The Institute of Medicine has issued stern new guidelines on guideline development. Apparently a good portion of the 2,700 clinical practice guidelines in the Agency of Healthcare Research and Quality's database are not based on a foundation of good evidence, do not acknowledge when the evidence is shaky, and their authors often have financial conflicts of interest.
The IOM calls for transparency and a rigorous evidence-based approach, and prohibition of COI in guideline chairpeople -- with COI defined beyond industry bedfellowship to include board certification in any subspecialty, since self-interested guideline-crafting could potentially boost revenues within a practice specialty. (Of course, this begs the question, who will actually write the guidelines, if not the experts?) JAMA 2011;305(18):1846-1848.