Time to Spare

Systematic racial bias in COVID-19 testing?

Tom Neill

Reading time: 3 minutes

This week, the government published their "Disparities in the risk and outcomes of COVID-19" report. Frankly, it doesn't contain enough data to be useful. That may be deliberate, or it may just be unfinished. Despite that, there are some snippets of evidence that would be consistent with a systematic bias against black people.

In my view, the key graph is this one. It shows the age-standardised death rate for different ethnic groups. It shows the baseline “normal rate” and how this has increased during the pandemic. This includes people who tested positive for COVID, but also some “non-COVID” unexplained increase.

Among black/black british men, the all cause (COVID and non-COVID) age-standardised mortality rate has almost quadrupled. This is compared to a less than doubling in the rate for white men. This, while terrible, in itself is not necessarily evidence of racial discrimination. As the government and the report has claimed, it could be caused by a combination of geography, comorbidities and occupational differences.

However, the “unexplained” death rate has barely moved for white men, but has doubled for black men. There are substantially more black men dying for reasons that are not clear. One potential explanation for this is that black men are being tested less often than other groups. And this explanation wouldn’t be surprising.

The main reason that has been suggested for this is because black people are disproportionately more likely to be deprived. However this doesn’t appear to be a deprivation issue. If you look at this graph of all cause mortality split by the deciles of deprivation, there’s not a clear difference in the unexplained increase in all cause mortality:

Further evidence

This graph of all cause mortality isn’t the only piece of evidence that suggests black people may find it harder to access a COVID test. For example, the share of people in intensive care who are BAME is much higher than the share of total COVID hospital admissions:

This is consistent with there being a higher barrier for getting a COVID test for black people. It’s the same reason why the world’s very best footballers tend to be quite short, and why Hollywood seems to ruin good books. If black people are less likely to get a COVID test or to be admitted to hospital, those who do get admitted to hospital, are likely to have significantly worse health outcomes on average.

Another piece of interesting evidence is the rate of infection among nursing staff of different ethnicities:

According to this graph, the share of black nurses infected with COVID is lower than the share of white nurses. Given that the pandemic hit London hardest, and London has by far the highest proportion of black people in the country, this is the opposite of what you would expect. If you pair this with other evidence from Runnymede that BAME nurses are less likely to have appropriate PPE, this is even more surprising. However, if there is systematic under-testing of black people, it’s exactly what you’d expect.

If there is a higher bar for black people to access a COVID test, the real infection rate among black people, when compared to white people, would be even higher than has been reported. It would also mean that black people would be less likely to get an accurate diagnosis and appropriate treatment.

Of course, there are other ways of explaining these statistics. There may be some mix of occupations, geography and comorbidities (e.g. diabetes, hypertension etc.) that align in a particular way to explain all these facts. Unless the government collects and releases some decent data, in particular on the rates of testing in each group, we won’t be able to tell. So while this is far from definitive proof that systematic racial injustice is causing more black people to die, I certainly wouldn’t want to stand up in parliament and declare that it’s not the reason.

More from Time to Spare