Factors associated with the outcomes of a novel virtual reality therapy for military veterans with PTSD: Theory development using a mixed methods analysis

Abstract

Background Multi-modular motion-assisted memory desensitization and reconsolidation therapy (3MDR) is a new psychological intervention for people with post-traumatic stress disorder (PTSD). 3MDR is immersive, delivered in a virtual reality environment, and emphasises engagement, recollection and reprocessing. Objective Through a theory-driven examination of data relating to 10 out of 42 UK military veterans taking part in a trial of 3MDR, the principal objective was to explore the complex interrelationships between people, interventions and context and to investigate how factors within these domains interacted in specific outcome typologies. Method Quantitative and qualitative data relating to 10 trial participants were derived from: researcher-assessed and self-report clinical measures; interviews; physiological recordings; words describing thoughts and feelings during therapy; and subjective unit of distress scores. Using a convergent mixed methods approach, data were tabulated using a person, intervention and context model. Participant summaries were grouped into outcome typologies, followed by an analysis of data convergence and divergence within each and an interpretation of identified patterns. Results Three outcome response typologies were identified: dramatic improvement, moderate improvement and minimal improvement. Within the person domain, factors associated with outcomes included walking capacity, commitment and ability to complete therapy, and levels of subjective distress. Within the intervention domain, factors associated with outcomes related to image selection and use, therapeutic alliance and orientations towards the tailoring of sessions. Within the context domain, factors associated with outcomes included reactions to the therapy environment. The patterning of secondary outcomes broadly corresponded with primary outcomes within each typology. Alongside patterned data differentiating aspects of the person, intervention and context domains, within the three response typologies data also existed where no obvious patterning was detected. Conclusions The model developed here may have novel value in evaluating a range of personalised interventions, but further work is needed before confident assertions can be made of who is likely to benefit from 3MDR specifically.