I liked her better before she healed

She healed. By any clinical measure, the case was a success. She came in three years ago with severe depression, significant anxiety, a history of relationships organized around her own damage. She did the work. She changed. By the end of the second year she was stable, functional, and building a life that would have been unrecognizable to the person who first sat across from me.

By the third year I was finding the sessions difficult to sustain interest in.

I want to be precise about what I mean, because the imprecise version of this statement sounds like a complaint about clinical success, which it is not. The success was real and I was glad of it. What I am describing is something that arrived alongside the gladness and that I did not have language for until I had seen it repeat across enough cases to recognize the pattern.

The client in crisis has a quality of presence that the stable client does not. The crisis produces a specific directness: the person who has exhausted their defenses, who has nothing left to protect because everything has already been exposed, who comes into the room with the most essential version of themselves because there is no energy left for the performance of any other version. That quality of presence is the thing I am trained to work with. It is also, in ways I have been reluctant to admit, the thing I find most compelling about the work.

The stable client is a different person. The stability is genuine and it is the point. The client who no longer needs to expose everything because she has rebuilt enough structure to protect herself again is doing exactly what recovery looks like. But the therapist who spent two years in close contact with the crisis version of that person now sits across from someone who is, in a specific sense, less available. The stability is a form of closure. The person who came in is not the person who is leaving, and what is leaving is something I find harder to be with than what arrived.

The clinical literature addresses termination grief as a process the client goes through. The client loses the therapeutic relationship. The therapist manages the client’s loss and maintains the clinical frame through the ending. What the clinical literature does not address is the thing the therapist is also losing, which is the specific quality of relationship that the client’s need created. The client in stable recovery no longer needs the therapist in the same way. The therapist, if they have been doing genuine work, was built for that specific need. The need ending is a loss for the therapist, and the profession does not have language for it that does not sound like a clinical problem.

I am not saying I wanted this client to stay sick. I am saying that the version of her I knew, the version I worked with through the worst of it, was more vivid to me than the version she became. The crisis version had a specific quality: the unfiltered observation, the directness that people have when nothing is working and so there is nothing left to protect. The stable version is kinder to herself, which is better, and less interesting to be with, which is also true.

I have thought about whether this preference is pathological. I have concluded that it is not primarily pathological. It is the natural result of what the work builds. The therapeutic relationship was formed in and by the crisis. That is the relationship I know. The person she became after the crisis is a person I met only briefly, at the end, before the relationship ended. I liked the person she was becoming. I knew better the person she had been.

The argument is not that this preference should influence clinical decisions. It should not. The argument is that the preference is real, and that it has clinical consequences when it is not examined. The therapist who does not know they have this preference will find reasons to extend treatment past its clinical usefulness. The therapist who does know it will notice the pull toward extending and can ask the harder question: whose need is being served by continuing?

I terminated her case at the appropriate clinical point. The case notes reflect a successful termination by all clinical measures. What the case notes do not reflect is what I noticed in the last few sessions: a quality of loss that was mine, not hers. She was ready to end. She was building a life that did not require me. I was the one sitting with the particular kind of grief that comes from losing a version of a person you will not see again, because the person who survived is not the same person as the person who almost did not.

The profession does not give therapists permission to call this grief. It would rather call it countertransference and process it toward resolution. The resolution protocol works for clinical material. This is not primarily clinical material. This is the cost of having been present with someone during the worst period of their life and then watching them leave it.

She left it. That was the whole point. The whole point was the loss I felt at the end. The fact that the loss was mine and not hers is the measure of how well the case went. I know that. I carry the loss anyway, which is what it costs to do the work at the level the client deserved.