Dialogic Practice Dialogic Practice


The Promise of Open Dialogue

Mary Olson, Ph.D.

Mary Olson, Ph.D.

Open Dialogue is an innovative, network-based approach to persons experiencing severe psychiatric crises and conditions.  Developed at Keropudas Hospital in Tornio, Finland, this way of working has garnered international attention for its outcomes with first time psychosis.  Noting the positive interest Open Dialogue has begun to attract in the U.S., publisher Marvin Ross, in a recent Huffington Post blog (11/11/13), argues that before making the global claim that Open Dialogue achieves better results than standard treatment, we need to do more research.  I agree.

At the University of Massachusetts Medical School, I am co-leading a research project on Open Dialogue with Douglas Ziedonis, MD, Chairman of Psychiatry and an internationally renown expert on implementation science, and Jaakko Seikkula, PhD, one of the original developers of Open Dialogue, now a professor of psychotherapy at the University of Jyväskylä, Finland. Our project is called “Preparing the Open Dialogue Approach for Implementation in the U.S.”  We have spent the past year developing the research materials — organizational and psychotherapy fidelity guides — that are the requisite scientific steps prior to undertaking a clinical study.  Our plan is to complete this first phase (developing these protocols) by next fall and then launch the clinical part. We intend to make our research tools freely available to other researchers in the U.S. and elsewhere. Our funding comes from the Foundation for Excellence in Mental Health Care.

As a scientist and a practitioner, I share the basic position that further research on Open Dialogue is necessary to determine whether (and how) it can be adapted to—and fully effective in–other countries. The last thing we wish to do is to give false hope to people who have had great suffering. At the same time, after reading Ross’s piece, it is not clear why he thinks more research is warranted. According to him, Open Dialogue produces results that are expected and happen everywhere.  Yet, he also says the initial data is deeply flawed and misleading due to poor or inappropriate scientific design.  Furthermore, he implies that Open Dialogue is being propelled by an anti-psychiatry ideology. Not only does Ross withhold praise (his title is “Don’t Be Too Quick to Praise This New Treatment”), his article may invite condemnation. Let me address the issues he raises: (1) expected trajectory; (2) research design; and (3) anti-psychiatry bias.

Expected Outcomes

Marvin Ross, referring to the medical textbook the Merck Manual says that the expected recovery rate, defined as “significant and lasting improvement,” for people diagnosed with schizophrenia should be around 30%.  In a complex argument based on his own calculations, his critique of Open Dialogue is that its outcomes roughly follow this natural, expected course. What he does not address is that our treatment-as-usual outcomes here in the States seem to be falling below this natural expectancy. The best longitudinal data we have is Martin Harrow’s naturalistic study (2007) that, after following people for fifteen years, suggests only 5% recover with standard care.  A recent study from the Netherlands (Wunderink et al., 2013), employing a randomized design, finds that, at a seven-year follow-up, only 17.6% of those who had received standard care after first time psychosis met the criteria for functional recovery. In any case, there is no scientific evidence to support Ross’s assumption that standard approaches are achieving expected recovery rates. (Accordingly, in addition to the above 30%, another 30% would have intermittent recovery, and the final group would struggle permanently.) If Open Dialogue outcomes do follow the Merck Manual bell curve, as Ross argues, its outcomes are better than those we know of standard care.

Lack of Randomization and Independent Assessment.

In order to respond to these criticisms, I have to give some background.

In the early eighties, when Jaakko Seikkula, and his Keropudas team—Birgitta Alakare, MD, Jukka Aaltonen, MD, Markku Sutela, MA and others—began to work together, their mandate was to deinstitutionalize a hospital ward population of long-term inpatients, many judged to be “incurable,” and to establish a community-based system instead.  What came to be called Open Dialogue was the result of this effort, which received a prize from the Finnish National Development and Research Center for Health and Welfare. The Keropudas team ultimately produced a transformation of an entire public psychiatric system from an old-style, “chronic” hospital into an acute, therapeutic facility. Open Dialogue consists, therefore, of two, interrelated features: (1) a particular kind of comprehensive, community-based, treatment system, and (2) a psychotherapeutic process of dialogue in open psychiatric meetings, which  Seikkula was the first to conceptualize in a unique way (Seikkula & Olson, 2003). Both of these need to be there to call something Open Dialogue.

It is easy to see that the methodological, or research, issues involved in studying Open Dialogue are challenging. The transformation and influence of an entire, community-based psychiatric system do not lend themselves to randomization. Randomization is the gold standard for medication trials and, at best, those of office-based therapies, in which a clear, causal variable,  (“X” causes “Y”), and a clear control group can be identified.

There are two legitimate difficulties with a randomized design when it comes to Open Dialogue at Keropudas Hospital. First, there is the real dilemma whether such a simple, linear, cause-and-effect formulation (X causes Y) can capture the complexity of Open Dialogue in a meaningful way, because this network-based, systemic practice generates a density of complex interrelationships and overlapping contexts (Seikkula & Arnkil, 2006, Chapter 9).

Second, there is no possible control group for the reorganization of an entire public hospital district.  Furthermore, once the new way becomes standard practice throughout an entire geographical catchment area, randomization also would not be an option. There might be the possibility of “mimicking” a randomized design in adapting open dialogue principles to other, new settings, but not during, nor after, the trial-and-error process of their genesis and crystallization. That said, the absence of randomization does not nullify the evidence. What it means instead is that a study is descriptive, rather than explanatory. That is, it can say Y happened, but it cannot say Y happened definitely because of X.

The other criticism that Open Dialogue lacks independent assessment is a valid concern, since there could be subtle and unwitting blind spots when those who invented the approach are also evaluating it. The team did have independent researchers examine elements of their research, such as their use of diagnostic categories. There is also a different qualitative research tradition at work here with different values, but that is a discussion for another time. More research is needed though, and the purpose of the UMass project. Everyone would agree that the outcome variables for Open Dialogue are objective:  disability vs. work or school, medication use, hospital days, and symptom intensity.

Before leaving the subject of the existing research data, a correction also is in order. None of these studies claim what Ross reports they claim: i.e., 80% of people with schizophrenia are so-called “cured” without medication under the Finnish method. Here is what the three key studies actually say:

In two, five-year, follow-up studies of Open Dialogue (Seikkula et al.,, 2006), 80% of those who had acute psychosis for the first time in their lives reportedly experienced functional recovery. That is, after five years, they were working, studying, or looking for a job and not on government disability.  Roughly 80% were also asymptomatic and not taking medication, though over a third had been exposed to antipsychotic drugs during their treatments.  In an earlier study, in 2002, the Finnish team looked at what happened with the subgroup of these people who were diagnosed with schizophrenia (Seikkula et al., 2003). While this group had somewhat more medication than those with the milder diagnosis of psychosis, 70% of them returned to full employment after two years.

The strong employment outcomes, which might seem improbable to us outside Finland, make sense if you know the local context.  Based on their ecological orientation, the hospital team has built a close, collaborative, mutually trusting relationship with the staff at their rural county’s employment office. When a person starts recovering from a severe crisis, they are encouraged to return to work and their other normal routines and can rely on a web of support, if they so choose, in resuming or finding a job.  For such contextual reasons, it is hard to know whether similar outcomes can be replicated outside of this small province of Western Lapland with people suffering similar, terrifying, symptoms elsewhere, say, in larger, socially isolating, urban environments. But it is important to find out. As stated, more research.

Anti-Psychiatry versus Need-Adaptedness

Open Dialogue is rooted in the democratic and humanistic reform of Finnish psychiatry, called “Need-Adapted Treatment,” which was pioneered by Professor Yrjö Alanen, MD. Need-adaptedness means employing all the available methods of psychiatry and mental health work on an as-needed, case-specific basis. The main format of Open Dialogue is the open meeting, which was originally Alanen’s idea. Here is what it means:

Open Dialogue provides an immediate response within 24-hours of the first contact to the crisis service. In advance of any decisions about hospitalization or therapy, the radically revised treatment meeting brings together the person in acute distress with all other important persons, including other professionals, family members, and anyone else closely involved. Everyone’s voice is heard and respected. Any decisions about medication and hospitalization are made with everyone’s input. The team that comes together at the start remains the permanent team whether a crisis last three weeks or three years. Transparency in Open Dialogue is also a main value. The professionals try to be as open and forthcoming as possible.  Their practice of transparency was further shaped by its cross-fertilization with the egalitarian, reflecting process work of the late Tom Andersen, MD and Magnus Hald, MD of Norway. Drawing on the writings of philosopher John Shotter and Tom Andersen, the influential social thinker Lynn Hoffman (2007) describes this approach as a “withness” versus an “aboutness” practice.  In other words, Open Dialogue emphasizes “being with” rather than “doing to.”

If possible, antipsychotic medication is avoided.  If not, it is used in as low doses as possible with an understanding of the risks involved. People have the option of tapering off and discontinuing the medication when they start feeling better. The Dutch, randomized-design study (Wunderink et al., 2013), mentioned earlier, makes a strong case for a strategy of discontinuation/tapering off of antipsychotics as fostering better long-term, functional outcomes. This strategy is consistent with the medication protocol of Finnish Open Dialogue (Seikkula, personal communication, 2013), which, in turn, may shed some light on their successes.  The current NIMH director Tom Insel, MD has been so persuaded by the Wunderick study that he wrote on his “Director’s Blog” (8/28/13) about rethinking standard, “one-size-fits-all” medication guidelines.

Open Dialogue in the U.S.

A scientific orientation and an argument for further investigation—are central to–not in tension with–Open Dialogue and its tradition of Finnish psychiatry. I was a Fulbright professor at the University of Jyvaskyla in 2001 when I first visited Keropudas Hospital and began to study Open Dialogue.   Since I returned from Finland in 2002, I have worked on generating a research study in the U.S. Jaakko Seikkula has given me his unwavering support. Over the past decade, Seikkula has traveled regularly every year, on biannual visits, to Massachusetts to further this academic and scientific partnership.

It has taken these many years to achieve tangible progress in the form of the above-mentioned UMass project.  Jaakko and I have also created a training program, the Institute for Dialogic Practice in Haydenville, MA, together with psychologist Peter Rober, PhD from Belgium and Markku Sutela and Birgitta Alakare from Keropudas. In tandem with UMass, we have been training select dialogical teams in community care that are part of new, experimental initiatives structured with provisions for outcome research and evaluation. One such team, led by Christopher Gordon, MD, is from Advocates, Inc. in Framingham, MA. We have included other practitioners in our training groups as well, and other working partnerships have formed,  such as Nazlim Hagmann, MD, Rebecca Ross, LICSW, and Will Hall, MA in New York City. Our trainees almost unanimously, and with great enthusiasm, report improvements in their clinical practice. I recently received a moving letter from a family seen by another graduate, Ross Ellenhorn, PhD, and his team, writing to thank me for teaching this approach. Doing training itself has turned out to be an invaluable kind of informal, ethnographic (participant-observer) research, since we have gained important insights.  Other researchers from our UMass team hope to do more systematic, formal interviews with our recent graduates, so we can learn more about their experiences adapting Open Dialogue and related dialogical practices to U.S. settings.

On a different topic, before I end, I would like to acknowledge the courageous efforts of the treatment reform activists in all this, whom Ross dismisses as “psychiatric contrarians.” The support of such leaders as Daniel Fisher, MD and Jonathan Delman, PhD, now both major consultants to our research, and Will Hall, formerly of the Foundation for Excellence, was key in terms of building the needed momentum to start the UMass project. Reform activists, peer specialists, and family members have also been central to another groundbreaking initiative, Parachute NYC, that is putting into practice the closely related Need-Adapted Treatment described above.

On a final note, despite its successes, Open Dialogue is not a panacea. If we follow the scientific evidence (and Ross’s own calculations), however, there is reason to believe Finnish Open Dialogue has achieved among the best outcomes in the world for psychosis and schizophrenia. At the same time, I am tempered by Gregory Bateson’s warning (1987) when he writes: “Behind every scientific advance, there is always a matrix, a mother lode of unknowns out of which the new partial answers have been chiseled (pp. 14-15).” He advises us to be careful, not to “rush in where angels fear to tread.” There is more to know and understand about this promising idea.  That is why carrying out careful research, hand-in-glove with training, is warranted. Incidentally, Bateson thought ecologically, contextually, and socially, and would have greatly appreciated the Finnish approach. So, while it may be too early to give praise. I’d add: “Don’t Be Too Quick to Disparage and Dismiss the New Treatment Either,” thus quashing further scientific inquiry and reinforcing the status quo.

FYI: The Institute for Dialogic Practice: www.dialogicpractice.net. (Though active and operating, we are redoing the website to make navigation easier.  It should be done by February.)


Bateson, G. & Bateson, M.C. (1987). Angels Fear:  Towards an epistemology of the sacred.  New York: Macmillan.

Harrow. M. (2007).  Factors involved in outcome and recovery in schizophrenia patients not on antipsychotic medication.  Journal of Nervous and Mental Disease. 195, 406-414.

Hoffman, L. (2007).  The Art of Withness.  In H. Andersen & D. Gehart (Eds.), Collaborative therapy: Relationships and conversatons that make a difference (pp. 63-79). New York: Routledge

Seikkula, J., Aaltonen, J., Alakare, B., Haarakangas, K., Keränen, J., & Lehtinen, K. (2006).  Five-year experience of first-episode nonaffective psychosis in open-dialogue approach: Treatment principles, follow-up outcomes, and two case studies. Psychotherapy Research, 16(2), 214–228.

Seikkula, J., Alakare, B., Aaltonen, J., Holma, J., Rasinkangas, A., & Lehtinen, V. (2003). Open dialogue approach: Treatment principles and preliminary results of a two-year follow-up on first episode schizophrenia. Ethical Human Sciences and Services, 5, 163–182.

Seikkula, J. & Arnkil, T. (2006).  Dialogical meetings in social networks. London: Karnac.

Seikkula, J. & Olson, M. (2003). The Open Dialogue Approach: Its Poetics and Micropolitics. Family Process, 42, 403-418.

Wunderink, L, Nieboer, R., Wiersma, D., Sytema, S., Nienhius, J.F. (2013). Recovery in Remitted First-Episode Psychosis at 7 Years of Follow-up of an Early Dose Reduction/Discontinuation or Maintenance Treatment Strategy: Long-term Follow-up of a 2-Year Randomized Clinical Trial.  JAMA Psychiatry. 70(9), 913-920. doi:10.1001/jamapsychiatry.2013.19.

Blog post originally appeared in Mad in America – Science, Psychiatry and Community on January 1, 2014

View Article

Blog Categories

  • No categories

Institute Faculty

Dialogic Practice Faculty The Institute is the only North American teaching facility that specializes in providing world-class training in Open Dialogue and Dialogic Practice delivered by the field’s leading experts and developers.

Institute Faculty