Research led by Professor Julia Hippisley-Cox in the University of Oxford’s Nuffield department of primary care health sciences, with collaborators across the UK, found that there are several health and personal factors which, when combined, could mean someone is at a higher risk from COVID-19. These include characteristics like age, ethnicity and BMI, as well as certain medical conditions and treatments.
The team turned their research into a risk prediction model called QCovid, which has been independently validated by the UK Office for National Statistics. It is thought to be the only COVID-19 risk prediction model in the world to meet the highest standards of evidence.
The work was commissioned by England’s chief medical officer Chris Whitty and funded by the National Institute of Health Research. Details of the development and validation of the tool have been published and the model has been fully published for transparency.
NHS Digital have now used this model to develop a population risk assessment. The risk assessment predicts on a population basis whether registered patients with a combination of risk factors may be at more serious risk from COVID-19, enabling the government to prioritise them for vaccination, and provide appropriate advice and support. These individuals will be added to the Shielded Patient List on a precautionary basis and to enable rapid vaccination.
This assessment is made possible for the first time by utilising the QCovid model from the Oxford-led team and emerging evidence about the impact of Covid-19 on different groups and who could be most vulnerable, which means further steps can be taken to protect those most at risk.
Up to 1.5m patients have been identified to date. Approximately 700,000 will have already been vaccinated as part of the over-70s cohort, and an additional 800,000 adults between 19 and 69 years will now be prioritised for a vaccination.
Hippisley-Cox said: “The QCovid model, which has been developed using anonymised data from more than 8m adults, provides nuanced assessment of risk by taking into account a number of different factors that are cumulatively used to estimate risk including ethnicity. The research to develop and validate the model is published in the British Medical Journal along with the underlying model for transparency. This will be updated to take account of new information as the pandemic progresses. I’m delighted that less than a year after being funded by the NIHR, the model is now being used to help protect people at most risk from COVID-19.”
Fred Kemp, deputy head of life sciences at Oxford University Innovation, said: “As a further example of how the University of Oxford is at the forefront of combatting the pandemic, OUI is proud to have supported the development and implementation of QCovid as a highly validated, evidence-based risk prediction tool that will enable prioritised delivery of vaccines to those most in need.”
Deputy chief medical Officer for England Dr Jenny Harries said: “For the first time, we are able to go even further in protecting the most vulnerable in our communities. This new model is a tribute to our health and technology researchers. The model’s data-driven approach to medical risk assessment will help the NHS identify further individuals who may be at high risk from COVID-19 due to a combination of personal and health factors. This action ensures those most vulnerable to COVID-19 can benefit from both the protection that vaccines provide, and from enhanced advice, including shielding and support, if they choose it.”
QCovid was developed using the QResearch database of anonymised electronic health records, a collaboration between Hippisley-Cox’s team in Oxford and primary are computer systems provider EMIS Health. The model included data from primary care, hospitals, COVID-19 test results and death registries, and was informed by a significant amount of patient engagement. It is the latest in a series of risk prediction models developed through the collaboration, which are widely used by healthcare practitioners to identify patients at risk of serious illness including cardiovascular disease, stroke, cancer and diabetes.
Commenting on the roll-out, Dr Shaun O’Hanlon, chief medical officer at EMIS, said: "EMIS is proud to have supported this important piece of research, which will enable the NHS to protect more vulnerable people, more quickly, from COVID-19. We thank all of the GP practices who have contributed anonymised patient data to the QResearch database in the 15 years-plus it has been in existence."
The independent validation from the Office of National Statistics is considered the ‘gold standard’ in quality assurance. The ONS has shown that the model performs in the ‘excellent’ range, and accurately identifies patients at highest risk from COVID-19. This shows the model is robust and meets the highest standards of evidence.
The development of the QCovid® model involved researchers from the universities of Oxford, Cambridge, Edinburgh, Swansea, Leicester, Nottingham and Liverpool with the London School of Hygiene & Tropical Medicine, Queen’s University Belfast, Queen Mary University of London and University College London. It was supported by the NIHR Oxford Biomedical Research Centre.
In related work from the University of Edinburgh, the QCovid model has been validated for use in the Scottish population.
An external validation of the QCovid risk prediction algorithm for risk of mortality from COVID-19 in adults: national validation cohort study in England
Vahé Nafilyan, Ben Humberstone, Nisha Metha, Ian Diamond, Luke Lorenzi, Piotr Pawelek, Ryan Schofield, Jasper Morgan, Paul Brown, Ronan Lyons, Aziz Sheik, Julia Hippersley-Cox
Pre-print published on 22 February 2021
To externally validate a risk prediction algorithm (QCovid) to estimate mortality outcomes from COVID-19 in adults in England.
Population-based cohort study using the ONS Public Health Linked Data Asset, a cohort based on the 2011 Census linked to Hospital Episode Statistics, the General Practice Extraction Service Data for pandemic planning and research, radiotherapy and systemic chemotherapy records. The primary outcome was time to COVID-19 death, defined as confirmed or suspected COVID-19 death as per death certification. Two time periods were used: (a) 24thJanuary to 30thApril 2020; and (b) 1st May to 28th July 2020. We evaluated the performance of the QCovid algorithms using measures of discrimination and calibration for each validation time period.
The study comprises 34,897,648 adults aged 19-100 years resident in England There were 26,985 COVID-19 deaths during the first time-period and 13,177 during the second. The algorithms had good calibration in the validation cohort in both time periods with close correspondence of observed and predicted risks. They explained 77.1% (95% CI: 76.9% to 77.4%) of the variation in time to death in men in the first time-period (R2); the D statistic was 3.76 (95% CI: 3.73 to 3.79); Harrell's C was 0.935 (0.933 to 0.937) Similar results were obtained for women, and in the second time-period In the top 5% of patients with the highest predicted risks of death, the sensitivity for identifying deaths in the first time period was 65.9% for men and 71.7% for women. People in the top 20% of predicted risks of death accounted for 90.8% of all COVID-19 deaths for men and 93.0% for women.
The QCovid population-based risk algorithm performed well, showing very high levels of discrimination for COVID-19 deaths in men and women for both time periods. It has the potential to be dynamically updated as the pandemic evolves and therefore, has potential use in guiding national policy.
National Institute of Health Research
University of Oxford material