Introduction

On 12 March 2025, stakeholders across the mental health research landscape gathered at the Science Gallery in London for an important workshop hosted by the Data & Digital Workstream of the NIHR Mental Health Translational Research Collaboration (MH-TRC) Mission. The workshop focused on a pressing challenge in the field: how to improve participant recruitment for mental health clinical trials and cohorts using data-driven approaches.

With around 50 attendees, 11 speakers, and 3 session chairs, the event was structured into four themed sessions. What made the day especially impactful was its firm commitment to centering the voices of those with lived experience of mental ill health—ensuring that the conversation around recruitment and data was grounded in real people’s personal experiences.

Centering Lived Experience: Stories That Resonate

The day began with opening remarks from Professor Ann John (Swansea University). The first formal speaker was Amy Chidley, a cross-sector Lived Experience Facilitator, who is involved with the MH-TRC Mission Midlands Demonstrator Site. Her talk, titled “My story, your data, our impact,” was both powerful and enlightening.

Amy’s personal account highlighted the complexities of mental health diagnoses. Amy’s postnatal symptoms led to a diagnosis of depression and anxiety and were treated conventionally, yet these interventions were at best ineffective and at worst catastrophic. She was labelled a ‘mystery’. Years later, research was unearthed which confirmed ‘psychiatrist misdiagnoses’ common in Ehlers-Danlos Syndrome (EDS), due to co-occurring dysautonomia, and crucially the exacerbating consequences of taking psychiatric medication. Amy thus delivered this critical message: data used in research is only as good as the accuracy of the diagnoses that underpin it. Misdiagnosis leads to misclassification in data, which in turn limits the quality and utility of mental health research, success of dissemination, and clinical approval and implementation of its findings.

Amy called on researchers to not only improve precision in recruitment of research participants but also to work toward better diagnostic processes. Her presentation highlighted that people with experience of mental health challenges want to collaborate in research, but the challenge is for their involvement to be informed by accurate diagnoses data to produce accurate data outcomes.

Later in the day, Jon Bartlett, who lives and works with a diagnosis of bipolar disorder, offered a similarly moving reflection. He described instances where inaccurate information in his medical records caused confusion and distress during clinical interactions. His story served as a further powerful reminder that behind every data point is a person, and that trust in research depends on getting the basics—like accurate data—right.

Session Highlights

Session 1: MH-TRC Mission Workstreams

The first session spotlighted several regional demonstrator sites and workstreams aligned with the MH-TRC Mission.

• Mood Disorders Network: Mary Wilson discussed plans to build a cohort of 2,000 people with treatment-resistant depression, recruited via primary care records, specialist referrals, digital registries like NIHR’s Be Part of Research, and social media outreach.
• DECODE Platform (Birmingham): Opeyemi Odejimi shared progress on this regional digital registry aiming to enroll 500 participants over two years. DECODE leverages both primary care and specialist clinic referrals to support future clinical trial recruitment.
• Early Psychosis Registry: Graham Murray introduced the Early Intervention Mission, which seeks to create the UK’s largest digital registry of people with early psychosis (target: 2,000 participants). Graham talked about the potential value to them of partnering with NHS DigiTrials, a national service designed to streamline participant recruitment.
• Count Me In (Liverpool): Oladayo Bifarin and Michelle Harvey presented on this innovative opt-out recruitment model being trialled in the Merseyside region, which aims to make participation more inclusive and less administratively burdensome.

Session 2: NHS Data Infrastructure

The second session turned to the broader infrastructure underpinning research data in the NHS.

• Sub-national Secure Data Environments (SDEs): Jim Hughes explained how 9 of the 12 regional SDEs are already operational, with the final 3 anticipated to be going live soon. These SDEs provide access to highly granular regional datasets—often with more detailed linkages than the national datasets—making them potentially powerful tools for localized recruitment strategies.
• NHS DigiTrials: Matt Sydes gave a compelling talk on the capabilities of NHS DigiTrials, including both the Feasibility Self-Service (FSS) tool, and the Recruitment service. One standout example of the latter was the Galleri trial, which recruited 140,000 participants in just 10 months—a world record. In addition, by reducing trial feasibility assessments from months to minutes, DigiTrials FSS is helping make feasibility analyses more efficient than was ever previously possible.

Session 3: Artificial Intelligence and Natural Language Processing

In the third session, the conversation turned toward the promise of AI and NLP in unlocking insights from unstructured data—like clinician notes or patient narratives.

• CogStack: Richard Dobson discussed this NLP platform’s ability to mine free-text data across NHS systems, which could be applied to help identify potential trial or cohort participants who might otherwise be missed using structured data alone.
• CRIS (Clinical Record Interactive Search): Rob Stewart shared how this platform at the Maudsley Hospital is enabling researchers to use both structured and unstructured data to gain a fuller picture of patient populations, which could also ultimately support smarter recruitment strategies.

Session 4: Recruitment in Primary Care

The final session explored how to work more effectively with primary care data systems, especially via the Clinical Practice Research Datalink (CPRD).

• Tim Williams reflected on lessons learned from CPRD’s involvement in a limited number of mental health trials to date. He acknowledged challenges in timelines and emphasized that despite digital infrastructure, real-world recruitment via CPRD still requires coordination with individual GP practices—highlighting the need for streamlining.

 

Key Takeaways

1. Keep People at the Centre
   The most resounding message of the day was the importance of grounding recruitment strategies for mental health research in the experiences of people who are living with, or have lived with, mental health challenges. Accurate diagnoses, compassionate care, and respectful use of data are essential for earning patient trust, involvement, engagement and participation.
2. Leverage NHS Infrastructure
   From SDEs to DigiTrials, the UK’s growing ecosystem of NHS data services offers promising tools to reduce recruitment burdens on frontline staff—especially GPs and other healthcare professionals—and to improve trial delivery timelines.
3. Embrace AI and NLP Innovation
   Unstructured data, like clinician notes or patient narratives, hold enormous potential. Tools like CogStack and CRIS could be applied to help surface potentially eligible participants that traditional coded data might overlook, opening new doors in precision recruitment.

 

Final Thoughts

The workshop was a timely and necessary gathering of minds, reminding us that solving participant recruitment challenges in mental health research is not just a technical issue – it’s a human one. By listening to those with relevant lived experience, investing in better infrastructure, and embracing innovation, we can build a more inclusive and effective mental health research ecosystem. As data continues to reshape the future of healthcare, we must remember: data isn’t just numbers – it’s stories, lives, and opportunities to do help more people more efficiently and effectively.