Is NICE delivering the goods?BMJ 2004; 329 doi: https://doi.org/10.1136/bmj.329.7473.1003 (Published 28 October 2004) Cite this as: BMJ 2004;329:1003
- Nick Freemantle, professor of clinical epidemiology and biostatistics ()1
Those of us concerned with the ability of organisations such as the National Institute for Clinical Excellence (NICE) to influence clinical practice in line with their guidance will read this paper with great interest.1 But what conclusions can we draw from it? If NICE was an unqualified success, clinical practice in the NHS would reflect its guidance—so use of implantable cardioverter defibrillators would have gone up smartly, laparoscopic hernia repair would have stopped, and so on. This was demonstrably not the case.
In contrast with randomised controlled trials, where the intervention is under the control of the investigator, the quasi-experimental method necessarily used by the authors is weak in attributing cause and effect. So we cannot even conclude that changes that occurred apparently in line with the NICE guidance were actually caused by it, either in part or in whole.
Some may find it surprising that prescribing of (two of four) taxanes for cancer, and of orlistat for obesity were the only topics out of 12 surveyed where significant changes in the rate of use occurred after NICE guidance. Given that the manufacturers of these products are also very interested in increasing prescribing, and from informal accounts have worked hard to increase sales, it seems a big step to presume that changing use at around that time was caused by the guidance. Indeed, it would be much more convincing if there was evidence that practice had changed after publication of NICE guidance in the counterfactual direction to that which would result from marketing activity. Without such evidence many will remain, correctly, sceptical as to whether there is any real return from the substantial efforts and resources that go into producing NICE guidance.
NICE has recently woken up to the potential problems regarding the implementation of its guidance in the NHS and is appointing a board level “implementation tsar.”2 That person's task may seem unenviable—influencing clinical practice seems much more difficult than merely issuing edicts. Indeed, as A H Weiler reputedly said, “Nothing is impossible for the man who doesn't have to do it himself.” Achieving real change in clinical practice is clearly a necessary part of the remit of NICE. Without this vital step, the resources currently used to support the NICE enterprise would be better spent on care for patients. Other regulatory structures, such as the pharmaceutical benefits scheme in Australia, which limits access to reimbursement in the health service to pharmaceuticals that are judged to be good value for money, seem much more effective in achieving real change, and there is a lot to learn from the experience in other health systems.3 So, rather than give up on the task of modernising the way the NHS uses healthcare interventions, we should look at a variety of ways to make NICE more effective.4
As he left the BMJ, Richard Smith ably appraised the performance of NICE in a sentence or two, questioning in particular the extent to which NICE dealt with rationing and the breadth of clinical practice, kept an appropriate distance from politicians and the pharmaceutical industry, or was directly accountable to the public.5 With the publication of this paper, we might further question whether NICE was delivering the goods.
Conflict of interest: Nick Freemantle has received funding for research and consultancy from a variety of governmental, industrial, and charitable sources.