Intended for healthcare professionals

Head To Head

Should NICE’s threshold range for cost per QALY be raised? Yes

BMJ 2009; 338 doi: (Published 26 January 2009) Cite this as: BMJ 2009;338:b181
  1. Adrian Towse, director
  1. 1Office of Health Economics, London SW1A 2DY
  1. atowse{at}

    NICE has recently raised the threshold for end of life drugs. Adrian Towse argues it should consider doing the same for other treatments, but James Raftery (doi:10.1136/bmj.b185) believes that the threshold is already too high

    Decisions made by the National Institute for Health and Clinical Excellence (NICE) about whether the NHS should fund treatments are based on cost effectiveness. NICE methods guides refer to a threshold of £20 000-£30 000 (€22 000-€34 000; $30 000-$45 000) per quality adjusted life year (QALY).1 However, this is an arbitrary figure. Evidence on the public’s willingness to pay suggests that it should be higher. There is a lack of evidence on opportunity costs.

    Willingness to pay

    The Department of Health has commissioned research to help it understand what the public wants the NHS to pay for health gain in the knowledge that it has limited resources and pays for care from taxes. This estimates willingness to pay at £30 000 to £70 000 per QALY.2 On this basis NICE’s threshold range should double.

    The willingness to pay approach is usually challenged on two grounds. Firstly, it is argued that it is difficult for people to make these informed choices and that surveys asking people to state their preference for A versus B are hypothetical. Studies revealing preferences from the decisions they made3 do not really mirror the situation facing the NHS. These points have truth in them, but we routinely accept similar “stated” (hypothetical) approaches to assess the relative value of different health states when estimating QALYs. Also, willingness to pay is being used elsewhere to justify public sector investment—notably in transport and the environment.4

    The second challenge is that willingness to pay is beside the point because the NHS has a fixed budget, set by parliament.5 The real issue is spending out of that limited budget. The opportunity cost of spending on A is that money cannot be spent on B, the next best way of getting QALYs somewhere else in the NHS. NICE’s current approach is to have a threshold based on opportunity cost. But what is B? What will be given up and what is its cost per QALY? NICE has been described as a threshold seeker6—that is, an organisation whose role is to identify “technologies in current use that are the least productive uses of current NHS resources, and . . . better value technologies that are not currently provided.” The problem is that we don’t know the answers.

    Opportunity costs

    Primary care trusts claim they can make better use of the money NICE requires them to make available for approved technologies.7 Sadly, however, most trusts do not know the cost per QALY of services they cut or add to their spending or of many of the health services they pay for.8 9 Primary care trusts who are low spenders on cancer could achieve health gains at around £19 000 per QALY if they spent more on the right interventions,10 but that does not mean they will. There is no evidence that cost effective interventions are being cut to make room for NICE approved technologies.11

    There are many problems with an opportunity cost approach. Firstly, we need to distinguish between what primary care trusts consider cutting in the short run and the resources they could make available over time if they stopped funding activities that are not demonstrably cost effective. But cutting services (or disinvestment) is not popular. Secondly, lack of information about which treatments are poor value for money has meant NICE has struggled with its disinvestment agenda. The Health Select Committee and others have called for more emphasis on disinvestment.5 12

    Opportunity cost thresholds will also differ by locality because of historical legacy, and new treatments will cost more in some areas because there are more patients. Without a national view on what to cut, the opportunity cost approach logically ends up with primary care trusts setting postcode thresholds.13


    NICE has not raised the value of its cost per QALY threshold since it first made public that it had one. It could be argued that the threshold should at least be increased for inflation since then, and also for the large increase in NHS real expenditure. Others have argued that improvements in NHS productivity (getting more from existing technologies) and an expected reduction in the future rate of growth of NHS spend may make the case for reducing the threshold. However, if the original threshold number is arbitrary then using these arguments to make the case either way does not seem sensible. What we want to know is the availability of funds relative to the opportunities to usefully spend them, and whether the funds are being spent efficiently. We do not know this. Until more evidence is available on the cost effectiveness of existing interventions policy pragmatism is called for.

    There is strong evidence that society’s willingness to pay for a QALY is above the current NICE threshold. Other countries that spend similar per capita amounts on health care are approving treatments that NICE is turning down, and we suspect the NHS is spending money on services that probably yield far less health gain per pound spent than NICE’s threshold. NICE is raising its threshold for end of life drugs based on willingness to pay 14 and could justifiably extend this to other treatments.


    Cite this as: BMJ 2009;338:b181


    • Competing interests: The Office of Health Economics receives research income and consulting income from the Association of the British Pharmaceutical Industry. It has recently completed a consulting project for a group of companies with oncology products.