25 Oct The Challenges of Understanding Covid-Related Risk: Part A
The COVINFORM project examines the impact of the COVID-19 pandemic across the EU and the UK. A key output of the project is the Risk Assessment Dashboard which allows users, such as researchers, academics and policymakers, to explore the COVINFORM Risk Score.
The COVINFORM Risk Score is a multidimensional index of risk built using data from a large range of sources, spanning different countries, different timespans and different collection protocols, with a particular focus on the intersection between risk and vulnerabilities (such as health, social and economic).
Creating the COVINFORM Risk Score has posed a variety of challenges. This is principally because all the different kinds of data used to create the score must be collected and processed in such a way that they accurately represent the information they are trying to capture, and must be directly comparable across geographic and temporal scales.
Trilateral Research led the development of the COVINFORM Dashboard and of the COVINFORM Risk Score, with data collection and processing undertaken by Trilateral’s DARSI team. In Part A of this article, we discuss what data the DARSI team used to calculate the CONVINFORM Risk Score. In Part B, we discuss the challenges faced, and how these were overcome.
What Data was collected for the Dashboard?
The Risk Assessment framework presented in the Dashboard is composed of four domains: Threat, Vulnerabilities, Consequences, and Resilience. Each of these domains are composed of multiple ‘indicators’, examples of which are given below. Indicators span epidemiological, social, economic and behavioural variables, and were defined following a deep literature review and expert consultations. The team mathematically combined these indicators to obtain the COVINFORM Risk Score, which provides an overall level of risk for each country.
- Threat: The risk object (i.e., SARS-CoV2) and its likelihood of development. This was calculated using a variety of indicators, such as:
- Disease prevalence
Higher case rates indicates higher risk, because the disease is highly transmissible.
- Population density
Higher population density indicates higher risk, because there is more chance of being exposed to and thus contracting COVID-19.
- Vaccination rate
Higher vaccination rate indicates lower risk, because vaccination lowers the likelihood of contracting severe COVID-19.
Lower temperature is believed to contribute to greater risk, because resultant behaviour (such as staying inside in areas of poor ventilation) may increase viral exposure.
- Disease prevalence
- Vulnerabilities: the presence of one or multiple vulnerabilities and how they contribute to the object at risk (health). Indicators included:
Areas with high proportions of people living in or at risk of poverty received higher risk scores. This is because poverty increases risks associated with COVID-19; for example, due to a higher prevalence of pre-existing health problems.
- Education level
A lower education level was associated with higher risk of severe COVID-19 across European countries. Thus, a lower education level corresponds to a higher level of risk.
- Digital skills
Low levels of digital skills tend to be more common in disadvantaged communities (e.g., the elderly and migrant populations), and can limit access to public health information. Thus, low levels of digital skills contribute to a higher level of risk.
- Trust in government
Areas that reported lower trust in their government received a higher risk score, because it is believed that trust in government influences compliance with public health measures.
- Consequences: The likelihood of negative consequences of the COVID-19 pandemic. Indicators included the number of COVID-19 deaths, income loss, lack of educational progress and food insecurity.
- Resilience: The ability to adapt and recover in the face of the pandemic. Indicators included government debt, economic performance, investment in healthcare and government corruption (which is known to negatively impact a country’s resilience).
Once the team had decided which indicators to use to calculate the COVINFORM Risk Score, they then had to collect and process the data. The challenges of this are discussed in Part B of this article.