Individual COVID-19 fatality risk (and the consequences for universities)
Many European universities plan to re-open as physical spaces in the autumn of 2020. How much danger will that create? What is an individual’s risk of severe illness and death? How should university leaders take the health of their staff into account when planning for the next academic year (the economic issues were discussed recently in a VoxEU article by Peter Dolton)?
In our judgement, the most reliable source of information on the risk from COVID-19 is a 2020 “Open Safely” paper written by Elizabeth Williamson of the London School of Hygiene and Tropical Medicine, with a large number of colleagues. It is a piece of technical epidemiological writing. We make an attempt to convey that paper’s substance in a way that we hope might be helpful to those enrolled and working in universities and the senior management teams who head them.
The first point to make is that the biggest influence on a person’s COVID-19 risk is age. The risk from age is actually an extraordinarily accelerating one (log-linear, as an approximation). There have been grumpy and inaccurate discussions of this issue of age and risk in UK newspapers where lobby groups attacked the idea that chronological age is what matters. Instead, the risk associated with so-called health conditions, with being a man, obese, non-white, or with having diabetes has been emphasized by the media.
Further, our two papers recommending a release out of the lockdown of the young have led to many older people reaching out to us voicing their beliefs that because they feel fit, they face little or no risk from COVID-19. Unfortunately, that belief is incorrect. Age is the crucial risk to bear in mind as shown by Williamson et al. (2020) (see Figure 1).
Figure 1 Deaths from COVID-19 in England and Wales since January 2020
Note: Number of deaths registered in England and Wales where COVID-19 was mentioned in the certificate – up to 15/05/2020
Examining anonymized health records of 17 million UK citizens, the authors assess who has died from COVID-19 while statistically adjusting for a large number of influences on people’s risk. Standard regression equations are then used to calculates hazard ratios. The results are the following:
People between 50-60 face a fatality risk from COVID-19 that is approximately 14 times higher than the risk faced by those around the middle of the age band of 18-40.
For individuals between 60-70 who are in good health, the risk is 30 times higher.
The fatality risk of men is double that of women.
Individuals with middle-aged spread, holding other things constant, face an about 50% higher risk.
Dramatically obese individuals’ fatality risk is doubled.
The same holds for non-white individuals as they face an almost doubled risk.
Most standard kinds of health conditions also increase fatality risk, roughly 2 to 3 times, in general. The risk of individuals diagnosed with cancer is higher than the risk of men in general, but it is still lower than the risk of simply being old.
Hence, the most prominent risk indeed stems simply from age. In essence, this is due to young ‘animals’ stronger immune-response to most viruses also applying to COVID-19.
Some newspapers have confusingly run headlines proclaiming obesity as a major cause of COVID-19 deaths. However, while keeping ones BMI below 25 is certainly advisable and obese people undoubtedly face a higher risk of dying from COVID-19, the main reason why obese people are dying much more is because obese people tend to be older.
A second worrisome concern is the likelihood of misleading triumphalism in Europe over the coming summer. Premature optimism had lulled people already during the so-called Spanish Flu in the years of 1918-1919. Figure 2, taken from Taubenberger and Morens (2006), depicts the three waves of deaths from the Spanish Flu. The first wave, which at the time seemed very large, peaked in early summer of 1918. When it died down, policymakers took a more relaxed approach as they believed it to be over. Unfortunately, a second wave appeared in autumn of 1918 and the number of deaths this time was five times higher. A third wave followed in the Spring of 1919 but most deaths occurred in late autumn and winter just after the pandemic was believed to be over.
Figure 2 Deaths in the UK from the Spanish Flu 1918-1919
University presidents should keep this in mind when planning universities’ autumnal schedules. Our hunch is that misleading optimism about having beaten COVID-19 may appear by the end of the summer and lead to calls to open up UK universities in a really fundamental way. While it may be true that COVID-19 will really be over by then, it is also worryingly possible that we are in for some form of a re-run of 1918. As early hotspots for COVID-19 were places like meat packing plants and skiing resorts, which are both characterized by low temperatures, it is possible that the coronavirus microbe may prefer cold environments. Studying the Spanish Flu virus, He et al. (2013) conclude that every extra 10 degrees Celsius reduced the spread of this virus by nearly 50%. Of course, COVID-19 is a different virus which might behave differently and much is still to be learned about COVID-19.
A third concern is that the virus might mutate. However, there is not much that can be done about that except to hope for the best.
What should university faculty know and do?
First, faculty members might want to appreciate the fact that most students are young and hence face a rather miniscule risk – current estimates suggest that it may be about the same as the risk of dying in a UK road accident in a normal year. Second, university staff older than 50 may need to consider their own relative risk, and might prefer to avoid front-line contact with students until a vaccine is found. Third and more encouraging, current evidence suggests that even without a vaccine the great majority of Europe’s population will live through the pandemic without serious illness. Absolute risk for most people appears to be fairly low, and there even remains a slight chance that the virus will disappear more quickly than standard epidemiological modelling has predicted.
Still, there is a long and menacing path ahead of us for the coming autumn and winter, where it pays for each individual to know about their own relative risk. For university presidents and their management teams, it is essential to recognize that the COVID-19 virus is much more dangerous to older members of university staff. This time, age matters.
Editor’s note: A shorter version of this article will appear in June in the Times Higher magazine in the UK.
He, D et al. (2013), “Inferring the causes of the three waves of the 1918 influenza pandemic in England and Wales”, Proceedings of the Royal Society B – Biological Sciences 280, article number 20131345.
Oswald, A J, N Powdthavee (2020a), “The case for releasing the young from lockdown: A briefing paper for policymakers”, University of Warwick, IZA paper 13113.
Oswald, A J, N Powdthavee (2020b), “Age, death risk, and the design of an exit strategy: A guide for policymakers and for citizens who want to stay Alive”, University of Warwick, IZA paper 13221.
Lourenco, J et al. (2020), “Fundamental principles of epidemic spread highlight the immediate need for large-scale serological surveys to assess the stage of the SARS-CoV-2 epidemic”, unpublished paper, Oxford University.