Telling the Truth About Terminal Cancer

Cancer accounted for more than 23% of all deaths in 1997. Although much progress has been made in reducing cancer deaths, the inescapable truth is that most patients with recurrent cancers will die as a result of the malignancy. This is important for several reasons: (1) two thirds of all cancers are preventable,¹ so much of the morbidity and mortality are unnecessary; (2) the cost of cancer care continues to increase, from $35 billion in 1990² to an estimated $50 billion in 1996³ ; (3) some physicians and patients are probably not taking advantage of the progress made in treating metastatic cancer or its symptoms; and (4) far too many patients are fighting a battle they cannot win, and not taking advantage of good end-of-life care.

In this issue of JAMA, Weeks and colleagues⁴ delineate some reasons that patients fight cancer and some of the personal and financial consequences of that fight. In this substudy of the Study to Understand Prognoses for Outcomes and Risks of Treatments (SUPPORT),⁵ the authors examined the prognosis, survival, treatment, and complication patterns of patients hospitalized for lung or colorectal cancer. Physicians were reasonably good at predicting who was going to die and when. However, 82% of patients overestimated their survival, and 59% were decidedly overoptimistic. The choices and consequences of these patients were strikingly different. Those who thought they were going to live for 6 months were 2.6 times as likely to choose "aggressive" anticancer therapy instead of "palliative" or hospice care, which is designed to relieve symptoms. These overoptimistic patients did indeed live longer than those who had more realistic aspirations. However, patients who overestimated their survival and received "aggressive" antineoplastic therapy had exactly the same median survival as those who received palliative care and were 1.6 times as likely to have a hospital readmission, undergo attempted resuscitation, or die while receiving ventilatory support.

Although these findings do not mean that chemotherapy is not worthwhile, patients must know their prognoses and the benefits of treatment to make good choices. For instance, chemotherapy for non–small cell lung cancer improves median and long-term survival, relieves symptoms,⁶ and is recommended by clinical practice guidelines for patients who are ambulatory.^{7 - 8} Raby and colleagues⁹ have expressed concern that physicians are not aware of current evidence showing slow but steady improvement in lung cancer treatment. The percentage of oncologists who indicated that they would accept chemotherapy treatment for lung cancer increased from 17% in the 1980s to nearly 67% in 1997, but is still not 100%.¹⁰ The case for chemotherapy treatment for metastatic colorectal cancer is just as compelling with improved quality of life and survival benefit compared with palliative care alone,¹¹ especially if treatment is started while patients are still well.¹²

The report by Weeks et al⁴ does not explain why patients did not benefit from chemotherapy. Perhaps these patients were too ill or had received prior treatment. The American Society of Clinical Oncology's lung cancer clinical practice guidelines note that documented benefits are restricted to patients with good performance status and first-time treatment.⁷ However, even with such little benefit, some patients still choose aggressive chemotherapy and will not settle for less. Slevin et al¹³ found that patients with cancer required only a 10% chance of symptom relief to accept chemotherapy with substantial toxic effect; their physicians and nurses (who were not facing death) would require a more than 50% chance. Meystre et al¹⁴ found that hospice patients viewed "aggressive" treatments such as blood transfusion as ordinary care and wanted them even when hospice nurses would have forgone them. Davies et al¹⁵ showed that although 40% of patients with brain tumors had received little stabilization from radiation and had experienced decrements in intellectual function, they would still opt for treatment. Such apparently "irrational" decision making shows the difficulty experienced by the dying patient.

Why do physicians provide treatment that does not improve survival and has unfortunate consequences? The article by Weeks et al⁴ only suggests where to get the answer—at the bedside, where it is possible to see the degree to which realistic information is transmitted and the degree to which it is accepted. The situation with cancer-treatment decisions is much like discussions about do not resuscitate (DNR) orders. In SUPPORT, physicians most often chose not to engage in discussions about DNR orders.⁵ The complaint that such discussions are too time-consuming was partially debunked since such discussion can be held successfully in 16 minutes.¹⁶ A more likely explanation is that physicians find it too uncomfortable to tell patients that they cannot effectively fight their patients' cancer because it means that clinicians have failed.¹⁷ It is much easier simply to give another round of chemotherapy. Does this mean that clinicians must sit by the bedside, hold patients by the hands, and tell them that there is no way to control their cancer? The answer is yes, when it is the truth. In the same conversation, physicians should assure patients that they will not be abandoned but will be helped to live for as long and as well as they can.

Weeks and colleagues⁴ recommend that to achieve the goals of supporting patient values and minimizing futile therapy, physicians need to change what they tell patients about their prognoses and be sure that patients understand it. This is the foundation of informed consent and the ethical principle of respect for patient autonomy. The authors recommend that patients who know their prognoses can make choices about their care that are consistent with their wishes, but this is not enough. Physicians are obligated to initiate patient dialogue, ask what patients want to know, provide estimates of survival and the likely effect of therapy, and discuss all treatment options, including palliative care.

Telling patients the truth about their illness is legally complex. Informed consent requires that patients be told their diagnosis, their prognoses, and the risks and benefits of all proposed treatment options including nontreatment.¹⁸ Courts use 2 standards by which to evaluate the adequacy of informed consent: the reasonable-physician standard, which requires information usually given by other physicians in a similar circumstance, or the reasonable-patient standard, which requires information a reasonable patient needs in order to make an autonomous choice. Under either legal test, sufficiency of information is qualitative and subjective. As the majority of courts move away from the physician standard to the patient standard of disclosure, studies indicate physicians are more willing to share the complexities of information with their patients.¹⁹ However, physicians may not realize exactly the types of information patients need.

Patients do not always hear what physicians are telling them.²⁰ As demonstrated by Weeks et al,⁴ patients left on their own tend to be overoptimistic regarding their prognoses. Physicians sometimes withhold the truth from their patients to give patients more hope for the future.²¹ Patients often do not ask questions regarding their illness, and the physical and mental stress of a terminal illness impedes patient capacity for informed consent, causing otherwise rational people to make irrational choices.²² Patients in stress often defer the decision-making process either to their physician or to an immediate family member, and neither physicians nor families accurately predict the treatment the patient would have wanted.^{23 - 24} Irrational choices lead to further dissatisfaction with the medical system, causing increased stress, financial strain, and risk for malpractice claims.²⁵

Increased patient participation in medical decisions may have significant impact on physician attitudes and behaviors. In particular, physicians are sometimes conflicted regarding their ethical role of beneficence and their role to respect the right of patient autonomy. Physicians could be less inclined to use therapeutic privilege as a means of withholding information to promote hope for their patients,²⁶ which could be accomplished by allowing them to make treatment choices for reasonable goals. Physicians are also less likely to allow patients their right to waive access to vital information by requesting that the physician make the decision on behalf of the patient. That is not to say that physician values play a lesser role in informed consent. Although patients value autonomy, they also seek physician preference as part of their own decision-making process.²⁷ Patients still ask, "What should I do, doctor? What would you do?"

Truth about prognosis might have a significant effect on family members making decisions on behalf of patients with terminal cancer. The majority of states allow surrogates to make decisions on behalf of incapacitated patients, using patient values and preferences as the foundation for substituted consent. However, surrogate decision makers do not always do this. Surrogates may consent to far more aggressive treatment than they would want for themselves. Families often request that "everything be done" for the patient, thereby placing the physician in an ethical dilemma in defining "everything."²⁸ Accurate end-of-life information could change a surrogate's role from doing everything to doing the right thing.

For now, simply telling patients the truth about their terminal cancer is a significant step in the right direction. The next step should be to study, in actual practice, how much information is given, how much is received, and the effects on decision making for those at the end of life.