Dose by Design: The Long Road From Playtest to Prescription

A game in a clinical trial does not get a review score. It gets an outcome measure, a protocol number and a file full of consent forms. That is the strange new reality for a handful of UK studios who have found themselves building interactive software for NHS trusts, research teams and rehabilitation services — and discovering that the gap between a lovely build and a prescribable product is wider than any early-access roadmap.

The pitch sounds familiar: a game that helps with anxiety, a game that supports attention, a game that makes repetitive physio exercises less grim. The difference is what happens next. Instead of a store page and a launch trailer, there is a ethics application, a device classification question, a data protection impact assessment and a clinical trial that may take years to report. This is the quiet corner of the UK games industry where the word “engagement” means something closer to adherence, and where “fun” is a means rather than the whole point.

The Build Is Not the Hard Part

Most studios can make something that feels good for twenty minutes. The clinical pathway asks for something that works for eight weeks, in a patient’s front room, on a tablet that may be five years old, with a therapist checking in once a week. Suddenly the design constraints are not about retention curves but about session length, fail states, readability, safeguarding and whether the thing can run on the NHS’s more modest hardware.

Then comes the regulatory fog. Software that claims to diagnose, treat or manage a condition can fall under medical device rules, which means a conversation with the MHRA about classification, conformity and UKCA marking. Even before that, NHS teams often want a Digital Technology Assessment Criteria, or DTAC, pass — a checklist covering clinical safety, data protection, interoperability and usability. It is not glamorous. It is also the difference between a pilot that ends with a report and a product that can be bought, supported and used at scale.

What Evidence-Based Actually Means

In games, “evidence-based” often gets used loosely. In clinical research, it has a narrower meaning: a pre-specified question, a control condition, validated outcome measures and a result that someone else can scrutinise. For a game aimed at anxiety, that might mean randomising participants to the game plus usual care versus usual care alone, then measuring change on a standard scale such as the GAD-7. For ADHD, it might involve parent and teacher ratings, not just self-reported focus. For rehabilitation, it could be range of motion, repetition counts or a timed walk.

The awkward problem is blinding. You cannot easily give someone a placebo version of a game without them noticing. Researchers sometimes build a sham build — similar in look and time commitment, stripped of the hypothesised active ingredient — but that is expensive and philosophically messy. A game is not a pill. It is a bundle of feedback loops, narrative, sound, pacing and habit. Working out which bit is supposed to be doing the clinical work is half the design job, and it often starts before a line of code is written.

This is also why the UK’s research infrastructure matters. NIHR funding, academic partnerships and NHS trusts with digital health teams can give a studio access to patients and clinical expertise it could never buy. The trade-off is time. A trial may run for two years and report in a third. Studios used to six-month milestones have to learn to sit still.

Playtesting With Patients, Not Players

A conventional playtest asks whether the controls feel right and whether the loop is clear. A clinical playtest asks whether a tired parent can set it up after a hospital appointment, whether an older patient can read the text without their glasses, and whether a therapist can explain it in five minutes without a manual. The recruitment is different too: ethics committee approval, informed consent, safeguarding protocols and clinical supervision. You are not testing a game on “users”. You are testing it with people who may be vulnerable, in a setting where a crash is not just annoying but potentially distressing.

The best teams treat this as design research rather than compliance theatre. They watch how a physio actually hands over the tablet. They notice that the game is too loud in a shared ward, or that the reward animation is too long for someone with fatigue. They learn that “daily streak” language can feel like a reprimand to a patient who missed a week because of a flare-up. These are small interface decisions with clinical consequences.

The App Graveyard Problem

Britain already has a long queue of wellness apps that were going to change everything and now sit unmaintained on an app store. The reasons are rarely dramatic: funding ran out, the original developer moved on, a mobile operating system update broke the build, or the NHS trust that piloted it could not find a budget line for renewal. A prescribed game needs a maintenance plan as much as a launch plan. It needs clinical ownership, a support route and a realistic answer to the question: who pays for version 2.3 in three years’ time?

That is why the procurement conversation matters as much as the trial result. NICE evaluation, NHS supply frameworks and local commissioning are not the sort of things that make a headline, but they determine whether a game becomes a tool or a pilot project. If a clinician cannot prescribe it without inventing a new workflow, it will not stick. If a trust cannot update it without a procurement fight, it will quietly disappear.

The Triage

The promising thing about the current wave is that it is less naive than the last one. Studios are talking to clinicians earlier, building accessibility in from the start and treating evidence as a design constraint rather than a marketing badge. The less promising thing is that the UK’s clinical pipeline is still slow, fragmented and heavily dependent on short-term funding.

So the question is not whether a game can be good enough to help. It is whether the system can be good enough to keep it alive. The real test of a prescribed game is not the launch-day download graph. It is whether a therapist can hand it over in a clinic in, say, Salford or Swansea, trust it to work on the kit in front of them, and still trust it six months later. That is a different kind of score, and the UK studios taking it seriously deserve more than a shrug.