With the development of more effective treatments, cancer has become curable for some, and for others, a chronic, progressive illness that people live with and, with their health care providers, manage over time. One consequence of this change is that physicians and patients must communicate with each other in ways that previously were unimportant. Communication research has focused largely on decisionmaking at the end of life, in particular, on the single issue of a “do-not-resuscitate” (DNR) decision. As important as that is, even more important for many patients is a decision to stop chemotherapy or other active treatment, but this decision has yet to be fully studied.
Patient preferences have an important role in shared medical decision-making. Published guidelines regarding end-of-life care strongly endorse a patient’s right to participate in health care decisions (Teno et al., 2001a). For example, the American Society of Clinical Oncology (ASCO) calls for physicians to speak truthfully to cancer patients and families about prognosis, treatment options, and advance care planning (ASCO, 1998). Despite widespread endorsement by professional guidelines, the Patient Self-Determination Act, and court rulings, there is still significant concern about whether patients’ preferences are honored along with persistent claims that they are “trumped” by physicians. The evidence to support this claim is scant and derives in large part from misinterpretations of SUPPORT results and studies that asked for the perceptions of nurses and physicians in training (e.g., a report that one in two health care providers believe they had provided overly aggressive medical care to a dying person) (Solomon et al., 1993).
The SUPPORT results were widely reported and have had a lasting impact that does not necessarily represent their most accurate interpretation. US News and World Report headlines were, “Doctors Don’t Listen” and “…Doctors Don’t Talk About Bad News” —the implication being that physicians were ignoring individuals’ informed preferences. Half of the patients in SUPPORT with colon cancer who voiced a preference to avoid resuscitation did not have a DNR order (Haidet et al., 1998), but even so, these patients were not resuscitated against their preferences (Hakim et al., 1996). Similarly, a review of those deaths with an advance directive found only one case in which an advance directive was trumped at the request of the family (not by a physician) (Teno et al., 1998). Whether it was ethically defensible to delay the death of this unconscious patient so that his daughter could come to grips with the decision can be debated.
The larger area of concern is not that patient preferences are being ignored but rather questions regarding the timing of communication and interpretation of the intended meaning of “hopelessly ill.” In a qualitative