In their otherwise excellent review Dr Samanta and his colleagues (March 2003 JRSM¹) assert that ‘clinical guidelines are systematically developed, evidence-based, clinically workable statements that aim to provide consistent and high quality care for patients’, and that because ‘guidelines from NICE... may be seen as the crème de la crème of authoritative bodies' it follows that their pronouncements are increasingly likely to inform the standard of care in negligence’. This is a dangerously illogical assumption to make: for all the diversity of lay and medical expertise available to NICE, the content of their guidelines is becoming increasingly proscriptive, and, in some cases, advocates a radical departure from pragmatic medical practice. The recent furore over the use of ultrasound locating devices for placing central venous catheters is a case in point.^{2, 3, 4} There are any number of criticisms that could be made of this document and its practical implementation,⁵ but with regard to the quasilegal nature of the guidelines, several important questions arise. Who owes the duty of informing doctors of new guidelines? Who is responsible for their execution—the hospital, the doctor, or both? Who is responsible for failing to follow the guidelines—the doctor, the hospital (either primarily, or vicariously), or both? Does obtaining specific consent from the patient vitiate the pseudo-legal requirement to follow the guideline? What is the legal recourse open to doctors who do not accept that the guidelines are ‘systematically developed, evidence-based or clinically workable’? Finally, should the courts be asked in every case to determine whether the guidelines are reasonable and rational before relying on them as the standard of acceptable medical practice?

NICE and CHAI are valuable agencies for improving the quality of care in the NHS, but their authority depends on the soundness of their decrees. It is ultimately for the courts to decide, case by case, whether such decrees should inform the standard of care in negligence. Guidelines should continue to inform, rather than prescribe, medical practice.