Unfortunately, the general quality of communication between health care providers and patients with life-limiting disease is suboptimal. Studies show that the discussion of bad news frequently does not meet patient needs or falls short of expert recommendations (9, 10, 11). Both physicians and nurses tend to underestimate and not elicit cancer patients’ concerns (12), and commonly do not attend to patients’ affect or even recognize their emotional cues (13, 14). Rather than using facilitative communication techniques, such as open-ended questions or empathic responses when inquiring about psychosocial issues, they often block discussion of these issues by changing the subject or not attending to patients’ emotional states (15). Even in a hospice setting, one study revealed that only 40% of patient concerns were elicited (16). As a result, patients with cancer tend to disclose fewer than 50% of their concerns (15, 16), which leads further to physicians’ inaccurate assessments of patient distress (17). Two large studies of audio-recorded oncology visits with terminally ill patients found that physicians only dedicated a small percentage of their time to health-related, quality of life issues, including psychosocial concerns, frequently missed opportunities to address issues that seemed to be most important to patients, and did not often check for patient’s understanding (18, 19). Finally, physicians rarely talk with seriously ill patients about their goals, values, or even treatment decisions (20, 21, 22). A significant gap exists between the idealized model of provider–patient communication at the end of life and the reality of practice.

Why is the “state-of-the-art” so poor? First, health care providers are not selected for their communication skills. Expertise in cognitive areas is not always positively correlated with empathy or an interest in understanding another person’s experience. Medical education emphasizes cognitive teaching techniques and cognitive material rather than the psychoemotional–spiritual aspects of care. Second, until recently there has been little training regarding communication skills in general, not to mention communication about these difficult topics. For example, in a survey of over 3200 oncologists, few had any formal training in end-of-life care or communication skills. Oncology programs are not alone in devoting little attention to this subject. At both the medical school and residency level, inadequate attention is given to care of the dying (23). Among graduating students at two medical schools, only 48% said they had adequate role models for how to discuss end-of-life issues. At another school, 41% of medical students on a medicine rotation and 73% on a surgery rotation had never observed a staff physician talk with a dying patient. Happily, educational interventions have improved this situation; over 60% of medical students now report “adequate” training on end-of-life care. Unfortunately,
these interventions continue to focus on cognitive aspects of care. Students report that death is still viewed as a loss not to be discussed and that they are discouraged from showing their emotions (24). Finally, physicians have difficulty inquiring directly about the emotional status of dying patients because of their feelings about the patient or their own mortality. We will particularly focus on this issue.

Considerable evidence suggests that physicians’ personal feelings toward their patients are important to the doctor–patient relationship (25), and many have suggested that physicians’ emotional responses to their dying patients may interfere with their care (26). Physicians dealing with dying patients are not objective observers. They are active participants whose beliefs and feelings influence the interaction. For example, a study of surrogate decision making found that physicians’ predictions of their patients’ wishes regarding life-sustaining treatment were closer to their own choices than to the choices expressed by their patients (27).

Caring for the dying may elicit significant stress in physicians and a variety of reactions, including guilt (“If only I’d convinced him to get that screening colonoscopy.”), impotence (“There’s nothing I can do for her.”), failure (“I messed up. I’m a bad doctor.”), loss (“I’m really going to miss this person.”), resentment (“This patient is going to keep me in the hospital all night.”), and fear (“I know they’re going to sue me.”) (28). According to Spikes and Holland, many physicians have unconscious feelings of omnipotence and troublesome responses stem from a physician’s need to preserve his or her image as a “powerful healer” who can master any situation (26). Feeling that he or she has failed the dying patient, a physician may respond by acting defensively, wishing that the patient would die (to avoid dealing with the patient), or by treating too aggressively (to ensure that “everything has been done to save the patient”).

Empathizing with a dying patient often evokes anxiety about physician’s own mortality. Physicians respond by withdrawing from terminally ill patients, avoiding threatening topics (15, 29), employing blocking behaviors that distance them from addressing affective concerns of patients (30), or falsely reassuring them that “everything is OK (26).”

Empirical data support these claims about physicians’ anxieties regarding death. Physicians score higher on death anxiety scales than other professional groups (31). They also find caring for terminally ill patients stressful. For example, in a survey of 598 oncologists, 56% reported being burned out and 53% attributed these feelings to continuous exposure to fatal illness (32). When caring for dying patients, physicians often report sadness, helplessness, failure, disappointment, and loneliness (33). These feelings, particularly if unrecognized, may affect patient care. Residents who are more burned out endorse more negative attitudes and behaviors related to patient care (34). Conversely, residents who report better personal well-being score higher on empathy scales (35). A study of 25 pediatric residents explored the relationship between their orientation toward death and their response to a clinical vignette. Residents with a high death threat and anxiety scores were more likely to adopt avoidance and denial strategies for dealing with the vignette (36).

Although these issues are profound, awareness of their own emotional responses to caring for dying patients can help physicians begin to focus more objectively on the effect of their behavior on the patient.

Talking to dying patients is just like, and completely unlike, all other communication with patients. Whether one is explaining the implications of hypertension, or talking about impending death, basic principles of good communication are useful. The primary difference between these communication tasks is the meaning of the conversation to the patient and the provider, and the attendant level of emotional significance. When the situation is more likely to make the patient (or physician) feel vulnerable, sad, or inadequate, one should focus extra attention on the task. In this section, we will address basic communication skills that are universal to all encounters.

A little effort spent on advance preparation can have a tremendous impact on the quality of the encounter. Whenever possible, important medical information, particularly bad or sad news, should be delivered during a scheduled meeting. This allows patients to prepare themselves for the type of information they will hear and to make sure that appropriate family members or friends are present. It also allows the physician to allocate the necessary time to the encounter and to come prepared with basic medical information and anticipating the most likely questions regarding treatment options, prognosis, and resources for support and guidance.

Communication best occurs face-to-face. Telephones accentuate physical communication difficulties and there is no opportunity to employ the benefits of nonverbal communication. Given that over 50% of communication is nonverbal, both parties operate at a disadvantage if they cannot see each other. The physician should sit at eye level and within reach of the patient. If possible, one’s pager or cellular phone should be turned off, or at least put on a quiet mode, and one should avoid interruptions.

Increasingly, we encounter non-English-speaking patients. One must absolutely employ the assistance of an interpreter in such settings. However, it is equally important to avoid using family members as interpreters. Not only does this run the risk of faulty translation or reinterpretation of the physician’s statements, it also places family members into the uncomfortable position of being the physician’s and patient’s spokesperson. The common practice of using bilingual young children as translators is particularly problematic. Most hospitals and health care facilities in regions with high numbers of immigrants employ professional translators or maintain lists of language skills among facility staff members.

Regarding the dialogue itself, considerable data exist from the medical and psychological literature to support certain general techniques that allow more accurate assessment of anxiety and depression and increased disclosure of concerns (13, 37). One should maintain good eye contact, ask open-ended rather than closed-ended questions, focus on the patient’s concerns as well as the agenda for the visit, respond to the patient’s affect, ask about the patient’s life outside of medicine, attend to psychosocial issues, and ensure that nonverbal behavior signifies attentiveness (Table 48.1) In contrast, disclosure of concerns is inhibited by closed-ended or leading questions,
focusing on physical aspects of illness, and offering of advice and premature reassurance (13).

One core precept is not to assume that one knows what is on the patient’s agenda. For example, while many patients will want to discuss end-of-life issues, approximately 25% will not. This may have to do with cultural particularities or with how individuals cope with illness. Physicians are not very good at predicting which patients want more and which patients want less information. Instead of assuming one should ask. For example, on a first visit one could say, “I want to touch base with you about how you want me to handle information we get about your illness. Some patients want to know everything that is going on with their illness, the good and the bad. Other people do not want as much information and want me to speak more generally. And some would really prefer I do not discuss bad news with them but want me to discuss these issues with their family. Which kind of person do you think you are?”

Another important precept of communication is to “ask before telling.” Patients often carry misperceptions or incomplete information obtained from the popular media, folklore, or friends and family. It is easier to deal with this information if it is discussed directly. Thus, it is usually helpful to ask patients about their understanding of their illness before educating them. Furthermore, one study of intensive care unit (ICU) family conferences observed that allowing families more opportunity to speak may improve family satisfaction (38).

According to dying patients, family members and health care providers, goals for communication at the end of life include talking with patients in an honest and straightforward way, being willing to talk about dying, giving bad news in a sensitive way, listening to patients, encouraging questions from patients, and being responsive to patients’ readiness to talk about death (39). Patients want physicians to achieve a balance between being honest and straightforward and not being discouraging. For some this requires leaving open the possibility that unexpected “miracles” might happen, discussing outcomes other than a cure that can offer patients hope and meaning, and helping patients prepare for the losses they may experience. Patients cope better when physicians emphasize what can be done, explore realistic goals and discuss day-to-day living (40). Although patients must receive adequate information to make informed choices, they wish to receive that information in an emotionally supportive way (41). Patients want to discuss emotional concerns but are frequently unwilling to bring them up spontaneously and may need to be prompted (42).

This may sound impossible. How can one be honest and be hopeful? How can one ensure informed consent and let patients decide they do not want to hear all the information? While many models have been proposed (43, 44, 45, 46, 47) they have in common several principles:

Most difficulties in communication at the end of life are the result of inattention to affect. Affect refers to the feelings and emotions associated with the content of the conversation. Feelings such as anger, guilt, frustration, sadness, and fear modify our ability to hear, to communicate, and to make decisions. For example, after hearing bad news, most patients are so overwhelmed emotionally that they are unable to comprehend very much about the details of the illness or a treatment plan. Some studies have shown that emotion affects processing; people who are in negative moods may pay more attention to how messages are given than to the content of the messages (50). Thus when patients are experiencing high levels of negative affect and caregivers do not ameliorate this affect, patients may be less likely to receive the health care providers’ messages. Unfortunately, conversations between doctors and patients often transpire only in the cognitive realm; emotion is frequently not acknowledged or handled directly and physicians miss opportunities to do so (51).