Death Foretold: Prophecy and Prognosis in Medical Care (review)

Death Foretold: Prophecy and Prognosis in Medical Care. By Nicholas A. Christakis. Chicago: Univ. of Chicago Press, 1999. Pp. 296. $30.00.

"Why am I so weak, doctor?" "What can you do for me, doctor?" "Am I going to die?" These questions summarize three tasks expected of the physician. First task: diagnosis. "You are weak because you have lost blood from colon cancer." Second task: select and offer treatment. "We will perform surgery first, and because the cancer is also in the liver we will start chemotherapy." The third, and perhaps most difficult, is prognosis: "We'll talk."

Prognosis, like any other forecast, is fraught with error and subject to change due to many factors, including our chosen treatment. But prognosis should actually inform the tasks of diagnosis and treatment decisions. Our hypothetical patient with advanced cancer, asking the aforementioned questions, might view surgery and chemotherapy as a great idea if it meant five more years of quality life. But what if the prognosis is but a few months? Perhaps our patient would not want further treatments. Accurate prognosis is crucial to making wise decisions regarding medical care--especially when prognosis is poor.

So why is prognosis the difficult stepchild of the medical profession? Some answers are revealed in Nicholas Christakis' first book, Death Foretold: Prophecy and Prognosis in Medical Care. Christakis, who has devoted much of his young career to issues of prognosis and care of the dying, holds a dual appointment in the departments of Sociology and Medicine at the University of Chicago. He is also a practicing hospice physician (Horizon Hospice), applying his skills and knowledge in caring for terminally ill residents of the inner city.

Through a comprehensive review of the history of prognosis in medicine, he notes that our skill and interest in prognosis was greatest when we could do little to change the course of human illness. A sage physician at the turn of the century could give comfort and counsel to a family where a child lay dying from leukemia by explaining the diagnosis and what to expect. With no effective therapies available, he could spend more time considering prognosis. In contrast, a modern clinician would typically order the unpleasant procedure of a bone marrow examination, and then speak optimistically about the advances in leukemia treatment, offering little input on prognosis--this despite a significant mortality rate in childhood leukemias. With the ever-growing number of therapies, prognosis takes a back seat.

Prognosis is unavoidably a prediction about death. This is a disquieting subject for both patients and physicians. This discomfort means that the possibility of death is often glossed over, evaded, or--worse still--avoided altogether. Even when sincere attempts are made, the physician is likely to frame the prognosis in ways that suggest the doctor's expected outcome. This phenomenon is meticulously described by Christakis, a skilled and insightful writer and sociologist. His discussions about the meaning of the physician-patient relationship, the stakes at hand, and the biases developed by our culture, society, and its physicians are thoughtful and intriguing.

Through his own research Christakis grades physicians on their ability to accurately prognosticate. Not surprisingly, as a group we are bad at it. More alarmingly, we are bad at it in a biased, overly optimistic way. This bias, together with our need to maintain the "prophet-supplicant" pattern, may be a disservice to our patients, and lead to needless suffering.

Instead of helping a woman with metastatic breast cancer clearly understand her poor prognosis, we often take aggressive action: "The breast cancer has now spread to the liver, so we'll change the combination of chemotherapy to different agents. I have many patients who have responded to this regimen." This approach effectively prevents the patient from choosing a less aggressive option, such as hospice. This failure to perform the "third task" of a physician is at least as tragic as failure to diagnose correctly or...