Musings on the Yale Conference


On April 24, I had the pleasure of attending the conference “New Directions & New Hopes Call for New Practices in Clinical Psychiatry.” Jointly sponsored by the Yale Program for Recovery and Community Health, the Foundation for Excellence in Mental Health Care and the Connecticut Department of Mental Health and Addiction Services, this was a rich experience. Most of the talks will be posted online and I highly recommend viewing them.

On my drive home from New Haven, I turned off my radio and used the silence of my car to reflect on what I had heard. Most of these talks focused on psychosis – the course, personal experiences of recovery, long term outcome studies, the impact of early identification and well as social and psychological interventions. As the title of the conference suggests, the organizers were looking towards new practices. As a psychiatrist, I am often called upon to opine on the use of drugs and I spent my ride home wondering how to make sense of what I heard.

While all of the talks were important, I must admit to having an ongoing fascination with Martin Harrow’s study. Harrow and Thomas Jobe, followed a group of 129 individuals with early episode psychosis for 20 years in a naturalistic study. I highly recommend listening to this talk (a version of which can also be viewed at Mad In America Continuing Education). Although some data was reported in Anatomy of an Epidemic, Harrow and Jobe have analyzed this data set further and what they report is fascinating. A finding discussed in Anatomy is that those who stopped taking neuroleptic drugs had a much better overall outcome than those who maintained. But there is more to this study that is worth pondering. First of all, at two years, the group who had stopped drug still had on average as much psychotic symptoms on assessment as those who did not stop. The improvement is only seen at 5 years and this group continued to improve over time.

The group who remains on drug does not show this ongoing improvement. Harrow also looked at those who were in remission at each assessment point. He then compared relapse rate over the next interval and compared relapse rates among those on and off drug. The relapse rate was much higher among those on drug(this is a tough concept to convey in words so I highly recommend looking at his presentation). I often talk to colleagues about this study and I am intrigued at how people hear surprising findings and explain to me why they are not surprising at all. They will come up with some explanation that both fits our current paradigm of care and fits the finding. With Harrow’s study it is often something along the lines of “Well, we always knew there were people who recovered and those who do well are the ones who stop their medications.” But that is not what Harrow found. People most commonly stopped their medications before they were doing well. And why would there be a higher rate of relapse among those who maintained drugs (remember, this was comparing individuals who were all in remission)? Isn’t that supposed to be the whole point of remaining on the drugs?

I was particularly excited to hear Lex Wunderink talk about his study of long-term outcome since he was the only speaker who I had not heard before. Dr. Wunderink evaluated individuals at 7 years who had participated in an earlier study which compared two different approached to neuroleptic drug strategies: continuous use of drugs or intermittent use only when the person was symptomatic. At 7 years, he found that those who had initially been randomized to intermittent drug use were far more likely to be working and due to this fact, there was a much higher number of people in this group who were judged to be recovered. In addition, the higher rate of relapse observed at two years in the group who was assigned to intermittent drug treatment was no longer observed at 7 years; it appeared the drugs postponed rather than reduced the risk of relapse. When he divided his cohort between those who were on no or very low dose of drug vs. the others, he found the same thing; those on low does had a much higher rate of functional recovery and overall recovery.

I had the opportunity to ask Dr. Wunderink about his tapering protocol. He told me that there was no fixed protocol but that most people in the discontinuation group stopped the drugs over a few weeks. I wondered if their results might have been different if they had started the randomization at an earlier point; in his study people were treated for 6 months before randomization or if the taper had been slower. In my own practice, I am tapering over months and years but I honestly have no good data on which to base that since this has not been carefully studied. I also asked him how they dealt with the worry that individuals and their family might have about stopping medications. He told me that they spent a lot of time talking with people and developing good relationships with them. I asked him if he ever ran into situations where once a person developed psychotic symptoms again, they were no longer willing to resume their medications. He said yes, on occasion.

What I have been thinking about is that there seems to be a point, on average somewhere around 4 years, when something changes. People who have ongoing or intermittent psychosis, may begin to recover. In the Harrow study, there was something about those individuals who never took neuroleptic drugs that allowed them to remain in the community and out of treatment even though at times many of them still had psychotic symptoms. In the Wunderink study, there was a higher rate of relapse early on in the group who stopped drug but then the relapse rate flattens. With our current treatment paradigm,  clinicians would recommend that a person with ongoing psychosis take neuroleptic drugs and most consider a single relapse as indicating that the person requires long term drug. In fact, it could be considered negligence to not make that recommendation.

But these studies suggest that a single relapse might not tell us much about outcome or the balance of long term benefits and risk of drugs. Most of our current treatment models hold the minimization of psychotic symptoms as the primary goal. As we look to developing new practices in clinical psychiatry, as the conference title recommends, maybe this is the counter intuitive step we need to take; while we are developing treatment programs for early episode psychosis – and these are proliferating around the world – maybe we need to hold off on having as the major goal, the complete elimination of psychosis. Maybe we need to focus more on functional improvement, trying to figure out what matters most to the young person, address the worries and concerns of his family, help the person remain in the community but hold off on the strict recommendation that we need to target first and foremost psychosis.

Most of these talks will be posted on Mad In America and many of the authors will have courses on Mad In America Continuing Education. Watch them and let’s keep the conversation going.


Mad in America hosts blogs by a diverse group of writers. These posts are designed to serve as a public forum for a discussion—broadly speaking—of psychiatry and its treatments. The opinions expressed are the writers’ own.


  1. Thanks so much for this article.

    I have always been confused by what was exactly meant by remission -particularly as my experience of watching recovering has been from seeing small increments of change over a long period of time – and thus at what level or with what criteria a subject was considered to be in remission?

    I also wondered whether the non-medicated subjects in Harrow’s studies took medications after relapse?

    • I also wonder if Wunderink is continuing to f0llow the people in his study – I am hoping that he is and that he is seeing even more people recover as time passes. I wonder if the people who refused to go back on medication after their relapse , if they ended up in the group that finally recovered or the group that didnt recover.

  2. Sandra,

    This is very interesting – thank you for sharing.

    The only critique is that it seemed that the medical model is still dominating a lot of the way these people are viewed. This may be coming more from other conference participants than from you. I was making the following translations in my mind while reading:

    “course” – what people do with their lives
    “recovery” – healing, improvement, or maturation
    “relapse” – having extremely distressing experiences again, in some form
    “symptoms” – distressing/anomalous experiences
    “psychosis” – distressing / incapacitating non-consensus experience of of one’s reality / relationships.

    I hope these ideas make sense to you.

    The “illness/recovery” medical model of psychosis implicitly assumes the following sequence, which is similar to what would occur for many physical illnesses:

    Step 1 – Person was well/ok.
    Step 2 – Person becomes ill/has a first episode (of psychosis).
    Ideal Step 3 – Person “recovers” to their pre-illness state (of being non-symptomatic and/or functional).
    Even if it does not mean this exactly, that is what the word “recovery” implies, basing the meaning on its Latin roots.

    However, we know from studies like John Read’s that many people who become psychotic are extremely disturbed and traumatized emotionally before breaking down – i.e. the victims of neglect, abuse, trauma – long before they have a “first episode.” Therefore, the recovery notion that they are non-ill, then become “ill”, then may “recover”, does not fit people’s emotional experience. In other words most of these people’s “pre-illness” states are precariously weak, immature and “ill” in many senses, and a person needs much more than a “recovery” to that state in order to be truly well. In fact, a recovery to that state is likely to lead them to become unwell again. Rather, what they likely need is maturation, healing, and relationship building of a kind they have likely never experienced in their life to that point.

  3. In my private practice I work with a number of people who have been labeled as having a “first break” and who have been trying anything to avoid labels and neuroleptics. As you well know psychosis is complicated with many different etiologies and trying to create sanctuary for folks going through unusual experiences can be very challenging. It feels increasingly important that these individuals be given a chance to process their experience without drugs, even if that means ongoing, or recurring “episodes.” And the reason is very much along the lines of what Harrow and Wunderink have discovered. If you give a person a chance, there is a place for integration and recovery. Long term with the neuroleptics, the chance goes down.

    You said that many folks dismiss these findings because they use the argument that these people were likely not very sick to begin with. But that begs the question- how do they know who will recover and who won’t? And seeing that long term use definitely worsens prognosis, isn’t it far better to attempt non-drug interventions?

    I think the big challenge is creating that support and sanctuary for those experiencing psychosis- and it is far easier to sedate a person than to support them through that process. Creating the template for support is far more expensive and complex. But if we viewed an episode like drug rehab, then there would be centers where people could go, often for a month or more, paid for by insurance, that would support reintegration without drug intervention.

  4. Dr. Steingard, Is this the direction you have always wanted to go in? I would see different things to appreciate about your ongoing development of clinical perspectives in either case, although would feel assured the same about your judgment and responsible interest taken in your own reasons for taking the actual path you took. Heading into the article with your recommendation to watch the videos that feature in your musings likewise reassures me about their value for me, the time investment, the convenience of having your reflections to consider–as it were–in retrospect. I love your indefatiguable willingness to put emphasis on our needing to welcome and carefully attend to important competing views about ourselves as survivors, our problems and hopes as modern people, and the issue of our having the belief in ourselves that we should have.