Dramatherapy in Mental Health Services: Wards, Forensic Units and Evidence

Dramatherapy is more established in secure and inpatient mental health than in any other UK setting, and it became established there long before there was controlled evidence to justify it, which is the honest starting point for anything else this page says. Wards and forensic units employed dramatherapists because the work appeared to reach people that nothing else on the timetable reached, and the research question was asked much later, in a setting that is close to the hardest place in the health service to run a trial.

I have never worked in an inpatient or secure mental health setting, and nothing on this page comes from having been in one. My settings are two care homes and a stroke rehabilitation unit, and the distance between those and a psychiatric intensive care unit is not a matter of degree. So the descriptions here are drawn from the profession’s published standards, from the statutory register and from the reviewer who checks these pages, and the one thing I can offer at first hand is adjacent rather than central.

That adjacent thing is a conversation I had in a corridor, and I have thought about it more than most of the reading. An activities coordinator I met through a shared training day had previously worked on an adult mental health ward. She told me that the hardest part of the job was not the work itself but the paperwork around it, because the form she completed at the end of a session offered one box, and into that box went both the afternoon she spent playing cards with a man who had not spoken for three days and the ninety minute group run by a registered therapist with an assessment behind it. Same box, same category, same weekly total. She said something I have quoted since: “If the form cannot tell the difference, the commissioner cannot either, and then the first thing cut is whichever one costs more.” She was not complaining about her own status. She was pointing at a measurement problem that runs right through this field.

Where an efficacy claim appears below it carries a tier: Tier 1 supported, Tier 2 promising but limited, Tier 3 practice and anecdote. The reasoning behind the labels, and how to check one, is on reading arts in health research.

Where the work is offered

Dramatherapy in UK mental health services appears in adult inpatient wards, psychiatric intensive care units, low, medium and high secure forensic services, community mental health teams, early intervention services in places, and prisons. Secure and inpatient services are the centre of gravity.

That distribution is a historical fact as much as a clinical one. Secure services have been employing arts therapists for decades, in part because they were among the few services with the time, the estate and the staffing model to sustain a weekly group over a long period. Community mental health teams, by contrast, are organised around shorter contacts and higher caseloads, and a twelve month group is a difficult thing to build inside that. So the profession clustered where the conditions suited it, and the general picture on wards is covered more widely on arts in mental health inpatient units.

Availability is patchy in the way everything in this field is patchy. In July 2026 the Health and Care Professions Council register listed 6,103 arts therapists across all four protected titles¹, with dramatherapists a minority inside that total. A trust may have one dramatherapist covering several units, or none, and the presence or absence has more to do with a historical post surviving a reorganisation than with any assessment of need. The profession’s own body is the British Association of Dramatherapists, which publishes practice standards and a members directory². What the modality is, in general terms, is set out on dramatherapy, and the regulated field it belongs to on art therapy and art psychotherapy.

Why secure settings in particular

Four things about secure and long stay settings make this kind of work fit them, and none of the four is about evidence. They are about the shape of the situation.

The first is time, and specifically the wrong kind of it. People in secure services are frequently there for years rather than weeks. A long admission contains an enormous quantity of unoccupied hours, and unoccupied hours in a locked building are not neutral; they are actively corrosive. Something that reliably fills ninety minutes a week for two years is a different proposition from something that fills ninety minutes a week for a six week admission.

The second is engagement. A substantial proportion of the population in these services has a long history of being asked to talk about themselves by professionals, often in circumstances where the answers had consequences. Talking therapy in that context can be a very hard sell, and not irrationally so. A method that lets somebody work through a story, a role or a set of objects, without having to make a direct first person disclosure to a member of staff who also writes reports about them, is reaching a genuine obstacle rather than dressing up a preference.

The third is that group work is doing a rehabilitative job as well as a therapeutic one. Turning up, tolerating other people, taking a turn, managing frustration when somebody interrupts, coming back the following week after a bad session: those are the capacities a person will need to live outside, and they are hard to practise anywhere else in a locked building.

The fourth is simply that the work is portable. It needs a room, a therapist and some objects. It does not need a kiln, a music technology suite or a budget for materials, which matters more in these settings than anybody likes to admit.

None of that is an argument that it works. It is an argument that it fits, which is a different claim, and confusing the two is how a plausible practice acquires an unearned reputation.

What the work looks like inside a locked setting

The clinical shape is recognisable from any other dramatherapy, and the constraints around it are not. Almost everything distinctive comes from the fact that the group members live together and will still be living together tomorrow.

Risk assessment comes before anything. Who can be in a room together, what is in the room, what happens if somebody stands up and moves towards the door, what the therapist does if an alarm sounds elsewhere in the building. Objects that are unremarkable in a community setting are not unremarkable here, and a therapist’s box of materials is agreed with the clinical team rather than assembled to taste.

What can and cannot be enacted is the boundary the whole practice depends on. Working through a role, a story, a myth or an arrangement of objects is one thing. Staging the actual event, whether that is somebody’s offence or somebody’s trauma, is another, and the difference is treated as fundamental rather than as a stylistic choice. The reasons are obvious once stated: the person may be overwhelmed, the group contains other people with their own histories, and the material may be legally live. A therapist who cannot hold that boundary under pressure has no business running a group in a secure unit.

Interruption is normal and has to be planned for. A group can lose a member to an incident elsewhere on the ward, to an unexpected escort, to a tribunal date, or to somebody becoming unwell in the room. What happens next is part of the work rather than an interruption to it: the group is closed properly, people are brought out of whatever they were in, and the therapist does not simply leave the building.

The therapist’s own supervision is not a formality. Clinical supervision is a registration requirement across the arts therapies, and in forensic work it does a specific job, because a practitioner holding highly distressing material weekly, inside an institution with its own pressures, is at genuine risk of absorbing both. The scaffolding around all this, which is the thing that separates a therapy group from an activity, is described for the visual modality on what happens in an art therapy session, and most of it transfers.

Therapy and occupation, and the box they share on the form

The distinction a patient and a family most need is the one the timetable is least able to give them: whether the thing offered on Thursday is a clinical intervention or an activity. Both are legitimate. They are arranged differently, delivered by differently qualified people, and answerable to different standards.

A therapy group has an assessment behind it, written goals, notes in a clinical record, clinical supervision, a registered practitioner and a planned ending. A ward activity has none of those and does not need them. It is open to whoever comes, it exists to make a long day bearable, and on many wards it is the single most humane thing on offer. I run the community equivalent myself in care settings and I would defend it against anybody. It is not therapy, and calling it therapy does it no favours in the long run, because the moment it is described in clinical language it becomes answerable to clinical evidence it was never designed to produce.

The coordinator’s remark about the form is the practical version of this. When both things are recorded identically, three consequences follow. Families cannot tell what their relative is being offered. Services cannot demonstrate what the specialist post is contributing that the general one is not. And when a budget round arrives, the more expensive of two apparently identical lines is the one that goes. That sequence is not a hypothetical. It is the ordinary mechanism by which arts therapy posts disappear from services that still believe in them. The same line is drawn from the other direction on art therapy against an art class.

Why this page quotes no effect size

There is no effect size for dramatherapy in mental health services on this page, and that is a decision rather than a gap somebody forgot to fill.

For a figure to appear here it would need to arrive with all of it: a systematic review or an adequately powered randomised controlled trial, a named design, a named outcome in a named population, an effect size, a confidence interval, a certainty rating, a sample size and a year. Nothing meeting that description has been verified and locked in this site’s source document for dramatherapy in inpatient, secure or forensic settings, so nothing is quoted. Producing a number from a sector summary or from memory would be precisely the failure the rest of the site documents.

The consequence is Tier 3, and the direction is the whole point. This is Tier 3 of the kind that means the adequately powered controlled study has not been done, not the kind that means it was done and came out null. Read as a verdict it would be simply wrong.

It is worth saying why the file is empty here, because the obstacles are unusually severe rather than merely typical. Recruiting a randomised sample inside secure services is hard, and consent in a detained population is a serious ethical problem rather than a form. Blinding is impossible; everybody in the room knows what is happening. Choosing a comparator is genuinely difficult, because the obvious one, an attention matched activity group, is itself a plausible intervention rather than an inert control. Admissions run for years, so a meaningful follow up period is long and expensive. The workforce is tiny, which makes multi site recruitment slow. And the outcomes clinicians in these services care about, such as whether somebody can tolerate a group at all, are not what standard symptom scales measure. Every one of those is a reason for absent evidence that has nothing to do with whether the work helps anybody.

The borrowing trap, and the trial that is not about this

There is one large, well conducted, directly relevant looking controlled trial in this general space, and the most important thing to understand about it is that it is not a trial of dramatherapy.

MATISSE was a three arm, rater blinded, pragmatic randomised controlled trial of group art therapy as an adjunct for people with established schizophrenia, recruiting from secondary care in four UK centres across 15 sites. It randomised 417 people: 140 to group art therapy, 140 to activity groups as the active comparator, and 137 to standard care alone, and followed up 355 (85%) at 2 years. The intervention was weekly group art therapy plus standard care for 12 months, in groups of up to eight, 90 minutes each, delivered by registered art therapists to nationally agreed standards. The co primary outcomes were global functioning (GAF) and mental health symptoms (PANSS) at 24 months³.

Comparison at 24 months	Adjusted mean difference	95% CI
Art therapy against standard care, GAF	-0.9	-3.8 to 2.1
Art therapy against standard care, PANSS	0.7	-3.1 to 4.6

There was no difference between the three arms on either primary outcome, and no secondary outcome favoured art therapy. The single secondary difference that was found ran the other way, with the activity group arm showing fewer positive symptoms at 24 months than the art therapy arm. A secondary analysis found that attendance at art therapy groups was not associated with improvement⁴. And the figure that must always travel with the result: only 86 of the 140 people randomised to art therapy, 61%, attended even one group (73 of 140, 52%, in the activity arm). That is what makes it a pragmatic test of referral rather than of therapy received, and it is the honest limitation to state rather than a reason to set the finding aside.

On this site that is Tier 3, of the null controlled evidence kind, not the nobody has looked kind, and those words are used deliberately. It is a stronger statement than an empty file, and it is a statement about a narrow thing: referral to 12 months of weekly group art therapy, as an adjunct, for people with established schizophrenia, on those two outcomes. It says nothing about individual art therapy, nothing about other diagnoses and nothing about first episode psychosis. The full account is on art therapy for psychosis.

Now the trap. It is extremely tempting, in a page about dramatherapy on a ward, to reach for MATISSE, because it is large, it is British, it is in the right kind of service and it involves a registered arts therapist running a weekly group. Reaching for it would be wrong in both directions. Using it as support is impossible, since it found nothing. Using it as refutation of dramatherapy is worse, because dramatherapy was not in the trial, was not the intervention, and shares with the tested intervention only a regulator and a room. Two modalities on the same statutory register are not interchangeable evidence for one another any more than two drugs in the same cupboard are. Transferring a result across a boundary it never crossed is the single commonest error in this literature, and doing it in the sceptical direction feels rigorous while being exactly as unfounded as doing it in the promotional direction. The same discipline applies to art therapy for trauma and PTSD, where the population overlaps heavily with forensic services and the evidence is again its own separate question.

Consent and capacity when somebody is detained

This is the hardest consent problem anywhere on this site, and it has three layers that get collapsed into one.

The first layer: detention is not consent. Being held under mental health legislation authorises detention and specified treatment. It does not make a person available to everything else a service offers. A detained patient can decline a dramatherapy group, can walk out of one halfway through, and can decline again the following week, and none of that should be recorded as a refusal of treatment or held against them in a review.

The second layer is capacity, which is decision specific rather than a global status. A person may have capacity to decide about a therapy group and not about something else, or the reverse, and it can change week to week. That is a formal assessment carried out properly, not a judgement made in a doorway by whoever is holding the clipboard.

The third layer is the one that goes wrong most often, and it is about images, recordings and audiences. A photograph of a session, a filmed excerpt for a service’s annual report, a piece of work shown at an event, a performance to other patients or to staff: each of those is a publication, and each needs consent of its own that is separate from consent to attend. In a secure setting the stakes are much higher than elsewhere, because the population is identifiable, the association is stigmatising, some people are subject to reporting restrictions, and a recording outlives the admission by decades. Somebody who lacks capacity cannot consent to being described, photographed or recorded, and a relative usually cannot consent on their behalf for publication. The whole framework is set out on consent and safeguarding in arts in health, and in this setting it is a live constraint on practice rather than a paperwork exercise.

Staffing reality, and what to ask

What a service can offer is set by whether a post exists, and posts in this profession are few, historically contingent and easy to lose.

A single dramatherapist covering several wards can realistically run a small number of groups a week, plus assessments, notes, supervision, multidisciplinary meetings and the reporting that keeps the post alive. That is the arithmetic. It means most people on most wards will not be offered dramatherapy, not because they were assessed as unsuitable but because there is nobody to deliver it, and it means a service losing one person can lose the entire provision overnight. Training routes and why the pipeline stays narrow are covered on training as an art therapist.

If you are trying to understand what a relative is actually being offered, five questions get you most of the way. Is there a registered arts therapist on the team, and in which modality. Is the session on the timetable a therapy group or an activity session, and who runs it. Was there an assessment before the offer was made, and are there goals. What happens if the person declines, and will that be recorded as a refusal of treatment. And what would happen to any photograph, recording or piece of work produced. A service that answers those five clearly is a service that has thought about them.

One last thing, and it is the thing the corridor conversation left me with. The absence of research about a practice is not the same as the absence of judgement about it, and the people who have spent twenty years watching these groups on locked wards have judgement worth listening to even though it is not evidence. Hold both of those at once. And whatever a ward offers or fails to offer, no arts therapy is a substitute for the treatment somebody is admitted for; it sits alongside it, and any service describing it otherwise has stopped being careful.