Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
7 Ethical Issues and Value in Oncology ETHICAL ISSUES WHEN considering insurance COVERAGE BASED ON VALUE IN THE TREATMENT OF CANCER Dr. Daniel Brock of Harvard Medical Schoolâs Division of Medical Ethics began with his description of value in cancer care and treatment: âThat the benefits in life extension and improved quality of life are obtained at a reasonable cost comparable to other typically-funded treatments and at a reasonable cost per quality-adjusted life-year (QALY).â Many cancer drugs fail to meet this standard, which is in essence a cost-effectiveness standard, he said. How should resources be distributed to different health care needs and patients? The value answer to that question is based on cost-Âeffectiveness, meaning that health care resources should be allocated to maximize health benefits from available resources to the population served, Dr. Brock explained. One way to approach cost-effectiveness is to limit the cost allowed for a unit of benefit. While the National Institute of Health and Clinical Excellence (NICE) in Britain uses a particular threshold, the United States spends about twice as much per capita on health care as what the United Kingdom spends, so there is no reason to use Britainâs thresh- old. Roughly $100,000 per QALY is a threshold that many economists often apply, and it fits with the $6,000,000 to $7,000,000 for the value of a statistical life often used in health and safety regulations. Such a cost per 69
70 assessing and improving value in cancer care QALY cap represents one way to determine how to distribute health care resources, but many new cancer drugs cost two to three times this threshold of $100,000 per QALY. Causes of Poor Cost-Effectiveness and Value Among Cancer Treatments What are the causes of poor cost-effectiveness and value in cancer treatments? There are multiple causes, Dr. Brock said. First, the intellectual property protections and monopoly pricing afforded by 20 years of patent protection enables pharmaceutical manufacturers to set any price they want or believe they can get for a new drug. Second, though Medicare is the biggest purchaser of cancer drugs in this country, it is explicitly prohibited under part B from negotiating prices with drug companies. As a result of this prohibition, pharmaceutical manufacturers know that they will not be denied coverage by Medicare on grounds of cost. Most private insurance companies tend to follow Medicareâs coverage decisions, meaning that they do not apply any cost-effectiveness standard either. Therefore, manufactur- ers have no incentive to develop cost-effective drugs and no incentive to price drugs in a cost-effective manner. Third, most cancer patients have health insurance, meaning that they tend to be concerned only with their out-of-pocket costs. With out-of-pocket costs making up just a fraction of the total cost to the health care system of an insured patientâs care, there is little incentive for patients to reject care that is not cost-effective. Finally, some oncologists receive substantial income from using these new drugs for patients, and many physicians feel an obligation to do what is best for the patient no matter what the cost. All in all, said Dr. Brock, treatment decision makers (physicians and patients) have little incentive to weigh the true costs of care against the benefits, payors are largely precluded from negotiating for lower costs, and manufacturers have monopoly pricing to charge as much as they can get. Regarding monopoly pricing, the usual justification has been that it is necessary so manufacturers will invest in research and development. This is a sound justification, said Dr. Brock. For some new cancer drugs, however, a different justification has been offeredâthat the prices reflect the value of the benefits of these drugs. In the setting of monopoly pricing, it is impor- tant that cost-per-QALY standards are caps rather than a price justified in this way. By analogy, suppose you were drowning and someone were on a
ethical Issues AND Value in ONCOLOGY 71 dock ready to throw you a life ring. Can that person charge you a hundred thousand dollars or more for the life ring? Ethically speaking, the answer is clearly âno,â said Dr. Brock. To do so would be to exploit someoneâs des- perate circumstances. Therefore, the value benefit does not determine what price a service should have when one is not in a competitive market. Ethical Justifications for Spending That Is Not Cost-Effective After discussing these reasons for the failure of cost-effectiveness and value in cancer treatments, Dr. Brock moved to discuss ethical justifications for spending beyond what a cost-effectiveness threshold would dictate. Adhering to the adage that one can tell the justice of a society by how it treats its least well-off members, special concern or priority is often given to those who are worst off. Does this justify spending more on cancer patients who are near death? In other words, are those with the worst health (those nearest death) truly those in the most urgent need? Dr. Brock explained that he thought these were not the patients who were the most in need based on equity considerations. Assume for a moment, Dr. Brock said, that giving cancer drugs is equivalent to providing additional months or years of life. Who are the worst off with respect to months and years of life? They are those who will have had the least life if they are not treatedâthis is not the elderly cancer patient but the younger patients. Consider 65-year-old patient A who will die in 1 month without a liver transplant and 35-year- old patient B who will die in 4 months without a liver transplant. Patient A is clearly more urgently ill, like the elderly advanced cancer patient, but patient B is worse off in the relevant sense because patient B will have had many fewer years of life if he or she doesnât get the transplant. There- fore, the most urgent are not the worst off, said Dr. Brock, and therefore urgent, dying cancer patients do not deserve special priority for medical interventions. Dr. Brock asked how would one want to allocate the money for their health care over the course of a lifetime? One would not want to ask that question with only 3 months to live because that would bias the answer. From an unbiased point in their lives, it is probably not the case that people would want to allocate as much as $300,000 per QALY at the end of life. Put another way, should there be a special priority given to life extension over other health needs? The so-called rule of rescue coined by Al Jonsen (Jonsen, 1986) is often applied here. It states that we are unwilling to let an
72 assessing and improving value in cancer care identified person in peril die or suffer great harm when we could rescue that person, even if it may be very costly. In an advanced cancer case, however, many expensive new drugs do not really rescue dying patients but only provide a small chance of life extension. This is not the type of benefit that the rule of rescue involves. Another factor that influences resource allocation is the difference in how people think about real versus statistical lives. When saving statistical lives, by improving cancer screening for instance, one does not know and will likely never know whose lives will be saved. There are many cases in which improving rates of cancer screening would save more QALYs than many expensive chemotherapeutic treatments, but the benefits due to can- cer drugs receive more attention. While we may be more moved by lives we can identify, statistical lives that we cannot identify should still receive as much moral and ethical importance as real lives, Dr. Brock said. Similarly, one should not place less value on future health benefits compared to acute benefits just because they come in the future. Investing in screening rather than acute interventions may be less appealing because the benefits will not be apparent until years later, and they are discounted because they are not immediately apparent. Determining value by cost-effectiveness with stan- dard discounting of health benefits in this way gives undue importance to acute cancer interventions at the expense of prevention, and it gives undue weight to expensive, last-chance treatments. Briefly touching on coinsurance, Dr. Brock described new fourth-tier insurance policies that typically require co-pays of a percentage of the total cost, as opposed to a fixed cost. Often these co-pays are on the order of 20 percent, and for an advanced-stage cancer patient with a $100,000 annual cost that amounts to $20,000 per year. Many patients are not able to afford this. So, while it may discourage some low-value uses of health care, coin- surance does so unethically because it prevents those who cannot pay from receiving care but only discourages or inconveniences the same health care use among those who can. What Is Needed to Improve Value? Dr. Brock said, âThe most fundamental thing that is needed is willing- ness toâIâll use the R wordâration even last-chance care at the end of life. What is needed is a willingness not to cover very cost-marginal interven- tions, in particular at the end of life.â It may not come easily or soon, and it will require major cultural changes in the thinking of those in health care,
ethical Issues AND Value in ONCOLOGY 73 as well as the establishment of a national health care system to avoid the current variations in care. In conclusion, Dr. Brock suggested smaller steps to improve value in cancer care. First, covering last-chance therapies only in clinical trials or for registry patients would provide us with more evidence of effectiveness. Authorizing CMS to negotiate drug prices would help reduce costs. Finally, transforming the comparative effectiveness program under consideration in Congress into a cost-effectiveness program would be a much larger step in the right direction. CLINICIANâPATIENT COMMUNICATION ABOUT CANCER THERAPY: ETHICAL ISSUES Dr. Neil Wenger of the University of California at Los Angeles deÂ- scribed his understanding of value in cancer care as âthe net benefit of the treatment in terms of the goals of the patient, accounting for the negative effects of the treatment across all patients and considering the cost of the treatment. This includes the value of having a treatment, the opportunity costs of forgoing other treatments or of closure, and potential alternative uses of health care resources.â This description of value focuses on the out- comes for the resources invested, is predicated on conserving and preserving resources for other valued activities, and incorporates elements of fairness, Dr. Wenger explained. Moving on from this description of value, Dr. Wenger turned to an exploration of the clinical, ethical realities and challenges that affect value. He began with a clinical case (Box 7-1). To operationalize value in cancer care means selecting to use some treat- ments and not using others. This requires transparency, an appeal mecha- nism, and enforcement, Dr. Wenger said (Daniels and Sabin, 2002). It is also important to recognize that in order to compute the value in cancer care, we are generally dependent on outcome data from clinical trials. To the degree that cancer treatments are used outside of conditions estab- lished in their randomized controlled trials (RCTs), the cost-Âeffectiveness data generated by the RCTs do not apply. There is a tremendous pressure to rescue in clinical cancer care, as evident in the case (Box 7-1) and âinno- vativeâ or off-label use of chemotherapeutic agents is common. While the RCT of erlotinib clearly showed a 2-month survival benefit and a 9 per- cent response rate for stage IIIB/IV non-small cell lung cancer (NSCLC) patients compared to less than 1 percent in the control arm (Shepherd et
74 assessing and improving value in cancer care Box 7-1 A Clinical Case A 55-year-old woman is transferred to the intensive care unit (ICU) after complications following resection of brain metastasis from her non-small cell lung cancer. She is a lawyer. She married late in life and has a young daughter. She never smoked. The patient and her husband traveled 40 miles, bypassing numerous other large hospitals with cancer programs, to receive cancer care at a particular academic center. Twenty months ago, the patient was diagnosed with Âunresectable non-small cell lung cancer. She responded to six cycles of combina- tion chemotherapy and then received maintenance Â bevacizumab treatment. One year after presentation, a full-body PET (positron emission tomography) scan showed no tumor activity. Eighteen months after diagnosis, the patient was admitted to the hospital with altered mental status; an MRI scan at that time showed two frontal brain metastases. She underwent surgical resection of the brain metastases on hospital day 11 with post-op MRI imaging showing evidence of residual tumor. The patient had a complicated post-operative course including intracerebral bleeding, a low blood platelet count, and deteriorating mental status on day 16 in the hospital. The ICU team was concerned that the aggressiveness of her care might be inappropriate. On hos- pital days 28 and 32, she underwent neurosurgical procedures to drain cerebral fluid collections associated with the tumor. On hospital day 47, the patientâs respiratory function worsened due to pulmonary infection and the ICU team was concerned that she would require mechanical ventilation. The primary oncologist discussed the patientâs dire progno- sis with the patientâs husband, and he requested any available treatment. The primary oncologist explained that administration of Â erlotinib was an option, but he did not recommend this treat- ment, citing evidence that it provides only a 10 percent chance of response with an average 2-month life extension. The patientâs husband requested continued ICU care so that his wife might receiveâand perhaps benefit fromâerlotinib treatment. On hospital day 53, imaging showed further intracerebral tumor progression. On hospital day 58, the patientâs condition was stable enough to begin erlotinib treatment by nasogastric tube. On hospital day 62, she had worsening clinical status, hypertension, and renal failure. The relationship between the patientâs husband and the ICU team became increasingly strained. Discussions between the patientâs husband and the primary oncologist then turned to comfort care, and on hospital day 64 the care goal was changed to palliation. The patient was moved out of the ICU and died comfortably the following day. SOURCE: Wenger presentation, February 10, 2009.
ethical Issues AND Value in ONCOLOGY 75 al., 2005), this patient did not fit the inclusion criteria for this trial. The patientâs performance status (Oken et al., 1982) was far too low, and she had symptomatic brain metastases. While one might say the oncologist got the data from the RCT right, these data should not have been applied to this patient in this clinical situation; the $37,000 cost of the mean benefit shown in the erlotinib RCT (Carlson et al., 2008) is a tiny fraction of the expense required to maintain this patient to receive the medication. Physicians have a powerful motivation to rescue. Dr. Wenger quoted Al Jonsen (Jonsen, 1986) saying, âOur moral response to the eminence of death demands that we rescue the doomed,â and then added, âThis rescue morality spills into medical care where our ropes are artificial hearts. Should the rule of rescue set a limit to rational calculation of the efficacy of technology?â Cancer patients are willing to pay the priceâalthough they rarely pay all or even most of itâto achieve a personal rescue. Seriously ill patients with metastatic cancer are willing to accept a far greater burden for a small chance of benefit. In one study cancer patients were willing to withstand substantial side effects and risks for a 1 percent chance of cure, while physicians required a higher probability of benefit (10 percent), and nurses and the general public required a higher probability still (50 percent) (Agrawal and Emanuel, 2003). This tremendous demand for rescue at the end of life sets off a cascade of aggressive care, often without a discussion of prognosis or next steps for care or palliation in the event of clinical decline. When this decline arrives, the health care system meets it with intensive care, organ support, and, not infrequently, undignified suffering prior to the patientâs death, which leads to pathological bereavement, poor health care practitioner morale, and high costs. The virtue of a public decision-making mechanism to guide care in rescue situations would be its impartiality and balance of interests. Our individual intuition that we should rescue may not be the correct response from the societal perspective, Dr. Wenger said. Additionally, there is not a clear right that people have to be rescued. A question posed by Dr. Ramsey, Dr. Wenger said, is how willing are we to restrict access to marginally The erlotinib clinical trial inclusion criteria required patients 18 years or older; pathologic evidence of NSCLC; prior combination chemotherapy; the absence of other malignancies, cardiac disease, and gastrointestinal or ophthalmologic abnormalities; the absence of symptomatic brain metastases; and an Eastern Cooperative Oncology Group (ECOG) performance status measure between 1 and 3, meaning included patients would at least be capable of performing some self-care.
76 assessing and improving value in cancer care beneficial cancer therapies because they are too costly for what they do (Ramsey, 2007)? Dr. Wenger reframed the question to ask: To whom does this responsibility fallâinsurance carriers? Physicians? In a statement endorsed by multiple professional medical associations entitled Medical Professionalism in the New Millennium: A Physician Charter (Charter), the fundamental element of a physicianâs professional responsibil- ity is the principle of primacy of patient welfare. This principle is founded on dedication to serving the interests of the patient and an altruism that builds trust, which is central to the physicianâpatient relationship (ABIM Foundation et al., 2002). Market forces, societal pressures, and administra- tive exigencies are not to compromise this principle, Dr. Wenger read from the Charter. Following this principle of the primacy of patient welfare, the very next principle is the principle of social justice, Dr. Wenger said, which requires that physicians promote the fair distribution of finite clinical resources and provide health care based on wise, cost-effective management while meeting the needs of individual patients. The Charter continues, âThe provision of unnecessary services not only exposes oneâs patients to avoidable harm and expense, but also diminishes the resources available for othersâ (ABIM Foundation et al., 2002). So, this Charter indicates that at least some of the responsibility to limit costly, ineffective treatment falls to clinicians. Numerous studies both in the United States and internationally have shown that patients sometimes want cancer treatment that clinicians might not be willing to accept for themselves due to high toxicities and low probability of benefit at the end of life (Matsuyama et al., 2006). Further- more, while a patientâs prognostic estimate and their desire for aggressive, life-Âsustaining treatment both decrease as the end of life nears, patientsâ p Â rognostic estimates are often unrealistically optimistic (Weeks et al., 1998). What responsibility do clinicians bear for these optimistic patient progÂnostic estimates? Many factors play a role: (1) oncologists tend to be optimists themselves, (2) prognostication is difficult, and (3) empathic discussions of prognosis, adverse health outcomes, and costs are difficult. Such discussions of prognosis infrequently happen in a timely manner, even though the evidence suggests that patients want this information and guid- ance in cancer care decisions. Physicians are able to provide the guidance patients want in cancer care decisions, Dr. Wenger said. An article coauthored by Tom Smith (Harrington and Smith, 2008), Dr. Wenger recalled, describes ways to help orient patients toward care at the end of life and outlines the questions that physicians might have patients ask them in order to elicit all the informa-
ethical Issues AND Value in ONCOLOGY 77 tion that might be needed to guide care and use resources appropriately (Table 7-1). Gruen and colleagues have presented the ways that socioeconomic factors influence individual patientsâ health and their relation to the phy- sicianâs core professional responsibility for patient care (Gruen et al., 2004; Figure 7-1). They posit that the physicianâs responsibilities go beyond shep- herding individual patient care, extending to obligations to ensure access TABLE 7-1â Helpful Questions for Patients with Advanced Cancer to Consider Asking Their Oncologists Questions About Treatment Questions About Prognosis What is my chance of cure? What are the likely things that will happen to me? What is the chance that this chemotherapy will make my cancer How long will I live? (ask for a range and shrink? Stay stable? Grow? the most likely scenario for the period ahead) If I cannot be cured, will I live longer with chemotherapy? How much longer? Are there things I should be doing? â Making a will? What are the main side effects of the â Making an advance directive? chemotherapy? â Deciding upon a durable power of Will I feel better or worse? attorney for health care who can Are there other options, such as hospice speak for me, if I am unable? or palliative care? â Addressing financial or family legal issues? How do other people make these â Appointing a durable power of decisions? attorney for financial affairs? Are there clinical trials available? â Establishing a trust? â What are the benefits? Family issues â Am I eligible? â Will you help me talk with my â What is needed to enroll? children? Spiritual and psychological issues â Who is available to help me cope with this situation? Legacy and life review â What do I want to pass on to my family to tell them about my life? Other concerns? SOURCES: Wenger presentation, February 10, 2009; Harrington and Smith, 2008.
78 assessing and improving value in cancer care Direct Broad Global Individual Access socioeconomic socioeconomic health patient care to care influences influences influences Domains of professional obligation Domains of professional aspiration FIGURE 7-1â Model of physician responsibility in relation to influences on health. SOURCES: Wenger presentation, February 10, 2009; Gruen et al., 2004. R01506 to care and toFiguresocioeconomic forces with direct influence on health effect 5-1 (shaded inner vectors, fully editable ovals, Figure 7-1). According to this model, the oncologistâs stewardship role at the grabbed from journal source original community level should be to participate in deci- sions regarding the appropriate use of cancer resources, to work with health care teams in deciding reasonable options while maintaining a continuity role, to participate in setting limits, to participate in quality improvement efforts, and to aid in policy decisions regarding appropriate care within oneâs catchment area, Dr. Wenger said. Dr. Wenger then returned to the patient discussed in his presentation, a 55-year-old woman with brain metastases who received erlotinib in the ICU with little chance of benefit. What was really needed, Dr. Wenger said, was the presence of an oncologist who would fill those rolesâguiding care, setting limits, and participating in quality improvement efforts. Dr. Wenger recalled a discussion with the patientâs husband (Box 7-1) after the patientâs death. He said, I needed to hear from our oncologist that the treatment should stop, that it would not help; not that there was something that we could do or might do. Although I desperately wanted my wife to survive, she and I would even have selected not to have the brain surgery if it was unlikely to help. They needed to say, âYouâre dying. This is what you should do.â My daughter never got to be with her mommy in the end. I still donât know what to say to her.
ethical Issues AND Value in ONCOLOGY 79 In conclusion, Dr. Wenger said, the physicianâs role in guiding cancer care is far broader than selecting the right chemotherapeutic agentsâit is to inform patients and surrogates about the course, risks, and benefits of treatment in an iterative, ongoing fashion. It is to discuss prognosis, if the patient is willing, and to provide continuity of guidance. It is to work with care teams in deciding reasonable treatment options, and to participate in deciding appropriate use of cancer care resources on the policy level. This includes guiding patients to receive care that best meets their goals within the constraints of policy-level decisions, discussing cost, and disclosing potential conflicts of interest. What if, Dr. Wenger asked, oncologyâs role in guiding and providing care Americans need was to delineate fair and appropriate costs for cancer care resources at the policy level? What if oncologists took the lead on a social exploration of when resource-based rules might be suspended for rescue? What if oncologists as a group demanded that the structure of rou- tine care yield data to improve treatment and explain the shortcomings of current standard approaches? DISCUSSION Dr. Patricia Ganz of the University of California, Los Angeles, com- pared the pressure to prescribe aggressive chemotherapy to the pressure to prescribe antibiotics for viral illnesses. In both situations, the clinician feels that it is simpler to provide a prescription than to leave patients with none. She mentioned an innovative solution pediatricians have developed to address the pressure they felt to prescribe antibiotics: prescription pads for supportive care therapeutic interventions to treat viral illnesses. By anal- ogy, oncology societies including American Society of Clinical Oncology (ASCO) should play an important role in making sure that palliative care can be made available in a similar way as an alternative prescription to give patients, instead of chemotherapy with little chance for benefit. The hospice benefit should also be made available to patients starting their first course of chemotherapy for advance-stage malignancies so that oncologists are not forced to say âI have nothing else to give or do for youâ before handing the patient off to another team. Dr. Ferrell emphasized the difficulties faced by those who lacked health insurance or had gone bankrupt due to large co-pays. Dr. Ferrell recalled the story of a colleague who had been an oncology nurse for 30 years prior to being diagnosed with her third primary cancer when she lost her job and
80 assessing and improving value in cancer care insurance because of her illness. She had lost nearly everything she owned, and the cost of her chemotherapy was over $10,000 per month. During an outpatient visit to receive her chemotherapy, Dr. Ferrell said, the colleague got up out of bed, had an episode of syncope, and collapsed. Staff in the clinic called a code, anticipating the need to resuscitate her, and one staff member began to wheel the crash cart toward the collapsed woman. On the floor, the colleague woke up and saw the clinic team running toward her with the crash cart. When she realized what was going on, she immediately began to try desperately to prevent the clinic team from opening the crash cart, mouthing the words as clearly as she could while still nearly incapaci- tated on the ground. She did this because she knew that breaking the seal on the crash cart would cost her another $1,000 out-of-pocket. This awful scenario is what our system has produced today, Dr. Ferrell concluded. Dr. Rossi said that he thought everyone at the workshop could agree that trying to assess value and determine whether we are getting value for the money is critical for the health care system. He asked the speakers whether there was an ethical requirement upon pharmaceutical companies to enter markets where there are differential standards of care purely to gain access to markets where a product is already being used off label. Dr. Brock thought that clinical trials of pharmaceuticals ought to benefit the commu- nity they are performed in, but it is ethically problematic for companies to put themselves in positions where the only way they can satisfy the study populations is to have a successful outcome and validate the existing off- label use of a drug. Dr. Max Coppes of the Childrenâs National Medical Center asked whether legal constraints and the pressure to practice defensive medicine force oncologists to make decisions that lead to low value. Dr. Wenger said there was no doubt this was the case. There is constantly a tension between doing the right thing and doing the defensive thing in clinical practice. That is why there is a need for standards of care established by the profession. Dr. Brock added that the problem of defensive medicine is not driven just by concerns over malpractice, but it is reinforced by a medical culture that teaches trainees to run unnecessary tests and drive diagnostic uncertainty as close to zero as possible. Ms. Alison Smith of C-Change asked whether advance directives could be used as a vehicle to proactively assess a patientâs vision of value. Dr. Wenger replied that advance directives are tools to help in guiding care and should be used by oncologists. Dr. Brock cited the SUPPORT trial (SUPPORT Principal Investigators, 1995), adding that completing an
ethical Issues AND Value in ONCOLOGY 81 advance directive is not enough. In addition, physicians need to have more discussions about the end of life with their patients, and they need to adhere to the advance directives once they are written. REFERENCES ABIM Foundation, ACPâASIM Foundation, and European Federation of Internal Medicine. 2002. Medical professionalism in the new millennium: A physician charter. Annals of Internal Medicine 136(3):243â246. Agrawal, M., and E. J. Emanuel. 2003. Ethics of phase I oncology studies: Reexamining the arguments and data. Journal of the American Medical Association 290(8):1075â1082. Carlson, J. J., C. Reyes, N. Oestreicher, D. Lubeck, S. D. Ramsey, and D. L. Weenstra. 2008. Comparative clinical and economic outcomes of treatments for refractory non-small-cell lung cancer (NSCLC). Lung Cancer 61(3):405â415. Daniels, N., and J. E. Sabin. 2002. Setting limits fairly: Can we learn to share medical resources? New York: Oxford University Press. Gruen, R. L., S. D. Pearson, and T. A. Brennan. 2004. Physician-citizensâPublic roles and professional obligations. Journal of the American Medical Association 291(1):94â98. Harrington, S. E., and T. J. Smith. 2008 The role of chemotherapy at the end of life: âWhen is enough, enough?â Journal of the American Medical Association 299(22):2667â2678. Jonsen, A. R. 1986. Bentham in a box: Technology assessment and health care allocation. Law, Medicine, and Health Care 14(3â4):172â174. Matsuyama, R., S. Reddy, and T. J. Smith. 2006. Why do patients choose chemotherapy near the end of life? A review of the perspective of those facing death from cancer. Journal of Clinical Oncology 24(21):3490â3496. Oken, M. M., R. H. Creech, D. C. Tormey, J. Horton, T. E. Davis, E. T. McFadden, and P. P. Carbone. 1982. Toxicity and response criteria of the Eastern Cooperative Group. American Journal of Clinical Oncology 5(6):649â655. Ramsey, S. D. 2007. How should we pay the piper when heâs calling the tune? On the long-term affordability of cancer care in the United States. Journal of Clinical Oncology 25(2):175â179. Shepherd, F. A., J. R. Pereira, T. Ciulenu, E. H. Tan, V. Hirsh, S. Thongprasert, D. Campos, S. Maoleekoonpiroj, M. Smylie, R. Martins, M. van Kooten, M. Dediu, B. Findlay, D. Tu, D. Johnston, A. Bezjak, G. Clark, P. Santabarbara, L. Seymour, and National Cancer Institute of Canada Clinical Trials Group. 2005. Erlotinib in previously treated non-small-cell lung cancer. New England Journal of Medicine 353(2):123â132. SUPPORT Principal Investigators. 1995. A controlled trial to improve care for seriously ill hospitalized patients. The Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments (SUPPORT). Journal of the American Medical Association 274(20):1591â1598. Weeks, J. C., E. F. Cook, S. J. OâDay, L. M. Peterson, N. Wenger, D. Reding, F. E. Harrell, P. Kussin, N. V. Dawson, A. F. Connors, J. Lynn, and S. R. Phillips. 1998. Relationship between cancer patientsâ predications of prognosis and their treatment preferences. Journal of the American Medical Association 279(21):1709â1714.