Physicians' "compliance with treatment" in the context of consultation-liaison psychiatry: The role of "triangle" relationships and projective identification.

Transference and countertransference issues arising in the context of consultation-liaison (C-L) psychiatry could be more complex than originally assumed since they include reactions evoked within the frame of a unique "therapeutic triangle" of the patient, the physician, and the C-L psychiatrist. A clinical vignette illustrates how the projective identification process could mediate the relationships between the members of the therapeutic triangle through the different transferences and countertranferences interwoven in the setting of C-L psychiatry. This, if left unchecked, may result in the disruption of professional relationships and in jeopardizing the treatment of patients.

Introduction

Psychosomatic medicine has evolved since its beginnings in psychophysiology and psychoanalysis to become a subspecialty in the practice of psychiatry devoted to the psychiatric care of complex medical cases (Gitlin et al 2004). Consultation-liaison (C-L) psychiatrists also provide training to nonpsychiatric hospital staff, while the recently established European guidelines for training in C-L psychiatry recommend that specialists in C-L psychiatry should have a comprehensive understanding of issues of transference/countertransference in the relationship between physicians, staff, and patients, and should be able to integrate this knowledge into formulating a working hypothesis and a treatment plan (Sollner and Creed 2007).

One critical point in the dynamics that develop in the context of C-L psychiatry is that the patient is referred to a C-L psychiatrist by his physician, which means that the physician is also seeking the consultant’s medical advice and help. The consultant, may, therefore, have the character of an “object” for the referring physician as well, and may thereby attain the status of “third object”. Thus, within the frame of this aforementioned unique “therapeutic triangle” of the patient, the physician, and the C-L psychiatrist, transference and countertransference phenomena could be more complex than originally assumed.

Transference is classically defined as related to reactions and feelings of the patient toward the analyst based on earlier relationships and fantasies, whereas countertransference concerns the attitudes and feelings of the analyst toward the patient triggered by the patient and is derived from the analyst’s early relationships (Gabbard 2000; Blumenfield 2006). Transference and countertransference, however, are also developed in every close relationship. As Brenner (1982, p. 194–5) put it: “Every object relation is a new addition of the first, definitive attachments of childhood. Transference is ubiquitous; it develops in every psychoanalytic situation because it develops in every situation where another person is important in one’s life.” Accordingly, C-L psychiatry constitutes a frame where complex transference and countertransference issues amongst consultants, physicians, staff, and patients are prominent.

The main focus of attention in the relevant literature has been on transference issues that arise with patients and nonpsychiatric physicians or with patients and consultants (Blumenfield 2006; Lefer 2006), whereas countertranference phenomena usually are discussed with regards to the consultant’s efforts in establishing a diagnosis and in providing short-term treatment (Teitelbaum 1986; Blumenfield 2006). The C-L psychiatrist, though, has to deal with more complex transference-countertranference issues arising also between the physician and the consultant as well (Neuburger 2000). These phenomena are sometimes mediated by analogous phenomena derived from the physician’s relationship with his patient which, although equally important, have not attracted much attention. The present report aims to highlight some complex transference and countertransference issues within the C-L psychiatry setting and to show how the application of basic psychoanalytic concepts in clinical work in the medical setting has relevance in this area. The following vignette introduces some of the issues that will be considered.

Case report

The case concerns a man in his late thirties who was hospitalized in the hematology-oncology department of our hospital with the recent diagnosis of chronic myeloid leukemia. He was referred for a consultation because of noncompliance with the medical regimens suggested (ie, he kept “forgetting” to take his medicines, missing out a dose of his medication, or adjusting it to suit “his own needs”). Discussion of the patient’s noncompliance with his physician in our first meeting revealed that the suggested medical regiments were absolutely necessary for the treatment of leukemia and there was not any obvious side effect that could give reasons, partly at least, for noncompliance.

During the first interview the patient reported severe anxiety and depression symptoms, including fatigue, depressed mood, persistent dysphoria, irritability, feelings of worthlessness and guilt, insomnia, indecisiveness, and lack of the ability to negotiate his usual activities comfortably. As his history unfolded, it became evident that dysthymic features were present throughout his life. Repeated efforts to control the psychiatric interview were apparent in this first meeting, which included refusals to elaborate when requested to do so. For example, initially he spoke reluctantly and so softly that at times he was inaudible, while he refused sometimes to answer a question remaining silent or asking the consultant about irrelevant themes or about the consultant’s personal attitudes. Later on, he repeatedly interrupted the interview by asking for some water or wishing to go immediately to the toilet, and also by asking to postpone the interview because “he was expecting visitors”. When the consultant invited him to discuss this attitude, the patient admitted that although he had agreed with his physician about the present psychiatric consultation, he nevertheless felt embarrassed at having a consultation because of “mental malfunction”. “Besides”, he said in a clearly annoyed tone, “I am not as ill as my doctor claims. After completing some laboratory tests I will be discharged and will return to my family”.

The consultant checked for other possible medical conditions that could be involved in the patient’s mental state (eg, hyperleukocytosis, anemia, possible influence of different drugs used in the treatment of leukemia), and made an initial diagnosis of depression, recorded the available information, informed the hematologist about his patient’s condition, suggested the prescription of anxiolytic and antidepressive treatment, renewed his appointment with the patient for the next day, and asked for supervision at the weekly meetings of our C-L Psychiatry Unit.

At the meeting of the C-L group, the limited available data from this first interview were discussed and a hypothesis for a therapeutic approach was formed. The supervisor highlighted the patient’s intense transference reaction which led to rejection feelings towards his medical doctors that were expressed, at least in part, as noncompliance with the medical regimen. This was felt to be the result of denial and of his projecting his own aggressive feelings onto the doctor. He was blaming his doctor for “regarding his condition as more severe than it really was” and blaming him essentially for his suffering. He seemed to imagine his condition as a severe punishment by a cruel authority that came to be represented by his doctor who prolonged his discomfort and requested an embarrassing psychiatric consultation.

The discussion during the C-L meeting concluded by proposing a schedule for future psychotherapeutic interventions. However, the next day when the consultant went to the hematology department for his scheduled appointment with the patient, he surprisingly (and angrily) discovered that the patient had been discharged without the consultant being informed, as required; an uncommon practice in the collaboration between the two departments.

It seems that the first meeting between the consultant and the patient resulted in a good doctor-patient relationship because on the following day the patient telephoned the psychiatrist and asked him for a new appointment. In this session, the psychiatrist surprisingly discovered that the hematologist had not prescribed the medication that the psychiatrist had suggested even though the regular medication for the leukemia had been cautiously and correctly prescribed. The patient insisted that no one had suggested him an additional to leukemia treatment medicine and he refused to take antidepressant medication. It should be mentioned that it was the first time that this hematologist had omitted the prescribed medication suggested by a C-L psychiatrist for patients who also presented noncompliance with the medical regiments suggested.

Discussion

Several hypotheses and explanations about this lapsus on behalf of the hematologist could be posed. Although the diagnostic issues within psychosomatic medicine are an area of concern with many difficulties in diagnosis issues with the failure to recognize depression and initiate antidepressants not being rare in primary care (Barbui and Tansella 2006; Ruttley 2006; Wise 2008), a glance at the dynamics of this “triangle” relationship could help us to better address clinical issues and problems arising within the setting of C-L psychiatry. While many interesting aspects of this case come to mind, here the focus will remain solely on countertransference and projective identification issues. It should be mentioned, however, that projective identification as a concept here is primarily a tool for generating useful and sometimes uncannily insightful hypotheses, such as the present one, and it is certainly not a tool for observing facts.

The hematologist’s omission may well have had its origins in the complex field of the transference-countertransference relationships between patient, hematologist, and C-L psychiatrist. The patient had been recently diagnosed with leukemia and his psychiatric examination revealed that he was anxious and depressed. An initial diagnosis of mood disorder was made; the possibility, however, that his symptoms could be related to the medical condition or to the acute stress induced by the announcement of leukemia diagnosis could not be disregarded. With regard to the patient’s personality, although the patient did not fulfill the criteria for a personality disorder, his history provided evidence that he exhibited a long history of dysthymic symptoms.

It has been reported that the diagnosis of cancer in many cases provokes a crisis that leads to regression and to the extensive use of the defense mechanisms of splitting, denial, and projective identification (Risko et al 1996). It has also been reported that dysthymic subjects present higher rates on individual defenses of projection, passive aggression, acting out, and projective identification (Bloch et al 1993). It is possible that, in our case, the process of projective identification had been triggered in the transference relationship between the patient and his hematologist. For example, primitive aggression, helplessness, guilt feelings, self-hatred, and the need for punishment could be projected onto and into the doctor who then possibly becomes a persecutory object to be avoided. Hence, perhaps, the patient’s noncompliance. Interestingly, a similar process of projective identification seems to have been triggered in the “transference” relationship between the hematologist and the consultant as well.

Projective identification is an unconscious process by which, initially, as in simple projection, aspects of oneself are disavowed and attributed to someone else, but the process involves two additional steps (Ogden 1979; Goldstein 1991): a) the patient projects a self- or object representation onto the therapist, b) the therapist unconsciously identifies with what is projected and begins to feel or behave like the projected representation in response to interpersonal pressure exerted by the patient, and c) ideally, the projected material is “psychologically processed” and modified by the therapist, who returns it to the patient via re-introjection (Gabbard 2000). However, projective identification regularly occurs also in nonpsychotherapeutic situations (Gabbard 2000) as well as within the hospital setting (Gabbard 1989). In these instances, the projections may be returned in completely distorted forms, instead of being modified or contained (Gabbard 2000). It seems that this was the case in the relationships between patient, hematologist, and consultant in the case presented.

During the first step of projective identification, the patient seemed to disavow his own helplessness or his aggressive aspects and projected them onto the hematologist, who was regarded as “cruel” and “unbending”, through their transference relationship. We hypothesize that the physician then, unconsciously identified with the projected “bad” object and began to feel or behave like the projected hostile object representation in response to interpersonal pressure exerted by the patient. Next, the hematologist had two courses of action: a) he could have recognized that these feelings belonged to his patient and were simply projected onto him and so he could have proceeded to work them through and return the projected material to his patient “psychologically processed” and modified, or, b) he could return the projected material to his patient without any elaboration. That is, he could behave exactly as suggested by the emotion that he felt. No one, of course, would expect a hematologist to do deep self-analysis and “elaborate” the “patient’s projected persecutory bad object”. The physician simply acted unconsciously as he felt or rather, as his patient had felt. And he acted as his patient was acting: he did not comply with the treatment. Therefore, through his “transference” relationship with the consultant, he behaved towards the consultant exactly as the patient was behaving towards him. Accordingly, through his countertransference reaction, he returned to the patient all the projected hostility and aggressiveness without any elaboration. The patient was discharged without any chance to be treated for his depression. Consequently, through the hematologist’s projection onto the psychiatrist of the patient’s hostility that had been projected onto him, the psychiatrist also became “impotent and inefficient” in his treating capacity.

As Rodewig (1995) has mentioned, a physical ailment itself can have the character of a “bad” object: given the threat posed by dangerous physical illness, the ego has recourse to defence mechanisms such as splitting and separate projective identification of positive and negative object- and self-parts, projecting the omnipotent, idealizing desires onto the doctor and the negative desires onto the ailment itself. In a later stage, a de-idealization of the doctor sets in, and the latter is identified with the illness so that the illness is then bandied back and forth between patient and the therapist (Rodewig 1995), with the doctor becoming a sort of “bad” and “persecutory” object. Within the setting of C-L psychiatry, which involves the illness and three players, the situation has the potential of becoming even more complicated as the patient has the opportunity to split the object into a “bad” object (in our case, located in the physician) and a “good” object (in our case, located in the C-L psychiatrist). This situation clearly interferes with effective management as the patient’s conflict is now projected outside him/herself to be fought out between the two physicians.

This vignette illustrates the different transferences that are interwoven in the frame of C-L psychiatry and highlights the role of the consultant in acknowledging and unraveling the tranference-countertranferrence issues evoked by these complex relationships. In our case, the consultant initially felt (consciously) angry that his recommendations were ignored. After discussed these feelings in the C-L supervision meeting a new insight was gained. This resulted in a very helpful and cooperative discussion with the physician focusing on the dynamics possibly underlie the rapid discharge of the patient without informing the consulting psychiatrist and the physician’s omission to describe the suggested antidepressants. This discussion was proved very helpful, since the whole climate of the collaboration between the two departments was even much better in the following referrals.

On the contrary, if the consultant had not acknowledged the underlying process of projective identification, he would have returned the projected hostility back to the hematologist without any psychological processing and modification. Thus, the blame about this omission would have been placed exclusively on the hematologist or on the entire function of the hematology department, thereby disrupting the climate of good cooperation between the two departments.

The consultant’s task is to recognize and to deal with these countertransference issues and then to return the projected material modified to the physician providing the proper explanations and interpretations, thus enhancing the professional relationships between the two departments. With regard to the patient, the consultant’s task is to try to help the patient to “merge” the previously “splitted” aspects of the object (the “bad” object, located in the physician and the “good” object, located in the C-L psychiatrist, as the rather impressive patient’s call to the psychiatrist for a follow-up appointment indicates). As Grete Bibring (1956), one of the pioneers in the C-L psychiatry and psychosomatics discipline, said: “in the doctor’s work, psychological understanding is of profound importance. It evokes in the patient all his positive strength, his willingness to cooperate, and his constructive wish to get well and to do right by himself and by his doctor. Thus, the optimal psychosomatic condition is established that make the difference between a patient who wants to live and the apathy and sabotage of the patient who lets himself die” (Lipsitt 2001).

Projective identification develops not only in psychiatric patients, but also among people with normal and neurotic personality organizations (Hamilton 1996). As Hamilton (1996) said: “all of us, continually need to define and redefine ourselves in relation to others as we grow and change”. Thus, the cooperation of clinicians of other specialties is critical in the implementation of therapeutic alliances in order to early acknowledge projective identification and to diffuse this dynamic which, if left unchecked, can disrupt professional relationships and compromise the treatment of patients (Robertson et al 1996). Medical staff and nurses can work more effectively with patients when they have an understanding of early object relations theory, the defences of projective identification and splitting, and their own countertransference reactions to patients and their illnesses (Teising 1997). Participation of the nonpsychiatric staff in conferences and supervision meetings of the C-L psychiatry units could enable them to deal with these intense countertranference feelings and could also help the residents deal with the unique aspects of the liaison and consultation environment.