Psychoanalysis and Pharmacotherapy - Incompatible or Synergistic?

June 1, 1997
Leslie Knowlton
Volume 14, Issue 6

Is the rising use of psychotropic medication to treat anxiety and mood disorders incompatible with the psychoanalytic approach? As a psychopharmacologist and psychoanalyst who frequently provides consultation to analysts regarding medication for their patients, Steven P. Roose, M.D., has studied this question and presented his findings and opinions in various scientific papers, books and meetings.

Is the rising use of psychotropic medication to treat anxiety and mood disorders incompatible with the psychoanalytic approach?

As a psychopharmacologist and psychoanalyst who frequently provides consultation to analysts regarding medication for their patients, Steven P. Roose, M.D., professor of clinical psychiatry at Columbia University's College of Physicians and Surgeons, has studied this question and presented his findings and opinions in various scientific papers, books and meetings.

His most recent talk--part of a panel called "Are Therapists from Venus, Pharmacologists from Mars?"--was given in May at the American Psychiatric Association's (APA) annual meeting in San Diego. There, he specifically discussed sequencing medication and psychotherapy treatments.

In a telephone interview with the Psychiatric Times before that meeting, Roose, who is also codirector of the Late Life Depression Research Clinic at New York State Psychiatric Institute, shared his overall views on medication's role in psychoanalysis.

Q. What is the history of medication use in psychoanalysis?

Roose: In the late 1950s, when psychotropic medications first became available, the modality was mostly greeted with outright rejection by the psychoanalytic community. Medication was characterized as treatment that relieved symptoms but didn't affect underlying psychic conflicts then considered to be the etiology of psychological illness. It was reluctantly conceded that reduction of florid symptoms, albeit superficial, could be useful to the degree that it controlled behavior disruptive to development of transference, thereby facilitating analytic treatment. But still, it was strongly believed that psychoanalysis is a deep and curative treatment that should be left undisturbed whenever possible, and medication was considered an undesirable intrusion that should be supplemented only as a last resort. So even when effective, medications were to be considered, at best, a necessary evil.

Over the next decades, many studies showed medications worked well alone and even better in combination with psychotherapy, yet in those studies, the drugs were combined with either interpersonal, cognitive or supportive psychotherapy rather than dynamically-oriented treatment. But recently, a number of psychoanalytic authors, among them David Kahn, M.D., of the department of psychiatry at Columbia University and Shepard Kantor, M.D., at the New York Psychoanalytic Institute, have themselves concluded that combining medication and analysis may have a synergistic, beneficial effect.

Anecdotally, even Anna Freud, when visiting the New York Psychoanalytic Institute in the early 1980s, expressed surprise at the almost complete rejection she saw there of drugs being used during psychoanalytic treatment. She said she'd used psychoactive drugs with three patients in severe states of depression and in no case did the use of medication interfere with the progress of analysis. In fact, she said, the medications helped the analysis maintain itself during phases when patients might otherwise have needed hospitalization.

Recent evidence, however, indicates that medications are increasingly being used. In a 1995 study we did at Columbia University Center for Psychoanalytic Training and Research, where I'm a faculty member, we found that in the prior five years, about 18% of training analysts' cases were on medication, primarily for dysthymia or major depression. Even more importantly, 62% of those analysts had at least one patient on medication. Since no data on this topic have been reported prior to our study, it can't be known whether ours represents an increase in medication prescription with analysis, but I think it's reasonable to conclude that it does.

Furthermore, for 84% of patients treated with medication, analysts in our study concluded there was not only the expected therapeutic response to medication, but also a positive effect on analytic process as well3/4 results that contradict the belief that medication undermines the psychoanalytic situation. Since these training analysts represent a select group chosen by each local psychoanalytic institute and approved by the American Psychoanalytic Association, their use of medication represents clinical practice of leaders of the psychoanalytic community. And the overall analytic community's interest in the topic is reflected by the numerous sessions devoted to it at the American Psychoanalytic Association's scientific meetings.

So much has changed in the psychoanalyst's attitude, especially when it comes to schizophrenia, bipolar illness and depressions. But despite that open recognition, there's still competition between different models of the mind that strongly influences treatment choices and makes our eclecticism perhaps more apparent than real.

Q. How is that competition manifested?

Roose: One way is that clinicians artificially divide psychopathology by saying Axis I disorders are biologically hardwired, i.e., not derived from psychic conflict, and therefore not amenable to psychotherapeutic interventions. In contrast, Axis II disorders are seen as resulting from psychic conflict, and therefore amenable to interpretation, implying that psychological treatment and not medication should be the primary treatment modality. It should be recognized that such a division has no empirical basis and serves only to promote misguided competitions in our field. And, of course, such competitions also can harm the patient, who may not get needed medication because of theoretical issues.

Q. So when is it appropriate to medicate during psychoanalysis?

Roose: Considerations are similar in all disorders, but let's take depression as a specific example. Over the years, there's been an increasing divergence between psychoanalysts' concept of depression and diagnostic categories of affective disorder used in clinical psychiatry. Underlying the differences in diagnostic approaches are significant differences in theories of etiology.

But in clinical settings, we don't need resolution of theoretical conflicts regarding etiology to develop a system to guide medication use in combination with psychoanalysis. Rather, we must focus on phenomenology. Data establishing efficacy of medication in affective syndromes were collected in studies that included patients based on phenomenology of their symptoms, including chronicity, content and intensity, not the presumed etiology or meaning of the depression. Of course the depression had meaning to these patients, and like all patients, they presumably had elaborate conscious and unconscious fantasies to explain what was happening. Even if, in its origination, depression is brain state without content, as meaning-seeking and meaning-creating creatures, we're compelled to attach meaning and cause to every feeling. Although the psychic content coupled to a physiological tone may in fact produce inaccurate and misleading understandings, whatever the veracity of our self-interpretations, it's critical to acknowledge that identifying and understanding meaning isn't equivalent to establishing etiology.

Yet for the clinician struggling to decide on treatment for a depressed patient, meaning isn't the critical dimension. Rather, what should guide clinicians as to whether medication will be helpful to a specific patient is phenomenology of the psychopathology. If a patient meets diagnostic criteria for an affective syndrome, or any other syndrome, medication should be considered. It's not that every patient should be on medication, but in all such patients, it must be given serious consideration and if not recommended, there should be good rationale as to why not. It must be recognized that the effectiveness of a so-called biological treatment no more proves that there is a biological etiology to an illness than an effective psychological treatment proves there is a psychological etiology. Effective medication treatment, like effective psychotherapy, is not evidence that favors any theoretician; it only helps the patient.

Actually, many psychoanalysts have already come to this position regarding melancholic depression. They may still maintain they can determine both cause and dynamic meaning of the symptoms while concurring that rapidly effective somatic therapy is the treatment of choice.

Although the psychoanalytic perspective may no longer have significant relevance to melancholic depression or bipolar illness, psychoanalytic theory is still important, if not critical, when trying to understand the mechanism of transitory depressive affect states and the relationship of those states to experiences of hurt and maintenance of self-esteem and anger.

In addition, with or without medication, psychodynamically based therapy may have an important role in treatment of patients with chronic depressive states. Over the years, these patients have been labeled as having characterological depression, depressive neurosis and now dysthymic disorder, but regardless of label, the depressive affect permeates every part of the patient's life with widespread and profound effects on object relations, career, parenting and identity. Whatever the etiology of affect, the patient and analyst understand the mood as a consequence of fantasies, conflicts, drives and transferences. For these patients, the depressed affect is a fundamental part of their life narrative.

But again, it's important to emphasize that etiology of the depressive affect is not the crucial issue in terms of its impact on psychological development. By analogy, a significant physical malformation such as a missing limb is clearly not psychological in origin, but just as clearly such a defect will be a critical influence, albeit not the only influence, on the child's psychological development. Thus, as opposed to treatment for melancholia which is an episodic illness that has no demonstrated relationship to character structure or defensive operations and responds well to somatic treatments alone, treatment for a chronic dysthymic disorder may require understanding the dynamic meaning of the depressive affect and its impact on character structure, sense of self and interpersonal relations.

Q. What about timing of the use of medication in the course of psychoanalytic treatment?

Roose: Medication in combination with analysis generally involves consultation with a psychopharmacologist and often split treatment. Not surprisingly, this generates a unique set of technical and theoretical problems. Consultations are generally sought either during the initial evaluation for treatment or during the mid-phase of the analysis, and the different timings generally reflect different motivations for the therapist and often different meanings to the patient.

In the first case, the analyst presumably believes neither that illnesses can be separated into those with biological versus psychological etiologies and consequently require biological versus psychological treatments, nor that psychoanalysis is always preferable to medication. In fact, many psychoanalysts have long observed that intense and persistent dysphoric affects or high anxiety levels interfere with psychoanalysis to the degree they limit the patient's cognitive capabilities.

One advantage of consultation at the initial stage of treatment is that in most cases, strong and constant transference paradigms haven't yet developed, so disruptive effects of introducing a third person may be minimized. Also, it allows for consideration of whether treatments should be sequenced before combined, specifically whether in certain circumstances a medication trial should precede the initiation of psychoanalysis. This may also be approached from the other direction, that is, starting with a trial of psychoanalysis to reduce symptoms or establish a therapeutic alliance before committing to medication. One should be mindful of the long-standing principle in therapeutics that in order to evaluate efficacy and side effects, it's preferable to begin only one treatment at a time.

However, most consultations are sought after analysis has been well established, primarily because of persistence of symptoms of anxiety or depression, despite the presence of good therapeutic process. Many of these patients may indeed need medication. However, the major problems encountered are not diagnosis and treatment recommendations, but rather the often unexpressed meaning to both analyst and patient of turning treatment over to medication. If the analyst believes psychological treatment is always preferable to medication, which is only to be considered when psychoanalysis has failed, then the analyst may turn to medication consultation primarily out of frustration, and more importantly, with a tacit sense of defeat. Here, the pharmacologist's recommendation for medication, albeit correct, doesn't sufficiently address the problem, and a medication consultation shouldn't replace a more encompassing treatment consultation.

To the patient, the suggestion of medication may have multiple meanings, prominently one of failing to have worked effectively in analysis. They may think that if only they were good enough or had sufficient strength of character, they wouldn't need medication and could do it on their own. If the patient interprets recommendation for medication as proof the therapist is giving up, then outrage, despair and relief may all be expressed. One way a patient commonly expresses anger is saying, "If you think a medication is indicated, why is it only now being recommended? Why didn't you suggest this previously, especially since my symptoms have been chronic?" Actually, the answer to that can most often be found through analysis of the countertransference.

A variation on the mid-phase consultation occurs when the call to the consultant is initiated by the patient, often prompted by newfound conviction they have a biological illness and that medication will provide them with a faster and more dramatic result than psychoanalysis. Such behavior is invariably considered to be acting out. But whatever the motivation, it shouldn't obscure the fact medication might be indicated.

Furthermore, transference paradigms manifest during consultation represent important material to be brought back into therapy. For example, if a patient instantly feels that the pharmacological consultant is too cold, mechanical, superficial and only interested in symptoms, or alternatively, if the patient experiences the consultant as extraordinarily clear and feels he or she has an instant understanding of the illness, it's a reasonable assumption that these intense feelings represent part of transference that may not have been previously expressed. Therefore, interdicting consultation so as to prohibit acting out may deprive the patient not only of helpful medicine but also of treatment of significant transference material.

It's important to realize that effective administration of medication with psychoanalysis requires establishment of multiple therapeutic alliances, between analyst and patient, between patient and pharmacologist, and perhaps most importantly, between analyst and pharmacologist. Although dynamics are inherent in any consultation setting, a unique feature of this last relationship is that it's often cross-generational. More experienced senior analysts are seeking pharmacological consultations from a younger generation for whom they have often been teachers, supervisors and, not infrequently, significant transference figures. So if combined treatment is faring poorly, then raising dose or interpreting transference may not be the key intervention. Rather, it may be that attention to the relationship between therapist and consultant will correct the course of what is very often effective treatment.

Q. What do you see happening in the future?

Roose: Medication is now being used in combination with psychodynamic therapy with sufficient frequency so that it is possible to do prospective studies of medication's impact on therapy and vice versa, and equally important, what is the best way these combined treatments should be administered. The question is not only whether prescribing medication may interfere with an analytic treatment, but also whether the analytic situation interferes with optimal administration of medication. My concern here is that analysts doing excellent analysis may be constricted by the analytic situation, so that it might interfere with the prescription of medication. I don't know if this is so, but we have the opportunity to study these questions. It's no longer sufficient for me or anyone else simply to give their opinions. The issue is that a frequently occurring treatment like medication with analysis should be studied. We need to convert opinions into testable hypotheses and do well-designed studies that will give us data-based answers.

It's striking to me that during my generation the issue of medication with psychoanalysis has significantly changed. When I was training as a candidate, there was almost a don't-ask-don't-tell rule; if you had a patient on medication, you don't tell about it and the supervisor shouldn't ask. Within the past 20 years, there has been a dramatic change in the use of medication and it is probably the most frequently occurring parameter in current psychoanalysis. The issue needs to be studied and taught in the psychoanalytic curriculum and in psychiatric residency.