Background: People with psychosis have high rates of post-traumatic stress disorder (PTSD), which is associated with a poor prognosis. We aimed to investigate the effectiveness of a trauma-focused therapy integrated with cognitive behavioural therapy for psychosis (CBTp) in people with psychosis. Methods: STAR was a rater-blind, parallel-group, pragmatic randomised controlled trial conducted across five UK sites. We randomly assigned (1:1) adults with co-occurring PTSD and psychosis in secondary care to trauma-focused CBTp plus treatment as usual or treatment as usual only, using randomly varying block size, stratified by site. Trauma-focused CBTp is a flexible, individualised, formulation-based therapy integrating trauma-focused therapeutic components with CBTp, lasting 9 months. The primary outcome was PTSD symptom severity, assessed using clinician-administered PTSD scale for DSM-5 (CAPS-5), at 4 months (mid-therapy) and 9 months (primary endpoint). Secondary outcomes were (1) percentage of participants showing PTSD remission from diagnosis and clinically significant change, individual symptom clusters, post-traumatic cognitions, and dissociation; (2) psychosis symptoms; (3) mood disorders; (4) psychological recovery; and (5) social functioning, assessed at the same timepoints. We used linear mixed models in the intention-to-treat population at 9 months, incorporating all outcome data including at 4 months. Lived experience experts were integral throughout the research. This study was registered prospectively with ISRCTN, ISRCTN93382525. Findings: Between Oct 1, 2020, and March 20, 2023, we randomly assigned 305 participants (151 [49·5%] to treatment as usual and 154 [50·5%] to trauma-focused CBTp plus treatment as usual). The mean age of participants was 38·9 years (SD 12·4). 127 (42%) participants were men, 171 (56%) were women and seven (2%) were non-binary or preferred not to say; 232 (76%) were White. All reported repeated and multiple traumas. Of 154 in the trauma-focused CBTp plus treatment-as-usual group, 144 (94%) engaged with therapy and 146 (95%) received a minimal therapeutic dose. 267 (88%) of 305 participants attended a follow-up assessment (4 months, 9 months, or both); at 9 months, 241 (79%) participants provided primary outcome data and 169–248 (55–81%) provided secondary outcome data. There were significant effects on CAPS-5 scores (adjusted mean difference in PTSD symptom severity –8·67 [95% CI –13·41 to –3·94]; p=0·0003; Cohen's d –0·73) and on 22 (81%) of 27 secondary outcomes, unadjusted for multiple testing (Cohen's d ranging from –0·26 to –0·82), including all PTSD outcomes (except derealisation and depersonalisation); delusions, paranoia, and hallucinations in other modalities; suicidal ideation, depression, anxiety, and stress; and psychological recovery. There were no effects on voices, referential beliefs, substance use, or social functioning (Cohen's d ranging from –0·05 to –0·28). 62 (50%) of 125 participants in the therapy group showed PTSD remission compared with 25 (22%) of 116 in the treatment-as-usual group (odds ratio [OR; PTSD present] 0·11, 95% CI 0·03–0·32; p=0·0001) and 56 (45%) participants in the therapy group compared with 31 (27%) in the treatment-as-usual group had clinically significant change in CAPS-5 score (OR 4·45, 95% CI 1·59–12·44; p=0·0044; number needed to treat for PTSD remission was 4). No serious adverse event was unexpected and related to trial procedures (78 reported in each group), with the most common being physical illness or injury. Interpretation: Trauma-focused CBTp is safe, highly acceptable, and effective in people with co-occurring psychosis and PTSD. This underserved population should no longer be denied access to psychological interventions to treat their trauma sequelae. Funding: UK National Institute for Health and Care Research (Health Technology Assessment).
CORE (COnnecting REpositories)
CORE (COnnecting REpositories)