The Talk Therapy Assumption
If someone witnesses a gruesome death, survives an assault, or pulls a mangled body from wreckage, what should they do next? Talk to someone. That is the reflex, and it feels so obvious that we have built entire institutions around it: crisis intervention teams dispatch counselors within hours of mass-casualty events, employers mandate debriefing sessions after workplace trauma, and the entire architecture of acute PTSD prevention rests on the premise that processing trauma through language is the correct first response.
A 1980s puzzle game suggests otherwise.
The Cognitive Vaccine
Emily Holmes, a clinical psychologist now at Uppsala University, has spent nearly two decades developing what she calls a "cognitive vaccine" against traumatic flashbacks.
The theory begins with a basic property of human memory: the brain maintains separate channels for verbal-conceptual processing and visuospatial imagery, and flashbacks belong squarely to the visuospatial side. They ambush you as vivid sensory snapshots — the blood on the floor, the sound of impact, the expression frozen on a face — images, not narratives, and that distinction turns out to be everything.
If you could saturate the brain's visuospatial channel during the critical window when a trauma memory consolidates, you might disrupt intrusive image formation without erasing the factual memory of what happened. The person would still know what happened; they just wouldn't keep reliving it at 3 a.m.
Holmes needed a task demanding enough to flood that channel. She chose Tetris.
From Lab Bench to Lancet
The original 2009 experiment was small and used a film of graphic injuries as an analog for real trauma. Participants who played Tetris for ten minutes reported significantly fewer flashbacks the following week than those who sat idle. Those who played a verbal quiz game fared worst of all (Holmes et al., 2009).
The finding could have stopped there. It didn't. Holmes and her collaborators replicated the result across emergency cesareans, reconsolidation paradigms, and real emergency departments between 2015 and 2018, each study closer to clinical reality. At John Radcliffe Hospital, 71 motor vehicle accident survivors who recalled the crash and then played Tetris for 20 minutes logged 62% fewer intrusive memories than those receiving standard care (Iyadurai et al., 2018).
Then came the definitive trial, published in The Lancet Psychiatry in March 2026. The GAINS-02 study enrolled 99 UK healthcare workers exposed to traumatic events during COVID-19: witnessing deaths, performing futile resuscitation, delivering devastating news to families, all while reporting at least three intrusive memories per week at baseline.
Participants were randomized 2:2:1 into three groups, with the intervention group recalling a specific intrusive memory, practicing mental rotation of 2D and 3D shapes, then playing a modified version of Tetris for 20 minutes in a single guided session, after which they could repeat the exercise at home using a digital platform. The active control listened to Mozart and podcasts; the third group received standard care.
The results were stark. Zero versus ten. At four weeks, the Tetris group reported a median of zero intrusive memories while both control groups reported approximately ten; at six months, 70% of the Tetris group had no intrusive memories at all, with improvements extending to broader PTSD symptoms.
Why a Video Game Beats a Conversation
The mechanism is specific, not magical, and it hinges on a bottleneck in the brain's architecture. Tetris demands visuospatial attention: rotating shapes, tracking spatial relationships, anticipating how blocks will interlock. This engages the same neural resources the brain uses to consolidate visuospatial memories, and playing Tetris during or after memory recall creates a resource conflict that the brain cannot resolve by doing both at once. The factual record survives, but the sensory-perceptual component loses its vividness and eventually its involuntary trigger.
Verbal tasks accomplish nothing here, and may actively do harm, because talking about the event engages the verbal-conceptual channel while leaving the visuospatial channel free to keep cementing intrusive images. This is why the pub quiz group in the 2010 study had more flashbacks than the no-task control (Holmes et al., 2010). Critical Incident Stress Debriefing, the dominant post-trauma intervention for decades, is essentially a structured conversation, and a Cochrane review of 15 trials found it can worsen PTSD symptoms.
The Strongest Case Against
The most serious critique is about generalizability: the GAINS-02 sample of 99 is modest for a definitive claim, and all participants were UK healthcare workers with high baseline education, digital literacy, and motivation to engage with a research protocol. Combat veterans, refugees, and survivors of sexual violence face different trauma profiles and different barriers to engagement, and their response to a brief video game intervention might differ substantially. Holmes herself emphasizes the need for larger, more diverse trials before this can be considered standard.
There is also the question of mechanism versus engagement: the active control (listening to Mozart) matched time commitment but not visuospatial demand, so some portion of the effect could reflect the specific cognitive engagement of any absorbing gameplay rather than the theoretical mechanism of visuospatial interference, and only a head-to-head comparison with another visuospatial task that is not Tetris would settle the question.
What We Didn't Prove
The trial was not fully blinded. Participants knew whether they were playing Tetris or listening to music, though they were masked to which was the experimental treatment, and this matters because no physiological biomarker for flashbacks exists: all measurement relies on self-report diaries, and a participant who expects the game to work may unconsciously underreport.
Ninety-nine sufficed for a Bayesian adaptive design to reach its stopping criteria, but not to detect rare adverse effects or subgroup differences by trauma type. The intervention's durability beyond six months is also unknown.
The Bottom Line
Seven studies across seventeen years, from a 40-person lab experiment to a randomized clinical trial in The Lancet Psychiatry, converge on the same conclusion: playing Tetris after recalling a traumatic memory reduces intrusive flashbacks. The 2026 trial reduced them tenfold, and 70% of participants reached zero. The mechanism is real. It is grounded in cognitive neuroscience, not faith.
An estimated 70 million people worldwide have PTSD. Nowhere near enough trauma therapists exist to treat them. A digital intervention deliverable on a phone, by non-specialists, at scale, is not a curiosity. It may be the most important treatment innovation in trauma psychology since EMDR.
What You Can Do
If you have experienced something traumatic recently, the research supports a specific sequence: briefly recall the memory, then play Tetris or a similar game demanding visuospatial attention for at least 20 minutes. This appears most effective within six hours but may work days or weeks later via memory reconsolidation. Standard Tetris is freely available on most platforms. This is not a substitute for professional care, but it is an evidence-based first step.
If you work in healthcare, emergency services, or any role with routine trauma exposure, consider this approach as a self-care tool after difficult shifts. The GAINS-02 participants used the intervention independently at home after the initial guided session.
If you manage crisis response teams, this evidence challenges the default protocol of immediate verbal debriefing, because visuospatial tasks may protect against flashback formation while structured conversations may inadvertently facilitate it.