The world of medicine is quite properly focused on finding and applying the best available evidence to processes of care. Unfortunately, too often this results in the use of inappropriate hard targets and mis-applied guidelines. National quality groups, payors, government agencies, professional societies and medical care organizations (including my own) are eagerly looking for ways to improve care by standardizing care around evidence-based best practices. While I celebrate the replacement of eminence with evidence, I wish I saw more humility. The evidence is imperfect and provisional, and the resulting targets and guidelines should be understood as a place to start discussing appropriate care, not a way to end discussion of what is appropriate for a given patient in a given setting. I call this Practicing Safe Guidelines.
Several times a month there is published information that reinforces this. Today, for example, I read in the November 18, 2013 issue of JAMA that the guidelines we have been using for a decade, advocating the use of beta-blockers in cardiac patients undergoing non-cardiac surgery, are wrong for the majority of the patients to whom they are being applied.
Because of the known benefits of beta-blockers for patients with acute ischemic heart disease in a surgical setting, and supplemented by scanty and contradictory evidence, guidelines for a decade have encouraged the use of beta-blockers for all patients with heart disease undergoing non-cardiac surgery. The JAMA study by Andersen, et. al. found that there was no benefit from beta-blocker use preoperatively in patients with ischemic heart disease in terms of either major adverse cardiovascular events (MACE) or all cause mortality. There was a small subset that clearly did benefit: those with heart failure or recent MI. Patients with stable CAD and without CHF or recent MI did not benefit and were at risk for serious complications (bradycardia and hypotension).
In short, the guidelines were wrong for the majority to whom they were applied, and a clinician who had looked at the evidence and used beta-blockers selectively rather than following the guidelines would have been delivering superior care - but would have been labeled as out-of-compliance and criticized for not following current best practices.
My conclusions:
- Guidelines can be wrong. In fact, if one waits long enough, it is almost guaranteed that they will be found to be wrong in some sets of the population to which they are applied.
- Guidelines frequently are applied to a large and heterogeneous group, within which only certain subsets benefit. For others the impact is either neutral or negative.
- Guidelines should be advisory or informative, not decisive. The strength of the evidence behind the guidelines, the medical applicability to a specific patient, and the individual patient's context (personal history, life style, values and preferences) must determine to what degree the guideline should impact each individual decision.
Practicing safe guidelines is problematic in three ways:
- It is harder. Doing what one is told by following a policy or procedure is always faster and easier than thinking things through.
- It is riskier. If one follows the guidelines and things go wrong, it is easy to say: "I followed established best practices. It isn't my fault that things didn't work out." (Of course, if the guidelines were wrong or not applicable, it is MY fault if I follow them. Following orders is no defense.)
- It is punished. Institutions and individual clinicians (including me) are now subject to financial penalties if we do not follow the guidelines.
My advice to clinicians: before you use a guideline in clinical practice, examine it carefully and consider its flaws, weaknesses and the arguments against it. When you use it, make sure that it is actually pertinent to the patient you are treating.
My advice to patients: when a clinician suggests a treatment based on a clinical guideline, help them follow my advice to clinicians by asking them what the strengths and weakness and counter-arguments to the guidelines are, and ask if the guidelines were based on studies of patients like you.