TIFT #74: How We Make Clinical Decisions

Teaching a class of trainees and early career therapists got me to thinking about the question of clinical decision making. My co-leader shared the example of a client asking, “If you were in my position, what would you do?” I counted six ways a clinician might make such a decision, some better than others:

1. Follow common factors correlations to optimize results.
2. Stick with rules and taboos about what not to do.
3. Be a nice person and a good listener.
4. Follow the methods and rationale from one school of therapy.
5. Adhere to an evidence-based treatment manual.
6. Predict the results of different choices, based on a universal framework describing how people get into trouble and how they change.

Of course you can tell where I am going, but this is serious. Many therapeutic choices are, in reality, made on the basis of of one or more of the first five approaches and that is not always good for therapy. Let’s consider the possibilities one at a time.

1. Follow common factors:

Common factors such as the value of affect in sessions or the importance of a positive therapeutic alliance are supported by statistics showing correlations with a good outcome. Problems come when we try to answer specific issues like how to respond to the client’s question about what to do. The weakness is that common factors are theory-free, meaning they don’t provide any way to conceptualize why those factors are helpful or how they contribute. Nor do they provide clear guidance about a specific response. What they do accomplish is describing an optimal ambiance within which therapy should take place, but they don’t address mechanisms by which change happens or, specifically, what choices to make to achieve the goal of change.

2. Stick with rules and taboos

This actually happens more often than one might think. Before the class started, I asked the group what they wanted to get from the session, which was focused on building the therapeutic relationship. The first things that came up were “how to set the framework,” and “how to manage boundaries.” They were concerned more with how to say no rather than yes. One source of evidence that values, including prohibitions, are internalized in adulthood (and that the superego doesn’t stop developing), is the ominous power of “Thou shalt not!” to the ears of newcomers, highly motivated to become members of our profession. These new therapists didn’t remember the value of a strong bond with the client as much as they recalled the taboos about giving too much. They were most concerned with not “gratifying” the client’s wishes.

What I wanted to teach them, instead, was that having a clear idea about how things can go wrong allows one be more relaxed about give and take within the therapeutic relationship. In the end, my colleague described how he responded to the question. He answered it, but did so as a human, not an authority. If the client later shows a distorted understanding, such as that the answer was a command, it will be far easier to explore the transference than if the therapist had thrown cold water on the relationship with a non-response.

More importantly, the client had been expecting a stereotypically evasive response. What he got instead was a corrective emotional experience, one in which his prediction was contradicted by the therapist’s helpful and genuine answer. The client ended the session with a much more positive feeling about the therapeutic relationship, which could be predicted to strengthen and improve the results of the therapy.

The point is that old taboos learned early in training, such as not exploring the past and not gratifying the client’s wishes, can take on far too much power and lead to a therapeutic relationship focused on the negative. And that, according to common factors, will correlate with a degraded quality of the alliance and a poor result.

3. Be a nice person and a good listener

A highly sophisticated and experienced therapist once admitted that she didn’t really recall the complicated jargon and concepts she learned in psychoanalytic training. At this point in her long career, she went more with instinct and intuition. In this case, I trust that her instincts were excellent, but her comment illustrates a phenomenon that happens too often. Therapists don’t get strong enough training or supervision to apply the complex constructs and the concepts they read about for classes. The end result is too often that they fall back on social habits as a “nice person” and a “good listener.” 

Here are two ways that approach to decision making can, and probably does, do harm to clients: The first is a tendency to collude with the client to avoid painful emotions. That’s what nice people do in ordinary social interactions, but it is often counterproductive in psychotherapy. We know that in a general way because common factors research says experiencing affect is good for therapy. More specifically, modern knowledge of the neurophysiology of change by memory reconsolidation tells us that affect is our best clinical indicator that two of the three requirements are being met. First, affect tells us that the old response pattern has been activated and second, that a communication channel between the consciousness and the limbic system is open for transmission of new information.

4. Follow the rationale of one “brand” of therapy

In my own training I worked hard to follow the psychodynamic approach to therapy. I focused on identifying internal conflicts and helping the client resolve them. However, soon after entering practice, I encountered people who had been abused early in life. To oversimplify, their avoidance of painful memories didn’t seem related to internal conflict. Soon after, I began to see people suffering from alcoholism. The most sophisticated interpretations failed to change their behavior. The problem is that most of the 500+ schools of therapy, are limited and not a great fit for all problems. One solution is to select only clients who match one’s preferred modality, but too often, people who come for help have the (naïve) expectation that a therapist is simply a therapist and will be able to help them. How likely is it that the therapist will say, “No, I don’t work with people who suffer from your kind of difficulty.”

It would be nice if each brand of therapy had a definite range of problems, but matching the therapy to the problem is far from clear, nor do practitioners typically master enough therapies to cover the full range pathologies. In the end, therapists are left to struggle with the limitations of the therapies they have learned, each one finding a more or less satisfactory way to adapt what they know to what their clients bring. And clients are left with little but hope that they guessed right in their choice of therapist.

5. Adhere closely to a manualized treatment

How could you go wrong? Manualized treatments of 16 sessions have been proven to be better than sham treatment and as good as other manualized treatments. The difference with branded schools of therapy is that manuals are written by a few authors, based either on an existing therapeutic tradition or on the author’s own intuition and experience. Rather than being honed by generations of therapists, they are vetted almost entirely through clinical trials, which have limitations in number of sessions and the level of adherence to the dictates of the manual. At best, they offer the same self-referential rationales as branded therapies, and at worst, they simply give a cook-book recipe for what to do and little rational basis for solving a clinical problem like predicting the possible outcomes of answering the client’s question versus avoiding it

6. Make predictions of what will happen, based on a modern scientific understanding of why humans have problems of the mind and how they can change.

What Freud, Watson, and just about every other inventor of a therapy were missing was a scientific understanding of why and how the human mind comes up with maladaptive response patterns and how talk and relationship can lead to change. Freud mentions Darwin just once in his writing, and that is to say how long it takes for a new idea to become integrated into our thinking. He was so right. It was only in the latter half of the 20th century that it became clear that the brain’s information processing, based on the functioning of our mammalian predecessors, is designed for survival, identifying threats and opportunities and responding in a way that quiets the limbic neurons that triggered the response in the first place.

Integrating this knowledge with clinical observation, we can see that the mind’s information processing, long before the arrival of humans, and honed by natural selection, operates largely outside of consciousness. Furthermore, observing its operation in humans makes it clear that the system does not always come up with the best answer. Today’s maladaptive patterns might have made sense in a pre-human context but are often sub-optimal in the 21st Century. In addition, given our slow development, the mind learns to model how life works in a childhood context which regularly turns out to be less appropriate when applied to adulthood.

In this way, the theory of evolution gives an entirely new solution to a problem that troubled philosophers and psychologists for ages, why humans do things that are not in their own best interest. To us, this may seem obvious, but consider that in 1920, Freud’s answer to that question was to posit the “death instinct.”

The other crucial question for our field is how maladaptive response patterns can be changed. For most of the history of psychotherapy its greatest mystery was how therapy actually works. How is it that words and relationship can produce changes in patterns that have remained fixed for most of a lifetime? 21st century neurophysiology, at last, offers the explanation that enduring change in the identification and evaluation of threats can take place through the mechanism of memory reconsolidation.  When the old pattern is activated in the limbic system and is contradicted by new information, the old model can be updated in memory. In the end, the identification and evaluation of the threat is permanently changed and the individual’s response is modified. This is not the only change mechanism but it corresponds closely with those corrective emotional experiences that produce improvement in response patterns that don’t fade or require further reinforcement.

Clinical decision making in the light of recent science

Returning to my colleague’s example, combining knowledge of the infrastructure of psychotherapy with clinical observation, we could look at his decision like this: By asking what his therapist would do in his place, the client was testing his uncertain trust in the relationship. He was asking if his therapist could understand and connect with him emotionally. My colleague made a prediction that the gain in trust and depth of the relationship was more important at that point than any gain from analysis of the content of the question. Implicitly, he also made a prediction that any future attempt on the part of the client’s non-conscious problem solver to seek closeness or comfort at the expense of processing more layers of broken trust from earlier in his life, could better be dealt with later in the context of an improved therapeutic relationship.

Developing our skill at making accurate predictions regarding the pros and cons of a given clinical decision provides a more flexible, universal way to navigate the therapeutic space. Learning to do therapy should focus on bringing together existing clinical wisdom with a universal framework based on understanding the evolution of the mind and the neurophysiology of change. Putting the two sources together, it becomes practical to base clinical decisions on our best predictions of what will happen if we take this fork in the road or the other. And there’s a cherry on top: When our predictions turn out to be wrong, we learn even more.

Jeffery Smith MD

Photo credit,  Jens Lelie, Unsplash