The data
The STAR study (Study of Trauma And Recovery), run jointly by King's College London's IoPPN and South London and Maudsley NHS Foundation Trust, enrolled 305 participants and was published in *The Lancet Psychiatry*. The intervention: a nine-month protocol integrating trauma-focused work directly into cognitive behavioral therapy for psychosis (CBTp).
Outcomes were measured across 27 domains. Twenty-two showed statistically significant improvement. The primary endpoint — PTSD symptom severity — registered a moderate-to-large effect size. Secondary gains included reduced paranoia, fewer multisensory hallucinations, lower depression and anxiety scores, and decreased suicidal ideation. Disengagement from therapy sat at 6.5%, a strikingly low figure for a population historically considered treatment-resistant or fragile in session.
The control arm, receiving treatment as usual, cleared the PTSD threshold in roughly 20% of cases. The integrated protocol doubled that rate to 50%, with similar patterns emerging for complex PTSD.
Why this cohort was excluded — and why that changes now
PTSD prevalence in psychosis populations runs up to five times higher than in the general population. Flashbacks, hyperarousal, and avoidance behaviors become entangled with delusions and hallucinations; traumatic content often shapes the form psychotic symptoms take. Yet this group has been shut out of nearly every major PTSD trial. The clinical rationale was straightforward but unsupported by controlled data: direct engagement with trauma memories might destabilize psychosis.
The STAR results refute that concern at scale. Professor Emmanuelle Peters, the study's first author and lead of the PICuP Clinic at SLaM, described the protocol's core design principles as direct trauma memory work, sustained engagement focus, and flexible individualization. The combination — not any single component — appears to be the active mechanism. Low dropout rates suggest patients found the approach tolerable, which matters as much as efficacy when assessing real-world transferability.
Clinical signal for practitioners
For clinicians managing patients at the psychosis–PTSD intersection, the protocol's structure warrants close examination: integrated delivery rather than sequential treatment, a nine-month arc allowing gradual trauma processing, and explicit tailoring to psychotic symptom presentation. The 22-of-27 outcome improvement pattern suggests benefits extend well beyond PTSD symptom reduction into broader functional and emotional domains — a profile that purely disorder-specific interventions rarely produce.
What remains to be seen: long-term durability of gains beyond the immediate post-treatment window, scalability outside the specialized SLaM setting, and whether the integrated model can be competently delivered by clinicians without dual-specialist training. Until replication data arrives, the takeaway is clinical and specific — this population can and should be offered trauma-focused care, not excluded from it by inherited caution.