Guidelines have become an important vehicle for influencing clinical practice. Many local, national, and international societies now go through the process of identifying relevant clinical areas, formulating specific clinical questions, reviewing the applicable evidence, and formulating recommendations that they believe clinicians and their patients should follow.

Over the years, in recognition of the diversity of individuals required to produce optimal recommendations (content experts, methodologists, front line clinicians, patients’ representatives), guideline panels have grown in size. The resulting large and diverse panels present challenges for decision making, such as ensuring that all participants have a voice and can influence the results of the debate, ensuring transparency, dealing with disagreement, achieving consensus, and resolving situations in which consensus is not possible.

Guideline panels often use only informal processes to deal with these challenges. Informal processes are, however, vulnerable to the idiosyncrasies of small or moderate sized group interaction. Factors including time pressure; fatigue; lack of expertise in content, methods, or group leadership; and, most importantly, dominance by individuals with powerful personalities and intimidating reputations threaten the integrity of the process.

Those interested in the science of guideline development have developed two strategies to deal with these problems. The first uses structured approaches to collect, analyse, and summarise the relevant evidence and to use that evidence to produce and grade recommendations. These approaches are epitomised by the method suggested by the Grading of Recommendations Assessment, Development and Evaluation (GRADE) working group, which has developed an increasingly widely adopted structure for developing guidelines.1 2 3 4 5 6 The second relies on somewhat formalised processes to encourage a consensus to which all panellists can contribute more or less equally.7 …