doses plus a first booster dose. We prescribe a relative reduction in vaccination coverage among this group, where by **15th December**, uptake of the second booster reaches **50% of coverage** achieved by the first booster campaign.

In the scenarios we vary only the timing of the booster campaign:

-   Slow summer campaign: uptake starts **15th June**

-   Fast autumn campaign: uptake starts **15th September**

We note that some countries have already started a second booster campaign. 

- For nearly all countries, please **continue to use the scenarios as they are**. The scenarios ask for modellers to include a second booster dose open to ages 60+. In most countries currently offering a second booster, these scenarios are an expansion of the eligible age group

- The exceptions are **Greece and the Netherlands**, where second booster vaccination is already open to ages 60+. In these two countries:
   - In scenarios A & C (uptake starts in June), please model booster vaccination uptake among the 60+ **as it is continuing now**. 
   - In scenarios B & D (uptake starts in September), please model a **third booster (fifth dose)**, with otherwise the same parameters (uptake among 60+, starting 15 September, coverage reaching 50% of first booster dose by December)


##### Assumptions about the booster campaign

###### Shared assumptions

All modellers should include the following parameters when modelling the booster campaign:

-   Second booster is only recommended for population aged \>=60 years

-   Uptake among the target population reaches 50% of coverage achieved by the first booster campaign among that population

-   Uptake at 50% of the target population is reached by 15th December

###### Assumptions left to modeller judgement

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

-   The existing level of vaccination coverage reached in the first booster campaign

-   Vaccine effectiveness against both infection and severe disease, for both previous vaccinations and the second booster

#### Waning immunity

Waning immunity means protection against new COVID-19 infection. We take a similar general approach to waning immunity as in Round 0 (Euro Scenario Hub) and Round 13 (US Scenario Hub). We specify that immunity wanes to a **60% reduction** from baseline levels. The baseline is the level of protection reported immediately after exposure (vaccination or infection).

In the scenarios we vary the median time taken to reach this reduction of immunity:

-   Optimistic slow waning: a median time of **8 months** for immunity to wane by 60% from baseline

-   Pessimistic fast waning: a median time of **3 months** for immunity to wane by 60% from baseline

##### Assumptions about waning immunity

###### Shared assumptions

All modellers should include the following parameters when modelling waning immunity:

-   *New to this round*

    -   Immunity wanes by 60% (i.e. immunity after 3 or 8 months is 40% of baseline) across all scenarios

    - Waning against severe disease: this is now left to modeller judgement with a recommendation to include waning against severe disease - see below

-   *As the pilot round (text adapted from US Scenario Hub)*

    -   We prescribe the relative reduction in protection against infection after the waning period, where comparison is to the levels observed immediately after natural infection or vaccination

###### Assumptions left to modeller judgement

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

-   Waning of protection against severe disease
    - We strongly recommend including a **decrease of 20% protection within 3 months**. This is a combined estimate based on various recent studies, [see example](https://view-hub.org/sites/default/files/2022-05/COVID19_VEStudies_ForestPlots_Delta_Omicron.pdf).
    - However, this is up to modeller judgement and teams should continue to produce results even if they are unable to include waning against severe disease

-   The absolute baseline level of protection against infection or severe disease after exposure

-   Whether the baseline level of protection varies by the source of immunity (vaccination or natural infection)

-   Whether or how new exposure to infection or vaccination during the waning period increases the level of immunity. Teams can choose to bump individuals to a higher level of protection after repeat exposures (where exposure is vaccination or infection), but waning would still occur on a 3 to 8 month timescale after each new exposure

-   The distribution of waning immunity, including

    -   the shape and rate at which immunity wanes, as long as the median is 3 or 8 months

    -   whether immunity reaches a plateau or continues to wane after that time

For additional background on waning immunity, we suggest exploring the scenarios for [Round 0](https://github.com/covid19-forecast-hub-europe/covid19-scenario-hub-europe/wiki/Round-0) or the more extensive documentation for the [US Scenario Hub Round 13](https://github.com/midas-network/covid19-scenario-modeling-hub#waning-immunity).

#### Additional shared assumptions

We also ask modellers to assume the following:

-   No new variant of concern

-   No novel vaccine types or novel drugs that strongly impact burden

### Findings

See the [full report](https://covid19scenariohub.eu/report1.html) for our findings from Round 1.

## Round 0 {.tabset .tabset-pills}

### Scenarios

We ran a pilot round ("Round 0") over March-May 2022. Round 0 projections begin from `r scenarios[["round-0"]][["origin_date"]]`. 

```{r scenario-round-0, results = 'asis'}
cat(paste0("_", scenarios[["round-0"]][["scenario_caption"]], "_"))
cat("\n\n")
cat(scenarios[["round-0"]][["table"]])
```

### Assumptions

The details for this round identically match those of the [US Scenario
Hub](https://covid19scenariomodelinghub.org/viz.html), applied to the [32
countries](https://github.com/covid19-forecast-hub-europe/covid19-scenario-hub-europe/blob/main/data-locations/locations_eu.csv) of the European Hub.

-   [Round 0 overview](https://docs.google.com/presentation/d/1MiQsN0p-nvDF8km-OLjHOaEAqxMDwmxlOuDPu3POxnk/edit?usp=sharing)
-   [Full scenario details - US  Hub](https://github.com/midas-network/covid19-scenario-modeling-hub#round-13-scenarios)

### Findings

In Round 0:

- 20 independent teams joined calls and/or submitted models

- 8 teams started working on Round 0; 5 teams contributed results

- 5 countries had >2 model projections

As it was a pilot round, we provide only general, broad conclusions for Round 0.

- Scientific interpretation
   - Results were consistent with the speed of waning protection against infection as a more significant factor in future outbreaks, compared to a new variant with some immune escape
   - This matched with similar results from the US Hub Round 13

- Policy relevance
   - This indicates greater consideration for the timing of vaccination programmes relative to waning protection, rather than as a response to new variant introductions

- Interpret with caution:
   - Models were intended as experimental for the pilot round, and we found substantial variation between models as well as between countries and scenarios
   - Results may be biased by both submission constraints and scenario confounding