ions Supported Explicitly Models Patient Preferences No No No No No No No Current Trends and Future Directions The guideline community is eager to facilitate authoring of well-structured, computer-encoded guidelines that can be delivered to the point of care, and integrated into applications used by providers in the course of delivering care. While many such applications have been implemented 39 40, 41, there has been no standardized way to represent the medical knowledge upon which such applications operate. Because of this, it is difficult to share applications and knowledge bases. A number of organizations and research groups are engaged in developing approaches to computerencoded guidelines for the above purposes, but there is still little standardization to facilitate sharing or to enable adaptation to local practice settings. Nonetheless, considerable progress has been made. In March, 2000, an international workshop “Toward Sharable Guideline Representation” was held in Boston, organized by the InterMed Collaboratory, and sponsored by the US Army, the Agency for Healthcare Research and Quality (AHRQ), the National Library of Medicine (NLM), and the Centers for Disease Control and Prevention (CDC). The workshop included representatives from eight different countries, among whom were stakeholders from academia, government agencies, professional organizations, insurers, providers, and industry, who recognize the need for a standard sharable computer-interpretable guideline representation format. The workshop resulted in five task forces that are pursuing a sharable guideline representation standard through exploration of functional requirements of guidelines, modeling and representation, special needs of clinical trials, infrastructure and tools, and organizational framework for continuing these activities. Any one who is interested in joining these teams is welcome to do so through the www.GLIF.org website. The most important near-term goals are to move toward adoption of a common standard, and to create prototype authoring tools that can encode a set of medically sound, well-validated guidelines and then provide a mechanism for these guidelines to be linked seamlessly to the electronic health record. One can easily imagine busy clinicians using information conveniently made available at the places and points in time when that information can assist them with the work-up of patients or development of management plans. It is equally easy to understand why busy clinicians in an overworked medical system fail to utilize this same information if it is accessible in a less convenient or timely fashion. Development of shareable computerized CPGs will benefit health plans that wish to reduce costs through undesirable practice variation. Most importantly, patients will benefit from reduction in adverse events and medical errors, which will result from increased integration of decision support in clinical practice.
