by other groups. The authors postulated that the higher-quality scores of guidelines developed by government agencies reflect the fact that the production of high-quality guidelines requires substantial and sufficient resources and that government agencies have more resources available to do the work.
Smaller professional organizations often lack the internal resources, including staff capacity and expertise, required to produce guidelines. This is especially true when the organization produces both the systematic reviews and the guideline recommendations, two tasks requiring different skill sets. Even larger professional organizations can face resource constraints in this area. Some have suggested that, given these resource constraints, government is in the best position to produce clinical practice guidelines (Burgers, 2003; Hasenfeld and Shekelle, 2003).
Many of the criticisms directed at the U.S. system of guideline production in 1990 still apply today (IOM, 1990). These criticisms focused on conflicting clinical recommendations; failure to address certain topics; and incomplete public disclosure of the evidence surveyed, methods used, composition of the panel, and conflicts of interest. In addition, it remains true that, aside from the role that AHRQ plays in populating the NGC website, no independent entity exists in the United States to certify guideline quality or to develop national standards regulating the content or methods of guideline developers. The 1990 IOM report Clinical Practice Guidelines: Directions for a New Program sought to encourage more standardization and consistency in guideline development, and although the quality of clinical practice guidelines has generally improved since then, substantial inconsistencies in the methodologies and reporting language used still exist (Guyatt et al., 2006b; Shiffman et al., 2003).
The translation of systematic reviews into practice recommendations is not straightforward. The same information can be interpreted in different ways by different panelists, resulting in the provision of different guidance (Burgers and van Everdingen, 2004). Often, even when there is substantial consensus about what the scientific evidence says, there are disagreements about what the evidence means for clinical practice. Conclusions about clinical effectiveness can vary widely as a result of conflicting viewpoints, such as which outcomes are the most important and which course of action is appropriate given that the evidence is imperfect (Atkins et al., 2005b). This section highlights strategies that guideline developers have used to improve the reliability and trustworthiness of the information that they provide. It also examines methodological approaches, and how groups