A Clinical Guide to Supportive and Palliative Care for HIV/AIDS

Chapter 2. Overview of Clinical Issues


The past two decades have seen both the emergence of AIDS as a new, life-threatening infectious disease and its conversion from a rapidly fatal illness into a manageable chronic disease. This pattern has been most marked in industrialized countries where the promise of HIV-specific therapies has been realized for many individuals living with HIV. However, even in the era of "highly active antiretroviral therapy" (HAART), AIDS remains an important cause of morbidity and mortality in many young adult populations, and attention to palliative and end of life issues is an essential aspect of clinical care. In the early years of the AIDS epidemic, clinicians had to learn about palliative care by necessity, in the absence of any hope of curative therapy. With the advent of antiretroviral therapy and the ability to control HIV disease progression, it remains important to incorporate relevant aspects of palliative care in the comprehensive management of patients with AIDS. Rather than being "either-or," curative and palliative approaches to HIV care need to be "both-and," and one paradigm need never fully substitute for the other. This chapter will outline some of the important clinical issues in HIV palliative care, many of which will be addressed in greater detail in subsequent chapters.


Starting in the early 1980s, AIDS rapidly became the leading cause of death for young adults in the United States.1 With advances in AIDS care and HIV-specific therapy in the mid-1990's, mortality rates began to decline, and with the introduction of the protease inhibitors in 1996, the rates declined even more dramatically.2-7 However, the decline in death rates has since plateaued, and there remain approximately 15,000 deaths per year from HIV/AIDS.4 While the number of deaths from AIDS dropped by 25% and 42% from the preceding years in 1996 and 1997, respectively, these figures dropped to 17% and 8% for 1998 and 1999.4, 6 Moreover, the declines in death rates have not been uniform across all populations affected by HIV/AIDS, and decreasing mortality has not been as pronounced among African-Americans and Latinos as it has been among whites.4, 8 In addition, the incidence of new HIV infections is not believed to have decreased, and has remained stable at approximately 40,000 new cases per year.8 As a result of these trends, AIDS-related mortality continues to be an important phenomenon, and the number of patients living with HIV (i.e., the prevalence of AIDS) has actually increased.4, 8 (Figure 2-1.)

In the pre-HAART era, AIDS was a rapidly fatal, acute infectious disease, characterized by multiple typical opportunistic infections, swift decline, and death within months of diagnosis. The impact of disease-specific treatment on the natural history of HIV infection has now resulted in a much more variable trajectory of illness for many patients. For some, HAART has meant the possibility of full return of function and health, with AIDS experienced as a chronic condition that has little impact on daily quality of life. For others, treatment has meant the "conversion of death to disability," with the emergence of a chronic disease phase characterized by exacerbations, remissions, and eventual decline and death from the illness,9 as the survival time from AIDS diagnosis to death has lengthened.10 Thus, over a remarkably short period of time, the historical evolution of HIV disease -- for which the rapid disease course from diagnosis to death once resembled that of untreatable fatal cancers -- has shifted to a trajectory more typical of chronic, progressive illnesses such as congestive heart failure, chronic obstructive pulmonary disease, or hepatic cirrhosis, with much more variability in outcomes.11 (Figure 2-2.) Clinical assessment of prognosis and outcomes is thus much more complex than was previously the case.

As AIDS-specific mortality has declined, surveillance data and clinical studies have indicated that mortality among patients with HIV has proportionally increased as a result of co-morbidities such as hepatitis B and C, co-occurring malignancies (both AIDS-defining and non-AIDS-defining cancers), and substance abuse-related deaths.9, 12-16 In addition, for certain patients, even the benefit of HAART is not always attainable, due to lack of access to care, inability to adhere to effective treatment regimens, active substance use or other psychiatric illness, progressive viral resistance despite therapy, serious other co-morbidities, or unmanageable drug toxicities. For all these reasons, comprehensive AIDS care must continue to encompass end-of-life issues even as treatment continues to advance. Just as we must advocate for timely access to HAART for all patients, so too must we be ready to provide needed and appropriate palliative care even after HAART is no longer a life-extending option. The need to incorporate both palliative and curative approaches in HIV care is even more important than it was in the pre-HAART era: the availability of treatment does not give clinicians the luxury of ignoring the important issues posed by a chronic progressive illness and its management over time.17

Centrality of Palliative Care in HIV Care

In the early years of the AIDS epidemic, a cohort of physicians and other care providers found themselves confronted with a disease that they could not understand, let alone cure. It was a time in which the narrow medical model of care -- isolating specific elements of disease and then focusing on treating the pathology -- was totally inadequate in the face of the enormity of this disease and its effects on patients and their families. Paradoxically, it was a time in which clinicians learned the importance of being present, of accompanying patients through illness and being with patients and families as they grappled with the critical issues of death and dying. Clinicians learned how to focus on the goals of care, helping to empower patients even in the context of a life-threatening illness and discussing treatment decisions in terms of quality of life and care preferences of patient and family, rather than making decisions for their patients (as occurs with the more one-sided and hierarchical model that often predominates in the medical encounter). When disease could not be prevented or even treated, it was important to provide relief by treating pain, other symptoms, and the overall suffering caused by progressive illness. Patients' needs were understood to be multi-dimensional -- medical, psychosocial, and spiritual -- and they needed to be addressed in a multi-disciplinary model of care that did justice to the complexity of these needs, especially at the end of life. Families and significant others were also central to the context of care for patients early in the AIDS epidemic, and it was common for providers to attend funerals and memorial services as well as provide formal and informal bereavement services for survivors.

What is striking from this summary of salient clinical issues is that care for patients with AIDS in the pre-HAART era constituted comprehensive palliative care and that many of the elements of palliative care remain central to the routine care of patients with HIV even in the therapeutic era. Indeed, the widely used World Health Organization definition for palliative care serves as an appropriate and timely description of comprehensive care for patients with advanced HIV disease:

"Palliative medicine is the study and management of patients with active, progressive, far advanced disease for whom the prognosis is limited and the focus of care is the quality of life. [It is] the active total care of patients whose disease is not responsive to curative treatment. Control of pain, of other symptoms, and of psychological, social, and spiritual problems, is paramount. The goal of palliative care is achievement of the best quality of life for patients and their families." 18

What is also striking is the emergence of a new cohort of HIV care providers who did not live through that earlier era and who have learned about HIV care more within the therapeutic paradigm of HAART. Not surprisingly, this paradigm tends to be more strictly biomedical and less likely to incorporate end-of-life and palliative care issues since, thankfully, end-of-life issues are less inescapable than they previously had been. However, it is important to recognize that the curative vs. palliative dichotomy is a false one. HIV care has evolved over a short period of time with the rapid development of disease-specific therapy, although the tension between palliative and curative approaches has always been present. Since the early 1980s there have been several mini-"paradigm shifts" in which the hope of cure for AIDS or the reality of the limitations of HIV therapy have led prematurely to extreme optimism or pessimism, respectively, as shown schematically in Figure 2-3. The overall direction has been toward more curative or disease-modifying therapy, but the curve has oscillated back and forth, and even after the heady optimism of the late 1990's following the introduction of the protease inhibitors, few if any researchers are currently speculating about the likelihood of viral eradication or "cure" for HIV infection.

In fact, perhaps a more sober assessment would lead one to conclude that AIDS is a chronic, progressive disease that may be effectively managed in some patients but may still cause considerable morbidity and mortality in others. As such, while we eagerly embrace and seek to extend any benefits of HAART and other disease-specific therapies, we must also be prepared to anticipate and address the important aspects of palliative and end-of-life care that many (if not all) of our patients will be facing at some point during the course of their illness. Otherwise, we run the risk of isolating or unconsciously abandoning our patients when our therapies are no longer effective or of being ill-prepared to help and accompany them through the later stages of their disease. The more we focus on the therapy, the less we tend to focus on the patient. Indeed, some providers experience growing frustration with patients who do not adhere to HAART, and the conflict over adherence may becomes the main platform for interaction -- or avoidance -- between provider and patient. This conflict can eventually become dysfunctional, especially for patients who refuse therapy outright, which may threaten the basis of the relationship unless patient and provider can interact in ways that emphasize the totality of care and not merely adherence to recommended antiretroviral regimens.

Rather than the false dichotomy of either curative or palliative care, this guide is suggesting that both approaches need to be considered throughout the course of HIV disease. This integrated approach is represented graphically in Figures 2-4a and 2-4b. Rather than the simple and misleading framework of sequential curative and then palliative care (Figure 2-4a), a much more dynamic and integrated framework is required (Figure 2-4b) -- one which accounts for the changing needs of the patient throughout the course of the illness but does not rigidly choose one or the other domains.17,18

Clearly, in the early phases of HIV infection, curative or disease-specific therapies may predominate (e.g., HAART, specific prophylactic regimens to prevent opportunistic infections), and in the later phases of AIDS, palliative approaches become more important (e.g., management of pain, symptoms, and quality-of-life issues in addition to prolongation of life when possible). However, as noted above, this is not simply "either-or," but rather "both-and:" management of nausea and vomiting and gastrointestinal toxicity from HAART may be a key aspect of promoting good adherence with antiretroviral therapy, and treating CMV retinitis in a dying patient with daily intravenous ganciclovir or foscarnet may be an important quality-of-life issue even after HAART no longer has a role in reversing the course of illness. Thus, it is important to adopt an integrated model, which allows both curative and palliative elements to be incorporated in the comprehensive care of the patient.

As will also be emphasized throughout this guide, this integrated model of care is most fully realized when it encompasses all the contributions of an interdisciplinary team. The key role of the clinical team becomes even more important as patients approach the end of life. Once patients' physical comfort needs have been addressed, important emotional and spiritual work can often be done. The team -- including members from medicine, nursing, social work, and pastoral care -- can become an integral part of the patient and family's support system throughout the death and dying process. Patients and families experience the totality of illness (with medical, psychological, social, and spiritual dimensions), and the interdisciplinary team can best implement a biopsychosocial approach to care, regardless of which needs are expressed at any given time. The complexity of AIDS in all these dimensions requires this type of comprehensive, collaborative, and multi-leveled response.

For many clinicians involved in the first phase of the AIDS epidemic, it took a disease we could not cure to teach us the true meaning of healing. It is our hope that these lessons will not be lost as we strive to provide our patients with all the best that HIV therapy has to offer at the same time as we seek to help them and their loved ones navigate the complex trajectory of this disease and its changing impact on their lives.

Pain and Symptom Management

Since early in the epidemic, clinical studies have documented a high prevalence of pain and other symptoms in patients with AIDS.19-22 Pain in AIDS has been attributed variously to:

  1. The effects of specific opportunistic infections (e.g., headache with cryptococcal meningitis, visceral abdominal pain with disseminated Mycobacterium Avium complex [MAC] infection)

  2. The effects of HIV itself or the body's immune response to it (e.g., distal sensory polyneuropathy, HIV-related myopathy)

  3. The effects of medications used to treat HIV disease (e.g., dideoxynucleoside-related peripheral neuropathy, zidovudine-related headache, protease inhibitor-related gastrointestinal distress)

  4. The non-specific effects of chronic debilitating illness, along with other miscellaneous causes

Some of these findings on pain are from the pre-HAART era, and certainly the contribution of specific opportunistic infections to pain syndromes and other symptoms in AIDS has diminished over time as the incidence of such infections declined. However, it should be noted that in some instances the incidence and/or prevalence of pain may have actually increased over time. As is often the case with AIDS, the irony of decreased mortality rates is that by surviving longer some patients may thus be vulnerable to new complications and pain, as in the observed increasing prevalence of peripheral neuropathy which occurred with longer survival according to the Multi-Center AIDS Cohort Study.23 Further, in a recent example of the potential overlap of palliative and disease-specific therapies in AIDS, analysis of data from an ongoing observational cohort study found that the severity of HIV-related neuropathy was associated with plasma viral load levels -- suggesting that antiretroviral therapy itself might in fact be useful for treating or preventing this painful syndrome.24 In addition, while pain due to opportunistic infections may have diminished with the advent of HAART and more effective prophylactic regimens, the medications themselves may cause pain and other symptoms, e.g. the antiretroviral side effects alluded to above, which may compromise effective treatment unless the symptoms are also effectively palliated.

Despite the high prevalence of pain in AIDS, several studies have also demonstrated that pain in patients with AIDS is likely to be under-diagnosed and under-treated.21, 25 This failure to diagnose and treat may reflect both the general under-recognition of pain by most physicians and/or the additional reluctance to consider seriously any self-report of pain in patients with a history of substance use problems. Moreover, recent reports have documented that non-white race/ethnicity may be a risk factor for inadequate analgesia in general in medical settings, and that even the physical availability of narcotic pain medication may be limited in pharmacies serving poor urban neighborhoods where HIV infection may also be concentrated.26, 27

Regardless of the possible explanations for under-treatment of pain, the result is that patients with AIDS are at risk for significant pain and the resulting diminished quality of life -- an outcome which in most cases could be prevented with adequate pain assessment and management. As described in Chapters 4 (Pain) and 11 (Substance Use Problems), the science of pain management has advanced considerably in recent years. It is now fully possible to assess and treat pain effectively in patients with AIDS, including substance users, using standard measurement techniques, rational decisionmaking, evidence-based practice, and common sense.28, 29 These pain management tools should be as much a part of the pharmaceutical inventory of HIV care providers as antiretrovirals and prophylactic agents.

In addition to pain, patients with AIDS have been found to have a high prevalence of other symptoms, particularly but not exclusively in the advanced stages of the disease.30-38 Moreover, one recent study suggested that physicians frequently fail to identify and under-treat common symptoms reported by patients with AIDS.35 Symptoms have included a mixture of physical and psychological conditions, such as fatigue, anorexia, weight loss, depression, agitation and anxiety, nausea and vomiting, diarrhea, cough, dyspnea, fever, sweats, pruritus, etc. Table 2-1 lists common symptoms in AIDS by organ system. Table 2-2 summarizes the findings of several key studies that have examined the symptom burden in patients with AIDS in different populations. It is striking that these studies, conducted in the United Kingdom, Canada, France, and Italy, showed a remarkable consistency of symptoms across populations even with different selection criteria, different time periods, and varying methods for determining the prevalence of symptoms.

In addition to these studies, which were conducted mostly in late-stage patients admitted to hospice or hospital with symptomatic disease, a symptom prevalence study was conducted using a large national probability sample of over 3000 patients with HIV infection in the United States receiving care in 1996.39 For this sample, the prevalence of the most common ten symptoms for the preceding six months was as follows:

  • Fever, sweats, or chills (51%)

  • Diarrhea (51%)

  • Nausea or anorexia (50%)

  • Numbness, tingling, or pain in hands/feet (49%)

  • Headache (39%)

  • Weight loss (37%)

  • Vaginal discharge, pain, or irritation (36%)

  • Sinus infection or pain (35%)

  • Visual problems (32%)

  • Cough or dyspnea (30%)

Another study of AIDS outpatients receiving care in New York in the early 1990s found a mean of 16.7 current symptoms (using the Memorial Symptom Assessment Scale), of which the most common were worrying (86%), fatigue (85%), sadness (82%), and pain (76%).40 With these populations as well, the concordance of symptom types with those in the previously published studies is noteworthy, as is the high prevalence of symptoms in relatively non-selected populations of patients with HIV.

As with the findings related to pain, symptoms in AIDS may result from the specific effects of opportunistic infections (e.g., sweats, fever, in disseminated MAC infection, dyspnea in pneumocystis carinii pneumonia), from the apparent effects of the progression of HIV infection itself (e.g., weight loss, fatigue), from medications (e.g., antiretroviral-induced nausea and vomiting), or from the non-specific manifestations of late-stage illness (e.g., depression, fatigue, malaise). As with studies of the prevalence of pain, much of the observational knowledge regarding symptom prevalence in AIDS was generated in the pre- or early HAART era, and the specific effects of certain opportunistic infections may have diminished as these entities have become less common.3, 7, 41 Nevertheless, the challenges of symptom management have also grown with several factors:

  1. The prolongation of the chronic disease phase in some patients, with the resulting more protracted late-stage course described above, requiring more sustained attention to complex symptom management in chronically ill patients;

  2. The recognition of the wide cumulative range of drug toxicities and symptomatic sequelae which occur in patients on long-term antiretroviral therapy;

  3. The assessment of the contribution of drug toxicities vs. the underlying effects of disease progression, raising important clinical issues regarding the competing effectiveness and toxicity of different therapeutic options (e.g., balancing short-term quality of life vs. long-term prolongation of life in a patient contemplating a potentially toxic antiretroviral regimen, deciding whether disease-specific therapy may also in fact have a palliative impact on symptoms in a given situation, etc.);

  4. The emergence of co-existing co-morbid conditions that have complicated the management of patients with AIDS, including chronic viral hepatitis (especially but not limited to hepatitis C), and chronic co-morbid psychiatric illness and substance abuse in patients with AIDS.42-44

As with the management of pain, clinicians caring for patients with AIDS need to be familiar with the science and practice of palliative medicine, which has emerged as a fast-growing specialty receiving increased attention in the United States and elsewhere. Indeed, the United States is a relative newcomer to the field, with much of the scientific and professional contribution to palliative medicine having emanated first from the United Kingdom and Canada.29 The literature of palliative medicine has documented impressive advances in recent years in the elucidation of pathophysiology and treatment of many of the common symptom syndromes listed above, such as nausea and vomiting, dyspnea, fatigue, and weight loss.29, 45 Some of this work has included palliative care in the context of AIDS.46 Much recent work has also been done in the psychopharmacologic treatment of depression and anxiety and other psychiatric illness, including the treatment of such conditions at the end of life,47 and many treatment options exist for these conditions as well. These options will be discussed in more detail in Chapter 5, Constitutional Symptoms, and Chapter 10, Psychiatric Problems, respectively.

Although some complicated syndromes in symptom management, as in pain management, would benefit from the input of a palliative care specialist, in many cases primary HIV care providers can identify and treat a wide range of AIDS-related symptoms using standard palliative medicine strategies that will both enhance patients' quality of life and also maximize the likelihood of adherence to disease-specific therapy in cases in which this is still an option. Basic familiarity with both realms of care should be part of the clinical repertoire of all AIDS care providers. As is now true for other areas of pharmacologic management in AIDS,48 clinicians also need to be aware of and anticipate possible drug-drug interactions between antiretrovirals and other AIDS-related medications and the medications commonly used in palliative care, including opioids, benzodiazepines, anti-convulsants, and sedative-hypnotics (see Chapter 27 for more information on drug interactions).

As was noted in an editorial in the journal Pain in 1986:

"Up to the 19th century, most medical care related to the amelioration of symptoms while the natural history of the disease took its course toward recovery or death. By 1900, doctors and patients alike had turned to a search for root cause and ultimate cure. In the course of this new direction, symptoms were placed on one side as sign posts along a highway, which was being driven toward the intended destination. Therapy directed at the signposts was denigrated and dismissed as merely symptomatic ... [Yet] the immediate origins of misery and suffering need immediate attention while the long-term search for basic cure proceeds. The old methods of care and caring had to be rediscovered and the best of modern medicine has to be turned to the task of new study and therapy specifically directed at pain." 49

The same combined strategy is no less relevant to HIV care than to medical care in general.

Strategies and Goals of Care

As defined above, palliative medicine seeks to provide the best quality of life for patients and families using a model in which the goals of care are collaboratively developed with care providers according to the wishes of patients and families. This model emphasizes communication, collaboration, and the willingness to accept patient-focused outcomes as paramount and to respect patients' wishes regarding such important issues as quality of life or cessation of therapy. Table 2-3 summarizes some of the key aspects of the philosophy of palliative medicine.50 This palliative philosophy comes from a model of care that is distinctly different from the familiar, hierarchical, and physician-dominated model that characterizes much of contemporary medical care, especially for diseases that have potentially effective therapies.

The more collaborative palliative approach -- one in which uncertainty is shared, in which patients and their care providers work through difficult decisions with an ultimate inability to control the outcome, and in which there are many ambiguities and nuances and no clear "right answer" -- was one that defined much of AIDS care in the pre-HAART era. In this environment, clinicians learned to work together with patients and families to clarify goals of care, to determine the important issues related to quality of life and the end of life, and to be comfortable helping patients and families negotiate the complexities of progressive, incurable illness.

With the arrival of more effective disease-specific therapy, much of the focus has shifted to more curative or quasi-curative therapies for which the goals of treatment are more clear-cut, uniform, and physician-generated. This guide seeks to provide the rationale and the logic for incorporating the more collaborative palliative approach into the routine clinical decisionmaking and interactions of clinicians and their patients. The intent is not to eschew therapy or to downplay the importance of achieving certain objectively measurable outcomes such as directing therapy to achieve an undetectable HIV viral load or a sustained rise in CD4+ cells, but rather to encourage care providers to recognize that above all it is the patient who must live with the illness, and that the goals of care are for the patient and not the physician. For instance, an overly narrow focus on antiretroviral treatment protocols and the technical details of HIV care will neither do justice to the patient's condition nor ultimately promote a meaningful patient-physician relationship.

Focusing on the patient- and family-centered goals of care ensures that decisions will be made which do not violate important concerns of the patient and family. Curative, palliative, or both types of interventions can be offered, not unilaterally and driven by diagnostic or treatment algorithms but rather collaboratively and driven by the priorities and values of the patient and family. Some of this work takes time and may be more open-ended than the narrow focus on test results and medication adherence that has come to dominate much of our attention in routine office visits with patients on HAART. This collaborative approach will also pose a particular challenge to busy primary care providers who already experience significant time pressure and possibly increased numbers of longer-surviving HIV-infected patients in their practices. Nevertheless, if we are to be effective as care providers for patients and families affected by HIV/AIDS, it is fundamentally important to work within this collaborative, comprehensive framework.

Specific examples of some common clinical scenarios involving decisions about palliative and/or disease-specific care -- and the context for decision-making based on goals of care rather than on simple diagnosis-treatment algorithms -- include the following:

  1. The use of transfusion and/or psycho-stimulants (e.g., methylphenidate) and/or corticosteroids to treat fatigue in late-stage patients, with or without HAART;

  2. Aggressive anti-emetic therapy for protease inhibitor-induced nausea and vomiting, or alternatively, revision or even discontinuation of antiretroviral therapy if such side effects are unable to be managed successfully;

  3. Continued suppressive therapy with intravenous foscarnet or ganciclovir for CMV retinitis, or intravenous amphotericin B for azole-resistant candidiasis, even after the discontinuation of antiretroviral therapy and other prophylactic regimens in a dying patient;

  4. Palliative treatment of disseminated MAC (e.g., with anti-pyretics, narcotic analgesics, and corticosteroids) in patients with advanced disease unwilling or unable to take anti-infectives;

  5. Withdrawal of MAC or PCP prophylaxis in patients expected to die soon;

  6. Withdrawal of HAART after evident treatment "failure," with assessment of risk-benefit as well as symbolic and emotional value;

  7. Decisions to withdraw or forego artificial nutrition or hydration in a patient unable to maintain oral intake, involving assessment of patients' wishes, functional and mental status, quality-of-life concerns, and social and cultural values;

  8. Realistic vs. false hopes regarding decisions to initiate HAART or "salvage" regimens in late-stage patients (e.g., the possibility for a potentially disease-reversing course of antiretroviral therapy, vs. the possibility that the patient may not respond, with all the resulting therapeutic ambiguities and challenges for counseling patient/family about prognosis).

All of the scenarios above involve decisions that need to be informed both by medical evidence and risk-benefit analysis as well as by the priorities, values, and preferences of patients and families. In all these scenarios it should be clear that decisions must be based on the specific goals of care, including such factors as:

  • The relative values of quality of life or prolongation of life

  • The use of aggressive palliative care interventions to help mitigate side effects of other HIV-specific medications

  • The use of certain disease-specific therapies primarily for palliative or quality-of-life intent

  • The decision not to prolong life "no matter what," once a certain important threshold had been reached (e.g., progressive dementia or the inability to manage oral feedings or medications)

Clearly, this type of decisionmaking takes place in much more ambiguous and nuanced territory than the simple assessment of antiretroviral efficacy by measuring surrogate laboratory markers, but it is no less important and in many ways more challenging.

Table 2-4 lists common symptoms in AIDS along with their varied possible etiologies and examples of both disease-specific therapies and palliative interventions. One or the other or both types of approach might be appropriate in a given situation, depending on the patient's disease stage, functional status, symptom burden, risk and potential benefit of therapy, and expressed care preferences and goals.

Advance Care Planning

The focus on goals of care incorporates the concept of advance care planning, which includes both medical and psychosocial elements. Medical aspects may include:

  • Decisionmaking about the risks and benefits of specific therapies (e.g., antiretroviral therapy, chemotherapeutic regimens, etc.)

  • The importance of quality-of-life elements in these decisions (e.g., the choice of a potentially toxic, time-intensive intervention that might prolong life vs. more palliative measures that would improve short-term quality of life but not extend it)

  • Decisions about particular medical interventions (e.g., cardiopulmonary resuscitation, artificial nutrition or hydration, re-hospitalization of a patient being cared for at home)

Psychosocial issues that need to be addressed may include:

  • Family ambivalence or conflict about care plan decisions

  • Guilt or other emotional "unfinished business" that may affect decisions regarding care planning

  • Concerns that any limitation on curative treatment interventions may represent abandonment of the patient or lack of commitment or concern on the part of the care provider

(Psychosocial issues will be discussed in more detail later.)

Discussions about advance care planning should be ongoing, anticipatory, and re-visited on multiple occasions as needed over time. These issues should be "normalized" as much as possible and incorporated into routine care, rather than waiting until a crisis such as an acute life-threatening illness or the loss of decisional capacity to begin to address them. It may, in fact, be an indication of the discomfort that providers feel within the either-or framework of curative vs. palliative care that they may unconsciously defer these discussions until the patient is precipitously dying, implying that discussing goals of care and end-of-life decisionmaking is somehow inappropriate within the current therapeutic model. However, the approach to these issues needs to be positive, respectful, and focused on the values, concerns, cultural beliefs, and care preferences of the patient and family. The family and/or other involved individuals need to be included earlier rather than later in these discussions, and decisions about health care proxies and other surrogate decisionmakers need to be made clearly and communicated to the designated individuals.

Despite the complexity of these psychosocial issues in the setting of HIV/AIDS -- that patients are often young adults who are faced with a life-threatening illness early in their lives, who may have young families, and who suffer from the stigma and social vulnerability still experienced by people with AIDS -- patients have been found to be unlikely to have discussed these issues with their care providers.51, 52 This finding was even more pronounced among African-Americans and Latinos than whites.51, 52 The lack of discussions about advance care planning underscores the importance of clinicians pro-actively bringing up these issues -- and sensitively addressing barriers to discussions53 -- before they emerge in the setting of an acute crisis requiring an immediate life-and-death decision. Rather than being a sign of withholding treatment or imposing values on patients, anticipatory advance care planning is a way to empower patients and families to make decisions that are true to their basic beliefs, values, and concerns. The most effective way to achieve this kind of decisionmaking is in the context of an interdisciplinary team (physician, nurse, social worker, chaplain, and other disciplines as appropriate) which can help the patient and family process and act on information on many levels at once. It is not an accident that the philosophy and practice of hospice care explicitly incorporate an interdisciplinary team approach in the routine care of dying patients, and this framework should be no less essential for the comprehensive care of patients with HIV.

Prognostic Uncertainty and Palliative Care

Even as we have been emphasizing the importance of focusing on the goals of care in informing treatment decisions, it must be recognized that prognostication and the expectation of likely outcomes in the course of HIV disease are much less certain and uniform than they were in the pre-HAART era. Ironically, it is precisely as we are reminded of the importance of clarifying the goals of care -- now that there are choices, we need to ensure that decisionmaking incorporates these choices -- that we are also reminded that prognosis and the "natural history" of HIV infection are much less clear-cut than they were previously. While CD4+ counts and viral load assays are excellent measures of response to therapy and indeed of prognosis in general, the possibility of effective antiretroviral therapy -- or alternatively the lack of this possibility when there are no viable treatment options -- can completely alter prognosis for people with AIDS.

The National Hospice Organization's 1996 Guidelines for determining prognosis in certain noncancer diagnoses attempted to generate criteria indicative of likely less-than-six months' prognosis for patients with AIDS (see Table 2-5).54 While some of these clinical conditions may be useful prognostic markers, none of them would likely override the potential positive impact of effective antiretroviral therapy if this were still an option. Indeed, some patients have been referred to hospice, received palliative care, and expected to die, only to surprise themselves and their care providers with their miraculous recoveries (the "Lazarus Syndrome") from effective HAART. In these cases, forcing patients to choose an "either-or" approach would clearly be unconscionable. We must both be able to prognosticate as best we can based on evidence and the patient's specific treatment history and options, and be prepared to accept that our best estimates may be made irrelevant by the potential impact of therapy. This reality only makes the integration of palliative and curative approaches both more challenging and more necessary than ever before.

One issue that frequently arises related to the complexities of prognostication and clinical decisionmaking in the HAART era involves the discontinuation of antiretroviral therapy in a patient who is either not responding or felt to be unlikely to respond to treatment. Even though there is controversy about whether to stop therapy even in the face of apparent treatment failure (i.e., the concept of viral "fitness" and possible benefit of antiretroviral selective pressure on viral replication dynamics even in the setting of high viral loads and low CD4+ counts),55, 56 we must recognize that the benefits of antiretroviral therapy, even when effective, are not immediate, and that they must be evaluated in light of potential favorable impact to prevent future decline. Thus, it is reasonable to question whether it makes therapeutic sense to continue antiretroviral therapy in a patient dying of lung cancer or end-stage liver failure -- whether or not the medications could even be tolerated in this setting -- or in the obvious end stages of progressive symptomatic HIV disease. In these instances, antiretroviral therapy will not be likely to have any meaningful benefit and will probably only add to the therapeutic confusion in a patient who is clearly dying yet for whom aggressive therapy is being continued. However, in some cases, the patient may have such a strong emotional investment in continuing therapy that it is completely reasonable to continue it, although it should be clear that this is as much a psychosocial as a medical decision, if not more so. These examples underscore the importance of clarifying the goals of therapy, assessing the potential impact and likely risks and benefits of therapy, and working collaboratively with patients to determine priorities and treatment plans -- all aspects of a good patient-provider relationship that become obscured if too much of the focus is on the treatment instead of on the patient.

Psychosocial and Family Issues

AIDS has always presented unique psychosocial problems for patients, families, and care providers: a life-threatening illness affecting young adults, often with multiple infected family members, raising difficult issues of premature death, unfinished business, legacy, and survivorship. Guilt, shame, anger, and despair may compound the emotional challenges of coming to terms with this progressive, incurable infectious disease. Some of the behaviors linked to HIV infection remain socially unacceptable in certain contexts, and the association between AIDS, poverty, and racial-ethnic minority populations in the United States further compounds the vulnerability of many patients living with HIV/AIDS. Despite some advances in public education and awareness about AIDS, there remains a significant degree of stigma, fear, and prejudice regarding AIDS within the society as a whole.

In addition to all of the challenges that AIDS has always posed, there are new issues that have arisen in the HAART era that are particularly relevant to palliative and end-of-life care. With the long awaited impact of HAART -- beginning most notably in 1996 -- and the subsequent rapid decline in death rates from AIDS, the inevitability of short-term mortality has receded. This phenomenon has even resulted in the identification of new potential stressors due to a "'second life' agenda" in patients who must now prepare to go on living instead of preparing to die.57 While the decreased risk of death has clearly been a welcome relief for patients and their caregivers, it also has had the tendency in some cases to isolate those who are still dying with AIDS in the HAART era, at a time when people's expectations have been so dramatically lifted by the promise of effective antiretroviral therapy.17 As noted above, while death rates have decreased and the rate of decline has slowed, there are still over 15,000 people per year dying from AIDS in the United States, with the prospect that this number might continue to increase when the compounding co-morbidities of hepatitis B and C, substance abuse and its sequelae, cooccurring cancers, and other potentially fatal conditions are taken into account.

In the HAART era, the formerly grim scenario of AIDS as a uniformly and rapidly fatal disease is no longer the usual outcome -- dying from AIDS has now passed from "fate" to "tragedy." Although in the pre-HAART era death came swiftly to AIDS patients, the new drug therapies are now capable of extending life for many years -- at least in some patients. Unfortunately, the patients who do not benefit from the new drugs may be blamed for their "failure" to respond to treatment or to adhere to the treatment regimen. These dying patients are sometimes ostracized and seen as the "anomalies" who are not able to benefit from the purported lifesaving effects of effective antiretroviral therapy. A new form of guilt may be seen in both patients and care providers, faulting poor therapy choices or poor adherence as the "cause" of a patient's death. With the advent of HAART, there are both greater possibilities for therapeutic success and greater opportunities for therapeutic failure, regret, and guilt over perceived bad decisions or missed opportunities.

Given the importance of adherence to HAART as a major factor in determining the likely success of a treatment regimen, it is not surprising that so much of the routine patient-provider clinical interaction has become focused on adherence to therapy and the close monitoring of laboratory tests such as CD4+ counts and HIV viral load assays. However, this focus on adherence has tended to reframe the patient-provider relationship in a new and more limited way, outside of the context of the experience of life-threatening illness and preparation for possible death. As with much of medicine in general, the more narrowly one operates within a biomedical model, the less comfortable one tends to be with end-of-life issues, since death in effect represents a "failure to cure." Consciously or not, HIV care providers in the current era may feel inadequate to help patients deal with impending death, since their role is understood as one of curing or fighting back the disease. Tragically, providers may feel so out of place that they may withdraw emotionally from patients and become less involved in their care as they approach death, so that patients experience both the losses imposed by the illness and the potential loss of a trusted long-term relationship with their care provider.

In addition, even prior to the end of life, providers need to develop a keen awareness of their own feelings regarding patient choices. Providers may actually feel anger at patients over failure to adhere to antiretroviral therapy or for active substance abuse and other self-destructive behaviors. Re-examining the goals of care, involving the interdisciplinary team, and continuing to engage and provide follow-up for patients -- without judging their choices or inability to adhere with recommended therapy -- will help to maintain and enhance the clinical relationship.

The fact that the new generation of HIV caregivers has not had the experience of their earlier colleagues -- in which there was no avoidance of death and no immediate prospect of forestalling it with disease-specific therapy -- may also be a contributing factor in the difficulty that providers experience in addressing end-of-life issues. While we would hope that the experience of the early years of the AIDS epidemic will never have to be repeated, we can also hope to retain the lessons that it brought regarding the importance of accompanying patients through illness until death, of not seeing death as an automatic failure of the clinician, of appreciating the profound importance of "being there," and of understanding the role of the caregiver as "caring" and not necessarily "curing."58 As outlined above regarding palliative and disease-specific therapy, the strategy should not be "either-or," but rather "both-and," and providers need to be as skilled and competent in anticipating and addressing the end-of-life issues as they are in interpreting surrogate marker test results and recommending effective combination anti-retroviral treatment regimens.

Another area in which caregivers can have an invaluable impact is in the support and education of families as patients move through the last stages of their illness, especially for patients who remain at home. Ongoing visits by nursing and other staff, either at home or in an institutional setting, can be extremely helpful in addressing concrete concerns and questions about prognosis; the expected changes that occur toward the end of life; and some of the physical, emotional, and spiritual issues that families can expect. It may be helpful to describe these events in terms of the estimated prognostic timeframe covering the last months, weeks, and days of life, along with the responses that families may consider to help their loved ones through this process. Table 2-6 describes the guidance that the clinician can provide caregivers as the patient moves through the last stages of illness. As with other aspects of palliative care, a team approach and sustained, consistent interaction over time are key elements in a successful therapeutic relationship.

In addition to the myriad psychosocial issues outlined above, and the more clinical psychiatric issues dealt with in more detail in Chapter 10 (Psychiatric Problems), two other important psychosocial themes are (1) multiple loss and (2) grief and bereavement. While the phenomenon of multiple loss59 in the AIDS epidemic has long been identified for patients, providers, and communities, this is compounded in the current era because of the long-term relationships that have been built up over years of patient-provider interaction. Even though the absolute number of deaths may be declining, the impact of each loss may be as great or greater. This can result in a cumulative burden of grief that can begin to have deleterious effects on providers over time. It is important to recognize, anticipate, and respond to these effects, and to appreciate that "emotional health-maintenance" and self-care are not signs of wasteful self-indulgence but rather important aspects of maintaining our perspective and effectiveness as HIV care providers.

Grief both anticipates the loss and follows the loss, with reverberations into the future for all those who have been affected. With the current chronic disease-type trajectory typical of HIV infection (see Figure 2-2), with many exacerbations and remissions in the declining stages, families and caregivers frequently describe the sensation of being on a "roller coaster" because it is exhausting and stressful not to know when death is going to occur. This uncertainty can also give rise to heightened anxiety, ambivalence within the family (wishing that it would just "be over with" and then feeling guilt over this), and equally conflicted feelings of relief when the death finally does occur. These complex emotions require sensitive and skilled attention by both medical and mental health care providers, and are best addressed by the interdisciplinary team.

Bereavement is an important phase of the process of loss and recovery and is another area that unfortunately is rarely addressed by most care providers and by medicine as a profession. The loss of the patient that occurs upon death is also often accompanied by the family's loss of the relationship with the care provider, which again may have developed over years of close interaction with both patient and family. The care provider's involvement in follow-up and contact with the family during bereavement (ranging from writing a condolence letter to attending a funeral service to providing counseling or other clinical follow-up to family members) can be extremely helpful and healing for both the family and the clinician.60 Some or all of these practices should be incorporated into routine follow-up care for families and additional significant others after the patient has died, whether or not the HIV care provider is also the primary care provider for other members of the family (see Figure 2-4b). Many clinical AIDS programs perform periodic memorial services for patients who have died, attended by both professional care providers and families/loved ones, which is both a powerful expression of remembrance and a part of the process of working through grief for the survivors. Chapter 16: Grief and Bereavement and Chapter 20: Care for the Caregiver address these issues in more detail.


As we enter an era in which the therapeutic possibilities for AIDS continue to expand, it remains important not to lose sight of the critical issues in end-of-life and palliative care that remain central to the comprehensive care of patients and families affected by this disease. This book attempts to provide useful information regarding these issues in the hope that providers can move beyond the artificial distinctions between curative and palliative care and be able to provide optimal care to all patients throughout the course of the illness.

In the current system of medical care, primary care providers are best able to deliver integrated, comprehensive care over the continuum of illness in ways that combine biomedical and psychosocial approaches within an interdisciplinary model of care. The group of providers able to provide integrated care may include clinicians in primary care disciplines such as general internal medicine, family medicine, and pediatrics, or as has been evidenced in HIV care to date, subspecialists such as infectious disease physicians or oncologists who have been able to take on the full range of care required by the patient.

Regardless of the training, what is required is the willingness for us to go beyond the false dichotomies of curative vs. palliative care and recognize that the true role and responsibility of the physician and care provider are in accompanying the patient through the experience of illness, doing whatever can and should be done at each step along the way. Part of this task involves being familiar and current with the science of palliative medicine, just as with the science of HIV medicine. Part of it also involves acceptance of the physician's inability to always defeat death and acknowledgement of our own limitations and vulnerabilities as well as our patients' needs. As was written by Sheila Cassidy, a British hospice physician, in her memoir Sharing the Darkness:

"Slowly, I learn about the importance of powerlessness. I experience it in my own life and I live with it in my work. The secret is not to be afraid of it -- not to run away. The dying know we are not God. All they ask is that we do not desert them."61

Or, as expressed in the often-cited phrase attributed to Edward Trudeau (the Canadian physician in the early 20th century credited with developing the sanitarium system for tuberculosis care long before the advent of disease-specific chemotherapy), the goal of the clinician for the patient with an incurable, progressive and terminal illness is: "To cure never, to treat sometimes, to comfort always."

A condensed version of this chapter was previously published in the online publication Innovations in End-of-Life Care as Selwyn PA, Rivard M. Palliative care for AIDS: Challenges and opportunities in the era of highly active anti-retroviral therapy. Innovations in End-of-Life Care. 4(3), 2002. www.edc.org/lastacts.


  1. Centers for Disease Control. Mortality attributable to HIV infection/AIDS-United States, 1981-1990. MMWR 40:41, 1991.

  2. Centers for Disease Control and Prevention. Update: Trends in AIDS incidence, deaths, and prevalence- United States, 1996. MMWR 46:165-173, 1997.

  3. Palella FJ Jr, Delaney KM, Moorman AC, et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. N Engl J Med 338:853-60, 1998.

  4. Centers for Disease Control and Prevention. HIV/AIDS Surveillance Report. 12:1-44, 2000.

  5. Wong T, Chiasson A, Reggy A, Simonds RJ, Heffess K, Loo V. Antiretroviral therapy and declining AIDS mortality in New York City. J Urban Health 77:492-500, 2000.

  6. Chiasson MA, Berenson L, Li W, et al. Declining HIV/AIDS mortality in New York City. J Acquir Immun Defic Syndr 21:59-64, 1999.

  7. Egger M, Hirschel B, Francioli P, et al. Impact of new antiretroviral combination therapies on HIV infected patients in Switzerland: prospective multicentre study. BMJ 315:1194-9, 1997.

  8. Centers for Disease Control and Prevention. HIV and AIDS -- United States, 1981-2000. MMWR 50:430-4, 2001.

  9. Selwyn PA, Goulet JL, Molde S, et al. HIV as a chronic disease: long-term care for patients with HIV at a dedicated skilled nursing facility. J Urban Health 77:187-203, 2000.

  10. Lee L, Karon J, Selik R, Neal J, Fleming P. Survival after AIDS diagnosis in adolescents and adults during the treatment era, United States, 1984-1997. JAMA 285:1308-15, 2001.

  11. Lynn J. Serving patients who may die soon and their families. JAMA 285:925-32, 2001.

  12. Sansone RG, Frengley J.D. Impact of HAART on causes of death of persons with late-stage AIDS. J Urban Health 77:165-75, 2000.

  13. Puoti M, Spinetti A, Ghezzi A, et al. Mortality from liver disease in patients with HIV infection: a cohort study. J Acquir Immun Defic Syndr 24:211-7, 2001.

  14. www.cdc.gov/hiv, viewed 8/29/01.

  15. Kravcik S, Hawley-Foss N, Victor G, et al. Causes of death in HIV-infected persons in Ottowa, Ontario, 1984-1995. Arch Intern Med 157:2069-73, 1997.

  16. Valdez H, Chowdhry TK, Asaad R, et al. Changing spectrum of mortality due to HIV: analysis of 260 deaths during 1995-1999. Clin Infect Dis 32:1493, 2001.

  17. Selwyn PA, Arnold R. From fate to tragedy: the changing meanings of life, death, and AIDS. Ann Intern Med 129:899-902, 1998.

  18. World Health Organization. Cancer Pain Relief and Palliative Care, Report of a WHO Expert Committee. Publication 1100804. Geneva: World Health Organization, 1990.

  19. LaRue F, Colleau SM. Underestimation and undertreatment of pain in HIV disease: multicentre study. BMJ 314:23-8, 1997.

  20. Singer JE, Fahy-Chandon B, Chi S, Syndulko K, Tourtellotte WW. Painful symptoms reported by ambulatory HIV-infected men in a longitudinal study. Pain 54:15-9, 1993.

  21. Breitbart W, Rosenfeld B, Passik SD, et al. The undertreatment of pain in ambulatory AIDS patients. Pain 65:243-9, 1996.

  22. O'Neill W, Sherrard J. Pain in human immunodeficiency virus disease: a review. Pain 54:3-14, 1993.

  23. Bacellar H, Munoz A, Miller E, et al. Temporal trends in the incidence of HIV-1-related neurologic diseases: Multicenter AIDS Cohort Study, 1985-1992. Neurol 44:1892-1900, 1994.

  24. Simpson D, Haidich AB, Miller E, et al. Severity of HIV-associated neuropathy is associated with plasma HIV-1 RNA levels. AIDS 16:407-12, 2001.

  25. Breitbart W, Kaim M, Rosenfeld B. Clinicians' perceptions of barriers to pain management in AIDS. J Pain Symptom Manage 18:203-12, 1999.

  26. Todd KH, Samaroo N, Hoffman JR. Ethnicity as a risk factor for inadequate emergency department analgesia. JAMA 269:1537-9, 1993.

  27. Morrison RS, Wallenstein S, Natale DK, Semzel RS, Huang LL. "We don't carry that" -- failure of pharmacies in predominantly nonwhite neighborhoods to stock opioid analgesics. N Engl J Med 342:1023-6, 2000.

  28. Jacox A, Carr DB, Payne R, et al. Clinical Practice Guideline: Management of Cancer Pain. Rockville M.D.: Agency for Health Care Policy and Research, U.S. Department of Health and Human Services, Public Health Service, 1994.

  29. Doyle D, Hanks GWC, MacDonald N, eds. Oxford Textbook of Palliative Medicine, 2nd ed. New York: Oxford University Press, 1998.

  30. Fantoni M, Ricci F, Del Borgo C, et al. Multicentre study on the prevalence of symptoms and symptomatic treatment in HIV infection. J Palliat Care 13:9-13, 1997.

  31. Kelleher P, Cox S, McKeogh M. HIV infection: the spectrum of symptoms and disease in male and female patients attending a London hospice. Palliat Med 152-8, 1997.

  32. LaRue F, Brasseur L, Musseault P, et al. Pain and symptoms in HIV disease: a national survey in France. Abstract: Third Congress of the European Association for Palliative Care. J Palliat Care 10:95, 1994.

  33. Foley F. AIDS Palliative Care. Abstract: 10th International Congress on the Care of the Terminally Ill. J Palliat Care 10:132, 1994.

  34. Moss V. Palliative care in advanced HIV disease: Presentation, problems, and palliation. AIDS 4(S):S235-42, 1990.

  35. Fontaine A, LaRue F, Lassauniere JM. Physicians; recognition of the symptoms experienced by HIV patients: how reliable? J Pain Symptom Manage 18:263-70, 1999.

  36. Breitbart W, McDonald MV, Rosenfeld B, Monkman ND, Passik S. Fatigue in ambulatory AIDS patients. J Pain Symptom Manage 15:159-67, 1998.

  37. Wood CG, Whittet S, Bradbeer CS. ABC of palliative care: HIV infection and AIDS. BMJ 315:1433-6, 1997.

  38. Filbet M, Marceron V. A retrospective study of symptoms in 193 terminal inpatients with AIDS. Abstract: Third Congress of the European Association for Palliative Care. J Palliat Care 10:92, 1994.

  39. Mathews W, McCutcheon JA, Asch S, et al. National estimates of HIV-related symptom prevalence from the HIV Cost and Services Utilization Study. Med Care 38:762, 2000.

  40. Vogl D, Rosenfeld B, Breitbart W, et al. Symptom prevalence, characteristics, and distress in AIDS outpatients. J Pain Symptom Manage 18:253-62, 1998.

  41. Ledergerber B, Egger M, Erard V, et.al. AIDS-related opportunistic illnesses occurring after initiation of potent antiretroviral therapy: the Swiss cohort study. JAMA 282:2220-6, 1999.

  42. den Brinker M, Wit FW, Wertheim-van Dillen P, et al. Hepatitis B and C virus co-infection and the risk for hepatotoxicity of highly active antiretroviral therapy in HIV-1 infection. AIDS 14:2895-902, 2000.

  43. Sulkowski M, Thomas D, Chaisson R, Moore R. Hepatotoxicity associated with antiretroviral therapy in adults infected with human immunodeficiency virus and the role of hepatitis C or B virus infection. JAMA 283:74-80, 2000.

  44. Goulet JL, Gaughan D, Molde S, Selwyn PA. The impact of serious psychiatric disorders on the long term care of people with AIDS. J Urban Health 77:213-21, 2000.

  45. End of Life Physician Education Resource Center (EPERC). www.eperc.mcw.edu, accessed 8/29/01.

  46. Corcoran C, Grinspoon S. Treatments for wasting in patients with the acquired immunodeficiency syndrome. N Engl J Med 340:1740-50, 1999.

  47. Block SD. Assessing and managing depression in the terminally ill patient. ACP-ASIM End-of-Life Care Consensus-Panel American College of Physicians-American Society of Internal Medicine. Ann Intern Med 132:209-18, 2000.

  48. Piscitelli SC, Gallicano KD. Drug therapy: interactions among drugs for HIV and opportunistic infections. N Engl J Med 344:984-96, 2001.

  49. Wall PD. Twenty-five volumes of Pain. Pain 25:1-4, 1986.

  50. Fox E. Predominance of the curative model of medical care. JAMA 278:761-3, 1997.

  51. Wenger NS, Kanouse DE, Collins RL, et al. End-of-life discussions and preferences among persons with HIV. JAMA 285:2880-7, 2001.

  52. Mouton C, Teno JM, Mor V, Piette J. Communications of preferences for care among human immunodeficiency virus-infected patients. Barriers to informed decisions? Arch Fam Med 6:342-7, 1997.

  53. Curtis JR, Patrick DL. Barriers to communication about end-of-life care in AIDS patients. J Gen Intern Med 1997; 12:736-41.

  54. National Hospice Organization. Guidelines for Determining Prognosis for Selected Non-Cancer Diagnoses. Alexandria, VA: National Hospital Organization, 1996.

  55. Deeks S, Wrin T, Liegler T, et al. Virologic and immunologic consequences of discontinuing combination antitretroviral-drug therapy in HIV-infected patients with detectable viremia. N Engl J Med; 344:472-80, 2001.

  56. Frenkel L, Mullins J. Should patients with drug-resistant HIV-1 continue to receive antiretroviral therapy? N Engl J Med 344:520-2, 2001.

  57. Rabkin JG, Ferrando S. A "second life" agenda. Arch Gen Psych 54:1049-53, 1997.

  58. Selwyn P. Surviving the Fall: The Personal Journey of an AIDS Doctor. New Haven: Yale University Press, 1998.

  59. Nord D. Issues and implications in the counseling of survivors of multiple AIDS-related loss. Death Studies 20:389-413, 1996.

  60. Bedell SE, Cadenhead K, Graboys TB. The doctor's letter of condolence. N Engl J Med 344:1162-4, 2001.

  61. Cassidy S. Sharing the Darkness. Maryknoll, New York: Orbis Books, 1991.