"We are exposed to risk every day from a wide variety of activities, substances, and behaviours. That undercooked chicken, crossing the road or cycling to work, skiing or surfing, drinking too much. Covid adds another risk, and we need to assess the risk to ourselves in the same way. Humans are not good at assessing risk; we tend to underestimate or overestimate. We often rely on others for advice, and that may be helpful or unhelpful depending on how scientific we are. Risk is actually all about numbers. It can be calculated, and risks than then be compared; that often helps us put risk in perspective.
Generally, we expect risk to be presented as a comparison with normal activity or health, such as ‘five times the risk’, or as an absolute risk of an event such as ‘one in a hundred risk of dying’. Once we have numbers, we can then work out ways to reduce risk. For example, if we know that keeping distance reduces risk by 75%, or wearing an FFP2 mask reduces exposure by 95%, we should be able to work out the effect on our overall risk.
Humans are not good at assessing risk; we tend to underestimate or overestimate.
We have heard lots of statistics about risk from Covid-19, and they are not always helpful. Being told that Black ethnicity doubles risk doesn’t mean much unless we know what the risk was in the first place. Doubling a fatality risk of one in a million is still an extremely low risk. Doubling a fatality risk of one in ten is very serious.
Importantly, we noted at the beginning that age was a major factor. Once studies were published, particularly the OpenSAFELY study (Williamson et al., 2020), we could estimate the age factor. We had been worrying about relative risks of two or three times, until we found that the age effect between birth and 100 was around 10,000 times the risk. Age had to be taken into account for all risk assessments, as it was clearly very unlikely someone would be extremely vulnerable if they were very young.
We have been able to estimate the ‘age effect’ for individual conditions, and by giving an age estimate this can be very helpful in understanding the risk. We estimate that the risk of dying after infection with Covid-19 is one in twenty at age 85, so we define this level as ‘very high vulnerability’ and would regard this as similar to ‘clinically extremely vulnerable’. We have chosen age 70, with a risk of dying after infection at one in a hundred, as the level for ‘high vulnerability’. Below this we have two groups, up to age 49, and 50 to 69, who are ‘low vulnerability’ and ‘moderate vulnerability’. Age 50 is a risk of dying if infected with Covid-19 of around one in six hundred.
This means that younger people with conditions on the CEV list have a Covid-age lower than 85, in many cases lower than 70, which is reassuring.
I strongly recommend that anyone who is worried about their personal risk levels use the Covid-age calculator.
I strongly recommend that anyone who is worried about their personal risk levels use the Covid-age calculator. This can also be used as part of a risk assessment when returning to the workplace. The values are averages, so some clinical judgement is needed. Where conditions are not included, this is either because the risk is minimal, or because there were insufficient patients in the study group to give us the data we needed. For example, we consider that those with inflammatory bowel disease or inflammatory skin disease would have similar risk levels to inflammatory arthritis (lupus/psoriasis/rheumatoid arthritis) and the same values can therefore be used. There is more guidance on the website.
We were particularly concerned about kidney disease, as this is perhaps the one clinical condition with the greatest vulnerability, but it was originally left off the CEV list (except for transplant and then dialysis patients). We were only able to give estimates for groups, so we have a combined value for chronic kidney disease (CKD) 3, 4 and 5. Judgement should therefore be used, recognising that the value we give for someone with CKD5 would be an underestimate, while that for someone with CKD3 would be an overestimation.
We can also use the same approach to consider the effect of past infection or vaccination. We know that past infection reduces risk of illness by around 80%, which would in turn equate to a reduction in Covid-age of around 18 years. In the general population, one dose of the AstraZeneca vaccine is 76% effective in reducing infection which corresponds to a Covid-age of around 15 years, and one dose of the Pfizer vaccine, at 90% efficacy, represents a Covid-age of around 25 years. People can therefore adjust their Covid-age grouping accordingly, for example dropping from very high to high vulnerability, or high to moderate. It is important to recognise that people on immune suppressant medication or with a condition that affects immunity may not respond so well to past infection or vaccine, so it may not be appropriate to drop groups.
Finally, perhaps the most important risk factor is viral prevalence. As this drops, risk drops too. The drop from 1000 infections per 100,000 per week to 10 per 100,000 represents a one-hundred-fold drop, which in turn equates to a Covid-age drop of around 50 years when considering overall risk. We can therefore feel much safer generally as viral levels drop. If, however, your job involves treating patients who are infected with Covid-19, the risk won’t drop at all."
- The Covid-age calculator
- Download the tool for combining Covid-age, viral prevalence, immunity and job role (PDF)
- OpenSAFELY study
WILLIAMSON, E. J., WALKER, A. J., BHASKARAN, K., BACON, S., BATES, C., MORTON, C. E., CURTIS, H. J., MEHRKAR, A., EVANS, D., INGLESBY, P., COCKBURN, J., MCDONALD, H. I., MACKENNA, B., TOMLINSON, L., DOUGLAS, I. J., RENTSCH, C. T., MATHUR, R., WONG, A. Y. S., GRIEVE, R., HARRISON, D., FORBES, H., SCHULTZE, A., CROKER, R., PARRY, J., HESTER, F., HARPER, S., PERERA, R., EVANS, S. J. W., SMEETH, L. & GOLDACRE, B. 2020. Factors associated with COVID-19-related death using OpenSAFELY. Nature, 584, 430-436.
