Therapeutic Writing and Photovoice: Useful, Unregistered and Often Mislabelled

Expressive writing and photovoice are two well established practices that sit next to the registered arts therapies on a great many programme lists, and neither of them is a registered therapy, has a protected title, or was designed to do what the therapies do. Both are worth having. Both are routinely mislabelled, and one of them carries a consent problem serious enough that it should be the first thing anybody reads about it.

I do have a first hand stake in this one, because songwriting is the closest thing I run to expressive writing. In a care home group a few years ago a man in his eighties wrote four lines about his wife, who had died the previous winter. He had not said much in eight weeks of sessions. The lines were plain and very good, and the group set them to a tune, and everybody in the room understood that something had happened. Afterwards, while the chairs were being stacked, he asked me for the piece of paper back. Not for a copy. For the paper, and for the promise that it would not be sung again. I gave it to him and said yes, and then I went home and thought about the fact that I had already read it aloud to nine people, that two of them had suggested changes which were now in it, and that I had been planning, entirely cheerfully, to include it in the home’s newsletter as an example of what the group could do. Nothing in my training, such as it was, had ever raised the question of who owned that verse. I have run the group differently ever since.

The second thing that changed how I work happened later and involved a photograph. A resident’s daughter attended one session, took a picture on her phone of her mother singing with two other residents, and emailed to ask whether I would circulate it to the group and put it on the home’s Facebook page, because her mother looked happier in it than in anything she had seen for a year. It was a lovely photograph. I said no to the Facebook page and no to circulating it, and I offered instead to ask each of the three families individually whether they were willing. Two said yes. One never replied, which is its own answer. The daughter was not doing anything wrong, and her request was generous rather than careless. She just happened to be asking about three people, only one of whom she had any standing to speak for, and only one of whom could have told me what she wanted.

Where a claim about effect appears below it carries a tier: Tier 1 supported, Tier 2 promising but limited, Tier 3 practice and anecdote. What the labels require is set out on reading arts in health research.

What expressive writing actually is

Expressive writing, in the form that turns up in health settings, is usually a short, structured, timed piece of private writing against a prompt, done repeatedly over a small number of sessions. It is not creative writing tuition and it is not journalling in the general sense.

The recognisable format is a prompt, a fixed period of continuous writing, and no requirement that the result be shown to anybody, be any good, or be finished. Prompts tend to be either open (“write about something you have not put into words”) or oblique (“describe the room you slept in as a child”), and the obliqueness is doing work: it lets somebody arrive at material sideways. Groups often add a short reading back, always optional, and a deliberate closing that moves people out of whatever they have opened.

The distinctive feature, and the one that matters practically, is that people can and do this alone. There is no equipment and no gatekeeper. A person can pick up a notebook tonight. That is a genuine strength, and it is also the origin of the one standard caution in the field: facilitators generally warn against writing at length and unsupported about a traumatic event soon after it has happened. The reason given is not mysterious. Writing can open something at eleven at night in an empty kitchen, and there is then nobody in the room to help close it. A structured group with an experienced facilitator, a clear ending and somewhere to go afterwards is a different proposition from the same exercise done alone at speed.

Formats vary a great deal, and so do facilitators. Some are published writers, some are health professionals who have added a facilitation qualification, some are volunteers with a genuine gift and no training at all. There is no register that will tell you which you have.

What photovoice is, properly

Photovoice is a participatory method with a specific origin and a defined protocol, in which participants are given cameras, take the photographs themselves, discuss the images together, and produce an output aimed at people with power over their circumstances. It was built to change something. It was not built to produce a clinical outcome.

The protocol matters, because it is what distinguishes photovoice from a photography group. Participants, not a professional photographer, hold the cameras. The group then works through the images collectively, and the discussion is the method rather than a nice extra: what is in the picture, why it was taken, what it says about the conditions the person is living in, and what should be done. The output is typically an exhibition, a report or a presentation to commissioners, councillors, service managers or funders, with captions written by the participants.

That purpose changes what a project is accountable to. A photovoice project is accountable for whether it reached the people it was meant to influence, whether participants had real control over the images and the framing, and whether anything changed. Recast it as a therapy and all of that quietly falls away, replaced by an outcome questionnaire it was never designed to answer. It also strips out the political point, which for the people who developed the method was not incidental.

None of this makes it unsuitable for use with people who are unwell, and it is used in hospital arts programmes, hospices and mental health services. It means that a service describing photovoice as therapy has misdescribed the method, and that a reader assessing it against clinical criteria is measuring the wrong thing. The line between the regulated and unregulated halves of this field is drawn in full on art therapy against an art class.

Neither one is a registered therapy

There is no protected title for either practice, no statutory register, and no legal requirement that a facilitator hold any qualification at all. That is a fact about regulation and not a judgement about quality.

Set it against the regulated half. In the UK four arts therapy titles are protected in law: art psychotherapist, art therapist, dramatherapist and music therapist, and using one without registration is a criminal offence. In July 2026 the Health and Care Professions Council register listed 6,103 arts therapists across all four titles¹. If somebody offers you art therapy, one search settles whether they are entitled to use the word. If somebody offers you a therapeutic writing course, no search exists, because there is nothing to search.

Unregistered does not mean unskilled, and it certainly does not mean worthless. Some of the most careful practitioners I have met work in exactly these unregulated spaces, and some of the least careful things I have seen were done by people with impeccable paperwork. What changes is the check available to you. Instead of a register, you ask: what training does this person have, who supervises them, are they insured, how long have they run groups like this, what happens if somebody becomes distressed, and what happens to the work afterwards. Those questions are less convenient than a database and they are not nothing. The pillar page on art therapy and art psychotherapy sets out what the regulated route guarantees, and craft making and mental health makes the wider case that the unregulated half is worth funding on its own terms rather than by borrowing clinical vocabulary.

Where these turn up, and who runs them

Both practices appear in four broad places: arts on prescription courses, hospital and hospice arts programmes, community and voluntary sector projects, and research studies.

Arts on prescription is the most common route in England, where a social prescribing link worker attached to a GP practice connects somebody to a time limited course run by a local organisation. How that system is organised is described by the National Academy for Social Prescribing², and read as a description of arrangements rather than as evidence about outcomes. The practical detail is on arts on prescription and what a ten week arts course involves.

Hospital and hospice programmes are the second route. Writing in particular has a long presence in palliative care, where a person may want to leave something behind and where the material is by definition not neutral; that setting has its own page at creative work at the end of life. Care homes are the third, and I would say from my own work that writing based activity in a home is much more likely to be reminiscence shaped than prompt shaped: see arts in care homes. Museums and galleries run both, sometimes as part of a prescribed offer, covered on museum prescriptions.

The fourth route is research, and it is worth flagging separately because it changes the consent position. A photovoice project run as a study has an ethics approval, a participant information sheet, a formal consent process and a named investigator. A photovoice project run as a community activity may have none of those, and the participants may be doing something identical.

The consent trap in photovoice

This is the substantial part of the page, and the reason it exists. Photovoice creates identifiable images of real people who are frequently not in the room and have not agreed to anything, and then places them in front of an audience. Every element of that sentence is a problem.

The people in the pictures are usually absent from the process. A participant photographs their street, their kitchen, the queue at a clinic, their grandchild, the neighbour who checks on them. Those people did not join a project, were not briefed, and in some cases do not know a project exists. The consent that was obtained belongs to the person holding the camera, and it does not extend to anybody in the frame.

An exhibition is a publication. This is the point that is hardest to hold onto in a room full of goodwill, because an exhibition in a health centre foyer feels informal, temporary and small. It is not. It is public display of identifiable images of identifiable people, and once photographs go onto a website, a funder’s report or a social media account, the informality is gone entirely. A launch event with photographs on the wall and a local newspaper present is a publication with a press launch, whatever anybody calls it.

Capacity matters, and a relative usually cannot stand in. Somebody who lacks capacity cannot consent to being described, photographed or recorded, and consent for publication is not something a family member can generally supply on their behalf. This constrains projects in dementia care, in learning disability services and in some mental health settings very tightly, and the correct response is to change the project rather than to find a signature.

Withdrawal after the fact is close to impossible. This is the asymmetry that makes photovoice consent different from most consent. A person can withdraw from a study and their data can be deleted. A photograph that has been exhibited, printed in a report, posted online or shared onward cannot be recalled in any meaningful sense. So consent has to be obtained before the image goes anywhere, has to cover the actual intended uses rather than a general permission, and has to be checked again at the point of exhibition rather than collected once at week one and treated as settled.

Anonymisation is weaker than it sounds. Cropping a face out of a picture of somebody’s own front room, in a project publicly identified with a named neighbourhood and a named condition, protects very little. The context does the identifying.

The method’s own protocols exist because its developers understood all of this, which is worth saying plainly: the discipline is native to photovoice rather than imposed on it from outside. A project that has quietly dropped the protocol has usually dropped it in the name of being friendly and informal, which is exactly how it happens. The general framework, including what a service should have in place before any of this starts, is on consent and safeguarding in arts in health.

The same problem, smaller, in writing

Writing produces the same problem in a quieter form, and the quietness is why it gets missed. A poem or a lyric written in a group is a piece of somebody’s private life that has been made in public.

The questions are: who owns it, who may read it, who may reproduce it, and who may change it. In my experience none of those is settled at the start of a session, and all of them come up eventually. A verse written by one person, altered by two others, transcribed by the facilitator and then reprinted in a newsletter has passed through four sets of hands, each acting kindly. The person whose material it is may not have understood that any of that was going to happen, and may not feel able to object once it has.

The rule I now work to is simple and I wish somebody had given it to me sooner. Nothing leaves the room without an explicit yes obtained after the thing exists, not before. The yes can be withdrawn. If a piece has been worked on collectively, the person whose material it was still decides. And a request for the paper back is not a setback for the session, it is the session working: somebody found the room safe enough to put something into it, and then took responsibility for it. Where the person cannot give that yes because of capacity, the answer is no.

Why this page quotes no effect size

There is no effect size for expressive writing or for photovoice on this page, and that is a decision rather than an omission.

For a figure to appear it would have to arrive complete: a systematic review or an adequately powered randomised controlled trial, a named design, one named outcome in one 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 either practice, so nothing is quoted here. Producing a number from memory or from a summary would be the exact behaviour the rest of the site documents as the field’s central problem.

That puts efficacy claims about both at Tier 3, of the kind that means the study has not been done and verified here, not the kind that means the controlled evidence exists and came out null. The distinction is not pedantry. An empty file and a null result look identical in a one line summary and license opposite conclusions.

Two things are worth knowing about the surrounding literature, precisely so that neither gets mistaken for a figure about these practices. The nearest relevant body of work is the arts on prescription evidence base, and the 2024 systematic review is the honest summary of it: 7,805 records screened, 25 included (10 qualitative, 7 mixed methods and 8 quantitative uncontrolled before and after studies), with the verbatim finding that “No Randomized Controlled Trials (RCTs) were identified in the search”, and most interventions running 8 to 10 weeks³. Its meta analysis did report a statistically significant improvement in wellbeing, and that pooled result rests entirely on uncontrolled designs, which is why it stays Tier 3 here and is never called a trial finding. It is also a result about arts on prescription courses in general, not about writing or photography specifically.

The other is the WHO Europe scoping review, which is the document usually produced when somebody wants the field to look large. It mapped over 900 publications, comprising 200 plus reviews and 700 plus individual studies, and those reviews between them cover over 3,000 studies⁴. The counts nest rather than add: the 3,000 sit inside the reviews, which sit inside the 900. And a scoping review maps a literature rather than pooling it, so it carries no tier and supports no claim about effect whatsoever.

How to find one, and what to ask before you join

Most of these groups are reached through a link worker, a library, an arts organisation, a hospice or a hospital arts team, and most take self referral, so the practical barrier is usually knowing they exist rather than getting past a gatekeeper.

Before joining, four questions are worth asking, and none of them is awkward. Who is facilitating, and what is their background. What happens to anything I write or photograph. Is there an exhibition, a publication or a funder’s report at the end, and can I take part fully without appearing in it. And what happens if I find a session difficult, both during the hour and afterwards. A well run project will have ready answers, because it will have been asked before.

The last point is the one I would leave anybody with. The reason to be careful about labels here is not that the unregistered practices are lesser. It is that they are different, they answer to different standards, and describing them in clinical language sets them up to be judged by a test they were never built to pass. Writing and photography are among the few things in this field a person can start without permission, which is worth a great deal. And none of it replaces treatment: if something is being managed clinically, it stays managed clinically, and a course of any kind sits alongside that rather than in place of it.