President Obama's health care plan touts comparative effectiveness research as a means to undergird effective decision-making and the delivery of care. On the face of it, few can take issue with any measures that would improve the quality of care. However, I advise that we proceed carefully. In offering this recommendation, I draw upon my professional experience as a physician specializing in palliative and end-of-life care, the research and reports of other clinicians, and also my personal experience on the receiving end of care. Requiring physicians to make decisions through narrowly prescribed quality filters has consequences. There is an art to medical practice that extends beyond hard science and numbers and often requires swift, sound decision-making in an environment of incomplete evidence. Sometimes no formula or calculation based on quantified evidence and data can determine the appropriate action. A definition of high-quality medical care is too narrow when it relies only on empirical evidence gathered by randomized controlled clinical trials. Harvard clinicians Jerome Groopman and Pamela Hartzband recently warned in the Wall Street Journal of the dangers of applying the quality mantra too quickly and aggressively. (1) They cited examples in which the overzealous use of quality metrics increased patient deaths, led to the intimidation of physicians, and may have made it more difficult for the sickest patients to see doctors. Furthermore, a recent study observed that pay-for-performance measures applied to orthopedic surgery actually bore no relationship to patient outcomes. (2) I raise another concern. How will we value clinical judgment and support the decisions of doctors who practice the art of medicine in the murky world of incomplete evidence and second-guessing? This challenge is illustrated by the example of my aunt, who has moderately severe Alzheimer disease. She has lived in a family rest home with three other elderly, cognitively impaired women for many years. She and the other patients in this home live socially with a family, which includes small children who call her Granny. She can dress and feed herself, and she often leaves the home for supervised recreational activities and church. When we visit her, she is clean, well groomed, seems happy, and sometimes remembers who we are. About five years ago, at the age of ninety-six, she was rushed to the emergency room of a local hospital with fever, severe abdominal pain, and vomiting. The surgeon informed us that she had an inflamed gall bladder and was in danger of dying from infection. …