James Raftery: NICE: “inconsistent,” “in large part arbitrary and opaque,” according to friends

A strong critique just published points to logical inconsistencies in NICE’s consideration of social values, specifically in how it handles quality adjusted life years (QALYs). Since these are key to many of the most controversial decisions made by NICE’s appraisal committees, this matters. It matters all the more that the authors include Tony Culyer, who was “the founding vice chair of the National Institute for Health and Clinical Excellence and is a member of the NICE Citizens Council Committee and the NICE International Advisory Committee.” Besides being a leading health economist, he pioneered NICE’s approach to economic evaluation. Chris McCabe, another prominent health economist, has long advised NICE, notably through its decision support unit, which specialises in difficult cases.

The article shows that NICE’s proposed amendments to its methods would increase the upper range of its cost per QALY threshold from £30k to £50k on the basis of the four “special considerations,” but at the cost of being inconsistent in how these would apply. They also describe the proposed criteria as: “in large part arbitrary and opaque.” Since the paper is dense and relies on examples to illustrate inconsistencies, the following paragraphs provide a summary overview.

The background is that for NICE, up to 2009 every QALY was valued the same, regardless of age, gender, severity, or the like. This was the meaning of “A QALY is a QALY is a QALY.” Economic evaluation for NICE aimed at maximising QALYs. In theory, every service and technology should be used in the NHS to a level that the marginal cost per QALY was the same for all. Any special value attached to particular QALYs (such as for severity) has implications for all other QALYs.

In 2009 the first special consideration was introduced to deal with the problem of patients close to the end of their lives being denied drugs, usually for cancer (read more). Treatments could qualify for what was in effect an “end of life” premium if they met three criteria: short life expectancy, normally less than 24 months, the treatment offers an extension to life of at least three months and applies to a small population. NICE in practice interpreted this as allowing a cost per QALY threshold of £50k or a maximum weight of 2.5 from a starting point of £20k per QALY. The recent critique points to several inconsistencies, to do with the higher weight applying to unadjusted life expectancy and the arbitrariness of the cut-off points of 24 and three months. Examples are provided of how this could lead to similar cases being treated differently.

The second special consideration has to do with discounting. This refers to estimating the present value of future costs and benefits. NICE introduced another special consideration in 2011 to deal with its appraisal of mifamurtide for osteocarcinoma, a rare disease of children and young adults. Using the standard 3.5% discount rate would have led to NICE recommending against this drug. But using a lower 1.5% for benefits enabled NICE to recommend it (read more here and here).  NICE amended its guidance to allow for the lower discount rate in selective cases, where “treatment effects are both substantial … and sustained (normally over 30 years).” This creates problems again because of the arbitrary cut-off of 30 years. Examples are provided to illustrate the impact on two different diseases straddling the 30 years cut off.

The proposed use of absolute QALY shortfall is the third special consideration. This is proposed by NICE as a measure of wider social impact, part of value based assessment. This is criticised first for breaching the NHS commitment as outlined in its constitution that “access to NHS services is based on clinical need, not an individual’s ability to pay.” Further, the authors show it provides a poor proxy for severity.

Proportional QALY loss constitutes the fourth special consideration, again linked to value based assessment. Intended to replace the end of life weighting, the authors query whether and how it might do so.

The crunch is that any special considerations should apply not only to the technology being appraised but also to the technologies that might be displaced. The key idea here is that for the fixed budget NHS to fund a technology recommended by NICE, it must spend less money on some other technology. The cost per QALY of that technology foregone constitutes the opportunity cost.
However, applying special considerations to the technology being appraised means that the same consideration should be applied to the technology being foregone. Otherwise “NICE’s proposals will harm the health of unidentifiable patients, while privileging the identifiable beneficiaries of new health technologies” and “we show it is not possible for NICE to prioritise some patients without deprioritising others, and that this deprioritisation is not obvious.”

The article recommends that NICE should eliminate the arbitrary cut-off in the application of value weights and research and consult on the proposed revisions. It also recommends that NICE should specify how it would include the effect of any special considerations on opportunity costs.

Until then, the authors recommend sticking with “A QALY is a QALY is a QALY.”

James Raftery is a health economist with several decades experience of the NHS. He is professor of Health Technology Assessment at Southampton University. A keen “NICE-watcher,” he has provided economic input to technical assessment reports for NICE, but has never been a member of any of its committees. The opinions expressed here are his personal views.

Competing interests: The author has no further interests to declare.