Providing a prognosis for terminally ill patients

An internist, Toni Brayer, MD, had a recent entry entitled "Should the Doctor Say You Are Dying?" This cites the somewhat recent article in Journal of Clinical Oncology by Christopher Daugherty, MD, and Fay Hlubocky, MD, (they produce very interesting work, I remember seeing some of Dr. Hlubocky's work at ASCO last year). Dr. Brayer simplifies the conclusions a bit, in that they found that although 98% of oncologists who responded to a survey (n=716) did tell their patients about the terminal nature of their cancer, only 48% specifically offered a prognosis only when asked. This may sound alarming at first blush, but this fits with my practice as well. I think Dr. Brayer misses the point when she translates this survey study into a discussion about overtreating patients at the end of life.

I believe that a patient has the right to express a preference for knowing prognosis, and although most cancer patients do want to know their prognosis, not all do (see, for example: http://jco.ascopubs.org/cgi/content/abstract/19/7/2049). In my mind, if the patient knows they have terminal cancer (and a relative sense of their time remaining ... days, weeks, months, years), that would suffice for what they need to know to make informed health care decisions. In the end, prognostication is nothing more than an educated guess anyway, so I don't know that a more specific estimate is even very helpful. All of that being said, if a patient does ask for prognosis, I always give them my best guess — sometimes I am right, many times I am wrong. I have told families a specific prognosis independent of the patient, with the patient's blessing, as they may use that information to plan for caregiving or time away from work, for example. But reading Dr. Brayer's entry did make me feel defensive because I think she missed the nuanced and respectful approach that I believe most oncologists take with conversations about prognosis for terminal cancer.