A model that can calculate a person’s risk of becoming infected and then seriously ill due to COVID-19 has been shown to accurately estimate risk during the first wave of the pandemic in England.
The research is funded by the National Institute for Health Research, led by the University of Oxford and co-led by Queen’s University Belfast along with researchers from the universities of Cambridge, Edinburgh, Swansea, Leicester, Nottingham and Liverpool with the London School of Hygiene & Tropical Medicine, Queen Mary University of London, University College London, the Department of Health and Social Care, NHS Digital and NHS England.
The model, developed using routine anonymised data from more than eight million adults in 1,205 general practices across England, uses a number of factors such as a person’s age, ethnicity and existing medical conditions to predict their risk of catching COVID-19 and then dying or being admitted to hospital.
More nuanced information
The model has the potential to provide doctors and the public with more nuanced information about risk of serious illness due to COVID-19, and to help patients and doctors reach a shared understanding of risk, within the context of individual circumstances, risk appetite and the sorts of preventative measures people can take in their daily lives.
Throughout the pandemic, new information has emerged about what factors may influence whether someone is seriously affected by COVID-19, creating an opportunity to develop a detailed risk prediction model.
Researchers at the University of Oxford and collaborators across the UK have used anonymous data from primary care, hospitals, COVID-19 test results and death registries to determine which factors were associated with poor outcomes during the first wave of COVID-19.
This data was used to create a risk prediction model – QCovid – that provides a weighted, cumulative calculation of risk using the variables associated with poor COVID-19 outcomes.
The factors incorporated in the model include age, sex, ethnicity, level of deprivation, obesity, whether someone lived in residential care or was homeless, and a range of existing medical conditions, such as cardiovascular disease, diabetes, respiratory disease and cancer.
Two independent sets of anonymised data
This model was then tested in two independent sets of anonymised data, from January to April 2020 and from May 2020 to June 2020, to find out whether it accurately predicted severe outcomes due to COVID-19 during the first wave of the pandemic in England.
The research results, published in the BMJ, showed that the model performed well in predicting outcomes. People in the dataset whose calculated risk put them in the top 20% of predicted risk of death accounted for 94% of deaths from COVID-19.
Lead researcher Professor Julia Hippisley-Cox, a general practitioner and professor of clinical epidemiology and general practice at University of Oxford’s Nuffield Department of Primary Care Health Sciences, said: “Risk assessments to date have been based on the best evidence and clinical expertise, but have focused largely on single factors.
"The QCovid risk model provides a much more nuanced assessment of risk by taking into account a number of different factors that are cumulatively used to estimate risk.
“This model will help inform clinical advice so that people can take proportionate precautions to protect themselves from COVID-19.”
The researchers plan to regularly update their model as levels of immunity change, more data become available and behaviour in the population changes, so that the model could also be used to support risk stratification for public health purposes as the infection rate changes over time.