The algorithm, which was constructed using data from more than eight million people across England, uses key factors such as age, ethnicity and body mass index to help identify individuals in the UK at risk of developing severe illness.
It’s hoped that the risk prediction tool, known as QCOVID, will be used to support public health policy throughout the rest of the pandemic, in shaping decisions over shielding, treatment or vaccine prioritisation.
The research, published in The BMJ, was put together by a team of scientists across the UK, and has been praised for the depth and accuracy of its findings.
Data was drawn from 1,205 general practices in England, along with Covid-19 test results and Hospital Episode Statistics (HES) — a database containing details of all admissions, A&E attendances and outpatient appointments at NHS hospitals.
Six million patients were used to develop the prediction tool over a 97-day period, between 24 January to 30 April, while additional data from a further 2.2 million people validated the performance of the model over two separate periods during the first wave.
Known factors such as age, ethnicity, deprivation, body mass index, and a range of pre-existing conditions (comorbidities) helped estimate the probability and timing of hospital admission or death from Covid-19.
A total of 4,384 deaths from Covid-19 were reported in the development group. This number fell to 1,722 and 621 respectively for the first and second validation groups.
The five per cent of people predicted to be at greatest risk of Covid-19 accounted for 75 per cent of deaths over the 97-day study period, The BMJ paper shows. People in the top 20 per cent meanwhile accounted for 94 per cent of all Covid deaths.
Overall, the model was able to predict 73 per cent and 74 per cent of the variation in time to death from Covid-19 in men and women.
The researchers pointed out that the prediction tool estimates levels of risk, and does not provide explanations of which individual factors causally affect a person’s vulnerability to Covid-19.
They added that the model can also be recalibrated for different periods of the pandemic and has the potential to be updated regularly as the prevalence of Covid-19 within a population increases or decreases, or as the population’s behaviour changes.
“This study presents robust risk prediction models that could be used to stratify risk in populations for public health purposes in the event of a ‘second wave’ of the pandemic and support shared management of risk,” the researchers say.
“We anticipate that the algorithms will be updated regularly as understanding of Covid-19 increases, as more data become available, as behaviour in the population changes, or in response to new policy interventions.”
Mark Woolhouse, a professor of infectious disease epidemiology at the University of Edinburgh, said the research marked an “important landmark” in health authorities’ understanding of Covid-19, adding that it “paves the way for more emphasis on targeted, risk-based responses to managing the public health threat over the coming months and years”.
Dr David Strain, a senior clinical lecturer at the University of Exeter, described the research as “excellent” and said that it would help “to facilitate the protection of key workers that are at very high risk as we move into the second wave of Covid”.
“It has caveats though,” he added. “These data are derived from hospitalisations and deaths during the first wave. Behaviour (and thus risk of transmission) has changed significantly this time, some for the better (e.g. the widespread acceptance of masks) some for the worse.
“Notably, the shielding scheme that supported the extremely vulnerable has been discontinued. Any tool that was derived during this time of artificially lower exposure may under-estimate the risk of this population.”