I have recently posted two Dodo Bird articles, arguing that when it comes to psychotherapy, all models seem to get approximately the same results. Now I say that the Dodo Bird is Wrong! 

In my post, “Is Mental Illness A Disease” [], I pointed out that the DSM5 appears to be a scientifically bankrupt enterprise. By expanding the number of diagnoses, the reliability of an individual diagnosis falls and approaches zero. That means that there cannot be any validity to the diagnosis itself. That is because the validity cannot exceed the square of the reliability. If the reliability is .20 (the number can range from zero to one), then the validity, the question of whether the measure actually assesses what it is supposed to, cannot be greater than .04, or .2 x .2

Further, I pointed out that all patients get essentially the same treatment. If you are insight oriented, then defense and transference interpretations are offered, regardless of the diagnosis. If you are well trained in CBT, then cognitive distortions are the focus, and that is regardless of whether your patient has a diagnosis of depression or schizophrenia.

So there is little actual reason for us to diagnose patients. I do agree that perhaps major mental illness such as bipolar and schizophrenia deserve a diagnosis since most of those folks (but not all!) do better with medication. That is, the diagnosis is actually related to treatment. 

You can well argue that severity of depression is worth diagnosis, using the CES-D or the Beck, since those who score above 27 on either seem to do better when medication is added to the treatment. On the other hand, patients scoring between 10 and 26 are going to do just as well with psychotherapy, and, frankly, any antidepressant effect is very likely a placebo effect.

One thoughtful reader pointed out that by making any exceptions in my “no diagnosis” rant, I am undermining my own argument. True enough. But my argument was based on reliability, and a plethora of diagnoses make reliability pretty much impossible. I am not exactly saying that no diagnosis is reliable. (I do have to say, I love my readers, and particularly the ones who argue with me. All thanks.)

(But, what is the Dodo Bird verdict, you ask? Well it originated in 1938, an article by Saul Rosenzweig, who pointed out that all approaches to treating mental illness get the same outcomes. Model seemed to Rosenzweig to make no difference. He quoted the famous line in Alice in Wonderland, where the animals demand to know who has won the race. The dodo replies, “Everyone has won and all must have prizes.” This is a prophecy of the 1980s and 90s in California where the idea was to raise self-esteem so as to raise performance. The self esteem movement failed.)

In my following post, Mindfulness and Depression, [] I pointed out how group mindfulness training was shown to equal individual CBT in outpatient treatment of depression and anxiety. That raises the Dodo Verdict question, namely, are all treatments equal. Have all won? Must all have prizes?

Some think that this means there is no evidence that we can improve our outcomes. But some of you already know that is not true.

There is a clear and well-established exception to the Dodo Verdict. Feedback Informed Therapy, or FIT. When we simply track our outcomes for each and every session, the evidence is that we can improve outcomes by around 30% or more. This isn’t a change in the content of the therapy but rather a change in the process. At the beginning of each session, measure! 

NOTE: FIT is now a SAMHSA designated evidence-based treatment and requires no change in your theory of psychotherapy. You keep using all the skills that you have painstakingly acquired, just adding the FIT tools. Tracking the data will improve your outcomes by at least 30%. Tracking tells you when you are doing well and when you are getting into trouble. While we don’t get in trouble very often, when we do, that causes bad outcomes. 

We can reduce our bad outcomes by half, simply by tracking the clinical status and tracking the therapeutic relationship. Bad outcomes are only 7% or so, but we can cut that by half! 

That means, before each session, we ask the patients how they are doing, using some sort of reliable and valid measure. I suggest the CES-D, since it is in public domain. I use the OQ-45, which I like better, but to use it, you have to have a license. []

At the end of each session, you’d want to ask the client to rate the session. The therapeutic alliance is the best predictor of outcomes, yet very few therapists and coaches measure it! I have, naturally enough, a free tool for that, called the Session Rating Scale. It has ten items. You will do well to use it each session. [] If you will download that scale and look it over, you see that I am tracking the therapeutic alliance with a ten-item scale.

My colleague (and former student), Dr. Scott Miller, found that by simplifying those two processes, the pre-session symptom checklist and the post-session alliance rating, his students were more consistent in offering the measures. When he was my student, I pointed out to him that tests have three, not two aspects. Reliability and validity, of course, but the third has to be Feasibility or Practicality. If a test is too long or difficult, or if it costs money, few therapists will use it. For example, while the 20 item CES-D is quite feasible, the PHQ-9 is better as it is less than half as long.

Dr Miller took that to heart. He modified the OQ-45 symptom checklist by creating four items: Individual distress, relationship distress, and work performance distress. The fourth category is “Overall” or a global rating. The client simply makes a slash mark, or “/” on a ten centimeter line. Using a metric ruler, the therapist translates that into a score from 0 to 10.

At the end of the session, the client rates today’s session in the same way, on a four item measure, using a slash mark to indicate where he or she scores the treatment.

You can download those two measures here. [] For individuals, he gives his tools away for free; if you want to use them in an agency or a group, it is only fair and ethical (which you are) that you purchase them and help support his work.

Both Scott and I have trained thousands of therapists to use these simple tools. You can download a manual on FIT I developed here [Link] that explains how to use them, or you can go to Scott’s website and see about his trainings.

So while the Dodo Bird Verdict lives, there is an exception. Regardless of model of therapy, those counselors who track clinical functioning and alliance on each session will significantly outperform other therapists.

Some of you may have gone through my training, or listened to the CDs, or read the manual, and began to track your outcomes. Others may start after reading this.

But regardless, all are welcome to post their views. Share this page with your friends, so they can download the free tools. 

By |2016-11-26T15:50:49+00:00January 21st, 2015|Articles, Diagnosis in mental health, Enhancing Mental Health|8 Comments

About the Author:


  1. Barbara January 22, 2015 at 6:04 am - Reply

    Lynn, Thank you for your generous spirit and seeking to make the world a better place!
    p.s. I still find the”process” of diagnosing helpful in deciding how to approach the patient in therapy.

    • Dr.J January 22, 2015 at 6:03 pm - Reply

      Barbara, using a diagnostic scheme to help strategize treatment seems like a reasonable approach. You are a caring and careful therapist and I imagine you get good outcomes.

  2. Ofelia Sigalove January 22, 2015 at 8:45 am - Reply

    I like your article,Can you also share your treatment on Sxual addiction-Porno

    • Dr.J January 22, 2015 at 6:04 pm - Reply

      HA! Well, I have some strategies that sometimes work well. Good suggestion. I may tackle that soon.

  3. eric houghton January 22, 2015 at 10:19 am - Reply

    This article was very interesting to me. It seems to imply that anyone could be a therapist as long as they measure their progress, etc. I don’t say this facetiously because there is some basis for arguing that many more of us are needed to help out in this ailing world.

    • Dr.J January 22, 2015 at 6:01 pm - Reply

      In the 1960s there was disturbing research showing that paraprofessionals did as well as psychotherapists as did very well trained therapists. Of course, none of us want to admit that! Then look at the research I wrote about, showing that 6 weeks of training in mindfulness meditation produced as good results as highly trained cognitive therapists! I am glad I have a Ph.D. because some people call me “doctor.” But there are many pathways to health!

  4. Brandon January 26, 2015 at 11:02 am - Reply

    Great words here. I’ve been preaching the good gospel of FIT for a number of years now, and have trained a bit with Scott D. Miller (and I plan to continue…). The reality is that FIT will simply out narrate the modern (medical model) presupposition that diagnosis (“illness”) informs treatment (“cure”). That is, in the world of mental health, health wins out over non-health every time. And while some may find use of the DSM valuable as part of the process in their practice, I would encourage such clinicians to ask their clients whether or not they find the use of a DSM label helpful in their overall care. Rarely, will anyone actually say, “Yes, thanks for letting me know I’m depressed. Knowing I’m depressed (anxious, schizophrenic, bi-polar…) has been very helpful to no longer being depressed.” Even so, for those clients who do affirm the use of a DSM label, what is “appreciated” about its use isn’t the label itself nor is it “being” depressed (anxious, et al). What is “appreciated” is the hope that comes with anticipated outcomes. In other words, “Now that I know this about myself, I have hope that I can live a healthy(ier) life.” It’s about the outcome, never about the diagnosis. Yet, Feedback Informed Treatment doesn’t preclude the use of the DSM, it simply helps keep the main thing the main thing – that is, therapy is about the client. The more clinicians who seek to make therapy about the diagnosis (or some other “secret” clinical language or skill) the more distance is created between the client and their preferred outcomes. As has been indicated elsewhere, diagnosis is only of (at best questionable) value to third party payers.

  5. Gabrielle January 26, 2015 at 1:42 pm - Reply

    A very nice distillation of outcome research. I think there is something else we are not seeing. It is a strange phenomena that outcome hasn’t changed in 50 years. I’ve had a few different ideas about that, but lately I’ve been stuck on something that Heinz Von Foerster used to say. That tests test the test, not the thing being tested. Could it be that there is something about the construct of therapeutic trials that hasn’t changed in 50 years?
    Goiters are common where I come from. There was a time when people who drank milk didn’t get goiters. Turns out it had nothing to do with the milk, but the cows udders were washed with iodine, and iodine deficiency had caused the thyroid problems. I think something like that is going on with therapy outcome research. Nothing about humans could be that constant. I’d be keen to hear your thoughts on this.

Leave A Comment