Patient relevance: reframing lived experience in mental health research

Patient and public involvement is now woven into the fabric of health research. The National Institute for Health and Care Research (NIHR) defines it as research carried out “with” or “by” members of the public rather than “to”, “about” or “for” them, and asks applicants to describe it on the standard application form, where the answers are considered by funding panels and influence decisions.¹ The Wellcome Trust goes further for mental health specifically, stating that it expects lived experience to be “central to most projects” it funds.² In January 2024, The Lancet Psychiatry began asking authors to report whether and how people with lived experience shaped their studies, and to declare the absence of such involvement as a limitation.³

These are welcome developments. But a mandate is not a purpose, and the space between the two is too often filled by tokenism. The same Lancet Psychiatry editorial observed that the number of papers claiming collaborative methods had risen by more than 600% in a decade, often using “vague terms… to imply a greater amount of engagement than has actually occurred”.³ Ocloo and Matthews describe involvement that remains “patchy and slow and often concentrated at the lowest levels”, echoing Arnstein’s classic ladder of participation, whose bottom rungs are manipulation and tokenism rather than partnership.^4,5 I have seen both sides of this. I have watched researchers recruit a person with lived experience to a panel because they know it is valued, without being able to say why that person is there; and I have been the person who is there, sensing that no one, sometimes including me, was sure what I was for. When the reason is unclear, involvement drifts towards ceremony, and a contributor’s value is quietly assumed to be low.

The remedy, I want to argue, is not more involvement but a clearer reason for it, and we already have the word for that reason: relevance.

Researchers are trained to make studies valid: well designed, adequately powered, free of bias. These are necessary virtues, but validity answers the question “is the result true?”, not “does the result matter?” A trial can be impeccably conducted and still measure something its intended beneficiaries do not care about. Chalmers and Glasziou estimated that a large share of research investment is avoidable waste, the first source of which is whether a study even asks a question of genuine relevance to the people who will use the answer.⁶ When James Lind Alliance priority-setting partnerships have compared the questions patients and clinicians want answered with those researchers actually pursue, the mismatch is stark: across partnerships, drugs accounted for fewer than one in five of the treatments patients prioritised, but for the large majority of treatments tested in commercial trials.^7,8

I use patient relevance to name the principle that the purpose of involving people with lived experience is to keep research anchored to the outcomes that matter to patients. The activities this implies are not new: priority-setting partnerships, core outcome sets and patient-reported outcomes all pursue overlapping ends, and NIHR itself has recommended “relevance” as a measure of successful involvement.⁸ The novelty is in the framing. Patient relevance makes relevance the explicit criterion against which any act of involvement is judged, and supplies a decision rule for who should be in the room and why; it names the purpose the machinery of involvement has too often lacked. It is the answer a researcher should be able to give when asked why a contributor is present. Framing it this way matters for two reasons. For the researcher, it dispels the mystery: the contributor is there to ensure the study measures what is worth measuring. For the contributor, it converts a vague invitation into a defined task and dignifies it; people are far less likely to feel decorative when they understand that the relevance, and therefore the value, of the whole enterprise depends in part on them.

Relevance and validity are allies, not rivals. When patients help choose outcomes, the outcomes selected shift towards the things that affect people’s lives: a systematic review found that core outcome sets developed with patient participation were substantially more likely to include life-impact outcomes than those developed without (86% vs 62%).⁹ In psychiatry, an analysis of depression trials found hundreds of different outcome measures in use, many omitting domains, such as mental pain, that patients consider central.¹⁰

The clearest example I know is one I lived. I have bipolar disorder, and I took part in the Cognitive Remediation in Bipolar (CRiB) study, a proof-of-concept trial testing whether a computerised, therapist-supported cognitive remediation programme could improve cognition in people with bipolar disorder.¹¹ Its primary efficacy outcome was a cognitive measure: performance on timed tasks such as digit-symbol coding.¹¹ I could complete those tasks. What I could not do, at the time, was work, or reliably look after myself. I did not much care whether my processing speed improved by half a standard deviation; I cared whether I could hold down a job, manage money, and run my own life. The trial was measuring something real, but not the thing that, for me, separated an illness merely contained from a life regained.

When the programme was developed into a full trial, CRiB2, that emphasis changed (table 1). Following review by a service-user feasibility and acceptability panel, the team adopted improvement of psychosocial functioning, measured with the Functioning Assessment Short Test (FAST), as the trial’s primary outcome, elevating a measure that had been secondary in the pilot and stating explicitly that this reflected “the prioritisation of functional recovery as an outcome”.¹² The FAST captures six domains of everyday functioning: autonomy, occupational functioning, cognitive functioning, financial matters, interpersonal relationships, and leisure.¹³ These were precisely the domains I had cared about and could not name in the language of a test battery. The change was not a retreat from science. Cognitive impairment is among the strongest predictors of functional and occupational disability in bipolar disorder,^14,15 so a functional primary outcome does not abandon cognition; it asks the more relevant question of whether changes in cognition actually translate into changes in life.

A second lesson concerned how we measured. Cognitive assessment is itself cognitively demanding. As a participant I found the batteries long and tiring, and sustained testing erodes precisely the attention it is trying to quantify, so an exhausted participant produces not only a worse experience but noisier data. Lived-experience feedback into CRiB2 fed into a design intended to prevent participant over-burden, including telephone screening ahead of in-person assessment and a curated set of measures.¹² Here relevance and validity converged again: attending to the participant’s experience improved the quality of the data. I cannot prove these changes were decisive, but a trial built around outcomes that matter, and around participants who are not exhausted by it, is easier to justify to a funder and more likely to influence care once complete. CRiB2 proceeded as a 250-participant, multi-site randomised controlled trial.¹²

The CRiB example can make the process sound simple. It is not. The central difficulty of lived-experience involvement is that people closest to an illness often express their needs in terms that sound, to a researcher, vague or unmeasurable, while researchers, immersed in psychometrics and statistics, can struggle to hear what is actually being said. Bridging that gap requires translation, not transcription. Taking a contributor’s words at face value (“I want to feel normal”) can yield an unmeasurable outcome; dismissing them yields an irrelevant one. The task is to interpret.

I have come to use a simple, iterative sequence I call why-what-how-repeat (figure 1). It begins with why. Had CRiB’s participants been asked whether the study should be repeated and answered, “No, I don’t see the point of improving my attention scores”, the unhelpful response would be to take that literally and conclude the research was unwanted. The better response is to ask why. Because I don’t see what the task is for; what I want is to do better in life. What, specifically? To work, to handle money, to look after myself. In two questions the conversation has moved from an apparently anti-research statement to a precise functional brief. The researcher then supplies the how: those domains can be captured, validly and reliably, by an instrument such as the FAST.¹³ That proposal is taken back to the people with lived experience (the repeat step) and refined until it both matters to them and satisfies the methodologist. The repetition is not a formality: it guards against the translator hearing only what is convenient, and checks each interpretation against the people it is meant to represent. The aim is not to let either party prevail, but to iterate until relevance and validity are satisfied together.

Who should be in these conversations? In my experience, effective involvement draws on three distinct contributions (figure 2). The most important are those closest to the lived experience: people currently living with the condition, or caring for someone who is, in the thick of the struggles the research addresses. Their priorities are central because they are closest to the experience the research seeks to change and most resemble its intended beneficiaries, though such priorities can be state-dependent, and are better triangulated across several voices than taken from one. No one else’s relevance judgement simply substitutes for theirs.

The second contribution comes from the researcher with lived experience, who acts as the translator described above: someone fluent in both worlds, who can lead focus groups with those closest to the experience, convert their needs into researchable outcomes, and guide informed decisions without speaking over them. CRiB2 embedded exactly such a person, a role I went on to hold, noting that a researcher with a bipolar diagnosis “joined the study team at the application stage… from an expert-by-experience perspective”.¹²

The third contribution is the advocate, whose work faces outward: recruitment, public engagement, dissemination and, vitally, helping to ensure the sample is representative and equitably recruited, so that relevance extends to the under-served groups too often absent from trials.^16,17 Relevance and representativeness are distinct: the first concerns what is measured, the second whose priorities are heard, and a study can measure the right outcome for an unrepresentative few. Involvement is also associated with better recruitment, modestly overall (odds ratio 1.16) and more strongly where contributors have lived experience of the condition under study (odds ratio 3.14), though evidence on retention remains inconclusive.¹⁸ All three roles are necessary, but their order of importance runs from those closest to the experience outward.

Terminology matters here too. We label these contributors in ways that describe the person, such as “expert by experience”, “service user” or “lived-experience representative”, yet say nothing about why they are there; at best the terms are uninformative, and at worst, as people with lived experience have themselves argued, they can feel reductive or patronising.¹⁹ Patient relevance names the purpose instead of the person. I retain the word patient deliberately: it names a relationship to clinical care, and to the outcomes that clinical research must ultimately move, which is precisely the relationship this argument concerns. In other settings “lived experience” may read better, and I accept that “patient” is itself contested.¹⁹ Nothing in the framework depends on the label; what matters is the principle, for “nothing about us without us” carried within it, from the start, a claim about relevance as much as about rights.²⁰

Patient relevance is most powerful when it is built in from the start rather than bolted on at the end (figure 3). At the conceptualisation stage, before a grant is written, the researcher with lived experience should lead focus groups with those closest to the experience, using the why-what-how-repeat model to define the primary outcome, the measures, and the assessment procedures. Those decisions, and the reasoning behind them, can then be written directly into the grant application, where, as noted, funders already expect to see them.^1,2 Once funded, the advocate establishes equitable recruitment routes and a dissemination plan, and all three roles remain engaged through delivery, analysis and reporting, including as co-authors.¹²

Reporting deserves particular care. The value of the Lancet Psychiatry policy is not that it counts involvement but that it can make involvement visible; its risk is that it rewards the claim rather than the substance. Standardised reporting tools such as GRIPP2 help,²¹ but the more important discipline, as Scholz has argued, is to report not merely that people were involved but what their involvement changed.²² A relevance lens makes this natural: the question to ask of any study is not “was there involvement?” but “did it make the research more relevant, and how do we know?” The Lancet Psychiatry’s new Commission on Lived Experience in Mental Health Research suggests the field is ready for exactly this shift in emphasis.²³

Lived-experience involvement is not a passing requirement, and it should not be. But requirement without reason produces ritual, and ritual is what we have too often built. Reframing involvement around patient relevance, the simple insistence that research measure what matters to the people it is meant to help, gives the researcher a reason, gives the contributor a role and, because relevance and validity travel together, gives the science itself a better chance of mattering. That the measurable effects of involvement remain modest is not a reason to retreat from it, but a reason to give it the purpose it has lacked. Funders and journals could begin by changing a single question: not “did you involve patients?” but “did it make your research more relevant, and how do you know?” I have written elsewhere about living with bipolar disorder.²⁴ But the lesson here is narrower, and I think harder to dismiss: I have been the participant ticking the boxes and the contributor helping to redraw them, and the difference between the two is not effort or sincerity. It is purpose.