For those who work in the healing professions, ethical values play a special role. The specific content of medical ethics was first formulated centuries ago, based on the sayings of Hippocrates and others. The refinement of medical ethics has continued up to the present by practicing health caregivers, health professional and religious organizations, and individual ethicists. As medical technology, health care financing, and the organization of health care transform themselves, so must the content of medical ethics change in order to acknowledge and guide new circumstances.
Over the years, participants in and observers of medical care have distilled widely shared human beliefs about healing the sick into four major ethical principles: beneficence, nonmaleficence, autonomy, and justice (Beauchamp and Childress, 2008) (Table 13–1).
Table 13–1. The four principles of medical ethics
Beneficence is the obligation of health care providers to help people in need.
Dr. Rolando Bueno is a hard-working family physician practicing in a low-income neighborhood of a large city. He shows concern for his patients, and his knowledge and judgment are respected by his medical and nursing colleagues. On one occasion, he was called before the hospital quality assurance committee when one of his patients unexpectedly died; he agreed that he had made mistakes in his care and incorporated the lessons of the case into his future practice.
Dr. Bueno tries to live up to the ideal of beneficence. He does not always succeed; like all physicians, he sometimes makes clinical errors. Overall, he treats his patients to the best of his ability. The principle of beneficence in the healing professions is the obligation to care for patients to the best of one’s ability.
Nonmaleficence is the duty of health care providers to do no harm.
Mrs. Lucy Knight suffers from insomnia and Parkinson’s disease. The insomnia does not bother her, because she likes to read at night, but it irritates her husband. Mr. Knight requests his wife’s physician to order strong sleeping pills for her, but the physician declines, saying that the combination of sleeping pills and Parkinson’s disease places Mrs. Knight at high risk for a serious fall.
The modern array of medical interventions has the capacity to do good or harm or both, thereby enmeshing the principle of nonmaleficence with the principle of beneficence. In the case of Mrs. Knight, the prescribing of sedatives has far more potential for harm than for good, particularly because Mrs. Knight does not see her insomnia as a problem.
Autonomy is the right of a person to choose and follow his or her own plan of life and action.
Mr. Winter is a frail 88-year-old found by Dr. James Washington, his family physician, to have colon cancer, which has spread to the liver. The cancer is causing no symptoms. An oncologist gives Mr. Winter the option of transfusions, parenteral nutrition, and surgery, followed by chemotherapy; or watchful waiting with palliative and hospice care when symptoms appear. Mr. Winter is terrified of hospitals and prefers to remain at home. He feels that he might live a comfortable couple of years before the cancer claims his life. After talking it over with Dr. Washington, he chooses the second option.
The principle of autonomy adds another consideration to the interrelated principles of beneficence and nonmaleficence. Would Mr. Winter enjoy a longer life by submitting himself to aggressive cancer therapy that does harm in order to do good? Or, does he sense that the harm may exceed the good? The balance of risks and benefits confronts each physician on a daily basis (Eddy, 1990). But the decision cannot be made solely by a risk–benefit analysis; the patient’s preference is a critical addition to the equation.
Autonomy is founded in the overall desire of most human beings to control their own destiny, to have choices in life, and to live in a society that places value on individual freedom. In medical ethics, autonomy refers to the right of competent adult patients to consent to or refuse treatment. While the physician has an obligation to respect the patient’s wishes, he or she also has a duty to fully inform the patient of the probable consequences of those wishes. For children and for adults unable to make medical decisions, a parent, guardian, other family member, or surrogate decision maker named in a legal document becomes the autonomous agent on behalf of the patient.
Justice refers to the ethical concept of treating everyone in a fair manner.
Joe, a white businessman in the suburbs, suffers crushing chest pain and within 5 minutes is taken to a nearby private emergency department, where he receives immediate coronary angioplasty and state-of-the-art treatment for a heart attack. Five miles away, in a poor neighborhood, Josephine, an African American woman, experiences severe chest pain, calls 911, waits 25 minutes for help to arrive, and is brought to a public hospital whose emergency department staff is attending to five other acutely ill patients. Before receiving appropriate attention, she suffers an arrhythmia and dies.
The principle of justice as applied to medical ethics is newer, more controversial, and harder to define than the principles of beneficence, nonmaleficence, and autonomy. In a general sense, people are treated justly when they receive what they deserve. It is unjust not to grant a medical degree to someone who completes medical school and passes all the necessary examinations. It is unjust to punish a person who did not commit a crime. In another meaning, justice refers to universal rights: to receive enough to eat, to be afforded shelter, to have access to basic medical care and education, and to be able to speak freely. If these rights are denied, justice has been violated. In yet another version, justice connotes equal opportunity: All people should have an equal chance to realize their human potential. Justice might be linked to the golden rule: Treat others as you would want others to treat you. While there is no clear agreement on the precise meaning of justice, most people would agree that the differential treatment of Joe and Josephine is unjust.
In exploring the concept of justice, one area of concern is the allocation of benefits and burdens in society. This realm of ethical thinking is called distributive justice, and it involves such questions as: Who receives what amount of wealth, of education, or of medical care? Who pays what amount of taxes?
The principle of justice is linked to the idea of fairness. In the arena of distributive justice, no agreement exists on what formula for allocating benefits and costs is fair. Should each person get an equal share? Should those who work harder receive more? Should the proper formula be “to each according to ability to pay,” as determined by a free market? Or “to each according to need?” In allocating costs, should each person pay an equal share or should those with greater wealth pay more? Most societies construct a mixture of these allocation formulas. Unemployment benefits consider effort (having had a job) and need (having lost the job). Welfare benefits are primarily based on need. Job promotions may be based on merit. Many goods are distributed according to ability to pay. Primary education in theory (but not always in practice) is founded on the belief that everyone should receive an equal share (Beauchamp and Childress, 2008; Jonsen et al, 2010).
How is the principle of distributive justice formulated for medical care? Throughout the history of the developed world, the concept that health care is a privilege that should be allocated according to ability to pay has competed with the idea that health care is a right and should be distributed according to need. In most developed nations, the allocation of health care according to need has become the dominant political belief, as demonstrated by the passage of universal or near-universal health insurance laws. In the United States, the failure of the 100-year battle to enact national health insurance, and the widely divergent opinions on the 2010 Affordable Care Act, attest to the ongoing debate between health care as a privilege and health care as a right (see Chapter 15).
If the overwhelming opinion in the developed world holds that health care should be allocated according to need, then all people should have equal access to a reasonable level of medical care without financial barriers (ie, people should have a right to health care). In this chapter, therefore, we consider that the principle of distributive justice requires all people to equally receive a reasonable level of medical services based on medical need without regard to ability to pay.
Ethical dilemmas (Lo, 2009) are situations in which a provider of medical care is forced to make a decision that violates one of the four principles of medical ethics in order to adhere to another of the principles. Ethical dilemmas always involve disputes in which both sides have an ethical underpinning to their position. Financial conflicts of interest on the part of physicians (see Chapters 4 and 10), in contrast, pit ethical behavior against individual gain and are not ethical dilemmas.
Anthony, a 22-year-old Jehovah’s Witness, is admitted to the intensive care unit for gastrointestinal bleeding. His blood pressure is 80/60 mm Hg, and in the past 4 hours, his hematocrit has fallen from 38% to 21%. The medical resident implores Anthony to accept life-saving transfusions, but he refuses, saying that his religion teaches him that death is preferable to receiving blood products. When the blood pressure reaches 60/20 mm Hg, the desperate resident decides to give the blood while Anthony is unconscious. The attending physician vetoes the plan, saying that the patient has the right to refuse treatment, even if an avoidable death is the outcome.
In Anthony’s case, the ethical dilemma is a conflict between beneficence and autonomy. Which principle has priority depends on the particular situation, and in this case, autonomy supersedes beneficence. If the patient were a child without sufficient knowledge or reasoning capability to make an informed choice, the physician would not be obligated to withhold transfusions, even if the family so demanded (Jonsen et al, 2010).
Pedro Navarro has lung cancer that has metastasized to his brain. No effective treatment is available, and Mr. Navarro is confused and unable to understand his medical condition. Mrs. Navarro demands that her husband undergo craniotomy to remove the tumor. The neurosurgeon refuses, arguing that the operation will do Mr. Navarro no good whatsoever and will cause him additional suffering.
The case of Mr. Navarro pits the principle of autonomy against the principle of nonmaleficence. Mr. Navarro’s rightful surrogate decision maker, his wife, wants a particular course of treatment, but the neurosurgeon knows that this treatment will cause Mr. Navarro considerable harm and do him no good. In this case, nonmaleficence triumphs. Whereas patient autonomy allows the right to refuse treatment, it does not include the right to demand a harmful or ineffectual treatment.
The traditional dilemmas described in many articles and books on medical ethics feature beneficence or nonmaleficence in conflict with autonomy. In two famous ethical dilemmas, the families of Karen Ann Quinlan and Nancy Cruzan, young women with severe brain damage (persistent vegetative state) asked that physicians discontinue a respirator (in the Quinlan case) and a feeding tube (in the Cruzan case). Both cases were adjudicated in the courts. The Quinlan decision promoted the right of patients or their surrogate decision makers to withdraw treatment, even if the treatment is necessary to sustain life. The outcome of the Cruzan case placed limits on autonomy by requiring that life-supporting treatment can be withdrawn only when a patient has stated his or her wishes clearly in advance (Annas, 2005).
In 2005, the case of Terri Schiavo, for 15 years in a persistent vegetative state similar to the situations of Karen Ann Quinlan and Nancy Cruzan, made national headlines. In spite of multiple decisions of state and federal courts—up to the Supreme Courts of Florida and the United States—supporting the right of Terri Schiavo’s husband to discontinue Ms. Schiavo’s feeding tube, the US Congress, encouraged by President George Bush, passed legislation reopening the option of reinserting the feeding tube. Eventually, based on the precedents of the Quinlan and Cruzan cases, the courts prevailed and Ms. Schiavo died (Annas, 2005).
Overall, medical ethics has moved in the direction of giving priority to the principle of autonomy over that of beneficence.
In the late twentieth century, a new generation of ethical dilemmas emerged, moving beyond the individual physician–patient relationship to involve the broader society. These social–ethical problems derive from the new reality that money may not be available to pay for a reasonable level of medical services for all people. When money and resources are bountiful, the issue of distributive justice refers to equality in medical care access and health outcomes (see Chapter 3). Is it fair that some people are unable to receive needed care because they lack money and insurance? When money and resources become scarce, the issue of justice takes on a new twist. Should limits be set on treatments given to people with high-cost medical needs, so that other people can receive basic services? If not, might health care consume so many resources that other social needs are sacrificed? If limits should be set, who decides these limits?
Angela and Amy Lakeberg [actual names] were Siamese twins sharing one heart. Without surgery, they would die shortly. With surgery, Amy would die and Angela’s chance of survival would be less than 1%. On August 20, 1993, a team of 18 physicians and nurses at Children’s Hospital of Philadelphia performed an all-day operation to separate the twins. Amy died. The cost of the treatment was $1 million. The Medicaid program covered $700 to $1000 per day, and the hospital underwrote the balance of the costs. On June 9, 1994, Angela died; she had spent her brief life on a respirator in the hospital.
The new fiscal reality has spawned two related dilemmas.
1. The first involves a conflict between the duty of the physician to follow the principles of beneficence and nonmaleficence and the growing sentiment that physicians should pay attention to issues of distributive justice. In the case of the Lakeberg twins, the hospital and the surgeons adhered strictly to the principle of beneficence: Even a remote chance of aiding one twin was seen as worthwhile. The hospital could have balked, arguing that its funding of the surgery would be unfairly shifted to other payers. The surgeons could have declined to operate on the grounds that the money spent on the Lakebergs could have been better used by patients with a greater chance of survival. But, the surgeons could argue, who can guarantee that the money saved would have gone to better use?
2. The second category of social–ethical dilemma is the conflict between the individual patient’s right to autonomy and society’s claim to distributive justice. In the Lakeberg case, individual autonomy won out. The Lakeberg parents could have decided that spending $1 million of society’s money on a less than 1% chance of saving one of two infants was excessive and could indirectly harm other patients. On the other hand, would not most parents have done what the Lakebergs did?
Physicians take up the practice of medicine with a recognition that they have a duty to help and not harm their patients. Individuals claim a right to health care and do not want others to restrict that care. Yet the principle of distributive justice (recognizing that resources for health care are limited and should be fairly allocated among the entire population) might lead to physicians denying legitimate services or patients setting aside rightful claims to treatment.
The basis for the principle of justice is the desire shared by many human beings to live in a civilized society. To live in a state of harmony, each person must balance the concerns of the individual with the needs of the larger community. There is no right or wrong answer to the question of whether the Lakeberg surgery should have been done, but the surgery must be seen as a choice. The $1 million spent on the twins might have been spent on immunizing 10,000 children, with greater overall benefit. When health resources are scarce, the principle of justice creates ethical dilemmas that touch many people beyond those involved in an individual physician–patient relationship. The imperatives of cost control have thrust the principle of justice to the forefront of health policy in the debate over rationing.
Dr. Everett Wall works in a health maintenance organization (HMO). Betty Ailes came to him with a headache and wanted a magnetic resonance imaging (MRI) scan. After a complete history and physical examination, Dr. Wall prescribed medication and denied the scan. Ms. Ailes wrote to the medical director, complaining that Dr. Wall was rationing services to her.
Perry Hiler arrives at Vacant Hospital with fever and severe cough. His chest x-ray shows an infiltrate near the hilum of the lung consistent with pneumonia or tumor. Since Mr. Hiler has no insurance, the emergency department physician sends him to the county hospital. At the time, Vacant Hospital has 35 empty beds and plenty of staff. When he recovers, Mr. Hiler calls the newspaper to complain. The next day, a headline appears: “Vacant Hospital Rations Care.”
Jim Delacour is a 50-year-old man with terminal cardiomyopathy. His physician sends him to a transplant center, where an evaluation concludes that he is an ideal candidate for a heart transplant. Because the number of transplant candidates is larger than the supply of donor hearts available, Mr. Delacour is placed on the waiting list. After waiting 6 weeks, he dies.
When the emergency department called, Dr. Marco Intensivo’s heart sank. The eight-bed intensive care unit is filled with extremely ill patients, all capable of full recovery if they survive their acute illnesses. He has worried all day about another patient needing intensive care: a 55-year-old with a heart attack complicated by unstable arrhythmias. Which one of the nine needy cases will not get intensive care? Dr. Intensivo needs to make a decision, and fast.
The general public and the media often view rationing as a limitation of medical care such that “not all care expected to be beneficial is provided to all patients” (Aaron and Schwartz, 1984). Such a view only partially explains the concept of rationing. More precisely, rationing means a conscious policy of equitably distributing needed resources that are in limited supply (Reagan, 1988) (Table 13–2). Under this definition, only the last two cases presented above can be considered rationing. In the first case, Dr. Wall did not feel that the MRI was a resource needed by Betty Ailes. In the second, Vacant Hospital’s refusal to care for Perry Hiler was simply a decision on the part of a private institution to place its financial well-being above a patient’s health; there was no scarcity of health care resources. In the heart transplant and intensive care unit cases, in contrast, donor hearts and intensive care unit beds were in fact scarce. For Mr. Delacour, the scarcity was nationwide and prolonged; for Dr. Intensivo, the scarcity was within a particular hospital at a particular time. In both cases, decisions had to be made regarding the allocation of those resources.
Table 13–2. Two definitions of rationing
During World War II, insufficient gasoline was available to both power the military machine and satisfy the demands of automobile owners in the United States. The government rationed gasoline, giving priority to the military, yet allowing each civilian to obtain a limited amount of fuel. In a rural area, there may be a shortage of health care providers; in an overcrowded urban public hospital, there may be an insufficient number of beds; in the transplant arena, donor organs are truly in short supply. These are cases of commodity scarcity, wherein specific items are in limited supply.
The United States is a nation with an adequate supply of hospital beds and physicians in most communities; commodity scarcity in health care is the exception (eg, scarcity of primary care resources is becoming a reality). But a different kind of health resource is becoming scarce, and that is money. Those who pay the bills are insistent that the flow of money into the health sector be restricted. Most discussions of health care rationing presume fiscal scarcity, not commodity scarcity.
In summary, rationing in medical care means the limitation of resources, including money, going to medical care such that not all care expected to be beneficial is provided to all patients, and the fair distribution of these limited resources.
While fiscal scarcity is the more common form of resource limitation, commodity scarcity provides an instructive example of the interaction of ethics and rationing.
Mr. George Olds is a 76-year-old nonsmoking retired business executive with end-stage heart failure. He has good pulmonary and renal function and is not diabetic; thus, he is medically a good candidate for a heart transplant. His life expectancy without a transplant is 1 month. He has a loving family, with the resources to pay the $300,000 cost of the procedure.
Mr. Matt Younger is a 46-year-old divorced man who is unemployed, having lost his job as an auto worker 3 years ago. He has a history of smoking and alcohol use. He suffers a heart attack, develops intractable heart failure, and will die within 1 month without a heart transplant. He has good pulmonary and renal function and is not diabetic, making him a good candidate for the procedure.
Mr. Olds and Mr. Younger are in the same hospital and cared for by the same cardiologist, who applied for donor hearts on behalf of both patients on the same day. The cardiologist receives a call that one donor heart—histocompatible with both patients—has become available. Who should receive it?
In 1951, the first kidney transplant was performed in Massachusetts. But it was in 1967, when Dr. Christiaan Barnard sewed a living heart into the chest of a person suffering end-stage cardiac disease, that modern medicine fully entered the age of transplantation. Since that time, thousands of people have been kept alive for many years by transplantation of the kidneys, hearts, lungs, and livers of their fellow human beings. In 2010, 17,000 kidney, 2300 heart, 6300 liver, and 1800 lung transplants were performed in the United States. Transplants are truly life saving in most cases. Seventy to eighty percent of patients receiving heart, liver, or kidney transplants survive at least 5 years after transplantation (http://optn.transplant.hrsa.gov).
Transplantation of organs is both a medical miracle and an ethical watershed. It has generated debate on such questions as these: When are people really dead (so that their organs can be harvested for use in transplantation)? What is the responsibility of the families of brain-dead people to allow their organs to be harvested? Who pays and who is paid for organ transplants? Who should receive organs that are in short supply? We will focus only on the last of these issues.
The number of persons on the national waiting list for organ transplants rose from 16,000 in 1988 to 111,000 in 2010, yet the pool of organ donors has been estimated at 14,500. Even if all potential donors became actual donors, the number of organs that could be harvested each year falls far short of the required number. Nineteen patients in the United States die each day awaiting organs (www.mayoclinic.org/transplant/organdonation.html).
Transplantation presents a classic case of commodity scarcity: There is insufficient supply to meet demand. Explicit rationing, which is a system that determines who gets organs and who does not, is inevitable. For heart, lung, and liver transplants, rationing is all or nothing: Those who receive organs may live, while those who do not will die.
Given the supply and demand imbalance, which potential transplant patients actually receive new organs? In the early 1980s, the major heart transplant center at Stanford University excluded people with “a history of alcoholism, job instability, antisocial behavior, or psychiatric illness,” and required transplant recipients to enjoy “a stable, rewarding family and/or vocational environment.” Stanford’s recipients had a better than 50% chance of surviving 5 years, signifying that acceptance or rejection from the program was a matter of life and death. The US Department of Health and Human Services was concerned about Stanford’s selection criteria, which favored those middle-class or wealthy people with satisfying jobs. Moreover, the $100,000 cost restricted heart transplants to those with insurance coverage or ability to pay out of pocket. Both the social and economic criteria for access to this life-saving surgery raised serious issues of distributive justice.
Following the passage of the National Organ Transplantation Act of 1984, the federal government designated the United Network for Organ Sharing (UNOS) as a national system for matching donated organs and potential recipients (www.unos.org). According to the Task Force on Organ Transplantation (1986), organ allocation should be governed by medical criteria, with the major factors being urgency of need and probability of success. The Task Force recommended that if two or more patients are equally good candidates for an organ according to the medical criteria, length of time on the waiting list is the fairest way to make the final selection. In 2006, the US Department of Health and Human Services issued updated guidelines and in 2007 the Medicare program promulgated conditions that hospitals with transplant programs needed to follow.
Overall, UNOS follows these recommendations, placing potential recipients of organ transplants on its computerized waiting list. Recipients are prioritized according to a point scale based on severity of illness, time on the waiting list, and probability of a successful outcome. A serious attempt has been made to allocate scarce organs on the basis of justice criteria. But haunting the ethics of the prioritization process is the issue of ability to pay. In 2008, the average kidney transplant cost $259,000 and liver transplant $534,400. Persons needing a transplant are often rejected if they lack health insurance coverage (Laurentine and Bramstedt, 2010).
During the 1980s, technologic advances in medicine combined with the rapid rise in health care costs led to the belief that medical care rationing was upon us. The ethical issues raised by organ transplantation have thereby become generalized to all medical care. However, great differences separate the case of organ transplants from that of medical care as a whole.
1. Medical care in general is not a scarce resource; in many geographic areas, facilities and personnel are overabundant.
2. Whereas a nationwide structure is in place to decide who will receive a transplant, no such structure exists for medical care as a whole.
Dr. Ernest, who works in a for-profit HMO, wants to do her part to keep medical costs down. She prescribes low-cost amoxicillin at 20 cents per capsule rather than ciprofloxacin, which is priced at $1.50 for each dose. She teaches back pain patients home exercises at no cost rather than sending them to physical therapy visits at $75 per session. At the end of each year, she enjoys calculating how many thousands of dollars she has saved compared with one of her colleagues, who ignores costs in making medical decisions. Because of her efforts and those of other cost-conscious physicians, the HMO’s pharmacy bill goes down, and HMO management is able to lay off one physical therapist, thereby raising its profit margin.
While Dr. Ernest can be praised for attempting to reduce costs without sacrificing quality, her cost savings had no impact on overall national health care expenditures. Nor were the savings used to provide more childhood immunizations or to hire a physician assistant for a nearby rural community without any health care provider. In the United States, there is no structure within which to effect a trade-off between savings in one area and benefits in another. According to analyst Joshua Wiener. (1992)
In countries that have a socially determined health budget, cuts in one area can be justified on the grounds that the money will be spent on other, higher-priority services. This closed system of funding provides a moral underpinning for resource allocation across a range of potentially unlimited demands. In the United States, it is difficult to refuse additional resources for patients, because there is no certainty that the funds will be put to better use elsewhere (Wiener, 1992).
In the United States, persuading physicians to save money on one patient in order to improve services for someone else is as illogical as telling a child to eat all the food on the plate because children in Africa are starving (Cassel, 1985).
In order for health care providers like Dr. Ernest to make their cost savings socially useful, two things are needed: a closed system of health care funding, whether governmental through a global budget or private through a network of HMOs; and a decision-making structure controlling such funding that has the responsibility to allocate budgets to health care interventions in a fair manner.
For the purposes of the following discussion, let us assume that the United States is in a position of fiscal scarcity and that a mechanism exists to fairly allocate medical care resources from one individual or population group to another. Which ethical conflicts arise between beneficence, nonmaleficence, and autonomy on the one hand and justice (equitable distribution of resources) on the other?
Assume that Limittown, USA, has a fixed budget of $250 million for medical care in 2011. Limittown has three imaging centers, each with an MRI scanner that is used only 4 hours each weekday. None of the medical facilities perform bone marrow transplantation, a procedure that can markedly prolong the lives of some leukemia patients. In 2010, Limittown spent $5 million to pay for bone marrow transplants at a university hospital 50 miles away.
Limittown’s health commissioner projects that 2011 medical care expenditures will be $5 million over budget; she must implement cost savings. She considers two choices: (1) Two of the three MRI scanners could be closed, allowing the remaining scanner’s cost per procedure to be drastically reduced or (2) Limittown could stop paying for bone marrow transplantation for leukemia patients.
Is rationing the same as cost containment? We have defined rationing in medical care as the limitation of resources, including money, going to medical care such that not all care expected to be beneficial is provided to all patients, and the fair distribution of these limited resources. While the limitation of money going to medical care is cost containment, not all cost containment reduces beneficial care to patients. In the case of Limittown, both options for saving $5 million can be considered cost containment, but only denial of coverage for bone marrow transplants requires rationing. Consolidating MRI scanning at a single facility would allow the same number of scans to be performed but at a substantially lower cost. Rationing is associated with painful cost control (reducing effective medical care), but cost containment (see Chapter 8) can be either painful or painless (not reducing effective medical care) (Table 13–3). The extent of unnecessary care and administrative waste has led many health experts over the past three decades to conclude that the United States may not need to ration effective medical services (Brook and Lohr, 1986; Relman, 1990). Other health policy experts feel that rationing is likely to take place, and the issue is whether rationing is rational, based on the most effective medical interventions, or irrational, based on income or health insurance (Dranove, 2003). Whether or not rationing is needed today, advances in medical technology guarantee that rationing of medically efficacious services will be necessary in the future. But to maximize beneficence and autonomy without violating distributive justice, no rationing of beneficial services should take place until all wasteful practices are curtailed; painless cost control should precede painful cost control.
Table 13–3. Rationing and cost control
Lula Rogers is an 84-year-old diabetic woman with amputations of both legs; multiple strokes have rendered her unable to move, swallow, understand, or speak. She has been in a nursing home for three years during which time her medical condition has slowly deteriorated. Ms. Rogers’ son wishes to remove her feeding tube, but her physician and the nursing staff feel it is cruel to cause her death by malnutrition and dehydration. Ms. Rogers continues to live for 3 more years, costing $300,000.
A hotly debated issue is the amount of health care that should be provided to the profoundly and incurably ill. Were Lula Rogers’ caregivers right to prolong a life that had value to her? Or were they prolonging Ms. Rogers’ suffering and denying her a peaceful death? Should cost be a factor in such decisions, or should such matters of life and death be governed by autonomy, beneficence, and nonmaleficence alone (Luce, 1990)?
Twenty-seven percent of Medicare’s budget is spent on people in their last year of life, with almost half of those funds ($68 billion in 2009) spent in the final 60 days (Lubitz and Riley, 1993; Hogan et al, 2001) (Figure 13–1). In 2000, an estimated 67% of people who died had their last place of care in the hospital or nursing home; 33% died at home, with half of patients dying at home cared for by hospice programs (Teno et al, 2004). Patients in hospice programs have lower end-of-life costs than those not in hospice programs (Emanuel et al, 2002), and family members of patients receiving hospice care at home are more satisfied with the care than families of patients dying in hospitals or nursing homes (Teno et al, 2004). Thus, a strong possibility exists that reduced expenditures can go hand in hand with better care. If these savings could be transferred to more efficacious therapies for other people, then improving the care of the incurably ill could promote all four of the ethical principles—beneficence, nonmaleficence, autonomy, and distributive justice.
Figure 13–1. Medicare funds spent at the end of life.
We have seen that cost containment does not necessarily equal rationing and that eliminating administrative waste, medical waste, and unwanted interventions for the profoundly and incurably ill before rationing needed services best realizes the principles of beneficence and justice. However, if rationing of truly beneficial services is needed, the issues become even more difficult. If a health care system or program must compromise beneficence because of true fiscal scarcity, how can this compromise be made in a manner that yields the least harm and allocates the harm in the fairest possible way? In 2009 and 2010, federal legislation created a new structure for performing research on the comparative effectiveness of medical interventions (Benner et al, 2010)
Joy Fortune develops Hodgkin’s disease, or cancer of the lymphatic system; she receives radiation therapy and is cured. Jessica Turner is moribund from advanced metastatic cancer of the pancreas. She undergoes chemotherapy and dies within 3 days.
In the event of rationing, science is the best guide: The providing or withholding of care is ideally determined by the probability that the treatment will maximize benefits and minimize harm, that is by the criterion of medical effectiveness. Radiation therapy can often cure Hodgkin’s disease, but chemotherapy is unlikely to provide much benefit to people with very advanced pancreatic cancer. If rationing is needed and only one of these therapies can be offered, a decision based on the criterion of medical effectiveness would allow for the treatment of Hodgkin’s disease but not of terminal pancreatic cancer.
If intervention A increases person-years of reasonable-quality life more than intervention B, intervention A is more medically effective. The cost of the two interventions is not considered. Cost-effectiveness adds dollars to the equation: If intervention A increases person-years of reasonable-quality life per dollar spent more than intervention B, it is more cost-effective. Which is a better standard for rationing medical care: medical effectiveness or cost-effectiveness?
If money were not scarce, medical effectiveness (maximizing benefit and minimizing harm) would be the ideal standard upon which to ration care (ie, the less effective the therapy, the lower its priority on the list of treatments to be offered). But if money were not scarce, we would not need to ration. It is unrealistic to pretend that costs can be ignored (Garber and Sox, 2010). Suppose that bone marrow transplantation saves as many person-years of life by treating advanced cancers as does doxycycline by curing pneumonia. The former costs $150,000, while the latter can be obtained for $10. There is no reason to ration doxycycline, as its cost is negligible, whereas to make bone marrow transplantation similarly accessible is costly. Thus consideration of costs is essential as a means of deciding which services to ration.
Mrs. Smith’s breast cancer has spread to her liver and bone. She has been told that her only slim hope lies in high-dose chemotherapy with autologous bone marrow transplantation (HDC-ABMT), costing $250,000. Even with the optimistic assumption that HDC-ABMT has a 5% cure rate, screening mammography is eight times as cost-effective as HDC-ABMT in person-years of life saved.
In 1991, Dr. David Eddy (1991a) published a compelling article entitled “The individual vs society: Is there a conflict?” Dr. Eddy poses the preceding case of Mrs. Smith. If medical care must be rationed, it seems logical to spend funds on mammography rather than HDC-ABMT because the former intervention is more cost-effective. Dr. Eddy does not confine his analysis to cost-effectiveness, however, but moves on to the ethical issues.
Each of us can be in two positions when we make judgments about the value of different health care activities. We are in one position when we are healthy, contemplating diseases we might get, and writing out checks for taxes and insurance premiums. Call this the “first position.” We are in a different position when we actually have a disease, are sitting in a physician’s office, and have already paid our taxes and premiums (the “second position”). … Imagine that you are a 50-year-old woman employed by Mrs. Smith’s corporation. … [The company] is considering two options: (1) cover screening for breast cancer. … or (2) cover HDCABMT. … Now imagine you are in the first position. … as long as you do not yet have the disease (the first position), option 1 will always deliver greater benefit at lower cost than option 2. … Now, let us switch you to the second position. Imagine that you already have breast cancer and have just been told that it has metastasized and is terminal. … The value to you of the screening option has plummeted because you already have breast cancer and can no longer benefit from screening. …
Maximizing care for individual patients attempts to maximize care for individuals when they are in the second position. Maximizing care for society expands the scope of concern to include individuals when they are in the first position. As this example illustrates, the program that delivers the most benefit for the least cost for society (option 1) is not necessarily best for the individual patient (option 2), and vice versa. But as this example also illustrates, individual patients and society are not distinct entities. Rather, they represent the different positions that each of us will be in at various times in our lives. When we serve ourselves in the second position, we can harm ourselves in the first. (Eddy, 1991a)
Physicians generally care for patients in Dr. Eddy’s second position—when they are sick. But if the cost of treating those in the second position reduces resources available to prevent illness for the far larger number of people in the first position (who may not be seeing physicians because they feel fine), the individual principles of beneficence and autonomy are superseding the societal principle of justice. One could even say that choosing for individuals in the second position violates beneficence for those in the first position. On the other hand, if all resources go to those in the first position (eg, to cost-effective screening rather than highly technical treatment for those with life-threatening disease), injustice is committed in the other direction by ignoring the costly needs of the very ill.
Clearly, no ideal method of rationing medical care exists. The use of cost-effectiveness as a measuring stick raises ethical problems, and because of the difficulty in determining the cost-effectiveness of different interventions, has scientific limitations. All efforts should be made to control costs painlessly before resorting to the painful limitation of effective medical care. But if rationing is inevitable, a balance must be struck among many legitimate needs: The concerns of healthy people for illness prevention, the imperative for acutely sick people to obtain diagnosis and treatment, and the obligation to provide care and comfort to those with untreatable chronic illness.
The previous discussion of rationing medical care nationwide presumes a mechanism that redirects savings from interventions not performed toward more cost-effective services. In fact, such a mechanism does not exist nationwide. Only in specific medical care programs do we find a decision-making apparatus for allocating expenditures. One example is the Oregon Health Plan (Bodenheimer, 1997).
In 1994, Oregon added 100,000 poor uninsured Oregonians to its Medicaid program. To control costs, a prioritized list of services was developed, and the state legislature decided how many services would be covered. The prioritized list was based on how much improvement in quantity and quality of life the treatment was likely to produce. The final list contained 745 condition–treatment pairs, and the State of Oregon paid for items above line 574 on the list; conditions below that line were not covered (Kilborn, 1999). What are some of the Oregon Health Plan’s ethical implications?
1. The plan was more than a rationing proposal; its chief feature was to extend health care coverage to 100,000 more people. That aspect of the Oregon plan promotes the principle of justice.
2. Another positive feature of the plan was its attempt to prioritize medical care services on the basis of effectiveness, which, if rationing is needed, is a reasonable method for deciding which services to eliminate.
Other features of the Oregon plan must be viewed as negatively impacting distributive justice, or equal access to care without regard for ability to pay.
1. In 1996, 12% of beneficiaries reported being denied services because they were below the line on the priority list. Of those, 78% reported that the denial had worsened their health (Mitchell and Bentley, 2000). Medical services were rationed for Oregon’s poor but not for anyone else.
2. The plan targeted beneficial medical services in a state with considerable medical waste. In 1988, many areas of Oregon had average hospital occupancy rates below 50%. The closing of unneeded hospital beds could have saved $50 million per year, enough to pay for some of the treatments eliminated in the plan (Fisher et al, 1992). Oregon did not exhaust its options for painless cost control before proceeding to potentially painful rationing.
By 2004, the Oregon Health Plan had unraveled (Oberlander, 2006). The state entered a period of budgetary woes, new premiums and copays were instituted, and Oregon Health Plan enrollees responded by dropping out of the program. The rate of uninsurance climbed from 11% to 17%. But the bold experiment in rational rationing remains alive in the minds of health care policymakers.
Ms. Wilson is a 71-year-old woman with a recently diagnosed lung cancer. Obstructing a bronchus, the tumor causes pneumonia, and Ms. Wilson is admitted to the hospital in her rural town. She deteriorates and becomes comatose, requiring a respirator. By the eighth hospital day, she is no better. On that day, Louis Ford, a previously healthy 27-year-old, is brought to the hospital with a crushed chest and pneumothorax suffered in an automobile accident. Mr. Ford is in immediate need of a respirator. None of the six patients in the intensive care unit can be removed from respirators without dying; of the six, Ms. Wilson has the poorest prognosis. She has no family. No other respirators exist within a 50-mile radius (Jonsen et al, 2010). Should Ms. Wilson be removed from the respirator in favor of Mr. Ford?
Resources may be scarce throughout an entire nation or within a small hospital. Macroallocation refers to the amount and distribution of resources within a society, whereas microallocation refers to resource constraints at the level of an individual physician or institution. Macroallocation decisions may be more significant, affecting thousands or millions of people. Microallocation choices can be more acute, bringing ethical dilemmas into stark, uncompromising focus and placing issues of resource allocation squarely in the lap of the practicing physician. The microallocation choice involving Ms. Wilson incorporates all four ethical principles, which must be weighed and acted on within minutes: (1) Beneficence: For whom? This ideal cannot be realized for both patients. (2) Nonmaleficence: If Ms. Wilson is removed from the respirator, harm is done to her, but the price of not harming her is great for Mr. Ford. (3) Autonomy: Withdrawal of therapy requires the consent of the patient or family, which is impossible in Ms. Wilson’s case. (4) Justice: Should resources be distributed on a first-come first-served basis or according to need?
These are tragic decisions. Many physicians would remove Ms. Wilson from the respirator and make all efforts to save Mr. Ford. The main consideration would be medical effectiveness: Ms. Wilson’s chance of living more than a few months is slim, while Mr. Ford could be cured and live for many decades.
Less stark but similar decisions face physicians on a daily basis. On a busy day, which patients get more of the physician’s time? In a public hospital with an MRI waiting list, when should a physician call the radiologist and argue for an urgent scan, thereby pushing other people down on the waiting list? Situations involving microallocation demonstrate why, in real life, health care professionals are forced to balance the interests of one patient against those of another and the interests of individuals against the imperatives of society.
Don Rich is a bank executive who receives his care through a New York City HMO. He develops angina pectoris, which remains stable for over a year. An exercise treadmill test suggests mild coronary artery disease. Although this evaluation indicates that Mr. Rich’s condition can be safely managed with medications, he asks his cardiologist to arrange a coronary angiogram with an angioplasty or coronary bypass if indicated. He is told that the HMO has finite resources for such procedures and limits their use to patients with unstable angina or highly abnormal treadmill tests, for whom the procedures are more efficacious. Mr. Rich flies to Texas, consults with a private cardiac surgeon, and receives a coronary angiogram at his own expense.
Most people in the United States believe that health care should be a right. But how much health care? If every person has a right to all beneficial health care, the nation may be unable to pay the bill or may be forced to limit other rights such as education or fire and police protection. One approach to this problem is to limit the health care right to a basic package of services. (In the case of Don Rich’s HMO, angiography for stable angina pectoris is not within the basic package.) Any services beyond the basics can be purchased by individuals who choose to spend their own money. This solution creates an ethical problem. If a service that does produce medical benefit is not included in the basic package or is denied by an insurance company medical director, that service becomes available only to those who can afford it. Where should society draw the line between a basic level of care that should be equally available to all, and “more than basic” services that may be purchased according to individual ability and willingness to pay (Eddy, 1991b)? Unless the basic package covers all beneficial health services, the principle of distributive justice, that all people equally receive a reasonable level of medical services without regard to ability to pay, will be compromised.
Yoshiko Takahashi’s first heart attack came at age 59. It was minor, and she felt well the next day. Then came the real shock: because of her high blood pressure, her private insurance policy considers disease of the cardiovascular system a preexisting condition and will not cover costs for complications of hypertension. She demands to go home to limit her hospital bill. Twelve hours later comes the second heart attack, which is severe. She is readmitted to intensive care and remains in the hospital for 8 more days. Because of persistent pain, she is a candidate for coronary angiography, which she refuses on account of the cost. When she purchased the insurance, Ms. Takahashi had not understood its terms because her English skills were poor.
Decisions by physicians encompass only one aspect of resource allocation; the payers of health care have great power in the distribution of medical care. The policies of the private insurance industry, which covers the largest number of people in the United States, raise important ethical issues. In the case of Yoshiko Takahashi, the insurance company, rather than her physicians, largely determined what kind of medical care she received.
Private insurance may be experience rated (see Chapter 2), with premiums costing more for people or groups with a higher risk of illness. Under the practice of experience rating, people who need health care the most (because they have a chronic illness) are less likely to be able to purchase affordable health insurance. Many people feel that private insurers violate the justice principle because those most in need of services have the least chance of gaining coverage for those services.
Health insurance executives, however, have a different view, believing that private health insurance is fair. An advertisement sponsored by the insurance industry argued,
If insurance companies didn’t put people into risk groups [experience rating], it would mean that low-risk people would be arbitrarily mixed in with high-risk people. … and [low-risk people] would have to pay higher rates. That would be unfair to everyone. (Light, 1992)
According to this notion, it is unfair to force one person or group to pay for the needs or burdens of another. An alternative view, citing the principle of distributive justice, holds that young and healthy people should pay more in health costs than they use in health services so that older and less healthy people can receive health services at a reasonable cost. Even from the perspective of one’s own long-term self-interest, it makes sense to pay more for health care while young and healthy, and to pay less when advanced age creates a greater risk of becoming sick.
A much-discussed issue involves individuals whose behavior, particularly smoking, eating unhealthy diets, and drinking alcohol in excess, is seen as contributing to their ill health.
Jim Butts, a heavy smoker, develops emphysema and has multiple hospitalizations for respiratory failure, including many days on the respirator. Randy Schipp, a former shipyard worker, develops work-related asbestosis and has multiple hospitalizations for respiratory failure, including many days on the respirator. Should Jim pay more for health insurance than Randy?
Gene eats a low-fat diet, exercises regularly, but has a strong family history of heart disease; he suffers a heart attack at age 44. Mac eats fast food, does not exercise, and has a heart attack at age 44. Should Mac pay more for health care coverage than Gene?
One view holds that individuals who fall sick as a result of high-risk behavior such as smoking, substance abuse including use of alcohol, and consumption of high-fat foods are entirely responsible for their behavior and should pay higher health insurance premiums. Opponents of this idea see it as “blaming the victim” and argue that high-risk behaviors have a complex causation that may involve genetic and environmental factors including uncontrollable addiction. They cite a number of facts to support their position. The food industry spends billions of dollars each year on television advertising; the average child sees thousands of food commercials each year, most of them for products with poor nutritional value. The tobacco industry heavily advertises to teenagers. Illegal drug use is associated with poverty, hopelessness, and easy availability of drugs. Some evidence finds a genetic predisposition to alcoholism. To the extent that individuals are not entirely at fault for their high-risk behavior, it would be unfair to charge them more for health insurance. On the other hand, it seems sensible that users of tobacco and alcohol pay through taxes on those products.
The predicament of limited resources has been likened to a herd of cattle grazing on a common pasture. The total grazing area may be regarded as the entirety of economic resources in the United States. A smaller pasture, the medical commons, comprises that portion of the grazing area dedicated to health care. The herd represents the nation’s physicians, using the resources of the commons in the process of providing care to patients. Physicians, guided by medicine’s moral imperative to “do everything possible for the patient,” continually attempt to extend the borders of the medical commons. But communities outside the medical commons have legitimate claims to societal resources and view the herd as encroaching on resources needed for other pursuits (Grumbach and Bodenheimer, 1990).
Who decides the magnitude of the medical commons, that is, the resources devoted to health care? Physicians and other health care providers, whose interventions on behalf of their patients add up to the totality of medical resources used? The sum of individual consumer choices operating through a free market? Health insurance plans, watching over their particular piece of the commons? Or government, using the political process to set budgetary limits on the entire health care system?
Traditionally, physicians and patients have had a great deal to say about the size of the medical commons. In the United States, the medical commons traditionally has been an open range. The quantity and price of medical visits, hospital days, surgeries, diagnostic studies, pharmaceuticals, and other such interventions determine the total costs of medical care. This is not the case in other nations, where government health care budgets constitute a “fence” around the medical commons, setting a clear limit on the quantity of resources available. Some advocates of fence-building in the United States have considered parceling the medical commons into numerous sub-pastures, each representing an HMO or Accountable Care Organization (see Chapter 6) working within the constraints of fixed, prepaid budgets. Not all pastures would be equal in size, and the fences would have holes that allow patients to purchase additional services outside of the organized systems of care.
Ethical considerations play a role in both open and closed medical care systems. In the US open range, the principles of beneficence and autonomy have the upper hand, tending toward an expanding, though not equitable, system. Fenced-in systems, in contrast, balance the more expansive principles of beneficence and autonomy with the demands of distributive justice in order to allocate resources within the medical commons.
If the United States moves toward a more fenced-in medical commons, decisions will be needed about who gets what. Do all 90-year-old people with multiple organ failure receive kidney dialysis that may extend their lives only a few months? Are very low-birth-weight infants afforded neonatal intensive care even with a small chance of leading a normal life? Do individual physicians, interacting with their patients, have the final say in making these decisions? Should societal bodies such as government, commissions of interested parties, or professional associations set the rules?
Microallocation issues come down to daily clinical decisions about which individual patients will receive what types of care (Lo, 2009). Physicians and other caregivers may well recoil from the prospect of “bedside rationing,” believing that allocative decision making unduly compromises their commitment to the principles of beneficence and autonomy. Levinsky (1984) has argued that physicians must maintain their single-mindedness in maximizing care for each patient:
There is increasing pressure on doctors to serve two masters. Physicians in practice are being enjoined to consider society’s needs as well as each patient’s needs in deciding what type and amount of medical care to deliver. … When practicing medicine, doctors cannot serve two masters. It is to the advantage both of our society and of the individuals it comprises that physicians retain their historic single-mindedness. The doctor’s master must be the patient. (Levinsky, 1984)
Yet if physicians abstain from the arena of macroal-location decision making, who is to decide? Currently, these decisions are often made by medical directors of private insurance companies and the leaders of the Medicare and Medicaid programs. Studies have documented that such decisions vary from plan to plan, and even within a single insurance plan, a medical director may make different decisions on different days for similar patients (Light, 1994). If physicians refuse to accept two masters, then medicine will be granting allocation decisions to insurance company and governmental officials. The physician of the twenty-first century will continue to face individual patient responsibilities but will find it difficult to escape the obligation to balance the wishes of individual patients against the larger needs of society (Cassel, 1985; Morreim, 1989).
If physicians are to serve two masters (ie, to maintain their dedication to individual patients while at the same time responsibly managing resources), they need rules to assist them. These rules should operate at both a population and an individual level. At the population level, society should ideally decide which general treatments are to be collectively paid for through the process of universal health insurance. At the individual level, rules are needed to guide decisions about the prioritization of resources for specific patients. The workings of organ transplantation provide a model of how physicians can serve two masters: They do everything possible to procure organs for their transplant patients, but also accept the rules of the system that attempt to allocate organs in a fair manner (Benjamin et al, 1994). The modern health care professional is caught in a global ethical dilemma. On the one hand, patients and their families expect the best that modern technology can offer, paid for through private or public insurance. The imperatives of beneficence, nonmaleficence, and autonomy rule the bedside. On the other hand, grave injustices take place on a daily basis: An uninsured young person with a curable illness is unable to pay for care, while an insured, bedridden individual who had a stroke incurs vast medical bills during the last weeks of her ebbing life. Should not the physician at the stroke patient’s bedside be concerned about both patients? However this dilemma is resolved, the principle of justice will relentlessly peek at the physician from under the bed.
Aaron HJ, Schwartz WB. The Painful Prescription. Washington, DC: The Brookings Institution; 1984.
Annas GJ. “Culture of life” politics at the bedside—the case of Terri Schiavo. N Engl J Med. 2005;352:1710.
Beauchamp TL, Childress JF. Principles of Biomedical Ethics. 6th ed. New York: Oxford University Press; 2008.
Benjamin M et al. What transplantation can teach us about health care reform. N Engl J Med. 1994;330:858.
Benner JS et al. An evaluation of recent federal spending on comparative effectiveness research. Health Affairs. 2010;29:1768.
Bodenheimer T. The Oregon Health Plan: Lessons for the nation. N Engl J Med. 1997;337:651, 720.
Brook RH, Lohr KN. Will we need to ration effective health care? Issues Sci Technol. 1986;3:68.
Cassel CK. Doctors and allocation decisions: A new role in the new Medicare. J Health Polit Policy Law. 1985;10:549.
Dranove D. What’s Your Life Worth? Health Care Rationing… Who Lives? Who Dies? And Who Decides? Upper Saddle River, NJ: Prentice Hall; 2003.
Eddy DM. Comparing benefits and harms: The balance sheet. JAMA. 1990;263:2493.
Eddy DM. The individual vs society: Is there a conflict? JAMA. 1991a;265:1446.
Eddy DM. What care is “essential?” What services are “basic?” JAMA. 1991b;265:782.
Emanuel EJ et al. Managed care, hospice use, site of death, and medical expenditures in the last year of life. Arch Intern Med. 2002;162:1722.
Fisher ES, Welch HG, Wennberg JE. Prioritizing Oregon’s hospital resources. JAMA. 1992;267:1925.
Garber AM, Sox HC. The role of costs in comparative effectiveness research. Health Aff. 2010;29:1805.
Grumbach K, Bodenheimer T. Reins or fences: A physician’s view of cost containment. Health Aff. 1990;9:120.
Hogan C et al. Medicare beneficiaries’ costs of care in the last year of life. Health Aff. 2001;20:188.
Jonsen AR et al. Clinical Ethics: A Practical Guide to Ethical Decisions in Clinical Medicine. 7th ed. New York, NY: McGraw-Hill; 2010.
Kilborn PT. Oregon falters on a new path to health care. New York Times. January 3, 1999.
Laurentine KA, Bramstedt KA. Too poor for transplant: Finance and insurance issues in transplant ethics. Prog Transplant. 2010;20:178.
Levinsky NG. The doctor’s master. N Engl J Med. 1984;311:1573.
Light DW. The practice and ethics of risk-rated health insurance. JAMA. 1992;267:2503.
Light DW. Life, death, and the insurance companies. N Engl J Med. 1994;330:498.
Lo B. Resolving Ethical Dilemmas. A Guide for Clinicians. 4th ed. Baltimore, MD: Lippincott Williams & Wilkins; 2009.
Lubitz JD, Riley GF. Trends in Medicare payments in the last year of life. N Engl J Med. 1993;328:1092.
Luce JM. Ethical principles in critical care. JAMA. 1990;263:696.
Mitchell JB, Bentley F. Impact of Oregon’s priority list on Medicaid beneficiaries. Med Care Res Rev. 2000;57:216.
Morreim EH. Fiscal scarcity and the inevitability of bedside budget balancing. Arch Intern Med. 1989;149:1012.
Oberlander J. Health reform interrupted: The unraveling of the Oregon Health Plan. Health Affairs. 2006:w96–w105.
Reagan MD. Health care rationing: What does it mean? N Engl J Med. 1988;319:1149.
Relman AS. Is rationing inevitable? N Engl J Med. 1990;322:1809.
Task Force on Organ Transplantation. Issues and Recommendations. Washington, DC: US Department of Health and Human Services; 1986.
Teno JM et al. Family perspectives on end-of-life care at the last place of care. JAMA. 2004;291:88.
Wiener JM. Rationing in America: overt and covert. In: Strosberg MA et al, eds. Rationing America’s Medical Care: The Oregon Plan and Beyond. Washington, DC: The Brookings Institution; 1992.