ETHICS AND THE CLINICAL INTERACTION

Lorenz (2003) has argued that “[p]hysicians play an inescapable role in presenting and shaping health care decisions” (p. 2282). In shaping this role, the doctor undertakes to negotiate the core ethical principles of nonmaleficence, beneficence, autonomy, and justice. At the same time the doctor attempts to carefully address the suffering of his or her patient and those close to the patient, and to compassionately alleviate the physical and emotional distress of the patient.

The fundamental ethical principle of individual autonomy has been given increasing and appropriate importance in medical decision making. It has also become an important expectation of the broader community in considering important health care decisions. The primary status given to autonomy as the dominant ethical principle is particularly evident in the approach to health care decisions in Western cultures (Glick, 1997). Yet the care of the dying patient can confront the limitations of reliance on autonomy alone as the sole determining ethical principle under which doctors can negotiate decisions (Mann, 1998). The autonomy of a seriously ill patient is inherently compromised by the dependency illness creates, and the inevitable disabilities in many areas of life functioning. Although treatment efforts to maintain or maximize autonomy are important goals, there are also the major psychological tasks for the patient in adapting to the irreversible losses of previous autonomous roles and functions as illness progresses (Mann, 1998). Health professionals play a critical role in assisting the patient in achieving this adaptation (Hendin, 1994).

The importance of the emotional dimensions of the doctor–patient interaction is particularly evident when one considers that the communication between the doctor and patient is the primary tool for negotiating and exploring these ethical matters. This interaction provides the setting in which such discussions are shaped and examined—the climate in which the communication is conducted. The ethical dimension of a patient's choices and wishes at end-of-life care therefore cannot ignore the ethical aspects of the doctor–patient relationship.

COUNTERTRANSFERENCE AND TRANSFERENCE

At one level, the interactions between a patient and his or her doctor may be primarily viewed as being guided and influenced by a set of rational and readily observable needs and responses. The patient needs skilled and competent health care and guidance in a trusting relationship with a suitably qualified clinician. The clinician wants to provide this care with skill, care, and empathy as both a responsibility of the clinician's role but also as an understandable and compassionate response to the patient's predicament.

However, alongside these readily observable elements, the interaction and relationship between the doctor and patient is also governed by “irrational” forces that influence all relationships, but become especially important in understanding how doctors and patients behave toward each other at times of greatest stress and tension.

Countertransference and transference are helpful concepts for understanding these deeper aspects of doctor–patient interaction (Goldberg, 2000). Countertransference is a concept deriving from the early history of psychoanalysis, to describe the origin of some emotional reactions experienced by clinicians in the care of their patients (Sandler et al., 1992). The conventional use of the term countertransference describes the unconsciously determined responses of a clinician to the specific characteristics and behaviors of the patient based upon the doctor's previous patterns of significant relationships in his or her life. The term originated to describe factors mostly of an unconscious nature arising from the clinician's previous relationships. Unconscious responses are those that the clinician may be unaware of in themselves (even if apparent to their colleagues and others), or if aware of a response (e.g., despair about a dying patient), unaware of the factors within themselves that cause them to experience such feelings. The term has at times come to encompass the full spectrum of emotional responses (those of which the clinician is both consciously aware and those that are chiefly unconscious in nature) that arise toward the patient in the course of treatment.

Transference refers to the patient's reexperiencing of emotions and enacting behaviors in a current relationship based upon the patterns established in past significant relationships. These are also unconscious in nature and may be inappropriate to the relationship in which the emotions are currently experienced. Countertransference is the similar process but as experienced by the clinician toward the patient. The patient may, by virtue of their individual characteristics or their illness, have the capacity to trigger a set of responses within an individual clinician, which in turn is governed by the clinician's particular characteristics. The combination and interaction of external factors (e.g., the patient and illness characteristics) and internal factors (e.g., those that reflect the individual characteristics that the doctor brings to the clinical interaction) can have a bearing on the emotional reactions of doctors in the care of their patients (Marshall & Smith, 1995). Thus it is useful to consider the transference and countertransference as created by both parties in the therapeutic relationship.

Whether doctor or patient, an individual brings to the clinical relationship a set of expectations, established patterns of interactions, and wishes, many of which may not be in the individual's conscious awareness. For the patient this may include patterns of relating to:

• caregivers
• those in positions of perceived authority or power
• situations that demand greater trust and dependency
• situations that demand the relinquishing of control to others
• threats and fears provoked by serious illness.

For the doctor these patterns may include responses to:

• the role of the clinician in terms of being in a position of trust and power
• the intensity of the patient's needs
• the nature of the patient's illness—for instance, a disfiguring condition, a condition that fails to respond to treatment, or an iatrogenic illness
• illness that provokes strong feelings—for example, fear of contagion or disgust or guilt in the clinician
• individual attributes of the patients—their age, gender, or culture.

The capacity to recognize the potential origin of these feelings is an important skill for a clinician, as is the ability to consider the contribution of factors within the patient to such reactions in the clinician. Understanding and monitoring such countertransference responses can then become an important tool to gain a better understanding of the patient (Meier et al., 2001). Countertransference responses can influence a wide and important range of clinical issues ranging from the approach taken to the making of the diagnosis, the extent of investigations, and the treatment choices offered, as well as estimates of prognoses and communication of these matters, and potentially to the way ethical issues are identified and framed.

Countertransference enactment is a term used to describe actions that arise from these emotional responses of clinicians and when these emotions motivate clinicians' behavior toward a patient (Gabbard, 1995). Such enactments have potentially detrimental effects for the patient, in particular, but also for the clinician, because they distract from responding appropriately to the patient's needs as they address the emotional needs of the clinician for the relief of tension and distress.

Appreciation of the phenomena of countertransference also involves understanding the potential origin of those responses from within the patient. Projective identification refers to the process through which the doctor experiences toward the patient feelings that the patient may be experiencing toward himself or herself. These are projected onto the doctor, who develops identification with the feelings, so that these become feelings that the doctor experiences toward the patient as if they originated in the doctor instead (Gabbard, 1995). For example, a patient's despair and hopelessness may be experienced as similar feelings in the doctor (feelings of futility and despair) about and toward the patient, without their true origins being apparent. The clinician may then act toward the patient in a way that reflects, and confirms, the way the patient feels about himself or herself. A despairing and suicidal terminally ill patient may then induce in the doctor a feeling that the only way to help the patient is to hasten the patient's death. Doctors can then find themselves acting in a way that confirms the patient's view of themselves, that is, that there is no hope.

BOUNDARIES IN THE CLINICAL INTERACTION

The accepted boundaries that define the doctor–patient relationship provide the necessary conditions for ethically sound clinical practice to occur (Gutheil & Gabbard, 1993). These boundaries define the fiduciary and professional nature of this relationship, and are represented in the doctors' capacity to focus on the needs of the patient rather than their own needs (Hamilton et al., 1998). The boundaries are embodied in professional codes of practice and include requirements for confidentiality and other ethical standards. The boundaries recognize the vulnerability that accompanies suffering and illness and the necessity for the patient to invest trust in the relationship with his or her doctor. With such trust comes the expectation that the doctor will understand that such trust places him or her in a position of power and responsibility with the patient, and to understand the limits of the relationship (Gutheil & Simon, 2002). These boundaries enable the development of a secure and trusting relationship within which the patient's concerns can be explored, and set clearly defined parameters that protect the patient against the clinician's enacting of their own distress. These boundaries provide a crucial orientation for clinicians amidst complex emotional forces (Miles, 1994). The environment of the doctor–patient interaction then becomes safe for the patient, providing the optimal conditions to explore feelings, wishes, and needs by setting limits on the doctor's behavior and role.

The therapeutic boundary provides protection from potential damage that may arise if clinicians acted on their emotional reactions. Damaging actions may arise from feelings of affection, sexual attraction, and even love (as sexual boundary crossing) but also from feelings of frustration, anger, hatred, or disgust that might be experienced by the clinician toward patients or their diseases.

The boundaries of practice are also necessary to guide actions arising from motivations otherwise considered “compassionate” for the patient or others who are “burdened” by the patient's illness and disability. The history of medicine and psychiatry illustrates the potential for harmful actions by doctors against their patients, even on a wider systematic or societal level, under the guise of worthy and humane pursuits (Burleigh, 1994).

DOCTOR–PATIENT INTERACTION AND THE WISH TO HASTEN DEATH

The issues outlined above can be used to understand the factors that surround the wish to hasten death, or the request for assisted suicide, as a specific clinical dilemma in end-of-life care.

Psychological and social factors are increasingly evident as the chief determinants of the wish to hasten death, and of interest in assisted suicide, among patients with serious physical illness. This has been demonstrated in a range of clinical studies. These factors include levels of depression, hopelessness, lower levels of social support (including perceived levels of family support), greater concerns regarding the burden to others, and strength of religious belief (Chochinov et al., 1998; Ganzini et al., 1998; Breitbart et al., 2000; Suarez-Almazor et al., 2002; Kelly et al., 2003b).

The importance of perceptions of burden to others and social support are indicative of the critical role of the patient's social network and relationships in his or her experience of illness. Seriously ill patients are uniquely sensitive to everyone in their milieu (Pellegrino, 2002). The doctor's relationship and communication with a patient are important determinants of patient well-being, and are given a high priority by patients, their families, and by other health professionals (National Health and Medical Research Council, 2000; Wenrich et al., 2001).

Some noteworthy research findings highlight the importance of closely investigating the role of the clinician in the patient's wish to hasten death. Physicians (oncologists) less likely to support physician-assisted suicide (PAS) or euthanasia reported having had sufficient time to talk to dying patients about end-of-life issues and having received adequate training in palliative care (Emanuel et al., 2000). Portenoy et al. (1997) reported that lesser religious belief, less competence in symptom management, and diminished empathy in clinicians were significant predictors of willingness to endorse assisted suicide. This role of diminished empathy further points to the importance of considering the emotional responses doctors bring to their interaction with patients that may act as impediments to effective care. Kissane (2002) has also considered the potential impact of demoralization among clinicians (as well as their patients) in the development of the wish to hasten death. It is cautionary that doctors with less training in palliative care are more likely to favor euthanasia (Portenoy et al., 1997; Emanuel et al., 2000) and have more limited skill and confidence in discussing the distress of a dying patient (Horton, 2001).

Doctors often describe feeling ill equipped to address the emotional needs of terminally ill patients (Kelly et al., 2003a). In a qualitative study of a group of doctors currently caring for terminally ill patients, doctors whose patients had the highest level of a wish to hasten death described their patents' illnesses as protracted and lengthy. The availability of or use of palliative care services was described less often in this group, who also reported lack of satisfaction with the care they had provided (Kelly et al., 2003a). Both doctor and patient factors require further investigation and the development of models to explore the impact of the interactions that occur between doctor and patient when a patient is dying.

THE DOCTOR'S DILEMMA: WHOSE WISH TO HASTEN DEATH?

Bearing these broad clinical issues in mind, the patients' expression of a wish to die to their doctor needs to be considered in terms of its meaning within the doctor–patient relationship (Varghese & Kelly, 1999). When a patient is dying the doctor also faces a set of important challenges—a confrontation with the limitations of medical interventions, suffering that may persist despite their best efforts, and the stark confrontation with death. The distress among doctors working with seriously ill patients can lead to burnout (Ramirez et al., 1995). The confrontation with the limitations of medical interventions to reverse disease progression may present a specific challenge to the doctors' sense of effectiveness and self-esteem. It is perhaps the confrontation with death that can be most provocative and complex, because this presents all those around the dying patient with a reminder of the inevitability and universality of death and loss. The limited training for doctors in communication and in care of the dying and doctors' shortcomings in addressing these tasks has also been recognized (SUPPORT Investigators, 1995; Wenrich et al., 2001). Despite the high priority given to good communication in end-of-life care, such communication is clearly very difficult for many doctors to provide (Wenrich et al., 2001).

Clinical decisions in end-of-life care based upon evaluation of “unbearable” suffering, “quality of life,” dignity, and a range of inferred impressions are especially vulnerable to the vicissitudes of the doctor–patient relationship, particularly the reactions of each to confronting death. Such parameters of suffering nevertheless form the language of debate in euthanasia and assisted suicide. Research has identified the inclination for doctors to underrate the “quality of life” of their patients (Uhlmann & Pearlman, 1991) and to misinterpret patients' likely treatment choices (Slevin et al., 1990). The impact of depression on the wish to hasten death is often under-estimated by clinicians (Haghbin et al., 1998), even though the current evidence suggests this is among the key determinants of such wishes. Furthermore, the capacity to accurately predict outcome and prognosis in severe physical illness is very limited, and many difficulties are faced by doctors in discussing with patients the future course of illness (Lamont & Christakis, 2003).

There are well-described attributes of doctors that have direct bearing on their responses and vulnerability when caring for a dying patient (Gabbard, 1985). These include a heightened sense of responsibility, a tendency to experience guilt, high self-criticism and perfectionism, and need for control. Although some of these may be valued among clinicians, they can create vulnerability to loss of self-esteem and to a sense of failure and guilt when the rewards of clinical encounter (a sense of effectiveness and value) are not forthcoming. The doctor's excessive sense of responsibility can create a tendency toward omnipotent expectations to achieve a degree of control of illness and even death that is not realistically possible. Clinical situations such as the care of the dying patient can represent a powerful challenge to the sense of efficacy, value, and worth of a doctor, and can confront this particular sense of responsibility. As described by Hendin (1998), patients are not alone in their fear of dependency and loss of control when ill. Western societies and cultures place a great value on self-determination, autonomy, and independence (Glick, 1997; Mann, 1998), and doctors may be particularly inclined to give salience to and reinforce such values (Varghese & Kelly, 1999). The patient's fears may resonate strongly with the doctor's fear of death and fears of dependency, loss of control, and difficulty tolerating uncertainty. In such circumstances both may find themselves seeking various forms of relief from the distress and tension such threats pose.

The doctor is cast in a specific role in the setting of requests for assisted suicide. The physician assesses the decisional capacity and the level of suffering of the patient, assesses available interventions, and evaluates likely outcomes and also judges quality of life. Of these, assessment of the “rationality” of the patient's request proves most problematic. This may be particularly so when the clinician brings an unrecognized bias (based on the frequently identified characteristics of doctors) toward the domain of autonomy and self-determination as the sole and overriding principle. Although such a concern is indeed a matter of importance, it may also provide an avenue for the clinician's own overvaluing of control and autonomy such as to override other concerns, leading to a reluctance to challenge or carefully explore the basis of the patient's request for assisted suicide. The behavior or wishes are considered to be “rational” by the clinician because they conform to the values and common concerns of the clinician—the fear of death, wish for control, and accentuation of individual autonomy.

Amidst this effect of a sense of failure on the clinician's self-esteem, there may also be less recognized anger toward the patient whose suffering continues despite the doctor's best efforts. Situations that engender a greater empathic gap (e.g., differences in age, gender, culture, language or race, socioeconomic background) may create an even greater risk that these emotions will affect clinical decisions. A doctor overwhelmed by the emotions that accompany the progression of terminal illness may come to view suicide as the solution, as the patient's needs and the needs of others become merged and confused.

The discussion of the wish to hasten death then can become a setting for enactment of countertransference feelings by the doctor. The degree to which the patient's views are explored or even challenged may be more indicative of the subtle processes occurring as the doctor faces the death of this patient than by patient factors alone (Hendin, 1994).

The doctor's inclination, or lack of inclination, to examine the full range of psychological and interpersonal factors that have a bearing on a wish to die (or a request for assisted suicide) may then be affected by countertransference feelings toward the dying patient (Varghese & Kelly, 1999). The failure of the clinician to be alert to psychiatric morbidity such as depression, the inclination to minimize the psychosocial needs of the patient, or the failure to carefully assess such needs, may represent a countertransference enactment of feelings such as hopelessness and nihilism about the terminally ill patient. Such responses to the patient's psychological needs may also be a manifestation of being fearful of being overwhelmed by the distress of the patient and hence distancing oneself from such distress. The willingness and capacity of the clinician to encourage emotional expression may be indicative of these countertransference responses to the patient's distress. This can become clouded in the language of the debate surrounding assisted suicide, especially concerning the “right” to suicide. As Muskin (1998) has stated, the failure to explore the meaning and basis of the patient's request for hastened death is the real violation of the rights of a dying patient.

Many factors influence the degree of rapport between a doctor and patient. Where there are language, cultural, ethnic, or sociodemographic differences between a doctor and patient, these may interfere with this level of rapport and the responses made to the patient's suffering. Alternatively, the potential for overidentification (i.e., decisions on the basis of what one would want for oneself) can become a form of “pseudo-empathy” (Miles, 1994), perhaps exemplified when the predominant motivator is a clinician's sense of “pity” for a patient. Pellegrino (2002) has argued that compassion requires the containing framework of reason and principles, in order to protect the vulnerable from the actions that may be motivated by “compassionate” interests. “The patient's autonomy is submerged in the observer's emotion of compassion” (Pellegrino, 2002, p. 48).

A patient who is despairing and has a wish to die may engender in the doctor, by a process of projective identification, despair about the patient and a wish that they die. The process of projective identification (as outlined above) is exemplified in the interactions between the doctor and his or her patient that may underlie the physician-assisted suicide and euthanasia. An understanding of the interaction of the doctor and the patient in this way is helpful in appreciating the factors that may influence both the patient's wish to die and the doctor's response to this (Varghese & Kelly, 1999).

With both the patient and doctor facing the helplessness of advancing illness, the motivation for assisted suicide may represent for both an attempt to gain power and control over death itself through determining its mode and timing (Battin, 1998; Hendin, 1998). Laws and ethical codes that prohibit physician-assisted suicide provide the necessary safeguards against the enactment of these processes between the doctor and patient, providing landmarks for the doctor amidst the emotional disorientation that can occur when a patient is dying (Miles, 1994). The ethical and legal arguments in favor of PAS fail to address these intrinsic issues in the relationships between doctors and patients.

Miles (1994) has argued that the “taboo” against PAS prevents actions by the doctor that are based on the doctor's fear of dying, sense of helplessness, and efforts to reinstate a sense of power over the inevitability of death, and the sense of failure that incurable illness can bring for a doctor. In other words, the “taboo” is a constructive boundary to prevent countertransference enactment. Legal sanctions that dictate the boundaries of medical practice provide the framework and optimal set of conditions within which the care of a dying patient can be conducted amidst such forces (Miles, 1994). Without this therapeutic frame the doctor and patient cannot explore the full range of the patient's emotions and distress.

ADDRESSING THE HIDDEN FORCES IN THE DOCTOR–PATIENT RELATIONSHIP

How could these issues be addressed in the debate concerning euthanasia? The recognition of the broad range of emotional responses experienced by doctors in their work with patients as a fundamental and universal aspect of their interaction is a necessary step to improve the doctor's psychological well-being and to reduce the risk of potentially damaging “enactments” of such feelings on patients (Meier et al., 2001).

There is a need for models and tools that guide clinicians in undertaking the important tasks of caring for dying patients and remaining capably involved and committed to this care and its ethical requirements. Consultation with colleagues and regular clinical supervision are practices that that can prevent severe levels of clinician distress. Such steps can also provide tools to achieve the following: (1) to increase confidence in talking with patients and family, (2) to better identify and effectively respond to common clinical problems (including depression, anxiety, and suicidal wishes), (3) to assist in considering the meaning and basis of any expressed wish to hasten death or request for assisted suicide, and particularly (4) to consider its meaning as a communication within the relationship with the patient. There is a very complex set of tasks for all clinicians in the evaluation of requests to hasten death that extends beyond assessment of the individual patient to include attention to the psychological issues in the system of relationships that surrounds the patient, including other health professionals, social and cultural factors, and spiritual issues (Cohen et al., 2000).

Attention to the often intense emotions experienced by the doctor and provision of mechanisms to potentially affect these unaddressed or unacknowledged emotional responses becomes crucial (Meier et al., 2001). Establishing guidelines for the assessment and response to a request for assisted suicide may assist clinicians (Emanuel, 1998). When legislation has set requirements for medical assessment (Ryan & Kaye, 1996), these have relied on a very limited clinical assessment and consultation processes (Foley & Hendin, 2002; Kissane, 2002). Given the difficulties that exist for effective communication, the problems for clinicians in identifying and responding to the range of severe psychological morbidity and suffering in patients and families, their difficulties in assessing decision-making capacity (Ganzini et al., 2003), and low levels of confidence in determining prognosis (Ganzini et al., 2001; Lamont & Christakis, 2003) any legislation is rendered inadequate. The realities of the clinical landscape in which such requests arise have not, and perhaps cannot, be addressed in any legislation.

Moreover, guidelines regarding the clinician's response to a request for assisted suicide (Bascom & Tolle, 2002) must place such a request in the context of a broader range of tasks in communication with the patient and family and consideration of the doctor–patient interaction and relationship, alongside the clinical treatment issues to be addressed (e.g., assessment of depression; Back et al., 2002). Emphasis on the response to a request for assisted suicide as a single and perhaps one-off task is inadequate. The way the clinical assessment (such as assessment of depression) is conducted and the flow of communication are likely to be greatly influenced by the emotional impact this complex stage of treatment has on doctors in their relationship with their patients, alongside their previous training and attitudes.

The need for improved communication between doctors and patients during end-of-life care has been widely identified. The specific needs go beyond how to give “bad news” and must include talking about dying and maintaining hope at the same time as honest discussion (Wenrich et al., 2001). Communication skills must include listening skills, including the capacity to remain alert to both conscious and unconscious factors, and listening to one's own emotional responses and reflecting on their meaning. New models and frameworks are needed to assist clinicians to effectively engage in the care of the dying, interact effectively with patients, and respond to suffering. Understanding of demoralization as it occurs in end-of-life care can assist clinicians to develop frameworks for conceptualizing and responding to the distress of patients (Kissane, 2001). Chochinov's (2002) proposed model for “Dignity-Conserving Care” provides an approach for clinicians that helps define underpinnings of therapeutic interventions and ways of exploring the experience of dignity for patients. Guides for clinicians in discussing religious and spiritual issues at the end of life (Lo et al., 2002) have also provided approaches and tools for exploring this important dimension of the patient's experience, much of which has arisen from the recognition of the difficulties faced by clinicians in understanding patients' needs and the factors that influence decision making. One approach, which is described as making a “situational diagnosis” amidst such ethical dilemmas, encourages the clinician to view the range of individual and interpersonal factors, relationships with health professionals, and quality of communication, alongside the impact of psychiatric disorder in the patient when appraising an ethical dilemma (Lederberg, 1997).