Does the ethnicity of mental health research participants reflect the eligible population?

Ensuring diversity in research involvement, including sample participants, is a key component in achieving good science and crucial to delivering effective and equitable healthcare. Health research directly informs assessment and treatment approaches and guidance and policies; thus, findings should be applicable to the populations we serve. There is a strong body of evidence across different conditions and countries, showing that racially minoritised groups are often underserved (e.g., underrepresented) in health focused research including mental health. In recognition of the importance of research inclusivity for underserved groups, the National Institute for Health and Care Research (NIHR) have recently made it a funding requirement that research is inclusive and addresses health and social care inequalities.

To address some of the implications raised by these findings locally, the Building Race Equity and Diversity (BREaD) in Research Network, South London was established. BREaD has several key objectives including addressing inequalities and identifying and optimising approaches to embedding race inclusion, diversity, and equity in health research. The ever-expanding network brings together stakeholders across the local community and university and healthcare partners (a full list of these can be found at the end of the blog).

One key output of the network has been a recent study to examine the ethnic background and representativeness of participants in mental health research conducted at South London and Maudsley NHS Foundation Trust.

Why does representation matter?

We often hear that research representation matters but why? Inclusive and equitable representation in clinical research is essential (among other reasons) to:

• Ensure that findings apply to all segments of the population.
• Improve trust in healthcare research among minoritised ethnic groups.
• Reduce disparities in access to effective treatments and services.
• Identify specific health challenges and outcomes within different communities.

Past research on various conditions has shown that a lack of diversity in participant recruitment can lead to findings that may not be applicable across different ethnic and cultural backgrounds. Thus, ensuring inclusive participation is a key first step towards addressing systemic health inequalities.

Investigating representation in research participation

Our study focused on large research studies conducted within South London and Maudsley between 2012 and 2022. We surveyed the ethnicity of research participants and compared this data to the ethnicity distribution of the population of the Trust’s catchment area as well as to the local population using the latest Census data. We collected data from 22 studies with a total of 3,279 participants.

Our findings were encouraging: the ethnicity distribution of research participants largely aligned with that of the Trust’s patient population, as shown in the Table below.

High-level* ethnicity across South East London Census data, Trust clinical population and Trust research participants

*High-level ethnicity refers to the 5 broad ethnic categories provided by the UK Census.

**Numbers represent London boroughs covered by South London and Maudsley (Croydon, Lambeth, Lewisham, Southwark)

***These percentages exclude missing data (N=45845)

****These percentages exclude missing data (4.7%) (total N=3125).

As shown on the table, the distribution of most ethnic groups among research participants was similar to the local populations. The proportion of participants from Mixed backgrounds, although matching the census data, appeared to slightly underrepresent the number receiving care from the Trust, while some minority groups such as Gypsy, Roma, Traveller and Arab populations had very little representation – this may potentially relate to the limitations of the ethnicity categories offered by research studies for participants to select from.

One of the challenges we identified in our research was that some studies did not record participant ethnicity, highlighting the ongoing issue of incomplete demographic data in research. A second challenge was that different studies used variable and idiosyncratic ethnicity categories such as “Mediterranean” or “British” without further specifications, making direct comparisons more difficult.

Addressing the gaps in ethnicity data

Inconsistent, incomplete and/or absence of recording of participant ethnicity is a recognised problem; while guidance has been published on how best to record ethnicity in clinical research, standardisation remains an issue. Moreover, participants may also be reluctant to disclose their ethnicity due to various reasons including a lack of information and/or understanding about why the data are being requested, mistrust towards research and/or a lack of relevant options in classification systems.

To improve data collection and representation, we advise research teams to:

• Provide clear explanations to participants on why you are collecting data on participant’s ethnic background.
• Offer a comprehensive list of ethnicity categories that align with Census classifications but also offer participants the opportunity to describe their ethnicity themselves.
• Implement consistent and transparent data recording across all studies.
• Engage with a broad range of groups and communities to build trust and encourage participation and trust in research.

Looking ahead

Our study is the first of its kind. While our findings suggest that, in this case, research participation is broadly representative of local communities, there is a lot more work to be done. We cannot assume this finding is applicable to all services and settings – our study only examined a fraction of studies in one London NHS Trust and had limitations.

Ongoing monitoring is essential to improve inclusivity in research recruitment and maintain gains. Future research efforts should focus on ensuring that participant ethnicity and diversity data are collected employing more harmonised approaches across studies and settings. Research teams should also aim for increased outreach towards underserved groups as part of efforts to build trust in communities and share our commitment to inclusive research.

Ultimately, by prioritising inclusive research, we can build a stronger foundation for equitable mental healthcare that serves all communities effectively. This will require consistency, commitment and continued efforts – our work only marks the first step.

Partners in the BREaD in Research Network include:

• Lived experience research ambassadors;
• Voluntary and community based organisations e.g. Croydon BME Health Forum, Black Thrive, Caribbean and African Health Network,
• Health partners, such as South London and Maudsley NHS Foundation Trust, King’s College Hospital NHS Foundation Trust, Guy's and St Thomas' NHS Foundation Trust, King's Health Partners;
• Southwark Council;
• Lambeth Council: Health Determinants Research and Evaluation Network
• Institute of Psychiatry, Psychology & Neuroscience, King’s College London;
• King’s Clinical Trials Unit, NIHR King's Clinical Research Facility (CRF), NIHR Maudsley Biomedical Research Centre (BRC), NIHR Applied Research Collaboration (ARC) South London, NIHR Regional Research Delivery Network (RDN), and Clinical Academic Groups (e.g., Psychosis).

'Does the ethnicity distribution of research participants reflect the eligible population? Survey of participants recruited through a UK mental health Trust', (Aikaterini Dima, Amanda Brown, Tanya Shlovogt, Silian Martinez, Juliana Onwumere, Maria Antonietta Nettis, Kia-Chong Chua, Matthew Hotopf, Fiona Gaughran) was published in BMJ Open. Doi: 10.1136/bmjopen-2024-093269