I have recently been wrestling with differential diagnosis, particularly within the anxiety disorders.

Some diagnoses are relatively straightforward. Social anxiety disorder, for instance, has clear telltale signs. A disproportionate fear of social judgment, embarrassment, and performance situations usually points in a single direction, assuming the person is actually seeking treatment for these problems.

Other presentations are far less clean. Repetitive, distressing thoughts could indicate OCD, health anxiety, or generalized anxiety. The thematic differences matter, but when you look at gold-standard treatment, the overlap is striking.

Across these conditions, the therapeutic sequence is largely the same.

First, we establish a therapeutic bond. That means personal alliance and task alliance. Without both, nothing that follows really sticks.

Second, we provide psychoeducation about anxiety in general. The goal is simple. The patient and I need the same map of the world. We need a shared language and a shared understanding of the thing in itself, the disorder as it functions, not just how it appears.

Third, we build an idiosyncratic maintenance model. Here, we help the patient see that their solution to anxiety, usually safety behaviors, has quietly become the problem that maintains it.

Throughout this process, I am always looking for function. If I cannot make sense of the patient’s behavior, I know I have not understood the problem well enough.

Avoidance makes sense if I believe speaking up will lead to humiliation. Excessive handwashing makes sense if I believe I am personally responsible for infecting my loved ones. The behavior is not irrational. The belief system is doing exactly what it is supposed to do.

Once we have alliance, shared understanding, and a functional model that actually makes sense, we are ready to test beliefs.

Whether we call this exposure or behavioral experiments is, in my view, mostly beside the point. What matters is that patients willingly approach what they fear and stop doing the behaviors that keep fear alive. Conceptually simple. Practically hard.

There are, of course, techniques that are more diagnosis-specific. Scheduled worry time can be useful in generalized anxiety. We would not typically use that in OCD, where re-exposure strategies may be more appropriate. These distinctions matter, but they sit on top of a much larger shared foundation.

And that foundation is the therapeutic relationship.

This is where my thinking aligns strongly with a Rogerian stance. Change requires more than techniques. The patient needs to acknowledge the problem, speak openly about it, and believe that a different way of living is possible. Without psychological safety and genuine collaboration, none of the technical work matters.

I began this reflection with differential diagnosis, and as a psychometrician an clinician, I care deeply about diagnostic precision. Diagnosis matters for science and for communication.

But if you do not also master alliance, functional analysis, and collaborative testing of beliefs, your diagnosis is not worth the paper it is written on.