INTRODUCTION
One of the most significant changes in inpatient psychiatry has been the dramatically shortened length of hospital stay from months and years to ten days or less (Lieberman et al. Reference Lieberman, Wiitala, Elliott, McCormick and Goyette1998). As a result, the medical model has become the prevailing approach to inpatient treatment (Rabiner, Reference Rabiner1986). There is a literature which addresses issues of countertransference in inpatient psychiatry (Brown, Reference Brown1980; Lewis, Reference Lewis1979; Markowitz & Milrod, Reference Markowitz and Milrod2011); however, the focus has generally shifted from understanding the psychodynamics complicating and/or contributing to psychiatric illness to behaviourally oriented treatment planning and case management in inpatient psychiatry (Gabbard, Reference Gabbard2005).
During the era of long hospital stays, the concepts of transference and countertransference were well known and played a central role in patient care and the life of the inpatient therapeutic community. Transference and countertransference can still negatively impact patient care even in this time of short lengths of stay. In studying staff reactions to difficult patients, Colson et al. (1986) commented that treatment problems could occur from the multiple, strong, and often conflicting emotions that these patients evoke. Some staff may feel flooded, overwhelmed or immobilized with affect. Members of the treatment team are affected differently, which increases the potential for tension, disagreement, and splitting. Main’s (1957, p. 129) classic paper observed ‘… the practice of medicine like every human activity has abiding, unconscious determinants. We know that if human needs are not satisfied, they tend to become more passionate, to be reinforced by aggression and then to deteriorate in maturity with sadism invading the situation, together with its concomitants of anxiety, guilt, depression, and compulsive reparative wishes, until ultimate despair can ensue.’
We will present two case histories to illustrate how countertransference may impact the assessment, treatment, and management of the psychiatric patient, and then we will review the concept of countertransference and strategies to recognize and manage it. These cases were chosen for this paper because each one was presented at the hospital Morbidity and Mortality Conference at the request of the hospital’s Chief Medical Officer who was aware of ongoing treatment problems.
Case 1
A 56-year-old woman with a past psychiatric history of schizoaffective disorder was stabilized on risperidone and oxcarbazepine for five years. She was initially admitted to the internal medicine service with hyponatremia (Na = 108) likely secondary to psychogenic polydipsia. The patient experienced three generalized seizures as a result of the severely low sodium levels, and the beginning of her hospital course was marked by periods of agitated delirium. The patient’s hyponatremia was appropriately corrected by the internal medicine service. A psychiatric consult was requested to assess the need for inpatient psychiatric hospitalization once the patient’s delirium improved. The consulting psychiatrist found the patient to be euthymic with a circumstantial thought process and thought content lacking hallucinations and delusions. There were no symptoms of depression or mania. The patient was subsequently discharged home.
Two days later the patient was brought to the psychiatric emergency service by her family after she complained of auditory hallucinations, persecutory delusions and threatened to kill herself with a knife. She was admitted to the psychiatric inpatient unit. Over the next eight days, the patient’s condition deteriorated: she became labile, child-like, intrusive, grandiose, and was sleeping about two hours every twenty-four hours. The patient was placed on haloperidol 15 mg daily. However, she did not respond to the medication and she became increasingly agitated. During one interview, she choked the psychiatrist and kicked a nurse. The attending psychiatrist concluded the behaviour was secondary to an ongoing delirious process and sent the patient to the emergency department for a medical evaluation. The patient underwent a repeat CT of the head, which showed a hypodensity in the right hypothalamus and right medial temporal lobe. This finding raised concerns of an intracranial pathological process. However, an MRI revealed no significant findings, and the patient had no neurological signs to explain her behaviour.
The patient was medically cleared and transferred back to the inpatient psychiatric unit under the care of the same psychiatrist. The patient was started on haloperidol that was titrated up to 10 mg twice a day. She showed minimal improvement as she engaged in bizarre behaviours including eating crayons, drinking soap, and disrobing in public. She was consequently tried on orally disintegrating olanzapine 20 mg, which was titrated to a twice-daily dose. The patient still did not sleep, remained intrusive and disorganized with periods of waxing and waning disorientation. During this time, the patient’s behaviour escalated and she ripped a drawer apart in her room. The patient was unpredictable, easily agitated and placed on one to one supervision. In response to the patient’s escalating behaviours such as kicking a nurse and choking the attending psychiatrist, staff decided in the treatment meeting that the patient should be confined to her room because of her regressed and potentially aggressive behaviour. At this point the patient had been on haloperidol for a month followed by an additional month of olanzapine. Because of the lack of patient improvement, the treating psychiatrist felt that the patient was delirious and needed to be transferred back to the medical unit for further evaluation. Despite other diagnostic and treatment suggestions from colleagues during a morbidity and mortality conference, the attending psychiatrist still insisted the patient was delirious and continued to voice concerns about the patient’s potential for aggression. The attending physician then went on vacation. A covering psychiatrist consulted a senior attending who suggested the patient was presenting with a syndrome previously described in the literature as manic delirium (Swartz et al. Reference Swartz, Henschen, Cavenar and Hammett1982). He diagnosed the patient with mania and recommended the addition of a mood stabilizer. Still skeptical, the treating psychiatrist returned from vacation and considered ECT, but agreed to start the patient on valproic acid before ordering ECT. After adding the valproic acid, the patient’s mental status and behaviour began to improve. She was discharged on valproic acid 1000 mg in the morning and 1,250 mg at bedtime with a therapeutic level of 99. She was also continued on and discharged on olanzapine 20 mg twice a day.
Case 2
A 48-year-old man with a history of schizophrenia and cocaine dependence was admitted to the psychiatric inpatient unit from the internal medicine service after swallowing coins and intentionally reopening a healing abdominal surgical scar. The patient had a long history of self-mutilation including swallowing foreign bodies resulting in intestinal obstructions as well as inserting objects into his urethra. The patient was initially started on haloperidol 10 mg at bedtime for paranoid delusions and thought broadcasting. However, the patient continued to insert items into his surgical scar and even swallowed a small battery extracted from a radio headset. The patient’s haloperidol dose was then increased to 10 mg twice a day and he was started on a 200 mg haloperidol decanoate injection. Despite this, the patient continued to display poor frustration tolerance especially around the issue of medication administration. He began locking himself in his bathroom at night and started ripping out plumbing from the sink and toilet. Following these nightly episodes, the treatment team developed a behavioural treatment plan, which the patient did not follow as he continued dismantling things and inserting foreign objects into his body orifices. The patient was switched to a forty-eight hour trial of olanzapine 20 mg daily with no improvement. Rapid changes in medication were the result of an increased sense of urgency to bring the patient’s behaviours under control. Medication trials for this patient were brief and systemized trials were not implemented because of staff voicing increasing feelings of frustration and anger towards the patient especially in team meetings. The treating psychiatrist then attempted to logically bargain with the patient in an attempt to improve adherence to his behavioural treatment plan. The frustrated staff began urging the patient be placed in walking restraints. Feeling the staff’s pressure, the treating psychiatrist finally considered the idea of diagnosing the patient with cluster B character pathology and pressing criminal charges against him. Before proceeding, however, the psychiatrist reviewed the case with a senior attending who suggested starting the patient on chlorpromazine 300 mg twice daily. The patient had a dramatic response characterized by a substantial reduction in destructive behaviour and a significant decrease in the intensity and intrusiveness of his delusions.
DISCUSSION
An overview of countertransference
The psychotherapy literature describes two definitions of countertransference: classical and total. The classical definition identifies the clinician’s past neurotic conflicts as the source of countertransference while the totalist definition identifies the past and present experiences of the patient and clinician interaction as the source (Lakovics, Reference Lakovics1985). In discussing the evolution of the concept of countertransference, Gabbard (2005, p.21) observed that countertransference ‘[entails] a jointly created reaction in the clinician that stems in part from contributions of the clinicians past and in part from feelings induced by the patient’s behavior.’ Psychiatric inpatients may evoke intense countertransference reactions from clinicians and staff because of their severe psychopathology, a lack of staff understanding of their own reactions, and the absence of effective psychotherapy on inpatient units (Gabbard, Reference Gabbard2005). Main (1957, p.129) commented that ‘the sufferer who frustrates a keen therapist by failing to improve is always in danger of meeting primitive human behaviour disguised as treatment.’ He went on to state that ‘…one might almost say recalcitrant patients, treatments tend, as ever, to become desperate and to be used increasingly in the service of hatred as well as love…’ (Main, Reference Main1957 p.130).
Hate encompasses a major theme in discussions of countertransference. Patients with psychosis can be ‘irksome’ and pose significant emotional burdens on those who treat them and therefore the clinician is under greater strain to keep the hate latent (Winnicott, Reference Winnicott1949 pp 69, 72). Unconscious countertransference may generate well rationalized but destructive acting out by the therapist (Maltsberger & Buie, Reference Maltsberger and Buie1974). Countertransference hate is a mixture of aversion and malice with the aversive component being the most dangerous to the patient as this is what tempts the therapist to abandon the patient (Maltsberger & Buie, Reference Maltsberger and Buie1974). Clinicians may try to avoid countertransference feelings by withdrawal, acting out, excessive mothering, denial, avoidance of unpleasant issues, over permissiveness, acceptance of the patient’s distortions or application of authoritarian measures including drugs and electroconvulsive therapy (Savage, Reference Savage1961). Countertransference feelings of anger in staff have been observed in patients who are agitated, belligerent and disobeying unit rules (Lion & Pasternak, Reference Lion and Pasternak1973). Rossberg et al. (2008) found that specific personality traits may evoke various countertransference reactions. Patients who presented themselves as more socially withdrawn, help seeking, and in need of professional competency evoked fewer countertransference feelings of rejection and resistance from clinicians. On the other hand, patients who are perceived as domineering, cold, and vindictive evoked countertransference reactions of feeling overwhelmed, rejected, inadequate and less confident (Rossberg et al. Reference Rossberg, Karterud and Pedersen2008). Colson et al. (1986) found that different forms of psychopathology elicit different emotional reactions among different disciplines thus laying the groundwork for division among staff. Character pathology was strongly associated with anger whereas withdrawn psychoticism was strongly associated with helplessness. Violence-agitation was associated with helplessness for psychiatrists, fearfulness for social workers and nurses, and anger for activity therapists. Rumgay & Munro (2001) described various types of countertransference reactions and the clinical impact of these responses (Table 1).
Table 1. Manifestations of the clinician’s countertransference (Rumgay & Munro, 2001)
Countertransference in case histories
Feelings of inadequacy, helplessness, and anger are the predominant countertransference reactions observed in both of our case presentations. In the first case, mania with a delirious presentation is a decompensation not often seen today in psychiatric patients. It has been speculated that delirious mania is under appreciated because symptoms of mania are thought to have a more gradual onset associated with pharmacological remission while symptoms of delirium are conventionally seen as an organic brain syndrome (Bond, Reference Bond1980). Akathisia would be another diagnostic consideration in the first case especially in the context of increasing doses of antipsychotic medication. Distinguishing manic delirium from akathisia can be difficult, as outlined in Table 2 (Bond, Reference Bond1980; Sachdev & Kruk, Reference Sachdev and Kruk1994). The patient in case one had more symptoms of manic delirium and began to improve with the addition of valproic acid and no change in the dose of olanzapine. The intrusive and bizarre behaviours associated with this unusual presentation of mania may have induced countertransference reactions in the clinician of feeling overwhelmed and less confident with the diagnosis and management. Feelings of inadequacy in the first case (possibly related to fears of looking incompetent in front of staff and peers) led to the misdiagnosis of the patient. The clinician based much of the diagnosis on the idea that the patient had been highly functional and stable on antipsychotic medications for decades. Once the patient failed pharmacotherapeutic interventions, the diagnosis was then assumed to be organic. The diagnosis of delirium shifted the focus from a psychiatric to medical illness, which allowed the rationalization of treatment failure and thus dampened the countertransference feelings of inadequacy. This provided justification to transfer the patient to another clinician in another hospital.
Table 2. Differences in the characteristics of delirious mania and akathisia (Bond, Reference Brown1980; Sachdev & Kruk, Reference Sachdev and Kruk1994)
The theme of inadequacy also surfaced in the second case whereby the nursing staff’s frustration with treatment response and the hospital maintenance staff’s repeated repair visits doubled the pressure on the psychiatrist to find a ‘quick fix’. Both staff and the clinician experienced increasing anger towards the patient for not complying with behavioural interventions. With countertransference possibly unchecked, the psychiatrist considered escalating the patient’s observational status from 15 minute checks to walking restraints with a one to one staff member. The physician also considered the idea of diagnosing a personality disorder to explain the behaviour along with possibly filing criminal charges for destruction of property.
Unchecked countertransference reactions can also adversely affect patient care through the misuse of pharmacotherapy. The increased use of medication can be an indication of an unrecognized countertransference reaction particularly in patients who were initially treated effectively with modest doses (DiGiacomo & Cornfield, Reference DiGiacomo and Cornfield1979). In the first case, the clinician insisted on increasing doses of neuroleptics to treat the patient’s unusual presentation rather than dealing with possible countertransference reactions, which prevented the switch to a mood stabilizer. In the second case, the inability to control the patient’s violent behaviour and the failure of his delusions to respond quickly to medications prompted the treating psychiatrist to conduct inadequate short trials of two antipsychotics. Furthermore, countertransference reactions of helplessness may have led to the interpretation of the patient’s behaviour as defiance instead of severe mental illness. As a result, the patient was considered to have personality disorder pathology rather than a psychotic pathology.
It is important to emphasize that in both cases the patients were violent and they destroyed property. These actions may have provoked strong countertransference reactions of anger in both clinicians. Feelings of fear and anger in clinicians treating violent patients may impede effective management (Lion & Pasternak, Reference Lion and Pasternak1973). In the first case, the psychiatrist was choked, and a nurse was kicked. The second patient not only destroyed property but also represented a potential physical threat to staff. Colson et al. (1986) described patients who were demanding, manipulative, hostile, labile and sabotaging treatment as evoking a reaction of anger in the caregiver. Potentially dangerous patients further elicit feelings of helplessness (Lion & Pasternak, Reference Lion and Pasternak1973). In their paper on countertransference with psychotic patients, Semrad et al. (1952, p. 384) stated, ‘One of the crucial countertransference problems in psychotherapy with psychotics is to determine what should be the psychotherapist’s correct attitude toward the hostility of a psychotic patient against the therapist whether it is elicited by realistic or irrational reasons.’ In both cases, clinician countertransference reactions of anger and fear may have delayed the appropriate patient diagnosis, treatment and disposition.
Managing countertransference
The failure to acknowledge countertransference reactions can have an adverse effect on patient care and outcome. Recognizing feelings of inadequacy, helplessness, anxiety, and anger when interacting with violent and psychotic patients may result in a greater capacity to understand patients and develop a therapeutic relationship. Without identifying these feelings, ‘negative adaptations by staff members on inpatient units can vary from confusion to burn out and apathy.’ (Lakovics, Reference Lakovics1985 p. 137). When a clinician is able to identify countertransference responses in himself, a greater understanding of the relationship between patient’s symptoms, behaviours, and staff responses can occur (Lakovics, Reference Lakovics1985). Countertransference reactions may actually be reflections of patients’ personalities that can help clinicians understand the issues that a patient may face (Colson et al. Reference Colson, Allen and Coyne1986). When psychodynamic psychotherapeutic techniques are not part of inpatient treatment, the result may be unchecked countertransference among clinicians and staff.
Inpatient psychiatry should always include educating staff members to identify and understand the broad array of personal reactions that are a part of their work and how to use this information as an aid to understanding patients and planning their treatment (Colson et al. Reference Colson, Allen and Coyne1986). The sine qua non of treatment is to create a climate where staff members feel comfortable to risk self-disclosure; not as a forum for personal psychotherapy, but for understanding one’s personal reactions and thus helping the patient (Stamm, Reference Stamm1985). Stamm (1985) further recommended that staff members routinely review their countertransference in regular staff meetings. The more ‘emotionally toxic a patient is,’ the more important it is that feelings are neutralized through open and candid staff discussions (Stamm, Reference Stamm1985 p.439). Staff should also be aware that their feelings may be deeply rooted in the patient’s actions. Main (1957) described how he met weekly with a group of nurses to discuss the treatment of difficult patients. This was neither group therapy nor a staff meeting, and nurses were able to identify certain features of the difficult patients that evoked strong feelings in them. Brown (1980) described three different strategies to deal with countertransference. The first strategy is a daily multidisciplinary team meeting with all staff involved in patient care. Countertransference is explored insofar as it sheds light on current diagnostic evaluation and a possible impediment to treatment. A second setting in which countertransference reactions are examined is during a weekly meeting between the patient’s therapist and the primary member of the nursing staff assigned to the patient. Finally, countertransference reactions of unit staff are managed in weekly ‘group supervision’ meetings. The supervision is less structured, and its task is to focus on management difficulties. Emphasis is placed on review by staff of their interactions around specific clinical problems.
Recommendations from the UK Department of Health’s Adult Acute Inpatient Care Provision (Department of Health, 2002, p. 16) include that ‘on the ward, there are two teams working to two consultants and every three months, each team will have a workshop. The second team will cover for the first team to allow them to meet, brainstorm challenges and map out solutions. The approach allows staff to engage in reflective practice with minimal disruption for patients.’ This provides staff with the opportunity to discuss countertransference issues. Maltsberger & Buie (1974) concluded that the best protection from antitherapeutic acting out is the ability to keep such impulses in consciousness. Full protection requires that the therapist gain comfort with his countertransference hate through the process of acknowledging it, bearing it and putting it into perspective. The principles that apply to the therapist also apply to the entire staff. Countertransference is an issue that transcends the therapist-patient relationship and involves the unit staff-patient relationship as well.
Gabbard (2005) described a strategy that addressed countertransference reactions on an inpatient unit based on object relations theory. In summarizing his discussion, Gabbard (2005, pp 162–165) noted how: ‘an inpatient unit can provide a new and different form of interpersonal relatedness that facilitates the internalization of less pathological object relations. In the optimal milieu, staff members relate to patients as to avoid being provoked into responding, as would their internal object relations. Thus, the patient is confronted by a group of persons who respond differently from everyone else in the patient’s world. The success of this model depends on the treatment staff’s openness to the powerful feelings engendered by these patients. The persons in leadership positions on the unit must foster a noncritical, accepting attitude toward the various emotional reactions of staff members to patients. Despite increasing pressure in recent years to use staff meetings for documentation, treatment, and discharge planning, the staff members’ emotional reactions to patients could be discussed openly and with understanding. The attitude of the team leader is crucial in setting the tone in such discussions. The leader must model for other staff members by openly examining his or her own feelings and relating them to the internal object relations of the patient. The leader must also value and accept the expressions of feelings of other staff members and not interpret them as a manifestation of unresolved and unanalyzed conflicts. Members of the treatment team should be given the expectations that they will experience powerful feelings toward the patient that can be used as a diagnostic and therapeutic tool. A distinction must be made between having feelings and acting on them. Staff should be advised to note and discuss with other staff but to not act on feelings of a destructive or erotic nature.’
Lastly, the significance of a second opinion by another clinician cannot be overstated. Not only is a second opinion valuable in expanding or questioning diagnostic and treatment options, but it can facilitate the recognition of countertransference reactions affecting patient care. In the cases described above, obtaining a second opinion led to the successful treatment of the persisting psychopathology and prevented inappropriate discharge.
A limitation of this paper is the fact that the clinician’s and staff’s countertransference reactions are generally inferred from the staff and treating psychiatrist’s responses to the patients’ behaviour. Neither the staff nor the treating psychiatrist was actually interviewed to elicit their feelings about the patients, though staff responses to the clinical situation are consistent with the literature discussed above. These cases were selected because they were presented at a Morbidity and Mortality Conference after being identified by the hospital’s Chief Medical Officer (CMO) as problematic cases in need of treatment suggestions. In both instances, the CMO was asked to review the cases by the attending psychiatrist for treatment suggestions. Future studies would ideally attempt to rate the clinical symptoms of patients and interview all staff about their feelings toward the patient and how it might be affecting care. Papers cited earlier are models for such an approach (Colson et al. Reference Colson, Allen and Coyne1986; Rossberg et al. Reference Rossberg, Karterud and Pedersen2008).
