Large English trial finds Open Dialogue crisis care cuts psychiatric admission and re-referral, despite no effect on relapse

A NIHR-funded multisite cluster-randomised controlled trial — the ODDESSI trial — tested Open Dialogue, a relational, multidisciplinary-team-delivered model of crisis mental health care, against treatment as usual across 5 NHS mental health trusts and 32 GP-practice clusters in England, with 494 participants stratified by deprivation (Index of Multiple Deprivation) and followed for two years Primary Study (Pilling et al.1). The trial’s primary outcome, time to relapse, showed no significant difference between arms (hazard ratio 0.95, 95% CI 0.67–1.32). Open Dialogue was, however, associated with a significantly lower probability of psychiatric inpatient admission and of re-referral back into crisis or secondary mental health services, alongside better self-rated recovery and service satisfaction, with no safety signal.

This is the first large pragmatic trial to test Open Dialogue at scale in England, and the first to do so stratified by deprivation. Trial sites were London and southern England trusts only; transferability to services with a different population geography, and the workforce cost of training multidisciplinary “reflecting teams,” were not tested by this trial and remain open questions.

Even with a null primary endpoint, the reduction in psychiatric admission and re-referral is directly relevant to acute mental health bed pressure, and represents evidence for piloting or service-development interest in crisis-pathway redesign — not evidence for wholesale replacement of current crisis models, given transferability beyond the trial’s sites is untested.