Nicholas A Christakis

University of Chicago Press, £19, pp 296 

ISBN 0 226 10470 2

Rating:

When I visit a teaching hospital and review a seriously ill patient's case with trainees, I often ask something like, "What do you imagine this lady's life will be like in a few months?" Until recently, I was regularly met with an answer like, "I'm glad that's not my job." As a long term care physician, I found this to be a distressing response. What could possibly be of more importance to the patient? Doctors have focused so much on fixing the physiological wreckage that we have lost our focus on the lives our patients lead.

Nicholas Christakis has compiled and interpreted the available evidence about how modern medicine deals with prognosis. Without some estimate of prognosis, he argues, doctors cannot provide an effective

service to patients whose lives will be shaped and curtailed by disease. His book reads well, and it brings together an array of data that he has analysed and interpreted. In describing our current practices, he shows us how practitioners in medicine avoid prognostication, convey prognosis badly, and often believe that giving patients an accurate prognosis can destroy hope and shorten their lives.

Death Foretold presents evidence that doctors routinely give overly optimistic prognoses to patients who may be considering entering a hospice. Future research efforts will need to confront difficult but illuminating methodological questions about such prognoses. For example, what do doctors really believe is at stake when they make a prognosis and refer patients to a hospice? Christakis has elsewhere proposed a split between "forecasting," which is kept to the doctor, and "foretelling," which is shared with patients. He has proposed that doctors might be more accurate to themselves than they are to patients.

Recently, I have encountered house staff who no longer simply deflect my questions about their patients' likely futures. More often now, I hear them wonder, "How would I answer that?" Giving patients an accurate prognosis is an important, difficult, and worthy challenge for research and practice.

Joanne Lynn, director. 

RAND Center to Improve Care of the Dying, Arlington, USA