A randomised controlled trial of eye movement desensitisation and re-processing (EMDR) in forensic services and in prison

Dear editor,

In their editorial recently published in Medicine, Science and the Law, Crole-Rees et al., highlight the high rates of lifetime trauma experienced by people within criminal justice systems, and the negative consequences of this trauma for individuals and society. ¹ The authors point to the limited evidence base regarding the efficacy and safety of trauma-focused therapy in criminal justice settings. They note that while some favourable case reports and pilot studies exist about Eye Movement Desensitization and Reprocessing (EMDR) in prisons and forensic services, that “to date” no such randomised control trials of EMDR have been conducted. We write to correct this and to describe our randomised control trial which was published online in early 2024. ² This trial investigated the efficacy and risks of EMDR for people with PTSD and psychotic disorders who were in prison or who had been diverted from custody into forensic mental health care.

The study was a single-blind randomized controlled trial comparing EMDR therapy to routine care in 24 offenders with psychotic disorders and PTSD. The primary outcome was the mean change in Clinician Administered PTSD Scale (CAPS-5) score. Secondary outcomes included participant-rated PTSD symptoms, psychotic symptoms, social functioning, disability level, self-esteem, depressive symptoms, posttraumatic cognitions, complex posttraumatic difficulties, and adverse events. Blinded investigators assessed outcomes at baseline, and after 10 weeks and 6 months. An in-depth description of the methods can be found in the study protocol. ³

As Crole-Rees et al. ¹ observe, people in contact with the criminal justice system often have complex histories and multiple comorbidities. This was certainly true for participants in our trial. All but one of the participants described multiple lifetime traumas, with mean Life Events Checklist Scores of 31; high scores demonstrating that exposure to multiple traumas across many domains was the norm. Most had complex PTSD. Co-occurring diagnoses like substance use disorders, personality disorders, and mood disorders were common. All participants had come through the criminal justice system, and one-quarter remained in prison at the time of treatment (the rest had been diverted into forensic mental health care through secure inpatient units). Their index offences were significant, for example murder or attempted murder in a third of cases.

In terms of outcomes, mean CAPS-5 scores after 6 months were lower (better) in the EMDR group, 21.3 (SD = 13.3), compared with the control group, 31.5 (SD = 20.7). The point estimate [95% CI] difference, averaged over two measurement times, was 11.4 [1.3, 21.4], p = .028, favouring EMDR. Self-esteem increased in the EMDR group and depressive symptoms and disability reduced. After 6 months, 16.7% of participants in the EMDR group still met the cutoff threshold for a diagnosis of PTSD compared with 36.4% in the wait-list group; relative risk 0.46 (95% CI [0.10, 2.03]). EMDR treatment was well tolerated with mean post-treatment Acceptability/Adherence Scale scores of 53.5 (SD = 8.3) showing positive attitudes toward EMDR. No participants experienced serious adverse events and no significant differences in minor adverse effects existed between treatment and control groups although point estimates favoured EMDR. ²

Crole-Rees et al. ¹ also report the concern some clinicians and researchers hold about using EMDR in the prison environment, fearing it cannot guarantee the physical and psychological safety important for undertaking trauma work. Aligning with this, we found some prison and forensic service staff were resistant to support people enrolling in the trial, concerned that they might decompensate. These concerns, and the COVID-19 pandemic, which closed the recruitment sites to researchers for almost a year, meant that we did not achieve our target sample size. Nonetheless, statistically significant differences between groups emerged. We found that EMDR was safe, effective, and acceptable to our participants.

Crole-Rees et al. ¹ also recommend that qualitative work should be undertaken to understand the views of people in contact with the justice system about the delivery of EMDR in these settings. We have also published a companion qualitative study to our RCT dealing with this topic. ⁴ We examined the experiences of participants in prison or forensic mental health care treated with EMDR. A key theme was the belief EMDR ‘Fits the Forensic Setting’, bringing empowerment in places perceived as disempowering and increasing their hope for violence-free futures. We identified EMDR as particularly suitable for this group because it does not require in vivo exposure, or for people to verbally articulate their traumas, reducing cognitive and emotional burdens.

Overall, our findings support EMDR being an appropriate treatment for PTSD comorbid with serious mental illness in secure settings. While we recommend replication of our findings using larger-scale trials, we do wish to correct the statement that no trial-based evidence exists.