An eight-session imagery protocol for PTSD in which the patient rewrites the scene
- Source and design. In Yonsei Medical Journal (online March 2026), Kim and colleagues at Hanyang University, South Korea, randomised 86 psychiatric outpatients with PTSD 2:1 to imagery stabilization and rescripting therapy (ISRT, 59 patients) or to CBT without a trauma focus (27 patients). Recruitment ran from June 2016 to June 2019; registration KCT0001918. Both arms had eight weekly individual sessions of 60 minutes, delivered by the same two clinical psychologists.
- Protocol. Sessions 1–4 build stabilising imagery: grounding, a containment exercise, an image of a positive treatment goal, resource and distancing images, and modified imagery rehearsal for nightmares or intrusive images. Sessions 5–6 are modified imagery rescripting of the trauma memory, sessions 7–8 imaginal cue exposure to feared situations.
- Dropout was 28.8% (17 of 59) with ISRT and 25.9% (7 of 27) with CBT, p = 0.782. No participant worsened by more than 15 points on the CAPS-5; the largest increase after treatment was 5 points, in one ISRT patient.
- Completers at week 8. CAPS-5 fell from 29.8 to 18.4 with ISRT (d = 1.38) and from 31.2 to 23.2 with CBT (d = 0.89); between-group effect size 0.40, p = 0.174. Response 55.0% vs 40.0%, loss of diagnosis 57.5% vs 35.0%, remission 30.0% vs 20.0%. None of the between-group differences was statistically significant, after treatment or at six months.
After the fifth session of imagery stabilization and rescripting therapy (ISRT), the patient goes home with a recording of their own voice rehearsing a rewritten trauma memory, to be played at least twice a day. The script went through three versions in the session, each change chosen by the patient together with the therapist.
The session opens with one brief exposure: the patient tells the index memory in the present tense, as a film with a beginning and an end, while the therapist checks distress (SUDS) and the thoughts, emotions and sensations behind it. Then the patient changes the whole story from start to finish, replacing the hot spots, their images and the distressing thoughts, feelings and sensations. After the first version is rehearsed aloud, the patient is asked whether anything should be added. The second is usually longer and carries more positive emotion; after a discussion, the third is rehearsed and recorded. Unlike the original versions of the technique, ISRT does not make mastery or nurturing imagery essential. The therapist regards them as positive themes to reflect in the new script, and the manual stops there.
Four sessions of stabilisation come first, described in the group's earlier open trial: sensory grounding, a containment exercise in which a visual image is created to reduce distress and strengthen self-control, an image of a positive treatment goal, resource imagery from the patient's own coping or from supportive figures, visuospatial changes to intrusive images, and modified imagery rehearsal for nightmares. Sessions 7 and 8 take four feared situations – an activity, a place, people or a sensation – and the patient imagines each with eyes closed as a short film until SUDS falls below 20 out of 100. If it does not, the therapist looks for the reason and suggests altering the image towards something neutral or positive.
The comparison arm was CBT built on stress inoculation training, with a body scan and DBT emotion regulation skills added. Both treatments were given by two clinical psychologists who had co-developed ISRT. Among completers, 57.5% of the ISRT group and 35.0% of the CBT group no longer had a PTSD diagnosis after treatment, and the authors found no statistically significant difference between the groups on any measure.
One habit follows for practice: offer a protective figure or a rescue as a question, and write it into the script only if the patient takes it up, since the patient is the one who will hear the recording twice a day.
In ISRT the patient changes the picture of the memory, and the therapist's themes of mastery and nurturing stay optional in the new script.
The sample was small, 59 and 27 patients, and outcome assessors were not fully blind to the treatment arm. Both treatments were delivered by the same two clinical psychologists, who had co-developed ISRT. Follow-up lasted six months, and medication doses were held stable only during the eight weeks of treatment.