There are two general types of guidelines: path or algorithm guidelines, which include branch points and if-then statements to guide decision-making according to a standard of care, and boundary guidelines, which are used to define the appropriate use of a new (and generally expensive) technology. Guidelines are not intended to dictate a rigid approach to care; rather, they give options that clinicians should be aware of, even if they choose a different strategy for a particular patient. Guidelines are partly an outgrowth of the boom in scientific information that makes it difficult for individual physicians to keep up with medical advances.

Practice guidelines have been developed by government agencies (e.g., AHCPR), specialty organizations (e.g., American Society of Clinical Oncology), and cancer centers. The National Comprehensive Cancer Network (NCCN), a consortium of 15 leading cancer centers, has assembled expert panels to review evidence and develop guidelines on the treatment of 15 of the most common cancers (Marwick, 1997; McGivney, 1998). Many other hospitals and health care systems have developed their own guidelines (e.g., use of antiemetics, use of single daily dose antibiotics for infection), but they are generally not available to other institutions (see Table 6.4—for this review, cancer screening guidelines were not considered).

The Advisory Board Company, a private consulting group, has launched an Oncology Roundtable to provide cancer centers with information on "best" practices in the following areas: patient-focused oncology (i.e., ensuring convenient access, informed decision making, compassionate care, quality service); breast cancer management; prostate cancer management; and pain management (Advisory Board, 1998).

Disease management programs incorporate a systematic approach for the management of specific chronic disorders (e.g., asthma, diabetes), which often include adherence to clinical guidelines. The goal of these programs is to improve quality of care and outcomes, integrate and coordinate care, and track and manage costs associated with chronic illnesses. Memorial Sloan-Kettering Cancer Center has, for example, established 17 disease management teams to develop treatment pathways, track resource consumption, and identify appropriate patient education materials for cancers (McDermott, 1997). Nearly 100 clinical paths have been developed at the M.D. Anderson Cancer Center as part of its disease management program (Morris, 1996; Morris et al., 1996).

Practice guidelines are available for only a small fraction of oncology practice, but for some cancers, several guidelines are available. When to create a practice guideline can be difficult to gauge. The impetus to create a guideline often comes from evidence of widespread variation in practice; however, this is often a sign that there is little evidence upon which to construct a guideline. Guidelines based on sound evidence rather than expert opinion are most likely to succeed in influencing provider practice. Sometimes practice guidelines can be issued too late, after providers have already changed behavior in light of new evidence. A major recommendation of the 1979 National Institutes of Health (NIH) Consensus Development conference on treatment of primary breast cancer—that few Halsted radical mastectomies should be done—was found in subsequent reviews of medical practice to be moot since the procedure was being performed very infrequently (Kanouse et al., 1989). If the intent of a guideline is to inform and possibly change physician behavior, priority should be given to developing guidelines for which there is practice variation, despite good evidence to support a standard set of practices (U.S. Congress, 1994).



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement