Crossing the line: NICE’s value for money threshold
BMJ 2016; 352 doi: https://doi.org/10.1136/bmj.i1336 (Published 09 March 2016) Cite this as: BMJ 2016;352:i1336
For a publicly funded health service, deciding what services and treatments should be available to patients is not just a matter of getting the best value for every (scarce) pound spent but an ethical duty too. Choosing “what’s in and what’s out” quite literally involves decisions about life and death; the cost of getting a decision wrong can be measured not just in wasted resources but in avoidable deaths and lower quality of life. In this respect, the creation of the National Institute for Health and Clinical Excellence (NICE) 17 years ago was a watershed moment. For the first time there was an organisation required to advise the NHS explicitly on what it should offer to patients, drawing on evidence of the clinical benefits and the cost of obtaining those benefits.
But NICE—and the growing number of sister organisations in other countries—faced a problem from the outset. At what point (or threshold) were the benefits generated by an intervention to be deemed worth the cost to obtain them? This question not only required a generic measure of benefit (the quality adjusted life year (QALY) became NICE’s accepted metric) but also a view about what the threshold for cost per QALY actually represented.
Should the threshold be a quantification of society’s willingness to pay for a QALY? Or should it reflect the opportunity cost of the least cost effective treatment the NHS currently provides? For various reasons, economists prefer the second view. If the threshold were higher than the least cost effective treatment currently provided then, given a fixed budget, the NHS would have to stop providing such a treatment in order to allow less cost effective treatments to be provided – not an efficient use of the money available.
In fact, there is a link between the two views of what the threshold represents—the NHS budget. The amount spent on the NHS not only determines what the threshold should be but is also (somewhat imperfectly) an expression of the public’s willingness to spend their money on healthcare. As Culyer’s neat bookshelf analogy makes clear,1 for a given healthcare budget and the ability to line up in cost effectiveness order all the services or treatments (“books”) the NHS provides, the cost effectiveness threshold becomes evident (fig 1⇓). For some obvious practical reasons, this is not how NICE arrived at its threshold range of £20 000 to £30 000. (It chose a range rather than a single number to allow some flexibility in decision making to reflect uncertainties in cost and effectiveness data and the use of criteria other than cost effectiveness in making decisions.
Fig 1 Culyer’s cost effective bookshelf of treatments showing how the cost effectiveness threshold is related to the total healthcare budget and how an increase in budget from A to B might change the threshold1
The exact origin of this range remains somewhat shady, emerging as a reasonable benchmark from NICE’s early evaluations of various healthcare interventions.2 Compared with the World Health Organization’s suggested cost effectiveness range based on gross domestic product (GDP) per capita of a country (“very cost effective”) to three times GDP per capita (“cost effective”), NICE’s threshold is much less generous (fig 2⇓).3 On the other hand, recent analysis by Claxton and colleagues suggests NICE’s threshold is too generous relative to the cost effectiveness of treatments and services currently provided by the NHS.4
Whatever the starting point for NICE’s threshold, because the NHS budget has roughly doubled in real terms since NICE started using its threshold range and the NHS has also increased outputs relative to inputs (that is, increased its productivity) over time, the threshold should have more than doubled over the past 15 years—becoming more generous to less cost effective treatments (fig 3⇓). To the possible chagrin of drug manufacturers, this hasn’t happened—not least because moving the threshold goal posts would create problems. Previously rejected treatments could become cost effective as the threshold moves; what does NICE do then, and what incentives does this create for drug manufacturers?
Fig 3 NICE’s cost effectiveness threshold range adjusted to reflect real changes in the NHS budget and NHS productivity. The overall effect of this is to increase the threshold from £20 000 to £30 000 in 2000 to around £42 000 to £64 000 by 2015
But we shouldn’t get too fixated by the cliff edge nature of thresholds. As Helen Dakin and colleagues have shown, while cost effectiveness explains most of the decisions made by NICE, other criteria play a part, and there is little indication of a sharp accept-reject divide at the supposed fixed threshold of £20 000 to £30 000 per QALY.5 Even (especially?) economists are probably pleased that NICE does not operate a strict utilitarian policy.
Footnotes
Competing interests: I have read and understood BMJ policy on declaration of interests and have no relevant interests to declare.
Provenance and peer review: Commissioned; externally peer reviewed.