Co-Designed Mental Health Formulation Model Wins Praise but Stumbles on the Ward

A new psychological formulation model built hand-in-hand with the people it is meant to help has been judged acceptable by clinicians and service users alike, yet a study published in PLOS Mental Health reveals a sobering gap between enthusiasm and everyday practice. The research, led by Aneita Pringle of Anglia Ruskin University with colleagues from the University of Oxford and the University of Cambridge, tested the ‘Personal Narrative Model’, a framework for making shared sense of mental health difficulties that was co-developed with a Service User Advisory Group from the very beginning. Although participants described the model as valuable, accessible and potentially transformative, it was not meaningfully implemented on the acute inpatient ward where it was piloted, exposing deep practical, cultural and attitudinal barriers that continue to guard the territory of the medical model.

Psychological formulation is the process of applying psychological theory to build a collaborative understanding of a person’s difficulties, which then informs their plan of care. Unlike diagnosis, which asks what category a person fits into, formulation asks what has happened in a person’s life, what meanings they have drawn from those events, and how those meanings shape their current distress. It has long been recognised as a core competency by psychiatrists and clinical psychologists, and it aligns closely with the recovery model of mental health, which shifts the emphasis from narrow clinical outcomes to goals that are personally meaningful to service users. Yet despite years of government endorsement of recovery-oriented practice in the United Kingdom, routine activities such as acute care-planning vary substantially in how far they actually embody recovery principles, and service user involvement remains inconsistently integrated.

The problem the researchers set out to address was a striking one: almost no existing formulation frameworks had involved service users at the development stage. Drawing on principles of epistemic justice and the disability rights movement’s slogan of ‘nothing about us without us’, the team built the Personal Narrative Model collaboratively with clinicians, academics and a Service User Advisory Group. Two core advisory group members attended approximately twenty meetings during development, shaping the model’s framing, assumptions and language. The resulting framework is trantheoretical, strengths-focused and deliberately accessible, integrating political, social and biopsychosocial influences on distress. It incorporates the Power Threat Meaning Framework, which famously reframes the diagnostic question from ‘What is wrong with you?’ to ‘What has happened to you?’, alongside Hagan and Smail’s power-mapping model, strengths-based approaches, the Comprehend, Cope and Connect model and established team formulation guidelines.

Structurally, the Personal Narrative Model offers prompts for reflection rather than a fixed protocol, organised under headings such as ‘Strengths and Resources’, ‘Past Experiences’, ‘Meaning’, ‘Threat’, ‘Threat Responses’ and ‘Next Steps’. It is designed to be used one-to-one with service users or within multidisciplinary team formulations. The advisory group insisted the model be grounded in the realities of ward life, that its language remain accessible, and that strengths-based practice be prioritised. Three iterative versions were produced before testing. The model was piloted on a UK specialist inpatient ward of roughly ten beds for people diagnosed with a personality disorder, where staff received four ninety-minute training sessions covering the model’s development, formulation practice and the practicalities of implementation, including a role-play exercise simulating how the model might be introduced to a service user.

The evaluation used a two-phase exploratory sequential design guided by a critical realist framework. Phase 1, running from April to November 2019, involved a focus group with seven inpatient staff, including nurses, a clinical nurse specialist, a healthcare assistant and an occupational therapist, held one month after training. Phase 2, adapted to online delivery because of the COVID-19 pandemic, took place between May and November 2020 and comprised an open-ended survey with twenty-six participants from across the UK, including practitioners, practitioners with lived experience and people with lived experience of services, plus follow-up interviews with eight participants. All data were analysed using reflexive thematic analysis following Braun and Clarke’s six stages, with selected transcripts double-coded by two team members and NVivo 12 supporting data organisation. Ethical approval was obtained from the North West Liverpool Central Research Ethics Committee and the Anglia Ruskin University Faculty Research Ethics Panel, and all participants gave written informed consent.

The findings split into three overarching themes: the value of collaborative formulation, ‘selling’ and embedding change, and implementation barriers. On the first theme, reactions were strongly positive. One participant with lived experience called the model ‘incredibly useful and needed’, while a practitioner said she would be ‘so happy to use it’. Participants valued the way the model explored how people feel about their current treatment, something rarely discussed in therapy sessions, and appreciated that it encouraged thinking about a person’s own constructs in ways other models do not. One lived-experience participant described it as ‘almost like a little life tool’ for structuring thoughts. Others highlighted the whole-person approach: one practitioner contrasted it favourably with a medical model ‘based on numbers on a paper resulting from laboratory tests’, and both groups praised the balance of difficult past meanings with more positive interpretations, such as reframing the experience of bullying from ‘I am a failure’ to ‘it made me stronger to fight bullies and protect others’.

Yet the study also surfaced important caveats within that enthusiasm. Some participants warned that a strengths focus could feel invalidating for people who struggle to identify anything positive about themselves, particularly some clients with personality disorder diagnoses; as one practitioner with lived experience put it, for some clients the suggestion of any strength meets the response of ‘how dare you say that I’ve got strengths’. Participants also noted that engaging service users to explore strengths could be a hard, possibly distressing and lengthy process given long-held beliefs about self-worth, even if it remains essential for long-term recovery. These nuances matter because they show that even a collaboratively designed, recovery-oriented tool cannot simply be dropped into practice without attending to timing, support and the individual’s readiness to engage.

The second and third themes explain why a well-received model nonetheless failed to take root. Ward staff described care-planning as a demoralising tick-box exercise, with one saying she found forwarding care plans to other organisations ’embarrassing’ and another suggesting formulation could actually replace care-planning altogether. Participants proposed strategies for ‘selling’ the change: positioning the model as building on existing practices rather than replacing them, emphasising its value in improving confidence with risk, and making sure junior staff felt included. But the barriers were formidable. The medical model remains deeply embedded in inpatient settings, where, as one practitioner observed, staff are ‘busy putting out fires’ and ‘don’t have time to look for the source’, and where the prevailing attitude can be that patients ‘just need to check with the meds and they’ll be fine’. Nurses worried formulation was ‘another thing to do’ that psychologists should handle, and some participants described a culture in which pushing unpopular change could carry real professional costs, with one interviewee bluntly stating that sometimes ‘you need to shut up’ to keep your job.

Risk aversion emerged as a particularly consequential barrier. Staff feared that conversations about past experiences might destabilise service users or open up distress they felt unequipped to manage, and several suggested careful attention to timing and available support before introducing the model. This finding resonates with broader evidence: a systematic review by Read and colleagues found that only zero to twenty-two per cent of service users reported being asked about childhood adversity by adult mental health services, despite widespread recognition that many have experienced trauma. The authors also note that no participants in the focus group had actually adopted the model in practice by the time of the discussion, apart from one who intended to try it with new service users, an implementation gap the researchers interpret as the study’s most important finding.

The study has limitations that the authors acknowledge candidly: a single researcher conducted the qualitative analysis, samples were small and drawn from one inpatient site, and some survey wording, such as asking how likely the model was to ‘make improvements’, may have nudged participants towards positive responses. Even so, the Personal Narrative Model stands among the first formulation frameworks co-designed with service users from the outset, and its promising acceptability signals genuine appetite for more contextualised, less medicalised ways of understanding distress. The practical recommendations that emerge, including integrating formulation into existing care processes rather than adding it as an extra task, securing leadership support, tailoring training to staff’s existing knowledge, and planning carefully for risk conversations, offer a roadmap for researchers and practitioners hoping to embed psychological interventions in acute settings where change is hardest and, arguably, most needed.

Subject of Research: Acceptability and implementation of a service-user co-designed psychological formulation model in acute mental health inpatient care

Article Title: An exploration of acceptability of a collaboratively developed model of formulation

Article References: Pringle, A., Totman, J., Van Bortel, T., & Kaminskiy, E. (2026). An exploration of acceptability of a collaboratively developed model of formulation. PLOS Mental Health, 3(9), e0000718. https://doi.org/10.1371/journal.pmen.0000718

Image Credits: AI Generated

DOI: 10.1371/journal.pmen.0000718

Keywords: psychological formulation, mental health, co-production, service user involvement, recovery model, acute inpatient care, medical model, Power Threat Meaning Framework, implementation barriers, epistemic justice, PLOS Mental Health, qualitative research