Internet Explorer is no longer supported by Microsoft. To browse the NIHR site please use a modern, secure browser like Google Chrome, Mozilla Firefox, or Microsoft Edge.

New risk prediction model could help improve guidance for people shielding from COVID-19

Published: 23 June 2020

New research funded by NIHR could help to protect people most at risk from COVID-19, by developing a new way to predict who should be shielding from the virus.

Researchers at the University of Oxford and at the NIHR Oxford Biomedical Research Centre are working on a new risk prediction model, using routinely collected health data from 8 million adults across the UK.

As lockdown eases, this new approach could help to more accurately identify those most at risk from coronavirus, rather than using very broad categories such as age or the presence of certain health conditions.

The researchers are analysing patient data from a database called QResearch, to find patterns that predict who is most at risk. They are looking at factors including age, sex, ethnicity, deprivation, smoking status, body mass index, pre-existing medical conditions and current medications.

The aim is that algorithms developed from this data can then be used across different health and care settings - for example at a GP surgery to identify which patients would benefit most from shielding, or to help doctors and their patients decide how to reduce the risk for each individual.

The risk prediction model could also help to predict the possible impact of new policies on shielding and preventing infection, and help identify patients at high risk so that they can receive a vaccine quickly, if and when it becomes available.

The project was a commission from the Office of the Chief Medical Officer for England (CMO) to NERVTAG (New and Emerging Respiratory Virus Threats Advisory Group), funded by the COVID Fighting Fund, which distributed by the CMO and Chief Scientific Adviser using money from the Department of Health and Social Care.

The research team is led by Oxford University and includes researchers from the universities of Cambridge, Edinburgh, Queen Mary’s London, Swansea, Leicester and Nottingham with the London School of Hygiene and Tropical Medicine, University College London, NHS Digital and NHS England.

The research team are planning to use other datasets from across all four nations of the UK to validate their model and offer a unified approach to evidence-based risk stratification policy. 

Professor Julia Hippisley-Cox, Professor of Epidemiology and General Practice at Oxford University’s Nuffield Department of Primary Care Health Sciences, who is leading the study,  said:  

“Driven by real patient data, this risk assessment tool could enable a more sophisticated approach to identifying and managing those most at risk of infection and more serious COVID-19 disease. Importantly, it will provide better information for GPs to identify and verify individuals in the community who, in consultation with their doctor, may take steps to reduce their risk, or may be advised to shield.”

Chief Medical Officer for England, Professor Chris Whitty, said: 

“The level of threat posed by COVID-19 varies across the population, and as more is learned about the disease and the risk factors involved, we can start to make risk assessment more nuanced. When developed, this risk prediction tool will improve our ability to target shielding, if it is needed, to those most at risk.”


Related Article

Oxford leads development of risk prediction model for smarter COVID-19 shielding advice

Latest news