The question of why more people from black, Asian and minority ethnic (BAME) backgrounds appear to be at greater risk of dying of Covid-19 is one of the most urgent of the epidemic.

There is little doubt about the raw numbers: major studies by the Office for National Statistics, the Institute for Fiscal Studies and Public Health England have found significantly higher mortality rates among ethnic minority populations.

Another troubling study, published today, shows that pregnant women from ethnic minorities are much more likely to have been treated for coronavirus in hospital.

Is race the deciding factor?

Almost every serious analysis of Covid-19 deaths finds that BAME people are over-represented compared to what we would expect if the illness affected all Britons in the same way.

For example, initial figures published by the Office for National Statistics (ONS) last month found that black people were more than four times as likely to die as white people of the same age.

Scientists then have to decide whether it is really the ethnicity of patients that is driving increased mortality, or whether other factors are confusing the issue.

Adjusting for other factors

For example, if a particular ethnic group is more likely suffer from an underlying health condition that we know makes Covid-19 more dangerous, this could be the real cause of the raised risk, rather than their ethnicity.

Or if Covid-19 was more prevalent in major cities with large BAME populations, we might expect to see more risk of infection among those groups.

Several major studies have tried to adjust for as many of these “confounding variables” as possible – and have still found that there is increased risk for BAME groups, although it is often less dramatic than the raw figures suggest.

The ONS adjusted its figures to filter out the effect of the region where people lived, deprivation, household composition, socioeconomic status, education, and health and disability.

Once these factors were adjusted for, there were still disproportionate deaths among black and Asian people. Black people were about twice as likely to die from Covid-19 than white people. For some ethnic classifications – “mixed” and “Chinese” – the risks were about the same as for the white ethnic group.

Following a similar process, but using a different methodology, Public Health England came to this conclusion:

“After accounting for the effect of sex, age, deprivation and region, people of Bangladeshi ethnicity had around twice the risk of death than people of White British ethnicity. People of Chinese, Indian, Pakistani, Other Asian, Caribbean and Other Black ethnicity had between 10 and 50% higher risk of death when compared to White British.”

It’s immediately obvious that there is no simple answer to the question of how more risky the pandemic is for BAME people. It depends on which ethnic group you belong to, and it depends on how scientists allow for confounding factors.

Equally obviously, structural inequality could ultimately be to blame. If an ethnic group is at higher risk because of overcrowded housing, deprivation or poorer general health rather than strictly race, that is still a finding that many people will find troubling.

There is no way of filtering out confounding variables that everyone agrees on – and no method will ever be perfect.

The ONS, for example, is using data from the 2011 census – now nearly a decade out of date – to analyse the characteristics of local populations. The PHE analysis did not allow for the effects of occupation, existing health conditions or rates of obesity.

Higher risk of infection or death?

People often talk about BAME people being “more likely to die” from the coronavirus, but there are really two questions we need to answer.

Are people from ethnic minority backgrounds more likely to catch Covid-19 than white people in the first place? And once they have caught it, are they then more likely to die from it?

There is evidence that people from some ethnic minorities may be getting infected at higher rates than the white population: several studies show BAME patients are more likely to test positive for Covid-19, though this could reflect where testing has taken place.

And today, a study in the British Medical Journal, found that more than half of 427 pregnant women hospitalised with Covid-19 across the UK between 1 March and 14 April were from black or minority ethnic backgrounds.

When the scientists excluded patients from London, the West Midlands and the North West, 46 per cent were still from BAME groups. That’s considerably more than we’d expect given the ethnic make-up of these less urban parts of the UK.

The researchers suggest black women could be at particular risk – they estimate these women are 8.1 times more likely to be hospitalised with covid-19 than white women during pregnancy.

Is there a difference in outcomes for patients of different races once they are in hospital?

Several studies have found no strong evidence that infected people from BAME backgrounds are more likely to die, once you allow for confounding factors.

This analysis carried out for the government’s SAGE advisory group in late April found that BAME people were disproportionately likely to be admitted to hospital with Covid-19, but were no more likely to die, after adjusting for their existing health conditions.

“No difference was seen in risk of death from COVID-19 between ethnic groups,” the study concluded.

A study by Imperial College on one London NHS trust did find slightly higher odds of death among black patients compared to whites, but the finding was of “borderline statistical significance”.

So why are BAME people at more risk?

No one knows the answer for sure, but there are a number of plausible theories.

If the real risk factor is exposure to infection in the first place, it may be that we need to look at the differences in occupations between different ethnic groups.

The ONS says that “some ethnic groups may be over-represented in public-facing occupations and may therefore be more likely to be infected by Covid-19.

“For example, individuals in the Bangladeshi and Pakistani ethnic group are more likely to work as transport operatives than those in any other ethnic group.”

The IFS paper points out that some ethnic groups may be over-represented in health and social care, which may carry greater risk of exposure to the virus.

It also notes that the demographic profile of coronavirus victims was likely to change as deaths in hospitals dropped but remained high in residential care homes, where the population is overwhelmingly white.

Public Health England says BAME people may be more likely to be infected because they are more likely to have higher-risk jobs and to live in deprived areas and overcrowded households.

In line with other studies, the PHE report says increased risk of dying could be driven by higher rates of underlying illnesses like cardiovascular disease and diabetes.

Some critics said the report should have gone further in investigating the possible role of structural inequality in increased mortality among ethnic minorities.