Inquiry regarding Confounding Medication Factors in the ODDESSI Trial
Dear Dr. Pilling and colleagues,
First, congratulations on the publication of the multisite cluster-randomised ODDESSI trial (sciencedirect.com) in The Lancet Psychiatry. It provides crucial insights into the implementation of Open Dialogue within the NHS framework.
I would like to raise a question regarding a potential confounding variable that might elucidate the null result found in your primary outcome (time to relapse). Specifically, what were the baseline and ongoing medication rates—particularly for neuroleptics—for both the Open Dialogue and Treatment as Usual (TAU) groups?
As reported by Bergström et al. (2018) in the Finnish context, neuroleptic onset rates varied dramatically between Open Dialogue (20.4%) and TAU (70.1%). If the ODDESSI trial mirrored these clinical frameworks, it raises an important question: if TAU involves significantly higher medication loads yet yields no superior benefit regarding relapse prevention, could this suggest that standard medication pathways are being over-utilized without clear clinical advantages for this outcome? Conversely, if medication rates were similar between the two groups in this UK cohort, that would also alter how we interpret the structural effects of the Open Dialogue intervention itself.
Could you clarify whether these medication metrics will be published in a secondary analysis, or provide insight into how concomitant psychiatric medication was monitored and adjusted during the two-year follow-up?
Thank you for your valuable contribution to the field.
Kind regards,
Walter Keim, 22. September 2026 Netizen: http://walter.keim.googlepages.com
Folow-up 20. October 2026: Letter to the Editor, The Lancet Psychiatry “Confounding medication factors in the ODDESSI Trial” which is in progress and will be published. Online First.
