options and make specific recommendations wherever possible, based on the strength of available research findings as well as the perceived degree of clinical consensus among practicing clinicians. Treatments that have not been adequately tested in well-controlled trials, or treatments where there are conflicting reports about efficacy but which are consistent with expert opinion and generally accepted treatment principles, are recommended with a lower level of clinical confidence or alternatively, recommended to be applied only in specific clinical circumstances. These guidelines leave the ultimate judgment to the clinician, based on data presented by the patient and on the diagnostic and treatment options available. It is anticipated that the guidelines will be revised every three to five years to incorporate emerging research and clinical experience.

Despite the potential benefits of incorporating advances in clinical research into clinical care delivery, many barriers exist to the successful dissemination and adoption of evidence-based practice guidelines within the drug abuse treatment community. A number of factors may contribute to this situation. Chief among them is the heterogeneity in the background, training, and clinical perspectives of clinicians practicing within the addiction treatment community. With the notable exception of methadone maintenance, the relative paucity of clinically effective, medically based treatments for this patient population has helped foster a treatment culture in which many treatment approaches, including self-help and therapeutic communities, have flourished. Heavily influenced by both the experience and world views of recovering drug abusers, this segment of the treatment system has embraced a treatment philosophy and approaches to care that depend more on the motivational power of group support and on spiritual beliefs, than on methodologically sound studies of treatment effectiveness. In this context, guidelines based on data from clinical research, particularly research carried out in medical settings, may be seen as undermining treatment approaches less amenable to study by scientific methods.

Both the American Medical Association (Office of Quality Assurance, 1996) and the Institute of Medicine (IOM, 1992) have developed principles for practice guideline development and implementation. Not surprisingly, AMA recommends that guidelines be developed by, or in conjunction with, physician organizations. In addition, AMA recommends that guideline dissemination be coupled with a plan for measuring their impact on short- and long-term treatment outcome. Testing is important for guidelines in any field because of the potential for obtaining useful new information as well as avoiding unintended bad effects (Weingarten, 1997).

The IOM report on Clinical Guidelines for Practice (1992) recommended that guidelines should evolve as a result of a multidisciplinary process "that includes participation by representatives of key affected groups" who can identify, critically evaluate, and incorporate all important