A call for the recording of the race – not that races exist of course so ethnic background instead – of those who die from Covid-19. This seems entirely fair, we are short of information at present and more data will lead us to having more information in time.
However, we do need to be careful about that transformation of data into information. Getting the calculations wrong means that the info will be incorrect, will obscure more than it illuminates.
The government’s failure to record and publish real-time data on the ethnicity of Covid-19 patients is a scandal that is endangering lives, according to the chair of the British Medical Association.
Speaking to the Observer, Dr Chaand Nagpaul said: “This is not an issue that should require further campaigning. It would be a scandal if it requires further lobbying as data recording needs to start now, not tomorrow. When you have stark statistics like this, it is an instruction for government to act.”
Hospitals are not currently required to record the ethnicity of any patients who are admitted, fall critically ill or die. There have been more than 15,000 hospital deaths in the UK with the virus, but an independent study of the first 5,578 patients has shown that Covid-19 is disproportionately affecting ethnic minorities.
We would actually like to know. For we are aware that certain diseases have racial – sorry, ethnic – susceptibilities. Sickle cell anaemia is strongly, to the point of always, associated with a West African background. Cystic fibrosis is very much more common in a Northern European one. Various drugs work differently on those of different genetic structures – not something that should be a surprise but it does shock some all the same.
OK, new disease, sure, we want to know how it it varies across the human genome.
On Friday, the intensive care national audit and research centre revealed that 34.5% of Covid-19 patients in critical care were black, Asian or from an ethnic minority, despite accounting for only 13% of the general population.
While BAME workers represent 44% of the NHS workforce, they accounted for 68% of the 57 NHS staff known to have died with the virus. Every one of the 14 doctors reported to have died so far is from an ethnic minority.
And that’s not the way to do it:
The reason it isn’t is because the population is not evenly distributed across the geography of the UK. Nor is the disease evenly distributed. Nor even is the ethnic minority population evenly distributed – in fact, we can find areas where what is normally called the minority is the majority.
Agglomerating our numbers to the national level is therefore not producing information for us from the underlying data. It is obscuring it.
And:
Etc, etc.
It is not a surprise that a pandemic disease turns up more in the heavily populated urban areas. Nor is it a surprise that recent immigrant groups are in those heavily crowded urban areas – that’s been true of immigration into Britain since whenever, the last group it not being true of were the Saxons.
We need therefore to compare – if we want to gain information that is – the ethnic composition of the population with the infected, or requiring ICU, or death rate, in a specific area, not nationally. The same is true of the NHS workforce. That ethnic distribution I have no idea about but it wouldn’t exactly surprise to find that the non-white portion in London were greater than in the nation as a whole.
We would expect, given that geographical distribution both ways, the ethnic infection and death rate to be higher. Finding that it is does not therefore lead to this conclusion:
On Sunday, Sadiq Khan, the London mayor, is expected to draw attention to the “structural racism in our society” that is causing BAME communities to be hit so hard by the virus. He will pay tribute to ethnic minorities vastly over-represented in frontline roles – such as care assistants, supermarket workers and bus drivers – and who are therefore exposed to a much greater risk of catching coronavirus in the first place.
We should also include the data that works the other way. The ethnic minority population, being relatively recent in arrival, is also rather younger on average than the indigenous. From memory, the place in 2011 was something like 96% white among the over-80s, maybe 80% (memory rather fails but you get the idea) among 10 year olds, and significantly less than that in London. Covid-19 preferentially kills the over-80s, we should adjust our numbers for that too.
Only once we have done all of that will we have extracted useful information from our data.
And guess what? None of the race baiting politicians are doing that, are they?
PHE medical director Yvonne Doyle said: “This is a really important issue, and detailed and careful work needs to be done before we draw any conclusions.”
Yes, quite so.
Have been waiting for someone to do the figures. Many thanks.
It makes common sense that in all spheres of life (and in this case, death) as much information is gathered to allow us to make as rational and accurate decisions as possible.
But we all know that whatever data is collected will inevitably be used to confirm that the UK is nothing more than a racist, white privileged, patriarchal society and that more must be done for black, asian and ethnic minorities victims. Oh, and women.
Recording ethnicity when dealing with crime however, is racist.
Couple that with the ordained outcome and resulting hysteria and there is no need to spend the effort on racist data gathering. They’ve already reached their conclusions.
The Guardian mentions the 14 doctors, then recites a list of theories that have no connection to the economic circumstances of the 14 doctors. It’s as though the Guardian was selectively blind…
At a rough estimate, taking the BAME population of London as ~40% of 9M and the rest of the UK at ~22% (so as to get the overall UK average to ~24% as in the graphic above), assuming uniformity, if you counted how many people have anything at all, you’d expect ~28% to be BAME c.f. ~34% of critically ill Covid-19 patients and ~24% of overall population being BAME.
“Hospitals are not currently required to record the ethnicity of any patients who are admitted . . .”
Really? Not my experience when using the NHS, I’ve always been required to complete forms which ask for my ethnic origin.
But there’s always an option of “not telling”, same with gender. At least last time I went.
As an estimate of the number of NHS workers you’d expect to have died from Covid-19. The age profile of CV deaths seems to be a close match to that for all-cause mortality. So, if you count the fraction of people the ONS says die between ages of 20 and 65 (~14%) and take the mortality rate for CV to be the same, then, of 14,000 CV deaths (UK 2 days ago), you’d expect ~1,960 to have been aged between 20 and 65. The NHS employs ~1.4M people, all aged ~20-65. Approx 39% of British population (~69M) is between 20-65… Read more »
Been wondering about this, as reports suggest the same effect in Sweden and elsewhere.
But don’t use general population, use age-adjusted, which should be lower than the 13%. COVID almost exclusively kills over 65s, and over 50% of deaths are over 80. But the population gets whiter as it gets older. So it possible that deaths are even more disproportionate. That’s unsurprising since the co-morbidities are things that blacks, and Asians are more susceptible to.
Relative serum levels of vitamin D might be one causative factor.
Presumably the virus would be OK if we could show that it kills indiscriminately?
Racist virus bad, Diverse virus good?
If it were found that people of different races required different treatment that would be of practical use. This sounds more like race baiting.
Affecting elderly far more than working age, and blokes far more than women. When is Diq Can’t going to call for a study or something to be done about why these are the vulnerable groups.
In the US, cultural factors also enter into it. Black Americans are more urban than average Americans, and predominantly are in a culture where touching is more central. (Do they owe the rest of us “reparations” for contributing to the spread of this pathogen?)
It would be nice for writing papers if we routinely collected all data that might show contagion patterns. But the country was designed around citizen rights (including privacy and autonomy) and not enlightened Herd Management.
Random new COVID data points from US conservative blog posts:
(1) Nick Arama at RedState.com quotes Yahoo news, National Biodefense Countermeasures Agency says COVID is less robust in high temperatures and high humidity and is quickly destroyed in bright sunlight, supports the notion of a decline at the end of “flu season” though possibly resurgence in Autumn. (2) Gil Gutknecht at TownHall.com cites amateur analysis that patients who had measles/mumps/rubella vaccine + booster are more resistant. (This too will vary by culture.)