Guidelines: Upstream Issues and
In response to concerns about cost containment and the quality of health care and to the belief that variation in medical practice may be due in significant part to uncertainty, there has been a great deal of effort to generate large volumes of outcomes data and produce practice guidelines on their basis.
"Outcomes research," or that part of it called "effectiveness research," carried out to discern "what works," involves examining large amounts of data about rates of various outcomes given various treatments. It typically involves statistical analyses of outcomes data drawn from very large databases (such as hospital and insurance records). Consensus panels and input from experts complement such information. "Practice guidelines" are generated on the basis of these data and expert opinions and serve as recommendations for practice.
There have been criticisms of both the outcomes research and practice guideline movement and of particular studies or recommendations, and there is perceived to be a resistance to, or at least a lack of adherence to, the recommendations. Claims that physicians often ignore or override practice guidelines or data arising from outcomes studies raise the following questions: If there is such resistance, why? In particular, do clinicians have good reasons for discounting outcomes data and practice guidelines? How should we respond to this nonadherence? Can we make clinicians more accepting of these guidelines, either by responding to their concerns with counterarguments or by addressing "upstream issues" by modifying the way the practice guidelines or outcomes assessments are generated? Or is the practice guideline project simply misguided and doomed to failure?
A useful place to begin the discussion is this: One can imagine a number of self-interested reasons (or motives) why physicians might criticize or decline to follow practice recommendations. For instance, guidelines may be seen as a threat to clinicians' professional autonomy and thus resisted, whether in an