‘Beyond the data: Understanding the impact of COVID-19 on BAME groups’ is the controversial Public Health England report investigating how COVID-19 has affected BAME groups. It has been reported that this 69 page report was originally meant to be a part of the earlier published ‘Disparities in the risk and outcomes of COVID-19’ report, but was removed before the report was published due to concerns that it would add fuel to Black Lives Matter protests. If you get a chance to read ‘Beyond the data: Understanding the impact of COVID-19 on BAME groups’ in full I would recommend doing so. It is important to see what the UK government does not want you to know. If you don’t have time to read the full report, I give a summary below.
TLDR; Here is a summary of the summary. The report concludes that structural racism is the main cause of the disproportionately large number of BAME COVID-19 deaths. It makes this conclusion based mainly on discussions with stakeholders, as there is little high quality data that would enable a more robust analysis of why the risk of dying from COVID-19 differs so greatly depending on ethnicity. The report makes seven recommendations, some of which are pretty vague. For some of the recommendations, it would be hard to measure whether they had actually been carried out or not.
The report begins with an executive summary and makes seven recommendations, summarised below:
The next section of the report is a literature review. The authors review existing COVID-19 publications and data and try to answer four questions. I go through each question and the answer given by the report below.
This is to my mind one of the most important questions of the report. If we had reliable, unbiased COVID-19 testing data in which ethnicity was recorded, we could concretely say whether the probability of contracting COVID-19 differed by ethnicity. Unfortunately, a large scale randomised COVID-19 testing trial in which participants recorded their ethnicity has not been carried out. The testing data that does exist supports the hypothesis that positive test results are higher than expected for BAME communities, but is not randomised and in some cases comes from studies with a small number of participants. The report concludes that due to a lack of high quality data, conclusions on the relative likelihood of BAME individuals testing positive for COVID-19 can’t be drawn.
The literature on whether BAME individuals are more likely to become severely ill and/or be admitted to ICU is mixed. Some studies have found that the risk of severe illness can differ dramatically by ethnicity, whilst others have found that there is no difference in the risk of severe illness or ICU admission. It is worth noting that many of these studies came from hospital trusts, and therefore the number of patients being studied is often small and may be squewed by local population demographics. The report notes that the rates of cardiovascular disease and diabetes differ by ethnicity but doesn’t draw any direct conclusions from this. The report concludes that more research is needed to determine whether BAME individuals are more likely to become severely ill from COVID-19.
Here, the data is a bit clearer. Data from hospital trusts, hospital mortality data and the Office For National Statistics death data all find higher COVID-19 death rates in BAME ethnic groups relative to those of white British ethnicity. The report concludes that the evidence available supports excess mortality in BAME populations.
The report firmly states that structural racism is the main cause in the differences in death rates seen between different ethnic groups. The clear conclusion on the role of structural racism in COVID-19 outcomes is almost certainly why these 69 pages were removed from the original Public Health England report. Whilst I agree that structural racism probably is the main cause in the differences in death rates, I was a little disappointed that the report authors didn’t collect more evidence to suport their conclusion.
I am a scientist by training, and was a little surprised and alarmed at how unscientific the approach taken to making the report’s seven key recommendations was. It seems that stakeholder reviews were used to inform the seven key report recommendations. Almost a third of the report was dedicated to summarising the stakeholder reviews, without really explaining who the stakeholders were, how they were chosen or why they should be considered to be an authority on improving BAME COVID-19 outcomes. In addition, no information was provided on how the stakeholder sessions were run (eg. what questions were the stakeholders asked? Was it a Q&A style session or was there an open discussion? How were the quotes from stakeholders used in the report selected?). It is possible that the stakeholder reviews held such a prominent position in the report in part due to the lack of high quality scientific data found in the literature review.
In a similar vein, parts of the report were very speculative, especially when considering how COVID-19, ethincity and socioeconomic factors might interplay. To be fair to the report authors, the lack of data on how COVID-19, ethnicity and socioeconomic factors would have made a more robust analysis very challenging, if not impossible.
The report ends by restating the seven key recommendations. I agree with the spirit of the seven key recommendations, but am concerned that some of the recommendations are overly vague and difficult to measure. If we can not measure whether or not a recommendation has been implemented, governments are more likely to get away with ignoring or only partially engaging with the recommendation. Hopefully this is only the first of many reports investigating the impact of COVID-19 on BAME groups, and more robust conclusions can be drawn as more data is collected.