Guidelines and killer B's
ACP Journal Club. 1999 July-Aug 131:A13-A14. doi:10.7326/ACPJC-1999-131-1-A13
Clinical practice guidelines are “systematically developed statements to assist practitioner and patient decisions about appropriate health care for specific clinical circumstances” (1). Guidelines require both general (external evidence) information and specific (local circumstances) information plus good judgment. Previous editorials have described how to apply your eye and nose to the external evidence on which a guideline is based: your eye to see if the guideline developers tracked down and critically appraised all the relevant evidence and your nose to smell whether they updated their review within the past year so that it is still fresh. This editorial suggests that you also apply your ear to the guideline's specific instructions to listen for any “killer B's.”
We submit that the local applicability of a guideline depends on the extent to which it is in harmony or conflict with 4 local (sometimes patient-specific) factors that might usefully be thought of as potential killer B's:
1. Is the burden of illness (frequency in our community or our patient's pretest probability or expected event rate [PEER]) too low to warrant implementation?
2. Are the beliefs of individual patients or communities about the value of the interventions or their consequences incompatible with the guideline?
3. Would the opportunity cost of implementing this guidelines constitute a bad bargain in the use of our energy or our community's resources?
4. Are the barriers (geographic, organizational, traditional, authoritarian, legal, or behavioral) so high that it is not worth trying to overcome them?
First, is the burden of illness too low to warrant implementation? Is the target disorder rare in our area (e.g., malaria in northern Canada)? Or is the outcome we hope to detect or prevent highly unlikely in our patient (e.g., the pretest probability for significant coronary stenosis in a young woman with noncoronary chest pain)? If so, implementing the guideline may not only be a waste of time and money, it might do more harm than good.
Second, are our patients' or community's beliefs about the values or utilities of the interventions themselves, or the benefits and harms they produce, compatible with the guideline's recommendations? The values assumed in a guideline, either explicitly or implicitly, may not match those in our patient or our community. Even if the values seem, on average, to be reasonable, we must avoid forcing them on individual patients. This is because patients who have identical risks may not have identical beliefs, values, and preferences as those used or assumed in the guideline, and some patients may be quite averse to undergoing the recommended procedures. For example, patients with early breast cancer who have identical risks, when given the same information about chemotherapy, make different treatment decisions based on how they weigh the long-term benefit of reducing the risk for recurrence against the short-term harm of being nauseated and losing their hair (2). Similarly, patients with severe angina who are at identical risk for coronary events, when given the same information about treatment options, exhibit sharply contrasting treatment preferences because of the different values they place on the risks and benefits of surgery (3). Although the average beliefs in a community are appropriate for deciding, for example, whether chemotherapy or surgery should be paid for with public funds, decisions for individual patients must reflect their personal beliefs and preferences.
Third, would the opportunity cost of implementing this guideline (rather than some other ones) constitute a bargain in the use of our energy or our community's resources? We need to remember that the cost of shortening the waiting list for surgery is lengthening the one for family therapy. As decision making of this sort is decentralized, different communities are bound to make different economic decisions, and “health care by postal code” becomes inevitable, especially with full democracy.
And finally, are there insurmountable barriers to implementing the guideline in our patient (who might flatly refuse the investigations or intervention) or in our community? Barriers can be geographic (if the required interventions are not available locally), organizational (one of us visited a hospital with its accident and emergency rooms in the basement, its coronary care unit 7 floors and 45 minutes away, and a regulation prohibiting thrombolysis in the former!), traditional (“But we've always done it the other way!”), authoritarian (“But you've always done it my way!”), legal (fear of litigation if a usual but useless practice is abandoned), or behavioral (when clinicians fail to apply the guideline or patients fail to take their medicine). If major barriers exist, the potential benefits of implementing a guideline may not be worth the effort and resources (or opportunity costs) required to overcome them. Changing our own, our colleagues', and our patients' behavior often requires much more than simply knowing what to do. If implementing a guideline requires changing behavior, we need to identify which barriers are operating and what we can do about them. The effects of strategies for helping both clinicians (4) and patients (5) modify their behaviors have been summarized in systematic reviews carried out within the Cochrane Collaboration and have appeared in ACP Journal Club and Evidence-Based Medicine.
So, in deciding whether a valid guideline is applicable to our patient, practice, hospital, or community, we need to identify the 4 B's that pertain to the guideline and then decide whether they can be reconciled with its application (or whether we are facing one or more killer B's).
None of these B's (even when present as killer B's) have any effect on the validity of the evidence component of the guideline. Further, the only persons who are “experts” in the B's are the patients and providers at the sharp edge of implementing the application component. Because the producers of the Cochrane reviews recognize that they can never know the local B's, they confine themselves to summarizing the evidence com-partments of the interventions they have studied.
Finally, how might we respond when we are asked to join a local guideline-development group? Based on the foregoing, it is doubly dumb for one clinician or a small group of local clinicians to try to create the evidence compartment of a guideline. Not only are we ill equipped and inadequately resourced for this task, by taking it on, we steal time and energy away from operationalizing our real area of expertise: knowing the local (indeed, sometimes patient-specific) burden, beliefs, bargains, and barriers that are vital to determining whether the guideline applies at all to our patient, practice, hospital, or community and if so how. This editorial therefore closes with the recommendation to front-line clinicians: When it comes to lending a hand with guideline development, work as a “B-keeper,” not as a meta-analyst.
Editors' Note: This editorial is derived from material in Sackett DL, Straus SD, Richardson WS, Rosenberg W, Haynes RB. Evidence-Based Medicine: How to Practise and Teach EBM. 2d ed. Edinburgh, Scotland: Churchill Livingstone;2000;.
Irish Lake, Ontario, Canada
3. Nease RF, Kneeland T, O'Connor GT, et al. Variation in patient utilities for outcomes of the management of chronic stable angina. Implications for clinical practice guidelines. Ischemic Heart Disease Patient Outcomes Research Team. JAMA. 1995;273:1185-90.
5. Haynes RB, McKibbon KA, Kanani R, Brouwers MC, Oliver T. Interventions to assist patients to follow prescriptions for medications. Cochrane Review, latest version 24 Feb 1999. In: The Cochrane Library. Oxford: Update Software.