The Limits of Guidelines

February 15, 2010

Jerome Groopman published a provocative and thoughtful essay In the February 11, 2010 New York Review of Books  about the way one determines “quality health care.” Groopman’s focus is on clinical guidelines and just how prescriptive they can be. He makes a fundamental distinction between guidelines that can be applied in a standardized way (e.g., how to clean a catheter to reduce the likelihood of infection) and guidelines that may need to be varied based on a variety of patient characteristics( e.g., diabetes sugar control). He is concerned that much of what has been discussed at the federal level in the context of health care reform misses this distinction and could well be applied by payers rigidly. For example, there are proposals being considered that would provide incentives to physicians and hospitals based on how rigorously guidelines are followed, independent of individual patient characteristics. His message is terribly important and often overlooked.

In our rush to embrace knowledge and science, we may over attribute “perfect knowledge” to those who analyze data and publish guidelines for care. As Groopman points out, if one looks back over time, guidelines have often changed when it comes to the best approaches to take for certain treatments. Science with regard to the human body is not immutable and results of research on outcomes of various clinical practices can be misinterpreted or misunderstood. Patients have often been confused by these changes, as recently witnessed with regard to guidelines on mammograms, pap smears, PSA tests and the like. Many of these guideline changes are in fact more nuanced than they seem when reported by the press. And all include recommendations to consider patient history and other factors in determining the right course for any particular patient. Indeed, guidelines are often a review of the extant data, intended to be useful to patients and clinicians in personal care decisions, not prescriptive “how to” manuals. That distinction is too easily forgotten by policy makers.

Does this mean we cannot use guidelines to make medical care decisions, interpret data, recommend best practices and evaluate quality of care? No – guidelines are useful and important components of quality improvement and evaluation. What is important, however, is to not over-value those guidelines and over-legislate their implementation. Medicine, like much in life, still needs judgment and understanding of the applicability of the general to the individual. It is essential for clinical decision-makers to be able to bring an understanding of the strength of the science and a depth of understanding of the research to implementation. Too little reliance on what we know in science certainly leads to potentially bad outcomes for patient care, but so can too much. Policy makers beware: in the search for strategies to reduce health care spending, Groopman has something very important to say.