Physicians who care for terminally ill patients confront a range of complex medical and psychosocial challenges, and treating patients who are experiencing psychosocial distress is often a particularly troublesome clinical task. Although it is difficult to imagine any patient facing the end of life without emotional distress, physicians may not immediately be able to differentiate between "normal," appropriate, inevitable distress and more severe disturbances. In this paper, I use three cases to illustrate assessment and management of normal distress and grieving, clinical depression, and the wish to hasten death in the presence of psychological distress.
Although psychological distress is well documented in dying patients (5), it tends to be underrecognized and undertreated (6). Numerous factors act as barriers to recognition and treatment of psychological symptoms. First, both patients and clinicians frequently believe that psychological distress is a normal feature of the dying process and fail to differentiate natural, existential distress from clinical depression. Second, physicians may lack the clinical knowledge and skills to identify such problems as depression, anxiety, and delirium; this may be especially true for the challenging clinical context of terminal illness, in which many of the usual diagnostic clues are confounded by coexisting medical illness and appropriate sadness (7). Third, many patients and physicians are reluctant to consider psychiatric causes of distress because of the stigma associated with such diagnoses. Fourth, patients and clinicians often avoid exploration of psychological issues because of time constraints and concerns that such exploration will cause further distress (8, 9). Fifth, physicians are sometimes reluctant to prescribe psychotropic agents, which can cause additional adverse effects, and therefore may hesitate to diagnose a condition that they do not feel they can treat successfully. Finally, when caring for dying patients, physicians may feel a sense of hopelessness that can lead to therapeutic nihilism (10). These factors are reflected in the fact that antidepressants account for only 1% to 5% of all psychotropic agents prescribed for patients with cancer (11, 12).
Psychological distress is a major cause of suffering among terminally ill patients and is highly correlated with poor quality of life (13). More than 60% of patients with cancer report experiencing distress. Differentiating the distress associated with normal grieving from that associated with psychiatric disorders requires an appreciation of the clinical characteristics and prevalence of these entities.
Derogatis and colleagues (14) found that 47% of patients with varying stages of cancer fulfilled diagnostic criteria for psychiatric disorders. Of this 47%, 68% had adjustment disorders with depressed or anxious mood, 13% had major depression, and 8% had organic mental disorders (for example, delirium). Research has shown that patients with other terminal illnesses also have a greater incidence of psychiatric disorders than healthy persons (15).
Mr. Roberts, a 53-year-old man with end-stage pulmonary disease, is cared for at home by his wife and the local hospice program. He receives long-term oxygen therapy, is bedridden, and has been hospitalized twice in the past year for respiratory failure that required ventilatory support. Mr. Roberts is concerned about becoming a burden to his wife and children. The family’s income is barely enough to meet their needs. Recently, the hospice nurse has expressed concern about Mr. Roberts’s mental state because he has been asking repeatedly why he has to wait around to die. When directly questioned, he states that he has no intention of ending his life but that he is distressed by his helplessness and dependence. He says he feels like "a time bomb ticking." He spends his time watching television and trying to complete two woodworking projects, one for each of his two sons.
The physician who hears this report must assess the severity of and the possible interventions for Mr. Roberts’s distress. Is he depressed, or is he experiencing normal grieving that is part of the dying process? What is the appropriate threshold for diagnosing depression? In addition, the physician must confront the challenge of bearing with the patient’s distress and remaining present as a witness and an ally while the patient traverses this difficult passage. The clinical features of grief and depression are contrasted in Table 1T1.
Our knowledge of psychological disorders in terminally ill patients is derived predominantly from patients with AIDS and cancer and from geriatric patients. Relatively little published literature is available about psychological issues affecting patients with end-stage pulmonary, cardiac, renal, and neurologic disease. Therefore, the recommendations in this paper represent extrapolations from existing literature and expert opinion but lack specific evidence of efficacy in some patient populations.
The physician makes a house call to further assess Mr. Roberts’s condition. Throughout the visit, Mr. Roberts makes cheerful jokes about his condition—"Hey, Doc, not dead yet!"—and repeatedly refers to his death in a joking manner. Through questioning, the physician learns that Mr. Roberts is not sleeping well because he is short of breath and anxious about not waking up, that his appetite is poor, and that he has little energy. He reports that he does not want to see anyone except his family and that he lacks the concentration and focus to read. When asked whether he is depressed, Mr. Roberts replies, "Depressed? That word holds no meaning for me. Angry, yes. Fed up, yes. Worried about my family, yes. But depression? Never." He remarks on how much he enjoys his woodworking projects and how he worries that he will not have time to complete them. He speaks about his pleasure in visiting with his sons. Then he says, "This dying thing can’t be over quick enough or last long enough for me." Mr. Roberts reports that he is realistic about his prognosis, hopes for a few more good months, is trying to do as much as possible for himself, and is not suicidal. He says that joking has always been his way of coping with difficult situations.
Mr. Roberts’s case presents many of the common challenges in diagnosing depression (32). He has several of the neurovegetative symptoms of depression (difficulty sleeping, poor appetite, loss of energy, and diminished concentration). However, these symptoms may be caused or exacerbated by underlying disease. Mr. Roberts is also grieving as he anticipates his death. His withdrawal from persons other than his family is probably part of the normal grieving process, particularly because he continues to enjoy his visits with his children. Like other terminally ill patients, he expresses ambivalence about the prospect of death, simultaneously accepting and denying it (33).
The clinical interview is the gold standard for the diagnosis of depression (34, 35). Chochinov and coworkers (36) found that the single question "Are you depressed?" provides a sensitive and specific assessment of depression in terminally ill patients. A patient who responds affirmatively to such an inquiry is likely to receive a diagnosis of depression after a comprehensive diagnostic interview. The busy physician can use this question as a screening tool; for example, Mr. Roberts’s negative response is important evidence against a diagnosis of depression. Table 2T2 summarizes the indicators of depression that are most useful in the diagnosis of patients with terminal illness. The criteria in Table 2T2 differ from the traditional criteria outlined in Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV) (41), because the usual diagnostic features of depression do not specifically apply to patients with terminal illness.
When assessing these psychological symptoms, physicians must put the patient’s responses in context. For example, the patient’s illness may be grounds for realistic hopelessness. The patient may be helpless because of his or her physical condition. His or her role in life may have changed drastically because of illness and may result in a loss of self-esteem. A patient whose illness is related to behavior (for example, smoking) may feel a sense of guilt for causing the illness. However, when these symptoms are out of proportion to the patient’s actual situation, they are useful indicators of major depression.
The physician can also use his or her own emotional responses to patients as a diagnostic clue. Patients with depression often engender feelings of boredom, hopelessness, aversion, and lack of interest in their caregivers, mirroring the dysphoria, hopelessness, and self-criticism that are hallmarks of the patient’s experience of depression (42). Mr. Roberts’s physician notes that he enjoys his patient’s mordant sense of humor and is amused by his delight in shocking the hospice nurse with his jokes (further evidence that Mr. Roberts is not depressed).
Mr. Roberts’s distress is focused on real issues related to his illness—the burden of care on his family, uncertainty, and distress about loss of control. He retains the capacity to laugh, to be involved in his hobbies, and to enjoy his family. Therefore, the physician concludes that Mr. Roberts is not depressed and that his distress seems to fall within the rubric of what the DSM-IV calls "adjustment disorders" (41). The physician suggests that the hospice nurse continue to monitor Mr. Roberts’s mood and recommends a trial of an antidepressant only if he demonstrates new depressive symptoms.
Because no "bright line" separates depression from grief or adjustment reactions, the physician must assess whether the patient’s symptoms have reached the threshold for treatment. Psychological distress requires treatment even when it does not constitute a psychiatric diagnosis. Many treatments—listening, exploring concerns, reinforcing the patient’s coping strengths, and facilitating dialogue with family members—are part of the clinician’s responsibility. These interventions, which involve the physician as a healing agent (43), can markedly improve the patient’s adaptation. In general, because treatments for depression have become easier to use and tend to have fewer side effects than older medications, a strong case can also be made for a therapeutic trial of antidepressant medication when a diagnosis is in question. Treatment with psychostimulants (see the following discussion) can provide a relatively quick test of whether antidepressants are likely to be effective.
Although Mr. Roberts frequently expresses worry about the effect of his illness on his family, he continues to use black humor to cope with his illness and approaching death. He dies peacefully at home 3 months later.
Ms. Ferrone is a 78-year-old woman with metastatic breast cancer and bone and liver metastases who is receiving palliative chemotherapy. She is Roman Catholic and lives with her husband and 40-year-old retarded daughter. In the past 6 months, she has fractured two vertebrae and has been hospitalized for pulmonary embolism. She and her family have always wanted aggressive treatment, primarily so that she could continue to care for her daughter. Recently, however, she said that she is too sick to be of help to anyone and that she does not want further treatment. Her pain is poorly controlled with pamidronate, naproxen, and morphine; on a 10-point scale, her current pain level is at best a 3 and at worst a 7. She notes that pain often interferes with her sleep and that she cannot sleep past 4:00 a.m. Her appetite is poor, and she has lost interest in her hobbies.
Because of the change in her status, the physician carries out a depression assessment. When asked about her future, Ms. Ferrone responds: "My future is over. There is nothing good ahead for me. I worry about how much suffering is ahead, about my daughter, and about how my husband will manage. If it weren’t for my religion, I would call that doctor who kills people. I used to feel proud of being a good mother and wife. But I’ve lost that. All I can see is how much suffering I am putting people through. I can’t forgive myself for that." When asked whether she thinks she is depressed, Ms. Ferrone says that she is nervous and sad and that she feels that anyone would be depressed in her circumstances.
Several factors indicate that Ms. Ferrone is depressed. Physicians caring for terminally ill patients should consider the diagnosis of depression when a patient unexpectedly elects to discontinue treatment, is experiencing unrelieved pain, or demonstrates any of the neurovegetative or psychological symptoms of depression. In particular, the diagnosis of depression should be considered in geriatric patients who complain about memory problems or who demonstrate increased levels of somatic concern. A clinician can fully evaluate depressive symptoms in this clinical context by asking such questions as "How do you see your future?", "What do you imagine is ahead for you with this illness?", "What aspects of your life do you feel most proud of? Most troubled by?", and "Are you depressed?"
Ms. Ferrone’s responses to her physician’s questions suggest that she is feeling hopeless and depressed. She is unable to imagine anything positive in her future, feels unable to contribute, and believes that her presence is only a burden to others. Although her religious beliefs make suicide unlikely, she clearly has some wish to end her life. Although many patients and clinicians believe that depression is a normal feature of terminal illness, most terminally ill patients do not become depressed (44). Suicide ideation and feelings of hopelessness, helplessness, worthlessness, and guilt—all of which are present in Ms. Ferrone—are among the best indicators of depression in terminally ill patients. Anxiety usually coexists with depression, and some patients experience an anxious depression (45, 46). In addition, organic mental disorders (for example, delirium) caused by metastatic disease or paraneoplastic syndromes may mimic depression (47). A medical evaluation should be completed to assess possible organic contributors to depressed mood. Components of the appropriate medical evaluation will vary depending on the patient’s clinical situation.
The first step in assessing and treating depression is controlling pain. Uncontrolled pain is a major risk factor for depression and suicide among patients with cancer (48, 49). Sixty percent to 90% of patients with cancer experience pain during the last year of life (50—52), and more than 90% of patients with cancer pain respond to simple analgesic measures (53, 54). The Study To Understand Prognoses and Preferences for Outcomes and Risks of Treatments demonstrated that even hospitalized patients who had diseases that are not usually considered painful (for example, cardiac disease) were reported to experience moderate to severe pain (55).
The doses of Ms. Ferrone’s analgesic agents were increased, and her pain control improved. Although her mood brightened slightly, she continued to express hopelessness about the future and her other mood symptoms did not improve.
Major depression is a treatable condition, even in persons who are terminally ill. Because treatments are usually relatively benign, experts recommend that clinicians have a low threshold for initiating treatment. Trials of individual interventions demonstrate the effectiveness of psychotherapeutic interventions in relieving distress (56, 57), improving quality of life (58—60), and even prolonging life (61). The effectiveness of psychopharmacologic interventions in relieving depressive symptoms and alleviating psychological distress has been demonstrated in as many as 80% of patients (62). Although no controlled clinical trials have evaluated the efficacy of combined interventions, most experts recommend an approach that combines supportive psychotherapy, patient and family education, and antidepressants (63). However, few of the reported trials, including those involving psychopharmacologic agents, meet the most rigorous standards for evidence-based practice. In addition, most research has been done in patients with cancer and few of these interventions have been systematically evaluated in terminally ill patients.
Psychotherapy and counseling
When developing a treatment strategy, the clinician must actively question the patient to elicit concerns about death and the dying process, fears about the effect of illness on family members, and past experiences with loss. By addressing these concerns, the physician can help the patient connect with past strengths and assets and spiritual and religious resources, thereby enhancing self-esteem and coping ability. Sometimes, supportive therapy alone is enough to treat depression. Supportive therapy can be provided by a psychiatrist, psychologist, social worker, hospice nurse, or primary care physician, depending on time, interest, training, and the severity of the patient’s condition.
Terminally ill patients benefit from an approach that combines emotional support, flexibility, appreciation of the patient’s strengths, and elements of life review. This helps the patient develop a sense of closure and completion. The physician’s ability to convey the potential for connection, meaning, reconciliation, and closure in the dying process is thought to facilitate the patient’s ability to come to terms with impending death (64). However, patients with severe depressive symptoms may be too immobilized, hopeless, and dysphoric to effectively engage in psychotherapy; they may first need to receive appropriate antidepressant medication.
Psychostimulants, selective serotonin reuptake inhibitors (SSRIs), and tricyclic antidepressants are the mainstay of treatment for depressed, terminally ill patients (65). They are particularly useful for patients who are seriously ill and may be unable to engage in psychotherapy (66). Characteristics of these agents are described in Table 3T3. Although various new antidepressants have been introduced in recent years, they have not yet been evaluated for use in terminally ill patients.
Psychostimulants (dextroamphetamine, methyl-phenidate, and pemoline) deserve special consideration in treating depression near the end of life because they take effect quickly. In patients with a limited life span, these agents can reduce the distress of the patient and family and create opportunities for them to cope more effectively with the challenges of the dying process. Even patients who are extremely debilitated and fatigued may experience an improvement in mood and energy within 24 hours of starting treatment.
However, psychostimulants are not the drugs of choice for terminally ill patients who have relatively long projected life spans; they are best used in patients who have weeks or several months to live. For patients with severe depression who require urgent treatment but are expected to survive for several months or longer, it is often useful to begin treatment with a psychostimulant, add an SSRI after the patient has a therapeutic response, gradually decrease the dose of the psychostimulant, and increase the dose of the SSRI to a therapeutic level over 1 to 2 weeks. Despite psychostimulants’ efficacy and relative lack of side effects, many physicians are reluctant to prescribe them because of concerns about side effects and about liability under the Controlled Substances Act. However, these agents are well-accepted in the psychiatric literature for treatment of depressed, medically ill patients.
Selective serotonin reuptake inhibitors (fluoxetine, paroxetine, and sertraline) are often the first-line agents for treatment of depression in terminally ill patients when immediate onset of action is not essential. In general, paroxetine and sertraline are better tolerated by terminally ill patients because they have fewer active metabolites, which can accumulate and cause toxicity.
Although tricyclic antidepressants are still used in terminally ill patients, they are not as well tolerated as SSRIs because of their sedating and autonomic effects. As a result of the proliferation of new antidepressants, including many not mentioned here, the internist is best served by becoming familiar with the risks and benefits of a small subset of available antidepressants in each class, using them when clinically indicated, and referring patients who do not improve on these agents for psychiatric consultation.
Although Ms. Ferrone is reluctant to start medication because she feels that she should be "glad to go to God," she agrees to start oral methylphenidate, 2.5 mg, taken daily at 8:00 a.m. and 12:00 p.m. In 2 days, her family notes that that she has more energy, is sleeping better, and reports less pain. The methylphenidate dose is increased. Several days later, Ms. Ferrone reports that she is feeling less downhearted and that although she does not want any more aggressive treatment, she is looking forward to the holidays. In 10 days, Ms. Ferrone’s family feels that she has fully recovered from her depression. She enters a hospice program, and methylphenidate is maintained without recurrence of depressive symptoms. After her death, Ms. Ferrone’s family expresses their gratitude that "she remained herself until the very end."
In understanding and treating depression, the clinician must recognize that meanings and expression of depressive symptoms may vary across cultures (85). Some patients who strongly believe in an afterlife may struggle to view death as an opportunity to be closer to God, and others may fear hell and damnation. These beliefs influence the patient’s response to the crisis of terminal illness. Furthermore, because some cultures stigmatize psychiatric disorders, patients and their families may be reluctant to acknowledge depressive symptoms and to accept treatment. Often, the additional expertise of respected leaders from the patient’s cultural or religious background can be helpful in encouraging a patient to accept treatment.
Mr. Wyznyski is a 36-year-old man with AIDS who recently stopped receiving antiretroviral treatment because of side effects. He lives with his partner near his family. Mr. Wyznyski recently stopped working because he was too ill but has remained active in AIDS education. Ten years ago, he cared for his former partner, who was dying of AIDS. Mr. Wyznyski has been open with his partner, family, and physician about his intention to end his life if his suffering becomes unbearable. He has a severe peanut allergy and intends to consume a nut-filled candy bar if he becomes intolerably ill.
Over the past several months, Mr. Wyznyski has become blind because of cytomegalovirus-related retinitis and is wheelchair-bound because of peripheral neuropathy. His Kaposi sarcoma has proliferated, and he has lost 35 pounds. Nonetheless, he has continued to be active in his teaching activities. One day, Mr. Wyznyski comes to his regularly scheduled visit to say good-bye and to thank his physician for the care that he has received. He says that he plans to kill himself within a few days. Although Mr. Wyznyski has considered suicide as a theoretical option, he now seems to have an immediate plan.
When assessing a patient’s risk for suicide, the physician must keep in mind that rates of suicide are higher in patients with medical illness than in healthy persons and increase as illness progresses (86). Additional risk factors for suicide in terminally ill patients include advanced age, male sex, a diagnosis of cancer or AIDS, depression, hopelessness, delirium, exhaustion, pain, preexisting psychopathology, and a personal or family history of suicide (87—89).
Mr. Wyznyski, like many patients with life-threatening illnesses, has had frequent suicidal thoughts. Such thoughts occur in as many as 45% of terminally ill patients with cancer, are usually fleeting, and are associated with feelings of loss of control and anxiety about the future. However, in a small study of terminally ill patients with cancer, 8.5% expressed a sustained and pervasive wish for death to come quickly. Fifty-nine percent of these patients received a diagnosis of depression (4) and were found to have increased levels of pain and limited social support. It is not known whether treatment of depression results in a diminished desire for early death. However, in a study of depressed geriatric patients’ preferences for life-sustaining therapy, Ganzini and colleagues (90) found that more than 25% of severely depressed patients who initially refused treatment showed an increased desire for life-sustaining treatment after depressive symptoms improved. Table 4T4 outlines an approach to the assessment of suicidal thoughts. Even patients who present the desire for suicide as a "rational" choice should receive a comprehensive assessment. The approach to patients who want to hasten death has been reviewed elsewhere (91, 92).
When assessing the likelihood of depression in a terminally ill patient, the physician must both remain involved and rely on the expertise of other members of an interdisciplinary team. Several circumstances should prompt a referral to a psychiatrist (Table 5T5). Psychiatrists who are experts in the assessment and management of patients with severe medical illness can be found in most major medical centers as part of medical psychiatry departments or consultation—liaison psychiatry services. A psychiatrist can provide an in-depth assessment of the patient’s judgment, decision-making capacity, and mood. A social worker can provide critical information about the patient’s social network and coping. A chaplain can provide insight into the patient’s spiritual concerns.
Some clinicians believe that exploration of suicidal thoughts may exacerbate them; however, no evidence supports this theory. The current standard of practice suggests that patients who show pervasive hopelessness or a persistent desire to die should be referred to a psychiatrist for assessment and treatment because of their high risk for suicide (93). Hallucinations or delusions in depressed patients should be viewed as indicators of high risk for suicide (94). Organic mental disorders are also risk factors, especially among patients with AIDS (95). Although psychiatric hospitalization is rarely indicated for terminally ill patients, mobilization of social supports, attention to sources and meaning of suffering, affirmation of the person’s value, and treatment with antidepressants or antipsychotic agents can often help the patient want to carry on.
The physician explores Mr. Wyznyski’s decision to end his life now and asks the social worker and chaplain for their input. Although Mr. Wyznyski is initially angry that the physician questions his intention to kill himself, his anger dissipates and he begins to cry when his physician says, "We’ve been through this whole rotten disease together. I am not going to abandon you now. We both know that your time is short, but I want to help you have the best possible death you can have." Mr. Wyznyski says that he is tired of fighting and trying to be a role model. He is afraid of letting people down by giving up. He says, "I never thought I would say this, but death has become my friend. I don’t have the energy to get up and get dressed and try to put on a good face."
With Mr. Wyznyski’s agreement, the physician and social worker arrange a family meeting where these concerns are shared. Mr. Wyznyski’s family and partner explain that they have recognized how exhausted he has become but have been afraid to encourage him to slow down because they do not want to demoralize him. After the meeting, Mr. Wyznyski says that he feels relieved that he does not have to work so hard to keep up appearances. He gives up his teaching engagements, spending his days in bed, and dies 1 month later.
In this paper, I review three cases that illustrate the assessment and management of normal or appropriate grieving, the diagnosis and treatment of depression, and the assessment and management of suicidal ideation in terminally ill patients. Skillful management of depression relieves suffering and is a core element of the provision of comprehensive end-of-life care. Although treatment of pain and other symptoms at the end of life has improved, depression and other psychological symptoms and disorders remain troublesome for terminally ill patients. Many of these conditions can be easily controlled with state-of-the-art psychosocial treatments. Physicians who care for dying patients must be competent in this critical area of clinical practice.