PHILIP J. BOYLE AND DANIEL CALLAHAN
Moral Conflicts and
Is clinical medicine best thought of as a science or an art? 1 The question is an old one, now given a fresh life by the powerful movement to develop outcomes data and practice guidelines. The assumption behind the movement is that a combination of good data and well-crafted guidelines will save money and more effectively serve patient care. It is an assumption that is heavily biased toward construing medicine more as science than as art.
Such a bias does not always sit well with practicing physicians. Sinclair Lewis caught the flavor of physicians' unease in a conversation between two doctors in his now classic novel Arrowsmith, written in 1924:
Doctor, do you find you can do much with asthma? Well now, Doctor, just in confidence, I'm going to tell you something that may strike you as funny, but I believe foxes' lungs are just fine for asthma, and T.B. too. I told that to a Sioux City pulmonary specialist one time and he laughed at me—said it wasn't scientific . . . I said "But I get results, and that's what I'm looking for." ... I swear I believe most of these damn' alleged scientists could learn a whale of a lot from the plain country practitioners, let me tell you!
In a recent study carried out by The Hastings Center, we sought to study this time-worn issue in the context of the renewed effort to influence physicians' behavior by providing them with better science and helpful guidelines for their clinical practice. What is new, perhaps unique, about the issue and found in the outcomes enterprise is a shift from physician reliance on traditional biomedical reasoning to a reliance on probabilistic statistical reasoning. 2 We had noted that, for all of the technical skills being deployed to develop good data, something seemed missing. While some research has examined changing