Caring for Bereaved Patients: Title and subTitle Break"All the Doctors Just Suddenly Go"

. . . our sorrow lives in us as an indestructible force, only changing its form . . . and passing from pain into sympathy—the one poor word which includes all our best insight and our best love.—George Eliot, Adam Bede¹

Mrs A, a longtime patient of Dr M, is 77 years old and has been widowed for 2 years. Her husband, a well-respected public figure, died in December 1998 after a protracted course of diabetes, hypertension, coronary artery disease, congestive heart failure, end-stage renal disease, and, ultimately, renal failure. The husband, who was not Dr M's patient, had been cared for at home by Mrs A until his final 10-week hospitalization, which involved repeated admissions to the intensive care unit. One of their sons moved into their home temporarily to help care for his father, while another son and a daughter live nearby.

During the first year of widowhood, Mrs A visited Dr M more than usual—roughly every other month. Her visits were nominally to address somatic complaints (eg, insomnia, perpetual weeping). It was clear to both Mrs A and Dr M that bereavement was the major source of these problems, so much of the time was spent addressing that explicitly. Dr M offered a sleeping pill, which she declined. Mrs A began seeing a psychiatrist and attending a bereavement support group. She was interested in obtaining additional information about grief and bereavement, including written material and Web resources.

Mrs A and Dr M were each interviewed by a Perspectives editor in December 2000.

MRS A: Immediately following my husband's death there was constant pain. I did things, but it was very difficult; he was in my mind all the time. I was running videos of his last days in my head—everything that had happened in his care, and how he reacted, and what the doctors were doing with him—it was not a very good way to die. I also felt numb . . . . It's hard to recall what happened and why I made certain decisions. I was depressed and couldn't sleep well. And I cried. I've never cried as much as I cried for the first few months after he died. I still cry when I think about it . . . . I'll never get rid of that pain. I know there is anecdotal evidence that if people had cancer, it can recur after a spouse or a partner or a child's death. I've had cancer 3 times and I didn't want that to happen again. That's one reason I really worked at trying to get myself steady.

DR M: I saw Mrs A about 3 weeks after her husband died. I reassured her that grief resolution takes time. I told her that grief was like a long tunnel, which she had entered suddenly, and that she was now in the dark, but that she would eventually emerge back out into the light. But I said that just like when you come out of a tunnel, things are different on the other side. She seemed to understand. When I saw her in November 1999, she complained of insomnia and had a lot of other somatic complaints. I noticed the date and I told her that she might reexperience intense grief again, around the time of the anniversary of her husband's death. In the following year, I saw her only 3 or 4 times. She was no longer tearful in the office. She was beginning to go out more, and to travel, and she made some new friends.

Although they may not always recognize it, physicians care for many distressed, ailing, bereaved patients. Loss through death is a common² and extremely stressful^{3 - 4} experience. Bereavement heightens a person's risk of depressive syndromes^{5 - 6} ; sleep disruption⁷ ; increased consumption of tobacco, alcohol, and tranquilizers^{8 - 9} ; suicide attempts^{10 - 11} ; and mortality.^{12 - 13} A comprehensive recent review¹⁴ concluded that "the health of bereaved people in general is at risk (compared to their non-bereaved counterparts)." The authors continue that high risk "has by now been well established. There is no longer any doubt that the costs of bereavement in terms of health can be extreme."

As exemplified by Mrs A, widowed people visit physicians more than they had when they were married—even after adjusting for age, sex, and socioeconomic and health status.¹⁵ Bereavement tends to occur most often in later life,² when health and adaptive capacities may already be compromised. Consequently, the aging of the US population² implies that physicians will devote an increasingly large percentage of time to caring for grief-stricken patients.¹⁶

Normal, or uncomplicated, grief reactions are those that, though painful, move the survivor toward an acceptance of the loss and an ability to carry on with his or her life.^{17 - 20} Indicators of normal adjustment include the capacity to feel that life still holds meaning, a sustained sense of self, self-efficacy, trust in others, and an ability to reinvest in interpersonal relationships and activities.^{17 - 20} Despite her distress over her husband's death, Mrs A's grief appears uncomplicated: she accepts her husband's death, her grief symptoms have attenuated, she is involved with her family and has made new friends, she is engaged in civic pursuits, and she works to maintain her health.

In 1944, Lindemann¹⁹ described features of "morbid grief reactions" (eg, ruminations about the deceased, hostility) that he viewed as deviations from "normal" grief and that required more aggressive intervention. Consistent with Lindemann's observations, recent research demonstrates that bereaved individuals with high levels of complicated grief symptoms have substantially greater dysfunction than those with lower levels of these symptoms.^{16 ,21 - 25} Studies find that complicated grief symptoms: (1) form a coherent cluster of symptoms distinct from bereavement-related depressive and anxiety symptom clusters (ie, the underlying phenomenology of the symptoms indicates they constitute separate syndromes)^{21 - 26} ; (2) endure several years for some bereaved subjects^{21 - 22 ,26} ; (3) predict substantial morbidity and adverse health behaviors over and above depressive symptoms (eg, cardiac events,²² high blood pressure,^{16 ,22} cancer,²² ulcerative colitis,¹⁹ suicidality,^{21 - 22} social dysfunction,^{19 ,23 ,25 - 26} anergia,^{19 ,23 ,25 - 26} changes in food, alcohol, and tobacco intake,²² and global dysfunction^{16 ,22 - 24 ,26} ); and (4) unlike depressive symptoms, are not effectively reduced by interpersonal psychotherapy and/or tricyclic antidepressants.^{27 - 28} These findings revealed a need to identify and treat complicated grief as a psychiatric disorder distinct from major depressive disorder (MDD).

Responding to this perceived need, a panel of leading experts in psychiatric reactions to loss and trauma, depression, sleep disorders, and psychiatric taxonomy met to evaluate the studies just described and, if the evidence justified it, develop diagnostic criteria for complicated grief (Table 1) (details of the consensus conference on traumatic grief, as complicated grief was referred to at that time, are provided elsewhere^{29 - 30} ). These diagnostic criteria do not constitute an official psychiatric diagnosis and do not appear in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV), but the panel deemed the evidence to be strongly supportive of complicated grief as a separate psychiatric disorder (ie, distinctive symptoms, risk factors, course, treatment response, and outcomes).^{29 - 30}

The published refinement of the complicated grief criteria set²⁹ was found to be highly sensitive and specific—that is, the diagnostic algorithm correctly classified 93% of the predetermined "cases" (true positives) and 93% of the predetermined "noncases" (true negatives) of complicated grief.²⁹ In addition, those diagnosed with complicated grief according to this algorithm have been shown to have significantly greater impairment (eg, high blood pressure, functional disability)^{16 ,23} than those not meeting the proposed criteria. Additional (unpublished) analyses of the refined complicated grief criteria set indicated that 6-month symptom duration was superior to the 2-month duration specified earlier, with respect to the reduction in the number of false positives and to enhanced predicitive validity. For this reason, we present a modified version of the diagnostic algorithm²⁹ and use the case of Mrs A to illustrate its application (Table 1).

To receive a diagnosis of complicated grief, a bereaved patient must first meet the necessary conditions outlined in criterion A (extreme levels of 3 of the 4 "separation distress" symptoms, such as yearning for the deceased). If criterion A has been met, then criterion B (extreme levels of 4 of the 8 "traumatic distress" symptoms, such as numbness, feeling that part of oneself has died, assuming symptoms of the deceased, disbelief, or bitterness) must be met. If the bereaved patient's symptoms in criteria A and B endure for 6 months or longer (criterion C) and these symptoms are linked to substantial functional impairment (criterion D), the individual satisfies the criteria for complicated grief.^{16 ,23 ,29 - 30}

Despite the fact that their grief is uncomplicated, patients like Mrs A are often unprepared for how profoundly they are affected by their loss. They frequently feel surprised by how much turmoil and pain (sometimes described as psychic trauma) bereavement brings.^{26 ,31 - 32} Bereaved patients may report feeling incredulous about the death.^{22 - 23 ,26 ,31 - 32} Immediately after her husband's death, Mrs A described a mild state of shock and confusion. She had difficulty remembering the events culminating in her husband's death, in understanding her own actions, and she felt emotionally numb. C. S. Lewis³³ poignantly described this as an "invisible blanket between the world and me." By 2 years post-loss, Mrs A's initial numbness and detachment from others appear to have subsided (she describes numbness in the past tense; she has made new friends). Bereaved patients who appear disoriented, are in a quasi-dissociative state, functioning on "automatic pilot,"^{14 ,26 ,29} reflect their extreme difficulty in emotionally and cognitively processing the loss. Remaining markedly stunned or dazed at 6 months post-loss is a telling symptom of complicated grief.²⁹

Mrs A describes the pain of grief, including the symptoms of separation distress^{17 - 18 ,26 ,29 ,33} : intrusive, intermittent yearning and thoughts about the deceased. Even highly functioning people may become transiently distraught and disabled by a preoccupation with the loss.^{18 ,26 ,33} By 6 months post-loss, however, most bereaved people begin to experience an abatement of the acute separation distress symptoms.^{22 - 23 ,34} Dr M noticed that in the year following her husband's death, Mrs A was crying less as well as socializing and traveling more, suggesting she was neither depressed nor otherwise impaired by ruminations about her husband's death (and therefore she would not meet complicated grief criterion A).

MRS A: I haven't been [to the grave site]. I can't. The thought of it is painful. I have pictures of him all over and I can't dispose of his suits. I can't do that without the help of my kids.

Mrs A's reluctance to visit her husband's grave and to part with his possessions suggests an aversion to things signifying her permanent separation from him. Although some researchers suggest that avoidant coping reflects a patient's difficulty in accepting and adapting to the loss,^{19 ,35 - 36} others claim it may be adaptive in the long term.^{37 - 38} Analyses of symptoms that define complicated grief reveal avoidance to be one of its weakest indicators; hence, avoidance was omitted from the refined criteria set for complicated grief.^{29 ,39} Irrespective of whether bereaved people avoid reminders of the death, those who are unable to accept the death and make changes in response to new situational demands would appear vulnerable to social and occupational dysfunction^{29 ,35} (complicated grief criterion D). This does not appear to be the case for Mrs A, who appears able to accept the death and to function reasonably well.

MRS A: We haven't approached the way he died. It was absolutely disgusting . . . that pushing him constantly as to whether he wanted heroic measures of care. We had a lot of irritation. It was insensitive. He had made his wishes clear—he did not want valiant measures.

Anger and protest over a significant loss are a part of grief.^{18 - 19} Bereaved patients may even feel anger toward the deceased for perceived abandonment.^{18 ,26} Hostility is often directed at the deceased patient's physician or the health care system for failing to provide what they consider to be adequate care for their loved one.^{19 ,40} Mrs A is irritated about the care her husband received and might well meet complicated grief threshold levels for bitterness, although she does not appear consumed by rage as many patients are with complicated grief^{19 ,26} (see Table 2 for ways physicians might diffuse anger directed at them).

MRS A: I feel very guilty myself. I spent over 2 months in the hospital with him. But the particular night that he died, I didn't stay up there. I'm not sure I'll ever get over that guilt. You think about the things you did wrong in the illness or in the marriage, and there's nothing you can do about it now. Although I did many good things, obviously. But that one night is going to haunt me for the rest of my life.

Like Mrs A, surviving family members may feel passing guilt over what they did or did not do for the deceased.^{19 ,41} When pervasive self-reproach and survivor guilt become part of the clinical picture, the person may be experiencing depression²⁰ and potentially may be suicidal.^{10 ,19 - 21}

Mrs A complains of sadness, guilt, and insomnia. For her and most bereaved people, however, these depressive symptoms are usually transient and not numerous. Many survivors meet criteria for MDD in the first few months post-loss,^{5 - 6 ,22} with a minority having persistent depressive syndromes beyond the first year (eg, 42% at 1 month, 16% at 1 year).⁶

MRS A: There's nothing different about the phases. . . . I still find it very difficult to deal with, but it is ameliorating a bit. . . . The anniversary dates are all terrible . . . but the pain was not as bad this year as it was last year.

There has been a growing recognition^{4 ,42} that grief does not progress neatly through the proposed stages^{17 - 18} of (1) numbness and outbursts of distress and/or anger, (2) yearning and searching, (3) disorganization and despair, and (4) reorganization and recovery. The Institute of Medicine cautioned against the use of the term "stages" because such use "might lead people to expect the bereaved to proceed from one clearly identifiable reaction to another in a more orderly fashion than usually occurs."⁴ "Pangs of grief"¹⁹ —the intrusive, time-limited intense yearning and pining for the deceased—may come and go in waves for years after the loss.^{17 ,19} As in the case of Mrs A, these experiences typically attenuate in intensity and frequency, becoming more bittersweet than painful. For some individuals, however, grief remains chronic and severe. Intense grief (that meeting criteria for complicated grief) lasting 6 months post-loss and beyond has been shown to predict enduring dysfunction.^{16 ,22 - 24}

During the often intense last few weeks of life, the physician not only cares for the patient, but often for the spouse and family. However, after the patient has died, the family continues to need contact from the physician.

MRS A: My husband's doctor . . . as soon as [my husband] died, that was the end of him. That's one of the things that I object to: all the doctors just suddenly go . . . there's no support. If I felt like [my husband's] physicians had enough respect and affection for me and would call me occasionally, that would be nice. Dr M and I talked, but it's not his responsibility to support me; my husband's doctor should have been there.

Mrs A resents that her deceased husband's physician did not call her after his death. A telephone call, condolence letter, or visit shortly after the death is usually welcome.^{43 - 44} According to Bedell et al⁴⁴ : "A physician's responsibility for the care of a patient does not end when the patient dies. There is one final responsibility—to help the bereaved family members. A letter of condolence can contribute to the healing of the bereaved family. . . ." A follow-up contact with surviving family members acknowledges the loss, expresses sympathy and concern, and offers an opportunity to clarify questions about the patient's terminal care.

In the first couple of months post-loss, the physician might telephone to offer condolences and also to recommend a visit to evaluate and then monitor the survivor's health care needs.⁴³ The content of office visits might shift from ordinary practice to a discussion about the course of grief—as Dr M and Mrs A's interactions illustrated—symptoms indicating a need for professional intervention (eg, complicated grief, MDD, suicidality) and behavioral recommendations (Table 2).

Reluctance on the part of physicians to approach the deceased patient's survivors may stem from their perception that the family is angry with them, and perhaps from a sense of guilt and/or helplessness about being unable to prevent the death. In a study of reactions to terminal care, 30% of surviving family members reported dissatisfaction with the information provided about the cause of death.⁴⁵ Main⁴³ found that bereavement outcomes can be significantly influenced by communication and the quality of information given to survivors. Physicians who contact bereaved patients and express sorrow and concern may minimize the anger directed toward them.^{43 - 44}

The physician's discomfort or uncertainty about what to say or do when encountering a bereaved patient must be overcome in favor of taking active steps to help them. A list (Table 2) of comments and practices in communicating with and caring for grieving patients has been derived from a synthesis of discussions with widowed persons, participation in grief support groups, and suggestions offered by various Web sites.^{46 - 47}

MRS A: I started going to a support group about a week after my husband died, and I go to it still. And also, about a year ago, I met a very nice gentleman. I just did it at first because I thought this was a very sensible thing to do. I really didn't care to do it. But he's very good to me. I think one thing one misses tremendously is touching . . . everybody needs warmth from another person.

Research confirms that empathic friends may afford a great deal of comfort.⁴⁸ The benefits derived from developing new romantic interests⁴⁹ and participation in support groups⁵⁰ have also been demonstrated. Hence, encouraging these sorts of social activities would appear a sound practice.

MRS A: I wonder what happens to the regular person who's out there, whose whole way of life has to change, and who has no experience maintaining a household. If I had gone before my husband, it would have been a disaster . . . .

For women, a primary mechanism linking widowhood to depressive symptoms is financial strain, while for men, it is the strains of household management.⁵¹ Thus, attempts to minimize the sources of strain (eg, learning to cook or to manage money, possibly seeking employment) might reduce the risk of MDD and related mental and physical disorders (complicated grief, high blood pressure).

MRS A: I've kept myself very busy. I'm very involved in civic activities and am on a number of commissions nationally on alternative and integrative medicine and on breast cancer. I exercise daily.

Two studies of elderly subjects found that bereaved persons who maintained a busy, daily rhythm of activity had better sleep⁵² and fewer depressive symptoms than those with less active, structured schedules.⁵³ Mrs A's civic involvements and exercise regimen structure her day and provide her with a sense of purpose. Bereaved patients may derive similar benefits from staying involved and keeping regularly active.

Putting upsetting experiences into words, including disclosure about emotions in response to the death of a spouse, is associated with improved physical and mental health.^{54 - 55} Written and oral disclosure studies have even demonstrated a positive influence on immune function.⁵⁵ Based on these findings, physicians might encourage bereaved patients to express their thoughts and feelings about the loss (eg, in a journal).

MRS A: I went to a psychiatrist who unfortunately has now just died himself. He thought I did really well with handling this. I don't think I was ill. I didn't have that much of a depression. I was simply depressed.

Although Mrs A's distinction between "a depression" and "simply depressed" may appear subtle, it is an essential clinical determination. In the absence of a structured clinical interview, it is difficult to determine if Mrs A had MDD. Because we suspect she did not, based on what she has said, and also doubt that she met criteria for complicated grief (Table 1), we believe referral to a psychiatrist was not necessary in her case.

When psychiatric complications are suspected, primary care physicians must begin diagnosis and treatment or refer for expert consultation and intervention. While some argue for early intervention for MDD⁵⁶ regardless of bereavement status, Horowitz et al⁵⁷ recommend that diagnosis and treatment for pychiatric disturbance(s) among bereaved patients occur beyond a year after the loss. We recommend treatment for MDD or complicated grief lasting 6 months post-loss or beyond. The delay in treatment minimizes the identification and treatment of false-positive cases of MDD or complicated grief—cases that would resolve without intervention. Obviously, immediate attention from a mental health professional should be sought if suicidality is suspected at any time post-loss.

When enduring psychopathology exists, we believe that a psychiatric referral can be very helpful. However, bereaved geriatric patients may be reluctant to see a mental health professional, preferring to be seen by their primary care physician.⁵⁸ Primary care physicians who acquire the requisite expertise in the treatment of psychiatric disorders can be effective.

The results of an emerging body of literature on bereavement interventions suggest that treatment selection should depend on the patient's specific psychiatric diagnosis or diagnoses. For bereaved patients diagnosed with MDD alone, treatment should follow general guidelines,⁵⁹ including the prescription of selective serotonin reuptake inhibitors or tricyclic antidepressants. A randomized, placebo-controlled clinical trial of bereaved patients with MDD found nortriptyline alone had a 56% remission rate; nortriptyline in combination with interpersonal psychotherapy, 69%; and interpersonal psychotherapy alone, 29%.²⁷ An open-label trial of paroxetine, a selective serotonin uptake inhibitor, administered weekly over 4 months, demonstrated a 54% decline in symptoms of MDD.⁶⁰ Although a randomized controlled trial is needed to confirm the efficacy of selective serotonin reuptake inhibitors for MDD secondary to bereavement, MDD following the death of a loved one has been shown to be no different than other manifestations of MDD.⁶¹ Consequently, treatments of proven efficacy for MDD would be expected to work well for the reduction of bereavement-related depressive symptoms.⁵⁶

Results of studies documenting the reduction of grief-related symptoms (those targeting both earlier formulations of grief symptoms and complicated grief criteria, specifically) differ from those reported for bereavement-related MDD.^{27 - 28} For example, the randomized controlled trial by Reynolds et al²⁷ of interpersonal psychotherapy and/or tricyclic antidepressants found that these treatments did not significantly reduce symptoms of complicated grief. Randomized controlled trials of crisis intervention ⁶² and brief dynamic psychotherapy^{50 ,63} demonstrate significant reductions in grief symptoms, with support groups showing efficacy equal to that of dynamic psychotherapy.^{50 ,63} In a small randomized controlled trial, a behavioral therapy called "guided mourning" significantly reduced symptoms of "morbid grief."⁶⁴ Another brief psychotherapy in development, called "traumatic grief therapy,"⁶⁵ is designed specifically to ameliorate symptoms of complicated grief and incorporates elements of cognitive behavioral therapy. In pilot work, traumatic grief therapy had large effect sizes (2.2 and 1.5 in analyses of completers and intent-to-treat patients, respectively) for reducing symptoms of complicated grief, with significant declines reported for symptoms of bereavement-related depression and anxiety. With respect to pharmacotherapy, the open-label trial of paroxetine demonstrated a 53% decline in symptoms of complicated grief. Based on these findings, it appears that traumatic grief therapy and selective serotonin reuptake inhibitors may be the treatments of choice, given their efficacy for reducing the symptoms of both complicated grief and MDD. Randomized controlled trials are needed before these recommendations can be made conclusively.

There are several compelling reasons for physicians to actively engage in bereavement care. First, they already are involved in caring for bereaved patients and will become increasingly so as the US population ages. Empathic "aftercare" for bereaved patients demonstrates the physician's respect for the deceased and concern for surviving family members. It may soften the psychological blow of losing a loved one and reduce the family's sense of abandonment by the health care system. Enhanced efforts to discuss the medical decisions and care leading up to the patient's final moments may assist both surviving family members and physicians in attaining a greater sense of closure. The detection and treatment of psychiatric complications secondary to bereavement may reduce the morbidity with which they are associated. Most importantly, as the introductory quote from George Eliot suggests, physicians who aid grief-stricken patients are afforded the rewarding, quintessentially human opportunity of transforming a personal sorrow they inevitably will experience into sympathetic and supportive "aftercare."