The long-awaited ODDESSI trial – the first randomized controlled trial of Open Dialogue – has at last been published in The Lancet Psychiatry. While this is exciting in and of itself to those of us who have championed Open Dialogue for decades, some have expressed concern that the primary outcome — no significant difference in relapse rate between Open Dialogue (OD) and Treatment-as-Usual (TAU), what is called a null result — will be helpful to those wishing to dismiss Open Dialogue.
A look below the surface, however, tells — appropriately, for anyone who has followed the twists and turns of the other summer blockbuster, Homer’s Odyssey, on the page, stage, and screen — a more complicated story: ODDESSI is the story of a siege, ending with a Trojan horse, and a long journey home.
It is not cynical to say that study results can support a range of conclusions. This has always been true. In fact, it is the essence of science. Studies suggest their conclusions with the best story they can tell from the evidence they have found — a story that makes sense in the world it seeks to understand and, perhaps, change. This is why there is a bias toward publishing positive results; null results seem lackluster and uninteresting, while a positive result seems to say something new, with numbers about which we can feel certain, and stories they tell about which we can be excited.
The effort to adopt Open Dialogue practices is rooted, from a research perspective, in the reports from Tornio, Finland that told of dramatically improved outcomes for psychotic patients treated with Open Dialogue. The results of these studies called into question the paradigms of psychiatric illness and treatment upon which western medicine is based. According to the mainstream paradigm the reported numbers could not be true, and so they were dismissed as invalid in their study design (e.g; they were naturalistic rather than randomized, drew on the wrong research population, etc.) as well as in their analysis.
This, in fact, is the essence of Open Dialogue as well. In a crisis of understanding, we invite all perspectives, with open minds about what the problem might be, in the hope of gathering all the relevant information and finding a way forward together.
Such is the case with the null result from the ODDESSI trial. Life is not reducible to mere numbers, much as our desire for a clear and certain path forward might like it to be. Life is the story we live, not the numbers that aid in describing it.
This is also the essence of Open Dialogue. ODDESSI’s primary finding can be seen as a Trojan horse: a finding that is interesting precisely BECAUSE it appears to have the same relapse rate as treatment-as-usual. As Tomi Bergström, a clinical psychologist from Open Dialogue’s birthplace of Tornio, and author of the nineteen-year outcomes study that first showed that Open Dialogue’s effects could hold over decades, put it simply when the results came out: this study isn’t the end of a story; it’s a beginning.
However the secondary findings tell a deeper story: less than half as many people in the study were hospitalized with Open Dialogue as under treatment as usual. Reported comfort and satisfaction with the Open Dialogue process ran high: Staff retention held at 80%, against 100% attrition on the treatment-as-usual teams. Whatever other conclusions one might reach from the ODDESSI data, it unmistakably shows that there’s an alternative to treatment as usual that people on both sides of the equation want to be part of.
For me that story begins with realizing that it’s impossible to evaluate the study from within one paradigm: The parity in “relapse” rates between Open Dialogue and Treatment-As-Usual tells an entirely different story depending on whether you regard the criteria used to make a diagnosis as symptoms of an underlying medical disorder – a hypothesized but never confirmed position – or as behaviors or experiences from which there may be something else to learn. “Symptoms” that invite a full range of dialogue about their potential meaning may induce people to reach for support when they need it. The potential benefits for individuals as well as society are broad and deep: from dramatic cost savings to accumulated wisdom. The dramatic reduction in hospitalization is especially important in light of the high rate of suicide [pooled estimates put this at more than 20x the general-population rate, highest in the first month, elevated for a year or more — Harvard Review of Psychiatry meta-analysis, Forte et al.; a 2025 Lancet study addressed involuntary discharge specifically] in the period following a psychiatric hospitalization.
In evaluating the results it’s important to note that although the trial is nominally about Open Dialogue, what was tested specifically was Peer-Supported Open Dialogue (POD). The POD model incorporates Open Dialogue’s principles with peer-supported values and practices, along with mindfulness-inspired awareness. These additions help to cultivate awareness and acceptance, within the system, of the struggles and needs of everyone in it — not only of the person at the center of concern.
Early in my own learning about Open Dialogue, I felt that for it to thrive outside the community whose response to crisis Open Dialogue had transformed — in communities where there were people who harbored distrust if not anger at clinicians, and who did not feel represented in or by a system that purported to serve them — it would need to link arms with peer networks.
Criticism of ODDESSI has noted that network meetings were not the predominant mode of interactions in the study. It is important to note, however, that in the fidelity criteria for the elements and principles of Open Dialogue as proposed by Olson, Seikkula & Ziedonis, it was not network meetings themselves that were listed but a social network orientation: a focus not on finding problems within any one individual, but in the matrices of and between individuals. The same is true of the finding that social network size and density did not significantly change: fidelity was never about how large or dense the network became, but about whether the team and the person in crisis came to hold their situation as belonging to that network — a shift in orientation, not a metric a headcount could capture.
There’s a finding buried in Zindel Segal’s research on cognitive therapy that’s more striking than it first sounds. Cognitive therapy was built on the premise that you get well — your mood improves — by changing the content of distorted thoughts. But when researchers looked inside the actual mechanism — the black box of how that change happens — the mood didn’t begin to improve when the thought changed. It began earlier than that. It was when the person began to recognize that a thought was just a thought — not destiny, not a verdict, not a command. The change began with their relationship to thinking itself, before the thought’s content had changed at all. The cure showed up as a byproduct of that change in awareness.
The Western Lapland team found the same thing, decades earlier, from an entirely different direction — not in a person’s relationship to a single thought, but in a team’s relationship to an entire crisis.
This orientation had arisen quite naturally in Tornio as the clinicians – faced with an overwhelming and overwhelmed psychiatric crisis and treatment system – found that when they took their attention off of finding a problem to diagnose and treat in an individual, and even eventually off of diagnosing and treating social systems, but simply put their attention on fostering the conditions for safe dialogue, crisis situations that might have previously met criteria for psychosis, and then progressed to eventually meet criteria for schizophrenia, instead resolved. The five-year outcome studies found – inexplicably, to the conventional paradigm – that 82% of patients had no residual psychotic symptoms, and 86% had returned to full-time work or study.
This was not the same, however, as not having what might be called, from another perspective, “symptoms.” The difference was in how the symptoms were interpreted; they were not assumed to be in and of themselves a problem. They were only a problem if they in fact interfered with functionality. As Rai Waddingham put it in her own reflection on the study: each “relapse” can be seen as breadcrumbs, leaving a trail I can follow to gain greater awareness of the things that I find difficult in this world and the things that have contributed to my overwhelm.
In my quarter-century of studying, credentialing in, and practicing with the inspiration of Open Dialogue, one of my most important observations is that when people know a conversation is happening of which they are naturally a part, they want to participate. If that is true, then the question becomes — if they are not participating — what is getting in the way. When I have worked with families that are disconcerted by a member’s unwillingness or seeming inability to participate in a dialogue, the focus then turns toward how we can continue to invite – and make it safe – for all members of the family. Eventually, if we were able to do this well enough and long enough, even the most unwilling or antagonistic member would join.
Another of my significant discoveries was what I called the “crisis horizon.” That is, when we came to the time in the meeting when we would agree on when we needed to meet next – meaning; for how long can each member of the network tolerate the unease they felt about the situation as they perceived it — the conversation would turn to a practical examination of the specific fears at play. While sometimes these were very valid safety concerns, which could be addressed from a practical safety rather than psychiatric perspective, what often happened was a subtle shift into realizing that the concern wearing the crisis’s clothes was not the one actually driving it.
In one case, a family was in crisis over a member’s belief that aliens were coming. When we examined the family’s actual fear, what surfaced was this: they weren’t afraid of aliens. They were afraid he wasn’t going to have an IRA. Once named, that fear reorganized itself — real, and worth addressing, but not something we medicate or hospitalize for. The situation was no longer held as a crisis. He went on to turn his belief in the coming aliens into a thriving online business serving a worldwide population who shared his concerns, and the business skills he discovered along the way eventually carried him into other, more terrestrial ventures, from which he still profits.
Perhaps most of all — especially given the justified concern that nobody should be forced into a meeting with anyone they do not want to meet with — I learned that a social network perspective does not in fact require meeting with anyone. It only means an awareness that what we call mental health refers to how meaningfully, how safely, and how adaptively we experience ourselves within our significant social network. Open Dialogue does not prescribe what that network is; it only offers help in and for whatever that network is for any given individual or group of individuals.
I have worked with individuals in which we set up a chair or even circle of chairs to represent that network. We might place an iPhone on a chair – with a person at the other end of that call, or just imagined to be, or a photo of a person, or a teddy bear. I have had people refuse even the suggestion of an empty chair in the room — refuse the idea of it before any chair existed at all. That refusal became the thing to talk about, and what surfaced was not resistance to a piece of furniture but a refusal to accept someone else’s judgment of them. Once that was clear, the work stopped being about what was right or wrong — diagnosable — with the person in front of us, and became about the relationship itself. Whether an actual chair ever made it into the room after that, I honestly don’t remember. The idea of one was enough. Filling it was never the goal — and it turned out neither was the chair.
What happened instead was closer to a dialogical facsimile. I’d ask what the absent person might say to something that had just been said, and what might come back to them in return, building out an exchange between two positions with neither one physically or even symbolically present. Something in the room softened as we did this — not because anything had been resolved, but because a dialogue that needed to happen seemed, in some partial way, to have happened. Nothing was forced, then or at any point. The facsimile was enough to let something real move.
My point is that a social network perspective is not predicated on the particular form of meeting but on the orientation; that no person is an island, even if what they feel is a tide rising around them. This is of particular importance in evaluating Open Dialogue not as a fungible treatment technology but as philosophy and practice in transition from its place of origin to a wider world. Here, peer advocates help address a gap that Open Dialogue had spent a quarter-century bridging in Tornio: earning the trust of a community that had reason to be suspicious of the system’s methods and outcomes. Open Dialogue had invested decades toward learning to divest itself of the power that accrues to people in a helping role. Many have noted that almost everyone in Tornio has had direct or indirect contact with the mental health system, and come to have confidence in it. Outside of Tornio, however, where there was no reason to assume that members of a community would feel ready to trust, peer advocates could help to navigate the divide.
Mindfulness practice – the cultivation of self-awareness that, in addition to helping to achieve a level of self-awareness that aids in being fully and safely present in a crisis – could also be thought of as a form of community-wide peer support: a practice for managing the stresses of being human, with a lineage that goes back for millennia. Together, peer support, mindfulness, and dialogue itself form a broader base for holding psychiatric distress than any one of them could offer alone.
In this context, read fully, what initially appears to be a null result opens the door to an entirely different paradigm of care, based on an entirely different understanding of the problem, as well as of the human experience of which the problem is a part. While some may see the study design as a conciliation to mainstream psychiatry – perhaps necessary but a compromise nonetheless – it can also be seen as a door that lets in a cohort of findings that, once established within the walls of the old paradigm, have the power to change it: half as many hospitalizations mean dramatically reduced costs. Turnover on Open Dialogue teams ran 80% lower than on treatment-as-usual teams, meaning — beyond the direct savings on training and hiring — a long-term retention of experience, wisdom, and morale.
Odysseus doesn’t get home by proving anything. He gets home by telling his story to strangers who decide whether to help him — disguised even from his own wife until the story is told right. The Cyclops, the sirens, ten years lost at sea: nobody asks him to prove it happened. The story is simply held, by whoever is listening, not listening for whether the story is true, but listening for the truth of the story. That is a description of what happens in Open Dialogue.
From a treatment-as-usual perspective, no significant difference in time to relapse is the end of the story — a null result. But an Open Dialogue perspective is a different story altogether, closer to what Tomi Bergström said at the release event: this isn’t a question that’s been answered; it’s a story that’s beginning. ODDESSI’s publication in the Lancet is not an endpoint: it’s the beginning of a dialogue between people who are weary of fighting, and looking for a way home.
