July 8, 2020

expert reaction to paper from OpenSAFELY platform quantifying clinical and demographic risk factors for COVID19-related death

A report, published by Nature (previously posted as a preprint in May), quantifying clinical and demographic risk factors for COVID-19 related death.

Dr Peter Tennant, University Academic Fellow in Health Data Science, University of Leeds, and Fellow at the Alan Turing Institute, said:

“This large and important study confirms many of the observations we’ve seen elsewhere; that the risk of death from COVID-19 is higher in men, older people, Black and Asian people, people living in poorer areas, and people with pre-existing conditions.  Unfortunately, it can’t say whether this is due to a higher risk of contracting COVID-19, a higher risk of dying once someone has COVID-19, or both.  Nor can it say anything about why the risks are higher in people with these characteristics.

“Although the study presents a model of the ‘risk factors’ for death from COVID-19, and the authors draw some conclusions (e.g. about the reasons for the higher risk in Black and Asian people), the methods are not suitable for such interpretations.  This concern was raised by several epidemiologists when an earlier version of the research (i.e. a pre-print) was shared online.  In response, the authors now warn that the results should not be interpreted causally.  But without a causal interpretation, I’m not sure the results can tell us anything.  I would also warn against using the model to predict (i.e. identify) people at higher risk of death from COVID-19, because it was not subjected to the recommended levels of testing and validation.  It’s important that both of these limitations are widely recognised to avoid confusion and misuse.”

Prof Ben Neuman, Chair of Biological Sciences at Texas A&M University-Texarkana, and Visiting Associate Professor at the University of Reading, said:

“This extremely large study authoritatively confirms much of what smaller studies had proposed about the relative risk of death from COVID-19 based on age, biological sex, obesity, diabetes and other respiratory problems.

“This study only looks at risk of death, which simplifies the statistical problem considerably, but does not capture the range of other debilitating disease associated with COVID-19 even in groups that have a relatively low risk of death, such as children and younger women.

“I think the key to a study like this will be whether or how it translates into policy.  A risk is that if policies are aimed at limiting death, for example by keeping schools open, but do not take account of other serious economic and health consequences of prolonging the pandemic rather than ending it, would not serve Britain well in the long term.”

Dr Tom Wingfield, Senior Clinical Lecturer and Honorary Consultant Physician, Liverpool School of Tropical Medicine, said:

“The article by Liam Smeeth, Ben Goldacre, and team is a well conducted and important piece of research that uses readily available anonymised national data to provide vital insights into the clinical and social risk factors associated with dying from Covid-19 in the UK.

“The authors’ findings are consistent with others in demonstrating that the principal independent risk factors for death from Covid-19 in the UK are being: male, older, Black or Asian, and more deprived.  These risk factors may also be associated with a higher likelihood of having chronic illness, such as diabetes or lung, heart, or kidney disease, which are themselves associated with Covid-19 related death.  However, a key finding of Smeeth and colleagues’ research was that such illnesses and being more deprived only partially explained the association of being Black or Asian with increased rates of death from Covid-19.

“The broader take-home message underlying these findings is that the health of our nation and its citizens continues to be deeply influenced by social factors including poverty, race, and inequality, and Covid-19 is no exception.  It is not enough to identify these risk factors and do nothing; it is vital that we work together to address the root causes of ill health by improving social support, integrating health and social care, mitigating unemployment, reducing inequality, increasing coverage of sickness and disability insurance, and meeting the needs of underserved groups.  This has never been more pressing in the UK than right now in the wake of the combined social, economic, and health impacts of Covid-19 and national lockdown.”

Prof Keith Neal, Emeritus Professor of the Epidemiology of Infectious Diseases, University of Nottingham, said:

“This study replicates a lot of what we already know about COVID-19 mortality: well established risk factors are particularly increasing age, male sex, obesity (particularly BMI > 40), diabetes (poorly controlled), stroke/dementia and other less common serious co-morbidities.  The big difference to previous works is no association with hypertension compared to other studies.

“One of the main strengths of the study is the number of people included.

“Ethnicity data was missing for 26% patients which could influence the results although they come out similar to other studies.  Under representation from London is also an issue.

“The association of death with social deprivation and ethnicity may well reflect a greater risk of acquisition of infection with COVID-19 (from a variety of reasons).”

All our previous output on this subject can be seen at this weblink:

www.sciencemediacentre.org/tag/covid-19

‘OpenSAFELY: factors associated with COVID-19 death in 17 million patients’ by Elizabeth J. Williamson et al. was published in Nature on Wednesday 8 July 2020.

DOI: 10.1038/s41586-020-2521-4

https://www.nature.com/articles/s41586-020-2521-4

Original Preprint;

https://www.medrxiv.org/content/10.1101/2020.05.06.20092999v1

Declared interests

None received.