How Open Dialogue works
At McLean, we have adapted the Open Dialogue approach to our schizophrenia and bipolar disorders inpatient unit (Rosen K and Stoklosa J, Psychiatr Serv 2016;67(12):1283–1285), basing our adaptation on Dr. Olson’s 12 key elements. Although we have applied the technique to an inpatient unit, it can be used at any level of care and patient interaction, including family meetings, intake interviews, follow-up visits, and phone calls.
Team meeting format
Include two or more clinicians in a team meeting. When you meet with your patient, don’t do it solo. It’s best to bring two or more clinicians, such as a psychiatrist, therapist, nurse, social worker, community support, or a trainee such as a medical student, all of whom can become collaborators. The first part of the meeting will be an interview you conduct with the patient (and the patient’s network), while the second part of the meeting will be a discussion between the clinicians about what was heard. Including more than one clinician decreases the “expert vs patient” or “me vs you” feeling often produced by one-on-one meetings. Differences of opinion between clinicians can help defuse this tension and show the patient that multiple viewpoints might be valid. In the context of our case vignette, Mary would be able to hear different clinicians discussing their concerns and reasons for recommending medication changes, which sets the stage for shared decision-making.
Include social supports. In addition to having clinicians in the meeting, invite others, such as family members, friends, or other important people in the person’s life (clergy, teachers, neighbors). These supports may become partners in the treatment planning process, rather than just sources of collateral information. Having supports involved prevents patients from being isolated from the rest of their lives through stigma and secrecy.
Language and phrasing
Use open-ended questions. Start the interview with open-ended questions to allow the patient’s story to unfold. Consider asking, “Whose idea was it for you to be admitted to the hospital?” to help establish the patient’s level of commitment to being there. Ask, “How would you like to use this meeting?” Ideally, two-thirds of the meeting is based on listening to the patient and the patient’s network, and only one-third follows your own checklist of questions. This allows the patient and network to direct the focus of each meeting. For Mary, we can ask, “What might be a good outcome of your hospital stay?” to better align around her goals.
Respond to people with their own words. Try to use your patient’s exact words when asking questions or making comments, rather than paraphrasing or translating into psychiatric symptoms, because creating a common language is a main goal of dialogic practice. For Mary, who said, “I’m scared they’re coming to get me because I can hear their inner consciences talking everywhere,” you might respond, “Other people’s consciences— what were they saying?” (You can ask this rather than a paraphrased question such as, “What were those voices saying?”— recall that Mary never mentioned “voices.”)
Emphasize the present moment. Consider using what is observed and shared in the room rather than outside collateral information. For Mary, you might say, “You said it was your mom’s idea to come, and you haven’t been feeling good,” rather than, “Your mom left me a message saying you were acting strangely”—the message isn’t part of what Mary has shared with you.
Timing and flow
Elicit multiple viewpoints (regardless of whether multiple people are present). Invite everyone to speak so that you give a voice to all present, rather than allowing the most talkative person to take over. It’s not necessary to establish consensus as each person speaks, but rather notice that each person has a unique viewpoint. For those people not present but important to the patient’s support network, consider asking how they might respond if they were present. In Mary’s case, you might say, “I know your mom isn’t here, and you said it was her idea to come. What would she see as a good outcome to your hospitalization?” This invites more perspectives into the discussion.
Use a relational focus in the dialogue. Consider framing the patient’s symptoms in terms of relationships, rather than relying on diagnostic labels, which risks oversimplifying problems and causing patients to feel at odds with clinicians. Let’s say Mary’s mother is in the meeting. Rather than saying, “Mary is paranoid and isn’t taking the medication she needs for her schizophrenia,” you might say, “Mary says she’s been scared lately, and though she finds the medicine helpful, she feels stressed being frequently asked at home whether she took her medicine-— then, because she gets so stressed, she ends up stopping it.” This sort of explanation allows everyone to be a part of the solution, creates more common understanding, and avoids unhelpful blame.
Response to psychosis
Discuss your thoughts about patients by having “reflecting talks” with other clinicians. After you have interviewed your patient and the patient’s network, it is time to discuss your assessment with other clinicians in the meeting. To pivot into this mode, start by asking, “Do you mind if I share a few words with my colleagues now?” You can physically turn toward your colleagues and have a dialogue in front of the patient’s network. Pay special attention to your language, use the patient’s words while avoiding jargon, and employ exploratory, tentative phrasing like “I wonder” or “I’m curious about.” Openly share your treatment ideas and their rationale with your patient. For example, you might say to a colleague, “I was struck by Mary’s fears and all the inner consciences she described. I wonder if we might help the fear with a different dose of her risperidone medication—I’d recommend we add another milligram.” Your colleague could respond, “I also really resonated with her fear; in fact, I felt tense as I listened. It seems like a risperidone adjustment might be helpful.” This creates a space for clinicians to listen to themselves and their inner dialogues, as well as for the patient and network to listen without pressure to respond to a treatment plan. Afterward, you can turn back to the network and say, “Does anyone have any reflections on what you heard? What did you agree or disagree with?”
Be transparent. Discuss treatment options openly. This is where the “open” part of Open Dialogue comes in: Clinicians co-create treatment plans, including hospitalization and medications, transparently through shared decision-making in a team meeting.
Tolerate uncertainty. Consider an approach where everyone has a valid perspective, rather than some being more right than others. In this style, the clinician creates safety by making sure all perspectives are heard before formulating treatment recommendations. This allows the full story to be told, which in itself may be healing and lead to crisis resolution.
Conclusion: Why does Open Dialogue work?
In our experience, many clinicians are intrigued by the Open Dialogue approach, but are curious how a change in communication style could actually improve treatment outcomes as dramatically as some research suggests. Here are some hypotheses:
1. Transparency eliminates the patient’s fear of what clinicians are really thinking, leading to a more genuine connection. Connection and alliance are key ingredients to good outcomes.
2. By including the support network, open meetings enhance connection and communication between patients and families. This decreases loneliness and isolation— factors that lead to poorer outcomes.
3. Creating a common language helps patients understand their problems and communicate with their clinicians and network. Enhanced communication may lead to higher-quality information, bolstering our treatment decisions.
TCPR Verdict: The Open Dialogue approach is a nice way to prevent patients from being alienated from caregivers. While the evidence that it improves outcomes is preliminary, it might be worth implementing some of these techniques, since the downsides (if they exist) are likely minimal.