Scientific summaries of experience are essential to evidence-based medicine. Their development stimulates discussion and maintains competence. A valid summary that settles a medical question should become a practice guideline (that is, a "statement" to "assist decisions"). Summaries of insufficient or conflicting data, like studies with controversial results, should not become guidelines. Dr. Dans [1] suggests that guidelines should be "works in progress." I disagree. Summaries evolve continuously. Guidelines should change by replacement, as practice recommendations appropriately change. I also disagree that "the assumption has been that doctors (do) not need" summaries. Physicians have always integrated previous experience into ongoing practice. Scientific summaries and their guidelines are a simple way to make that professional process explicit and most consistent with epidemiologic principles. Rather than being "beacons into the still vast scientific unknown," they should summarize what is known.

I raise two basic issues. The first is the degree to which physicians can be professionals—acquiring, appraising, and appropriating experience to maintain the profession, to maintain and enhance individual competence, and to deliver quality care to each patient. The second issue is that science summarizes quantities, but there is more to a decision than scientific medicine because so many modes of valuation exist. We need to understand the values, attitudes, and preferences of physicians and patients.

I believe that these two fundamental issues account for the "ambivalence" about the current concept of guidelines. Guidelines are not about "a reasonable desire to ensure quality." They are an expression of distrust toward physicians, conflated with economic concerns. They will not solve existing problems with professionalism, competence, and economic allocation. The term "practice guideline" carries too much weight and should be circumscribed. Effort should be placed on professionalism and scientific summaries and on valuation and decision science.