No single group has the right to ignore a consensus of thoughtful opinion. Mollie Hunter (1922–2012) In this issue, we publish a consensus report from the Clinical Practice Committee of the Canadian Association of General Surgeons regarding Clostridium difficile. Medical use of the word “consensus” is evolving. Its 19th century meaning — the harmony between physiologic sys- tems — has faded in most areas of medicine, with the possi- ble exception of the field of genetics in which consensus may refer to conserved sequences of DNA. The specific physiologic term has been replaced in medicine by a vaguer popular meaning, often used to confer authority upon a statement or process — authority that it might not otherwise possess. Popular use is derived from the legal term consensus ad idem, which originated in Roman law as an informal contract based on mutual consent. The element of consent has persisted in its legal use, but the emphasis has changed in popular and medical use to denote agreement not just between parties, but general acceptance. While we have become increasingly precise about grades of medical evidence, interpretation of these units of evi- dence in complex pathophysiological systems remains less structured. The outcome is more often decided by regula- tors and insurers than by agreement among practitioners. The National Institutes of Health (NIH) distinguishes con- sensus statements from clinical practice guidelines (CPGs) as the synthesis of new information from recent research. Standardization policies by jurisdictions such as the Euro- pean Union are vague with respect to methods of consensus development, but they are prescriptive with respect to CPGs, stating they may be used in litigation as a standard of practice.2 This is particularly true of cancer care in Canada.3 The NIH Consensus Development Program attempts to address some of these issues, but topic selection is very restrictive and the conference approach is limiting.4 The Delphi procedure for deriving consensus is said to have originated as a method used by the military after World War II to brainstorm about the capabilities and future actions of the enemy. Developed later for public policy cre- ation, the procedure attempted to include and protect all points of view in iterative cycles of facilitated refinement. The nominal group method assigns fundamental aspects of complex problems to groups using techniques intended to prevent domination by a single opinion, with voting to determine priorities. The nominal group method was used recently to determine waiting time targets for pediatric surgery in Canada.5 The Delphi procedure is more often used for distant communication, and the nominal group method is favoured by consensus conferences. With modern communication systems, these procedures may be blended. Consensus ad idem implies agreement not only about the answer, but also about the question. CJS protocol for pub- lishing consensus-based articles will require the authors to be selected by a representative group or society. The sub- ject, topic or question addressed by the statement should be validated by a committee of that group. Composition of the report should use the procedures outlined here so that research tasks are shared productively among sub- groups and opinions, including a patient perspective if possible, are debated. The initial draft should be circulated among the general membership of the society, whose comments should be considered in revisions before confir- mation by a committee or executive of the sponsoring society. When the final draft is submitted to the journal, it will undergo independent peer review. Those accepted for publication will include a footnote confirming compliance with CJS protocol for consensus development. In the end, the weight accorded to a consensus report will depend on the validity of the process it uses to distill the thoughtful opinion it appraises.