Developing clinical guidelines: how much rigour is required?

Clinical guidelines that are rigorously developed play a fundamental role in improving healthcare and reducing unnecessary variations in practice. National guidelines are increasingly used by healthcare professionals, patients and commissioners; however, national bodies are unable to meet the demand for guidance on all topics. There are fewer resources available for guidance produced locally or by specialty groups, and it is necessary to achieve a balance between pragmatism and rigour while conforming to the widely accepted norms of what constitutes a good guideline. This paper introduces the key concepts around this topic with suggestions for those interested in developing their own guideline. An example of challenges encountered in generating high-quality clinical guidance is given in box 1. Box 1 Challenges in guideline development Professor Johnson runs a local developmental paediatrics service with eight other colleagues. All have different ways of managing children with PAVING syndrome. This was difficult for patients and staff and has led to disagreements on how certain patients should be managed. As a result, Professor Johnson developed a Guideline Development Group to look at the management of PAVING syndrome. The group identified 12 clinical questions (including diagnosis, exclusion of comorbidities, treatment modalities), searched the PubMed database and found some useful evidence that they used to formulate key recommendations. For one question about behavioural therapy, PubMed did not suggest any evidence so they informally arrived at a consensus among themselves and wrote up their guideline. On the back of this success, they applied for the guideline to be endorsed or supported by the Royal College of Paediatrics and Child Health (RCPCH). To their frustration, it was turned down on methodological grounds. Professor Johnson wrote to the RCPCH saying that he was "pretty peeved that the PAVING syndrome guideline had been rejected" for the College endorsement given all the work that had gone into writing it and "would the College mind being a bit more explicit in their guidance," to anyone who might consider doing this in the future?