February 3, 1998

Analytical Listening and the Act of Prescribing Medicine

presenter: Larry Sandberg, M.D.
discussant: Steven Roose, M.D.
reviewer: Elizabeth Schwarz, M.D.

Should an analyst medicate his patients or refer them out? If an analyst does medicate his patients, how does he listen effectively as an analyst and as a psychopharmacologist? How does medicating a patient affect the transference, countertransference and other aspects of the treatment? These questions, among others, were raised during this presentation. Dr. Sandberg presented his paper, which addresses the issue of assessing and treating patients both pharmacologically and analytically. He proposed that in order to do both a therapist must listen bimodally. In his paper he gives three clinical vignettes in which issues of medication and therapy arise. Dr. Roose then discussed both the clinical vignettes and how bimodal relatedness represents a fundamental paradigm shift in analytic treatment.

Dr. Sandberg begins his paper by pointing out the tension that has always been present between a medical model and analytic work. He then asks the important questions: How does the mind of a pharmacologically informed analyst work? And how is his listening influenced when he is medicating his patients?

In addressing the therapeutic aspects of analysis, Dr. Sandberg notes that there are basic assumptions in psychoanalysis regardless of the particular theories employed. These assumptions include the influence of early life experience on later life, the role of the unconscious, the centrality of transference, infantile conflict and the use of free association. Dr. Sandberg points out that the analyst listens—sometimes with free-floating attention, sometimes in a more focused manner—and tries to elucidate unconscious elements. The analysand begins to understand himself and accept responsibility for his choices. In analysis, symptoms are to be understood and the relationship itself is a therapeutic vehicle.

A psychopharmacologist, on the other hand, sees symptoms as something to be treated. Dr. Sandberg differentiates between a patient’s pharmacology and a patient’s biology. He points out that a patient’s biology includes much more than medication and a patient’s response to medication.

Dr. Sandberg highlights the tension between medicating and analyzing a patient: analysis explores and tries to understand a patient’s symptoms, whereas drugs are used to eradicate symptoms. Dr. Sandberg suggests that handling this tension effectively has to do with the ability of the analyst and patient to tolerate ambiguity. In order to simultaneously treat a patient pharmacologically and psychoanalytically, the doctor must listen to the patient bimodally, oscillating between both modes as necessary.

Dr. Sandberg presented three cases illustrating issues involving medication and therapy. The first case he titled “Talk Therapy as Resistance to Medication.” Dr. Sandberg presented the case of a graduate student at a competitive institution who had difficulty focusing and had anxiety that was circumscribed around work. The patient also had a restricted social life. The patient was of superior intelligence, with a history of a distant relationship with an absent, less educated father. The patient also noted he was bored with school in childhood. Initially, Dr. Sandberg understood that the anxiety was in part due to pining for and competing with his father. Upon further exploration, Dr. Sandberg found the patient had a history of attention problems in school and had been tested at a young age and recommended for stimulant treatment, which he did not receive. Dr. Sandberg hypothesized that rather than being constrained by oedipal anxiety, the patient had shown ambition in the face of a deficit, and that his choice of neuroscience reflected in part an attempt to master an organ he found defective. The patient was started on stimulant treatment and supportive psychotherapy, addressing narcissistic issues relating to the ADD diagnosis. After several months, psychotherapy was discontinued. The patient did well in his studies and eventually became involved in a relationship with a woman he hoped to marry. Dr. Sandberg pointed out that a conflict model had led him to be curious. Pursuing boredom as an affect, he learned additional information that led him to reevaluate his hypothesis. Analytic curiosity and technique helped to establish a more accurate diagnosis and to recommend treatment that in fact was not analytic.

The second case that Dr. Sandberg discussed was labeled “Request for Medication as Resistance to Talk Therapy.” The patient, a 40-year-old man with a history of severe depression since his teenage years, presented with a request for medication, not psychotherapy. The patient’s most recent depression had lasted for almost a year and symptoms included despair, poor sleep, impaired concentration and suicidal thoughts. He was married and had three children. Previously, the patient’s depression had served to shield him from deeper engagements. When he presented for treatment, however, the patient felt his capacity to function at work and home was deteriorating. The patient’s family history was significant for mood disorder. He was a child actor with a narcissistic mother who sometimes directed him in his work. When the patient was a teenager, his parents divorced. His mother remarried an actor close to the patient’s age, which is when the patient first became depressed. Dr. Sandberg started the patient on sertraline, and although his mood improved, he continued to feel inauthentic, disconnected, fearful and ineffective. It became clear to Dr. Sandberg and the patient that he was trying to perform and to avoid exposure in his treatment, as he did in other areas of life. After the patient’s mother died suddenly and he was unable to express feelings about the event, he agreed to start analysis. A negative maternal transference emerged, and it was revealed that his history included soiling, being spanked with a metal swatter and being slapped when he confused his lines while being directed by his mother. Through analysis he understood his feelings of being damaged, enraged and full of shame. The patient wished to be strong and available to his own son. During the treatment, the patient and analyst learned that characterological difficulties contributed to triggering depressions, and the depression had served defensive functions in the past.

The third case was labeled “A Cautionary Tale.” A 35-year-old businessman was referred to Dr. Sandberg for medication by an analyst. The patient had bouts of severe depression and was started on sertraline. In two months a tremendous change had occurred, and the patient ended his analytic treatment. Six months later, however, he became mildly depressed due to job stress. There were persistent problems at work around authority; at times the patient was not tough enough, and at times he was too tough. The patient entered therapy and then analysis with Dr. Sandberg. The patient was intellectually gifted, but his father thought he was stupid. When he was young his father would become exasperated when helping the patient. As a child he had retreated to his room, and as an adult he retreated to his office. His mood disorder had precluded analytic work previously. The medication treatment clarified the persistent neurotic conflict and removed the obstruction of the affective illness, which allowed these issues to be addressed.

Final Comments

Dr. Sandberg notes that the capacity of the analyst to analyze and to medicate is a partial theory, and the attitude is not reductionist. He feels the introduction of medication does not preclude understanding the inner world. He points out that the analyst listens to what is typically ignored: signs of the unconscious. Listening is made more complex when medication is introduced, as the pharmacologically informed analyst is also listening for familiar patterns and symptoms that may signal a potential responsiveness to medication. Dr. Sandberg notes that when first medicating a patient, there are gross oscillations between the two stances. In addition, the analyst must understand his own attitudes toward medication and how it plays out in the countertransference. For example, an analyst may medicate because of an inability to tolerate the patient’s affects. The decision of the analyst to use medication and the patient’s acceptance of the recommendation becomes a basis for transference and countertransference themes.

Dr. Sandberg points out that even when the treatment is split the analyst needs rudimentary knowledge of medications (that is, time of onset, side effects) to treat the patient effectively and to understand the repercussions on the analysis. In an ongoing treatment, nontransferential aspects of the drug need to be assessed periodically. Dr. Sandberg suggests that when a medication regimen and response are complicated, it may be best to split treatment. The issue of medicating one’s patients raises questions of analytic identity and the need to wear at least two hats. Dr. Sandberg notes that the mind cannot be neatly divided. At a given time, analysis, drugs or both interventions may be necessary. By understanding the scope and limits of medication and analysis, Dr. Sandberg feels that each field is sharpened. Dr. Sandberg concluded by cautioning analysts not to be caught in their own reductionism. He notes not only the crucial place of listening in healing but also the natural tensions between listening and looking, and the scientific and humanistic approaches.

Dr. Roose’s Discussion

Dr. Roose suggested that Dr. Sandberg’s approach represented a fundamental paradigm shift. Analysts are no longer talking about whether to medicate, but how best to medicate. He proposes that the bimodal relatedness model requires not only a shift but a theoretical revision.

Dr. Roose discussed each of the clinical vignettes and his interpretation of the salient issues. The first case of the graduate student with anxiety illustrated that it is important to consider a comprehensive differential diagnosis, including nondynamic diagnostic systems, when doing a consultation. This protects both the analyst and the patient from embarking on inappropriate treatment. He also proposed that there were significant interactions and interpretations that occurred between Dr. Sandberg and his patient that he felt accounted in part for the patient’s clinical improvement.

In the second case, Dr. Roose noted three pearls. First, it is best to begin one treatment at a time whenever possible. This case illustrates sequencing, as the treatment of depression allowed analytic work to occur. Second, he noted the importance of being able to evaluate residual symptoms. Dr. Sandberg understood that medicating the depression allowed symptoms of intrapsychic conflict to come to the foreground. These symptoms needed to be distinguished from the depressive symptomatology. Dr. Sandberg discerned that the residual symptoms were not due to a partially treated depression, and he ordered psychoanalysis as a second treatment. Bimodal relatedness requires bimodal knowledge of the field of pharmacology and psychoanalysis. Lastly, Dr. Roose pointed out that adopting bimodal listening adds an element of complexity to the genetic history. An analyst is faced with both the genetic inherited diathesis to an affective disorder as well as the early interpersonal and intrapsychic experiences that result in identification with a depressed parent. Medication may treat the genetic diathesis of the affective disorder, but interpretive work is necessary to address the effect of this patient holding the depressed parent close.

Dr. Roose noted that the third case, the businessman who had had a prior analysis and history of severe depression, illustrates how the patient’s depression was an obstacle to dynamic therapy. The analytic treatment was successful only after the depression was treated with medication.

After discussing the cases, Dr. Roose presented his view that a bimodal shift implies revision of theory. He pointed out that the linear causal relationship at the core of the ego psychological model, in which intrapsychic conflict causes affect, may need to be made circular. In patients who suffer from chronic depression or anxiety, the affect creates intrapsychic conflict. Pathological states generate rather than reflect intrapsychic conflict. Dr. Roose stated that a bimodal approach does not entail simply tempering the scientific with the humanistic aspects on an evenhanded basis. Rather, proper application requires not holding both models simultaneously and equally at all times. In some instances, there are times when one model is appropriate and others are not. Proper application of a bimodal model precludes that the pharmacologic and analytic treatments always hold equal weight. Effective treatment is dependent on the application and relative valence of each model at a given time.

Medication treatment and dynamic therapy come from conflicting points of view. Medication treatment values clarity, attempts to identify common pattern of illness, and concludes that the more direct the treatment the better. Dynamic therapy, on the other hand, searches for complexity, accepts ambiguity, and looks for what is unique in a patient, his history and the doctor/patient dyad. In analysis, the analyst constructs complex formulations and accepts that he can never know all. The pace of analytic treatment is slow. In addition to a clash of cultures, bimodal relatedness stimulates conscious and unconscious fantasies regarding how the analyst feels about action. Medical doctors often feel they know what is best for the patient and make recommendations. And the patient wants the doctor to have all the answers. For analysts, the acceptance of this position is an anathema, making it difficult at times for analysts to adopt a straightforward medical model. Dr. Roose gave the example of seeing a urologist for relief of urinary retention and wanting a magical, mechanical cure. He stated that it is acceptable for a patient to want a rapid cure. Part of the difficulty in adopting a medical model while analyzing a patient stems from the fact that Dr. Roose characterized Dr. Sandberg as a hardworking analyst struggling to make theoretical and technical ends meet. He proposes that the attempt to incorporate different modalities in the treatment of his patients is a revolutionary position that introduces vitality into the field.

Audience Discussion

A lively audience discussion followed the presentations by Dr. Sandberg and Dr. Roose that raised interesting questions and issues relating to medicating and analyzing patients. A number of participants noted the difficulty of going back and forth between the therapeutic modes, and the need to recognize how both aspects of treatment affect each other. One participant noted that the analyst must not only address the practical aspects of providing a dual treatment but also help the patient to deal with the dual nature of his condition. The next area of discussion was the belief of some audience members that the neutrality of the analyst is compromised by prescribing medication, which implies taking action. Lastly, a number of analysts commented on the impact of introducing medications on an analytic treatment from their own experiences. It was noted that medication can decrease affective symptoms and anxiety and can therefore influence defenses and resistances. Medication can paradoxically enable a patient to be able to tolerate and to express higher levels of anxiety and depression, thereby allowing a patient access to aspects of his life that had been buried.


Dr. Sandberg summarized that analysts have both a burden and a responsibility to have an integrated model. Even when the treatment is split, the patient cannot be split. The patient needs help to integrate the effect of drugs on his mental functioning and to understand the meaning of his analyst prescribing or not prescribing medication. He noted the importance of raising these questions and of the ensuing discussion.

Dr. Roose noted that there is an opportunity to study these issues and their impact on treatment. He feels that the decision to medicate or not is a practical one that is different for each person. He agreed that the process of discussion is valuable. He concluded that it would be useful if analysts could be comfortable talking about other issues with which they struggle in an equally open manner.