Exposure Therapy in Mixed Reality for Obsessive-Compulsive Disorder: A Randomized Clinical Trial

JAMA Netw Open. 2025 May 1;8(5):e2511488. doi: 10.1001/jamanetworkopen.2025.11488.

ABSTRACT

IMPORTANCE: Exposure therapy in virtual reality (VR) is well established for anxiety disorders. While early studies show promise for VR-based exposure and response prevention (ERP) in obsessive-compulsive disorder (OCD), a key challenge of traditional VR is the weak sense of presence, deemed crucial for effectiveness.

OBJECTIVE: To evaluate the feasibility and efficacy of ERP in mixed reality (MERP).

DESIGN, SETTING, AND PARTICIPANTS: This randomized clinical trial (RCT) included 2 outpatient treatment groups, MERP vs self-guided ERP (SERP) from March 15, 2022, to October 26, 2024. Participants were required to be 18 to 80 years of age and have contamination-related OCD (C-OCD). Exclusion criteria consisted of schizophrenia or bipolar disorder, severe substance use disorder, acute suicidality, and ongoing inpatient treatment. Patients’ symptom severity was assessed by the Yale-Brown Obsessive Compulsive Scale (Y-BOCS) before (baseline) and after the 6-week intervention period as well as 3 months after posttreatment (follow-up).

INTERVENTION: Six sessions of MERP or SERP.

MAIN OUTCOMES AND MEASURES: Main outcome was the severity of OCD symptoms in the past 7 days on the Y-BOCS. Sense of presence and subjective appraisal were the secondary outcomes.

RESULTS: Thirty-six participants fulfilled inclusion criteria. Twenty-four participants (66.7%) were female; mean (SD) age was 35.42 (14.03) years. Both groups had severe OCD symptoms, with major depressive disorder as the most common comorbidity (23 [63.9%]). Twenty-nine participants (80.6%) completed the postintervention and follow-up assessments. At baseline, patients in the MERP group had a mean (SD) Y-BOCS score of 26.94 (5.83), while the SERP group had a mean (SD) score of 24.22 (4.12). The decline in patients’ OCD symptoms (Y-BOCS) was similar in the MERP (mean difference, 3.15) and the SERP groups (mean difference, 1.47) (ηp2 = 0.002; 95% CI, -3.187 to 3.893). Significant improvements were observed within the MERP group from baseline to post treatment across several psychopathological variables, including OCD symptoms, with a medium to large effect (Cohen d, 0.584-0.931; 95% CI, 0.026-1.551). Sense of presence was moderate (mean [SD], ≤3.24 [≤2.07]). Subjective appraisal was heterogeneous.

CONCLUSIONS: This RCT on MERP in C-OCD provided important insights. While there were no significant between-group differences, changes in psychopathology within the MERP group suggest some level of effectiveness. However, the limited impact indicates that the full potential of MERP has not yet been realized, and future research should aim to improve immersion and overall effectiveness.

TRIAL REGISTRATION: German Clinical Trial Registry: DRKS00020969.

PMID:40392553 | DOI:10.1001/jamanetworkopen.2025.11488