Dear Editor,
A recent letter in this journal by Kakkar and colleagues noted how, during the first wave of the coronavirus pandemic, black, Asian and minority ethnic (BAME) inpatients were significantly more likely to be admitted to intensive care compared to white inpatients at Sheffield Teaching Hospitals.1 This is just one piece of a growing body evidence that illustrates how BAME individuals have been disproportionately affected by the novel coronavirus pandemic, with excess mortality due to COVID-19 in BAME populations in England now a well-established phenomenon.2 The precise makeup of factors responsible for this disparity remains unknown.
An historical under-representation of BAME patients in medical research3 has prompted concerns that COVID-19 studies may suffer from the same pitfall4. Numerous studies concerning potential vaccines and treatments for COVID-19 are already underway and it is essential they reflect the populations they hope to serve. The extent to which these studies are fulfilling adequate representation in their study cohorts is unclear.
Several weeks after the initial COVID-19 surge, we conducted a retrospective analysis of the ethnicity of inpatients enrolled onto the six COVID-19 interventional treatment trials at Imperial College Healthcare NHS Trust (ICHT) in London, UK (Table 1 ). Patient records were insufficiently granular to provide ethnicity data beyond ‘BAME’, ‘white’, and ‘unknown’.
Table 1.
Interventional COVID-19 trials at ICHT.
| Trial | Interventions | Indication |
|---|---|---|
| RECOVERY | Dexamethasone; azithromycin; tocilizumab; convalescent plasma; REGN-COV2 (monoclonal antibodies) | Suspected or confirmed COVID-19 |
| C19-ACS | Early acute coronary syndrome therapy | Suspected or confirmed COVID-19 |
| Gilead Moderate | Remdesivir | Moderate COVID-19 |
| Gilead Severe | Remdesivir | Severe COVID-19 |
| COVACTA | Tocilizumab | Severe COVID-19 |
| REMAP-CAP | Multiple domains including antivirals and immunoglobulin therapy | Severe COVID-19 |
In total, 179 patients were enrolled onto the six trials. The trial with the largest cohort was Recovery (n = 83), which recently demonstrated that amongst inpatients hospitalised with COVID-19, dexamethasone reduced 28-day mortality among those receiving invasive mechanical ventilation or oxygen at randomization.5 Ethnicity was not included as a subgroup in this analysis.
Of the total 179 patients, 61 (34%) were BAME, 80 (45%) were white and for 38 (21%), ethnicity was unknown. Of the 83 patients enrolled into Recovery, 24 (29%) were BAME, 40 (48%) were white and for 19 (23%), ethnicity was unknown.
To understand if BAME representation within the trials was reflective of the total cohort of inpatients with COVID-19, we consulted the best available standard of comparison: a retrospective cohort study of all patients hospitalised with confirmed SARS-CoV-2 infection at ICHT between February 25 and April 5, 2020. Of the 520 patients in this cohort, 209 (40%) were BAME, 196 (38%) were white, and for 115 (22%) ethnicity was unknown.6
There was no statistically significant difference between the proportion of BAME patients across all trials compared to the total inpatient cohort (chi square test, p = 0.089). However, there was a statistically significant difference between the proportion of BAME patients in the Recovery trial compared to the total inpatient cohort (p = 0.036).
The reasons for this observation of an apparent under-representation of BAME patients in the Recovery trial at ICHT are unclear. Existing evidence suggests that the reasons for an under-representation of BAME individuals in medical research are complex and include a range of subject, clinician/researcher, societal and cultural factors.7
Patients were eligible for enrolment into Recovery provided they met three eligibility criteria: (i) hospitalization, (ii) clinically suspected or laboratory confirmed SARS-CoV-2 infection and (iii) no medical history that might, in the opinion of the attending clinician, put the patient at risk.
It is plausible that BAME inpatients were more likely to have co-morbidities that clinical teams believed could put them at risk if they were enrolled. For example, type 2 diabetes is disproportionately prevalent in South Asians and dexamethasone is known to interfere with glycaemic control. We also know that BAME patients with COVID-19 are disproportionately admitted to critical care settings1 , 2 where staff are required to wear FIT-tested PPE. During the initial surge and before research was announced as a key pillar of the government's plans, clinical teams were understandably prioritised for FIT testing, meaning the local research team initially experienced difficulty in accessing and recruiting patients being treated on intensive care. The Recovery trial also permitted clinician consent if a patient lacked capacity, raising the possibility that recruitment may have been skewed towards older patients who, as a population, are more likely to lack capacity. At ICHT, white COVID-19 patients had a higher age composition than their BAME counterparts.6
Furthermore, the research team was not provided with non-English language patient information sheets, and there was minimal access to interpreting services. Mandatory reporting of ethnicity was not a component of the Recovery enrolment protocol and it was only after the intense initial surge that a more robust screening log, tracking reasons for exclusion, was introduced.
A Public Health England review into the disparities in COVID-19 outcomes highlighted how a lack of ethnicity data hindered the scope of analysis2 and one systematic review of the impact of ethnicity on clinical outcomes in COVID-19 published in June found that of 1518 COVID-19 studies registered on ClinicalTrials.gov at the time, only six were recording ethnicity data.8 The National Institutes of Health (NIH) in the US has required researchers to replicate the ethnic composition of their study population since 1994, and phase III clinical trials must include subgroup analysis to assess ethnic differences in treatment efficacy.7 Although this practice is recommended in the UK, it is not mandated.
Whilst there is a paucity of evidence regarding effective interventions to increase BAME representation in clinical trials,9 mandatory ethnicity recording in COVID-19 research studies in the UK would ensure inclusivity could be more accurately tracked. Furthermore, in 2018 the National Institute for Health Research (NIHR) launched the Innovations in Clinical Trial Design and Delivery for underserved groups project (INCLUDE), which has issued guidance to support inclusive COVID-19 research.10 Accordingly, the government, public health bodies and funders should commit to ensuring local research teams are able to invest in the necessary resources to maximize trial inclusivity.
The issue of BAME under-representation in medical research is complex and will take considerable time and effort to overcome, including addressing societal and cultural factors. But the disproportionate impact of COVID-19 on BAME individuals and a second surge in cases and hospitalisations mean there is an urgent need for the government, researchers and healthcare professionals to do everything possible to ensure inclusivity in COVID-19 research studies.
Declaration of Competing Interest
None.
References
- 1.Kakkar D., Dunphy D., Raza D. Ethnicity profiles of COVID-19 admissions and outcomes. J Infection. 2020;81(2):e110–e111. doi: 10.1016/j.jinf.2020.05.059. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Public Health England. Beyond the data: understanding the impact of COVID-19 on BAME groups. 2020.
- 3.Smart A., Harrison E. The under-representation of minority ethnic groups in UK medical research. Ethn Health. 2016;22(1):65–82. doi: 10.1080/13557858.2016.1182126. [DOI] [PubMed] [Google Scholar]
- 4.Treweek S., Forouhi N., Narayan K., Khunti K. COVID-19 and ethnicity: who will research results apply to? The Lancet. 2020;395(10242):1955–1957. doi: 10.1016/S0140-6736(20)31380-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.The RECOVERY Collaborative Group Dexamethasone in Hospitalized Patients with Covid-19 — Preliminary Report. N Engl J Med. 2020 doi: 10.1056/NEJMoa2021436. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Perez-Guzman P., Daunt A., Mukherjee S., Crook P., Forlano R., Kont M., et al. Report 17 - Clinical characteristics and predictors of outcomes of hospitalised patients with COVID-19 in a London NHS Trust: a retrospective cohort study. (29-04-2020), doi: 10.25561/78613.
- 7.Redwood S., Gill P. Under-representation of minority ethnic groups in research — Call for action. Br J Gen Pract. 2013;63(612):342–343. doi: 10.3399/bjgp13X668456. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Pan D., Sze S., Minhas J., Bangash M., Pareek N., Divall P. The impact of ethnicity on clinical outcomes in COVID-19: a systematic review. EClinicalMedicine. 2020;23 doi: 10.1016/j.eclinm.2020.100404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Treweek S., Pitkethly M., Cook J., Fraser C., Mitchell E., Sullivan F. Strategies to improve recruitment to randomised trials. Cochrane Database System Rev. 2018 doi: 10.1002/14651858.MR000013.pub6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Ensuring that COVID-19 Research is Inclusive . 2020. Guidance from the NIHR CRN INCLUDE project.https://www.nihr.ac.uk/documents/ensuring-that-covid-19-research-is-inclusive-guidance-from-the-nihr-crn-include-project/25441 Nihr.ac.uk.[cited 29 October 2020]. Available from: [DOI] [PMC free article] [PubMed] [Google Scholar]