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

Chapter 23. Medical Care in Advanced AIDS


Well-informed patients can optimize their quality of life through partnership with their physician in clinical decisionmaking. This collaborative process is particularly important for patients with advanced AIDS and their families. This chapter assists the clinician in striking the appropriate balance between disease-directed therapy and comfort-centered care in caring for people with advanced disease.

Specific clinical events, presented in Table 23-1 (PDF), should, at a minimum, prompt an explicit discussion of the goals of care, and may trigger a shift in emphasis from cure or disease management to palliation. Acknowledging and discussing with the patient these transitions in illness and the inherent implications is integral to quality patient care. Failure to do so adversely affects patient quality of life and often leads to hopelessness, depression and amplification of physical symptoms.

These discussions with patients are extremely important interventions and require sophisticated communication skills. (See Chapter 21: Patient-Clinician Communication.)

Quality of Life-Based Decision-Making for Patients With Advanced AIDS

Many patients receiving highly active antiretroviral therapy (HAART) experience marked improvement in their clinical condition. Others are not as fortunate. Some have irreversible deficits (usually neurologic, cognitive and/or motor) and live at a level of considerable disability or incapacity. 1-3 Others experience persistent treatment toxicities that make remaining on HAART difficult or impossible. Moreover, even for those who do experience improvements in clinical status, there is not necessarily an improved quality of life in areas other than depression.1 Over time, therefore, the burden of treatment may begin to outweigh its benefits. New therapies increase the chances of living longer, but often with diminished physical or cognitive capacity and onerous treatment-related symptoms. 3, 4 This is particularly true with second-line and salvage therapy. For some patients, living longer in this diminished state is less desirable than a shorter life.

Providers must, therefore, elicit the patient's view on the key physical, medical, emotional and spiritual elements that define a good quality life, the life worth living. Providers must recognize, moreover, that patients' perspectives may differ sharply from their own, and understand the ethical implications inherent in these differences. (See Chapter 20: Care for the Caregiver and Chapter 17: Ethical Issues.)

Sentinel Events Triggering Discussion About the Advanced AIDS State

The occurrence of the events presented in Table 23-1 (PDF) often indicates that death is likely in the foreseeable future and should cause the provider to explicitly review prognosis, goals, quality-of-life perspective and expectations with the patient and family.

Relative and Absolute Antiviral Drug Resistance

Antiviral drug resistance is more likely in patients with CD4 less than 200, in patients who are less adherent to the initial drug regimen, and in patients on second and third-line drug regimens.5

In a group of 70 protease naive patients with advanced AIDS started on HAART, the treatment failure rate was 40% and mortality rate was 16% in three months. Brechtl's study shows mortality was not associated with nonadherence and, while depression and weight improved, other aspects of quality of life did not.1

Freedberg developed a statistical model of AIDS survival which showed significant improvement in overall survival and quality-adjusted years of life with adherent HAART therapy but also revealed a relatively short survival for patients with advanced AIDS. A graphic representation of two hypothetical patients is shown in Figure 23-1. Estimated survivals are shown in Table 23-2.6

Taken together the Brechtl and Freedberg studies imply that the evolution of drug resistance is associated with increasing illness, a reduced quality of life and diminished life expectancy. In a group of patients for whom short survival can be an expected outcome, therefore, disease-directed therapy and palliative care can and should be provided simultaneously, with palliative care playing an increasingly important role in the care of these patients over time.

Second-Line, Salvage, and Investigational Therapy

There are data in a study by Durant, et al., suggesting that second-line therapy, particularly when guided by genotype testing, is valuable in increasing longevity.7 Indeed, for some patients viral suppression on second-line therapy may last for years. This same study, however, also indicated that a substantial number of patients progressed despite therapy or were intolerant of therapy. An AIDS Clinical Trial Group (ACTG) study of response to second-line treatment further reinforces that second-line treatment is a marker for advanced disease.8 Institution of second-line therapy should therefore prompt recognition that, while another remission is possible, treatment failure may be immediate and is probably inevitable.9 Thus, conversations about the end-of-life and the role of palliative care are indicated when second-line therapy is considered or initiated. This is even more relevant as patients undertake third and fourth-line therapy after multiple treatment failures.

The outlook for patients on salvage therapy or those participating in clinical trials following exhaustion of all conventional regimens is very poor. Initiation -- indeed, consideration -- of salvage or investigational therapy under these circumstances carries with it the imperative to reassess the goals and burdens of treatment with patients and to explore the benefits and risks of intensive palliative care. These discussions may inform decisions about stopping or declining further antiretroviral therapy or prophylaxis against or suppression of opportunistic infections, depending on the defined goals of treatment and the perceived effects of continued disease-directed therapy.

Clinical Ethics and Practice in Salvage and Investigational Therapy

The decision to continue second-line, salvage or investigational therapy has several ethical and psychological implications which can complicate discussions about palliative and disease-directed care.

Although cancer and AIDS differ in many ways, the experience with cancer can inform AIDS clinician-investigators.

Patients who enter investigational cancer trials participate in part because of protocol eligibility, such as disease and functional criteria. Many advanced stage cancer patients who make decisions to participate in cancer clinical trials are highly motivated and feel that "active" treatment is the best course for them. Patients are also influenced by their physicians and their own sense of altruism.10-12

These patients often expect a response to therapy, a reduction in symptoms and improved and increased quality communication with their physician.11 They equate therapeutic efforts with superior quality of life and do not consider any other options or quality of life ramifications. Palliative care, therefore, is not a consciously considered option for many patients with advanced cancer who are enrolled in cancer clinical trials, and is not consistently offered. Cancer patients who enroll in clinical trials overestimate their survival, making these patients more likely to choose putative life-extending therapy over palliative care.13

Phase I and Phase II trials are not trials for the patients' benefit. Rather, their goals are to define toxicity patterns, establish maximum tolerated dose, and complete trials in selected tumor types that could, if responses are seen, lead to advanced Phase II and Phase III trials. Responses measured in Phase I and early Phase II trials rarely convert into prolonged survival for patients enrolled in those trials.14

Furthermore, systemic barriers obstruct palliative care program participation by these patients. Impediments dissuade programs from informing prospective patients, thereby restricting access. The added regulatory barrier that de facto prevents patients receiving disease-directed therapy from even being referred for hospice care15 exacerbates dissonance between disease-directed and palliative approaches. Non-viable financial requirements argue against simultaneous care and are detrimental to clinical research, to informed consent, and to best patient care. For example, hospice programs are required to absorb costs associated with disease-directed therapy (like HAART) and the cost of treating patients' side effects from investigational therapy.

Consequently, when investigational therapy is completed, patients are often close to death. Both patient and loved ones have little opportunity to address end-of-life tasks. The physician and the patient and family have focused often on the disease-directed therapies to the exclusion of end-of-life issues and palliative care.16

The perceived dissonance between the goals of disease-directed therapy and palliative care leads to patient and physician reluctance to discuss concurrent application of both investigational therapy and palliative care. For the patient with advanced AIDS, participation in clinical trials should not be a barrier to effective symptom management or intensive emotional support. A desire for optimal quality of life should not preclude clinical trials participation.


Another sentinel event that appropriately triggers a patient/provider reassessment of the goals of treatment is non-adherence to therapy. Current HAART regimens are complex. Patients are routinely expected to ingest between 6 and 20 pills a day, often on bid or tid dosing schedules with significant dietary restrictions.

Patients experiencing virologic and clinical failure despite protease inhibitor therapy seem to fall into two categories: those treated over an extended period of time with a variety of antiretroviral agents, and those whose adherence to a HAART regimen is inadequate to sustain viral suppression.5

Adherence research in other diseases with significantly less demanding regimens has documented levels of fully adherent behavior as low as seven percent.17, 18 Furthermore, "... rates of compliance with different long-term medication regimens for different illnesses in different settings tend to converge to approximately 50%."19

However, it should be noted that adherence rates vary greatly depending on measurement methodology. Moreover, adherence distribution curves are often U-shaped (not bell-shaped), suggesting that close to one third of patients are highly adherent, roughly one third are functionally non-adherent and the remainder fall in the 20%-to-80%-adherent range. This renders meaningless the "approximately 50%" figure cited above.

Adherence rates with chronic conditions also decline over time when reinforcement is absent. An 80% rate of adherence, often adequate though not ideal in managing other diseases, is inadequate for effective HAART.20, 21 Several investigators have demonstrated a precipitous decline in HIV suppression with each 5% to 10% decrement in adherence.5, 22-25

Many factors have been examined as potential influences on adherence. These include the patient's age, education, income, gender, active or prior alcohol or substance use, depression, relationship to the health care provider, interactions with others, primary language, race and ethnicity, involvement with AIDS service organizations, use of reminders and cues, use of mechanical devices (a pill box, for instance), location of care delivery, routine, treatment side effects experienced, beliefs about treatment, access to care, venue of treatment, costs associated with medications and others.5, 26, 27 Several of these factors, while intuitively associated with greater or lesser adherence, do not bear out empirically. Furthermore, physician ability to predict adherence success or failure is demonstrably poor.

Current or prior substance use, poverty, illiteracy and alcohol or drug addiction, commonly thought to predict poor adherence, are not in fact, significantly associated with adherence.2, 25, 28 As Wright concluded in a 2000 study, "not adhering to treatment regimes is so widespread that no combination of sociodemographic variables is reliably predictive of patients' not following doctors' orders."23 Depression and active alcohol abuse are the only two factors consistently demonstrated to have a deleterious effect on adherence. Scrutiny of most other characteristics yields equivocal results and argues convincingly for approaching the issue of adherence with care and on a patient-by-patient basis. Several strategies to improve adherence have been studied and can be helpful in the clinical setting.24, 26, 29, 30


AIDS-associated cachexia syndrome reflects the uncontrolled inflammatory state induced by the virus.31 Oversecretion of inflammatory cytokines produces pathologic changes including protein catabolism and weight loss. Pharmacologic appetite stimulants can lead to weight gain and may enhance the patient's sense of well-being, but survival time is not lengthened.32, 33

In advanced AIDS, symptoms of hunger and thirst are reduced or absent. Patients report anorexia and even dysgeusia ("food tastes bad"). Caregivers frequently respond by forcing foods and, sometimes in concert with physicians, advocating for enteral or intravenous alimentation. Such supplemental feedings are toxic, do not prolong life and are not consistent with biomedical ethical guidelines or high quality of life. Some reasons not to artificially feed a patient with advanced AIDS are presented in Table 23-3.34

Eating is often equated to fundamental caregiving, and offers opportunities for communication and sharing. Health care providers can help caregivers and patients identify alternative activities such as life review, storytelling, family outings, cards and games, reading. Time is often better spent promoting communication and demonstrating caring than preparing unwanted or discomfiting meals. Explicit reassurance to the caregiver that the patient will not starve to death, and that suffering is not increased, is critical both for patient comfort and the reduction of caregiver anxiety.

AIDS-Associated Malignancies

Several cancers are now AIDS-defining illnesses including Kaposi's sarcoma (KS), non-Hodgkins lymphoma, and invasive cervical carcinoma in women. The diagnosis and staging of these malignancies are identical to diagnosis and staging in the HIV-negative patient. The prognosis, however, is worse.

  • Kaposi's Sarcoma

    Dermatological KS often responds well to treatment, whereas visceral KS, including the lung and GI tract, is not curable. Although some patients may obtain short-term control with HAART and systemic chemotherapy such as a doxorubicin-containing regimen, in general palliation is the primary goal. Absent regulatory and financial barriers, addressed in Chapter 18: Legal and Financial Issues, all patients with incurable KS should be referred to hospice.

  • Non-Hodgkins Lymphoma (NHL)

    The incidence of NHL is estimated at up to 2% per year in HIV-infected patients and 20% over three years in AIDS patients. In the HIV-positive patient, NHL is always high grade B cell histology, is always disseminated (Stage III or IV), and presents as either systemic, CNS or body cavity disease. Diagnosis of CNS lymphoma should prompt hospice referral as average survival is only 2 to 4 months. Radiation does not prolong survival and provides minimal palliation. Systemic lymphoma has been treated with modified regimens with an average survival of eight months. At the end of therapy, however, and certainly at relapse, patients should be referred to hospice or otherwise be allowed to benefit from intensive palliative care.35

Progressive Multifocal Leukoencephalopathy (PML)

Progressive multifocal leukoencephalopathy (PML) is thought to be caused by reactivation of the JC virus, a human papovavirus. The sudden deterioration of intellectual and physical function of PML heralds a median survival of 2 to 4 months. Recent trials have demonstrated no improvement with novel interventions.36, 37 While clinical investigation should continue, and some clinicians recently have opted to use empiric cidofovir even though this is of unproven efficacy, active PML marks the need for intensified palliative care.

Anecdotally, a number of patients described as presenting with PML have experienced greatly improved scans and significantly improved longevity following the initiation of HAART. However, patients with PML who do not experience immune reconstitution with antiretroviral therapy have very limited life expectancy.

Advanced Liver Disease

Advanced liver disease due to viral hepatitis is becoming a leading cause of death in patients infected with HIV. While treatment guidelines for the management of HIV-infected patients co-infected with hepatitis B and/or C continue to advance, not even the emerging new treatments will be effective in preventing mortality for patients with advanced cirrhosis and liver failure.

Past efforts at prognosis using the Child-Turcotte-Pugh classification have not been consistent because of the subjective nature of several indicators. A recent report used objective laboratory parameters to assign prognosis in advanced liver disease: the Model for End Stage Liver Disease (MELD).38 A revised scoring system uses three laboratory tests (INR, total bilirubin, and creatinine) to assign a prognosis in chronic liver disease.

This revised scoring system has been utilized to stratify patients for liver transplantation, which has recently been recognized in many centers as a feasible option for people with AIDS. The scoring system may also prove useful in determining which patients deserve a greater emphasis on palliative or hospice care. In fact, any individual with a high MELD score and on the waiting list for an organ is also a candidate for hospice.39 If the patient receives a liver, then the hospice benefit can be revoked. Most patients awaiting transplantation, however, will not receive a liver, will die, and would benefit greatly from aggressive palliative care.

Large hepatocellular carcinoma is an additional complication of chronic hepatitis and an absolute indication for palliative or hospice care.

Intensive Care Unit/Mechanical Ventilation/PCP

Merely being in an intensive care unit is an indication for palliative care regardless of diagnosis. 40, 41 AIDS is no exception. Effective palliative care can be provided in an ICU with proper training.41 Pneumocystis carinii pneumonia is still the most common AIDS-defining illness.42, 43 Acute respiratory failure occurs in 5% to 30% of patients. Intubated AIDS patients with PCP experience mortality approaching 50%.44

The attitude of patients with advanced AIDS toward mechanical ventilation at end-of-life has been reported.45 Eighty-seven percent of 57 patients responded that withholding mechanical ventilation in the setting of "futility" was acceptable. Therefore, while mechanical ventilation may be very appropriate for the patient with the initial diagnosis of AIDS and PCP, the patient with advanced AIDS and drug resistant virus may be quite comfortable discussing alternatives to hospitalization and ventilation.

Thus palliative care and even an end-of-life style of support for the patient and family in these circumstances should be considered routinely and provided aggressively.

Quality of Life in Advanced AIDS

A study has demonstrated that while asymptomatic patients with HIV had physical functioning equivalent to the U.S. general population, patients with symptomatic AIDS were impaired much like patients with other chronic illness (gastro-esophageal reflux disease (GERD), prostate disease, depression, diabetes) and those with advanced AIDS scored similar to patients with end-stage renal disease (ESRD).46 One difference worth noting is that all three groups of people with HIV in the study scored lower on the Emotional Well-Being Scale than all comparison chronic illness groups except patients with primary depression.

Symptom Management and Quality of Life

Patients with advanced HIV disease often face a vexing array of symptoms: nausea, fatigue, weight loss, pain of various origins and presentations, headaches, neuropathies, diarrhea, fevers, shortness of breath, confusion, loss of memory, AIDS-related dementia, and others. Symptoms also arise as a result of therapy. Many of the drugs used in HAART can cause substantial side effects. HAART's complicated drug regimens, opportunistic infection prophylaxis and treatment of other conditions at times give rise to complex and confounding drug interactions. Control of pain and other symptoms is discussed in Part II of this guide and the related ethical considerations are addressed in Chapter 17: Ethical Issues.

Emotional and Interpersonal Quality of Life

Some patients will ask their physicians for help in hastening death. When this occurs, it is imperative for the physician to understand why such a request is being made and to respond accordingly (Table 23-4, PDF).47 For some patients the desire to hasten death can be reduced by good pain and symptom management. Others may need treatment for depression, anxiety or other psychiatric conditions. Interest in quickening death is often a response to factors that are degrading the patient's quality of life and can be mitigated. The quality of communication between physician and patient is of paramount importance. (See Chapter 21: Patient-Clinician Communication and Chapter 22: Facilitating the Transition to Home-Based and Hospice Care.)

Ongoing communication between clinician and patient yields benefits beyond mitigating requests for assistance with dying. Patients who have regular conversations with their physicians about quality of life and preferences for care are more likely to complete advance medical directives, and their physicians are more likely to know their patients' relative preferences regarding pain control, symptom relief, and prolongation of life.3

Table 23-5 (PDF) offers some examples of questions intended to deepen the dialogue between the patient and his or her health care providers.48 These questions focus on particular domains in the psychosocial and spiritual arenas.

Hospice: The Interdisciplinary Model for Quality-of-Life Focus

It is important that health care providers be familiar with policies and procedures necessary to refer patients to hospice. (See Chapter 22: Facilitating the Transition to Home-Based and Hospice Care and Chapter 18: Legal and Financial Issues). Because of the difficulty in determining prognosis in advanced AIDS and because of the benefits of hospice care, it is not unusual for a critically ill AIDS patient who is seemingly at death's door to regain function and health while in hospice care, indeed in many circumstances because of hospice's intense, home-based interventions. With marked improvement, the patient no longer meets hospice criteria, and so begins a cycle of repeated graduation, deterioration, readmission to hospice (often following hospitalization), re-stabilization and regraduation. This paradigm is highly suggestive of a salutary relationship between the quality-of-life focus coupled with individualized care by hospice's highly-skilled team of providers resulting in periods of enhanced quality and perhaps quantity of life.

Medicare regulations can make it difficult or impossible to provide concurrent HAART and hospice services. Because the hospice benefit includes expenditures for medications out of daily patient per diem reimbursement, HAART can quickly bankrupt a hospice program. These issues must be considered along with the risks and benefits of HAART when contemplating referral to hospice care.


Patients with advanced AIDS benefit from competent and conscientious palliative care interventions. Routine discussions with the patient and regular reassessment of the goals and burdens of treatment enhance high quality care and well informed treatment choices. Careful attention to sentinel occurrences as triggers for additional discussion and reassessment further enhances care. Delivering disease-directed therapies and palliative care simultaneously, shifting the mix as the HIV disease biology unfolds, and attending to patient choice, comfort and quality of life are the cornerstones and the future of excellent care for patients with advanced AIDS.


  1. Brechtl JR, et al. The use of highly active antiretroviral therapy (HAART) in patients with advanced HIV infection: Impact on medical, palliative care, and quality of life outcomes. J Pain Symptom Manage 21:41-51, 2001.

  2. Rosenberg T. Patent laws are malleable. Patients are educable. Drug companies are vincible. The world's AIDS crisis is solvable. Look at Brazil. New York Times Magazine, January 28, 2001.

  3. Wenger NS, et al. End-of-life discussions and preferences among persons with HIV. JAMA 285:2880-7, 2001.

  4. Steinbrook R. Providing antiretroviral therapy for HIV infection. N Engl J Med 344:844-6, 2001.

  5. Paterson DL, et al. Adherence to protease inhibitor therapy and outcomes in patients with HIV infection. Ann Intern Med 133:21-30, 2000.

  6. Freedberg KA, et al. The cost effectiveness of combination antiretroviral therapy for HIV disease. N Engl J Med 344:824-31, 2001.

  7. Durant J, et al. Drug-resistance genotyping in HIV-1 therapy: the VIRADAPT randomised controlled trial. Lancet 353:2195-9, 1999.

  8. Albrecht MA, et al. Nelfinavir, efavirenz, or both after the failure of nucleoside treatment of HIV infection. N Engl J Med 345:398-407, 2001.

  9. Montaner JSG, Mellors JW. Antiretroviral therapy for previously treated patients. N Engl J Med 345:452-5, 2001.

  10. Daugherty CK. Impact of therapeutic research on informed consent and the ethics of clinical trials: a medical oncology perspective. J Clin Oncol 17:1601-17, 1999.

  11. Jenkins V, Fallowfield L. Reasons for accepting or declining to participate in randomized clinical trials for cancer therapy. Br J Cancer 82:1783-8, 2000.

  12. Yoder LH, et al. Expectations and experiences of patients with cancer participating in phase I clinical trials. Oncol Nurs Forum. 24:891-6, 1997.

  13. Weeks JC, et al. Relationship between cancer patients' predictions of prognosis and their treatment preferences. JAMA 279:1709-14, 1998.

  14. Emanuel EJ, Patterson WB. Ethics of randomized clinical trials [comments]. J Clin Oncol 16:365-71, 1998.

  15. Lynn J. Perspectives on care at the close of life. Serving patients who may die soon and their families: the role of hospice and other services. JAMA 285: 925-32, 2001.

  16. Schroeder SA. The legacy of SUPPORT. Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments [editorial; comment]. Ann Intern Med 131: 780-2, 1999.

  17. Gerber KE, Nehemkis AM. Compliance: The Dilemma of the Chronically Ill. New York: Springer, 1986.

  18. Cerkoney KA, Hart LK. The relationship between the health belief model and compliance of persons with diabetes mellitus. Diabetes Care 3:594-8, 1980.

  19. Haynes RB, Taylor DW, Sackett DL. Compliance in Health Care. Baltimore: Johns Hopkins University Press, 1979.

  20. Proctor VE, Tesfa A, Tompkins DC. Barriers to adherence to highly active antiretroviral therapy as expressed by people living with HIV/AIDS. AIDS Patient Care STDS 13:535-44, 1999.

  21. Wu AW. Adherence and health care utilization in HIV/AIDS: rational or rationalizing? J Gen Intern Med 15:891-3, 2000.

  22. Sorensen JL, et al. Medication adherence strategies for drug abusers with HIV/AIDS. AIDS Care 10:297-312, 1998.

  23. Wright MT. The old problem of adherence: research on treatment adherence and its relevance for HIV/AIDS. AIDS Care 12:703-10, 2000.

  24. Stephenson J. AIDS researchers target poor adherence. JAMA 281:1069, 1999.

  25. Sherr L. Adherence -- sticking to the evidence. AIDS Care 12:373-5, 2000.

  26. Laws MB, et al. Taking antiretroviral therapy for HIV infection: learning from patients' stories. J Gen Intern Med 15:848-58, 2000.

  27. Kemppainen JK, et al. HAART adherence in culturally diverse patients with HIV/AIDS: a study of male patients from a Veterans Administration Hospital in northern California. AIDS Patient Care STDS 15:117-27, 2001.

  28. Laurence J. Adhering to antiretroviral therapies. AIDS Patient Care STDS 15:107-8, 2001.

  29. Katzenstein DA, et al. HIV therapeutics: confronting adherence. J Assoc Nurses AIDS Care 8 Suppl:46-58, 1997.

  30. Wolfe GS. The case manager's role in adherence. J Assoc Nurses AIDS Care 8 Suppl:24-8, 1997.

  31. Baronzio G, et al. Proinflammatory and regulatory cytokine levels in AIDS cachexia in vivo. 13:499-502, 1999.

  32. Flynn NM, et al. Megestrol for AIDS-related anorexia -- Response. Ann Intern Med 122:881, 1995.

  33. Vonroenn JH, et al. Megestrol acetate in patients with AIDS-related cachexia. Ann Intern Med 121:393-9,1994.

  34. Gillick MR. Rethinking the role of tube feeding in patients with advanced dementia. N Engl J Med 342:206-10, 2000.

  35. Little RF, Y.R., Wilson WH, Systemic chemotherapy for HIV-associated lymphoma in the era of highly active antiretroviral therapy. Curr Opin Oncol 12:438-44, 2000.

  36. Greenlee JE. Progressive multifocal leukoencephalopathy. Progress made and lessons relearned. N Engl J Med 338:1378-80, 1998.

  37. Hall CD, et al. Failure of cytarabine in progressive multifocal leukoencephalopathy associated with human immunodeficiency virus. N Engl J Med 338:1345-51, 1998.

  38. Kamath PS, et al. A model to predict survival in patients with end-stage liver disease. Hepatology 33:464-70, 2001.

  39. Meyers FJ. Letter to the editor. N Engl J Med 340:860-1, 2002.

  40. Faber-Langendoen K, et al. Dying patients in the intensive care unit: Forgoing treatment, maintaining care. Ann Intern Med 133:886-93, 2000.

  41. Danis M, et al. Incorporating palliative care into critical care education: Principles, challenges, and opportunities. Crit Care Med 27:2005-13, 1999.

  42. Benfield TL, et al. Prognostic markers of short-term mortality in AIDS-associated Pneumocystis carinii pneumonia. Chest 119:844-51, 2001.

  43. Forrest DM, et al. The incidence and spectrum of AIDS-defining illnesses in persons treated with antiretroviral drugs. Clin Infect Dis 27:1379-85, 1998.

  44. Forrest DM, et al. Validation of the modified multisystem organ failure score as a predictor of mortality in patients with AIDS-related Pneumocystis carinii pneumonia and respiratory failure. Chest 114:199-206, 1998.

  45. Curtis JR, Patrick DL, Caldwell ES, Collier AC. The attitudes of patients with advanced AIDS toward use of the medical futility rationale in decisions to forgo mechanical ventilation. Arch Intern Med 160:1597-601, 2000.

  46. Hays RD, et al. Health-related quality of life in patients with human immunodeficiency virus infection in the United States: results from the HIV Cost and Services Utilization Study. Am J Med 108:714-22, 2000.

  47. Back AL, Wallace JI, Starks HE, Pearlman RA. Physician-assisted suicide and euthanasia in Washington state. JAMA 275:919-25, 1996.

  48. Block SD. Psychological considerations, growth, and transcendence at the end of life. The art of the possible. JAMA 285:2898-905, 2001.