In March this year, NICE’s board agreed a proposal to introduce Quality-Adjusted Life Years (QALYs) as the measure of value for Highly Specialised Technologies (HST – products to treat very rare conditions), with a limit of £100 000 per QALY, below which the treatment would automatically be funded. In response to consultation input, the proposal also includes a mechanism for QALY weighting of up to threefold depending on the health gain experienced by patients – so a technology costing £300 000 per QALY could qualify, if the health gain was sufficiently high. The expected publication of two HST decisions that were in progress before the changes, prompted us to review the decisions to date, and consider the impact of the changes.
What decisions have been made through the HST process in England?
Including the two pending publications, 6 drugs have been assessed through the HST process, and all of them have been accepted for funding (although note the pending decisions are not final until publication!). Three are under Managed Access Schemes with data collection.
HST evaluations do not publish a cost per QALY number, but describe a cost consequence analysis in which the cost is compared with a catalogue of health outcomes, including but not limited to QALYs. The cost consequence framework has the advantage of breadth and flexibility, but can be criticised for lack of transparency, as the relative importance of the various domains is left to the discretion of the Committee. In the introduction of an explicit cost-per-QALY for use in decision-making, NICE and NHSE appear to be signalling a shift in emphasis, towards the more quantitative, standard measures of value.
Comparison to Scotland
These drugs were mostly appraised through Scotland’s ‘ultra-orphan’ process, which is similar to NICE’s HST appraisal. Only one was accepted for funding – a low rate of acceptance of non-cancer ultra-orphans that was highlighted by the recent Montgomery review in Scotland. We now expect adjustments to the Scottish Medicines Consortium (SMC)’s process, aiming to bring access to these drugs in line with the original intentions for the ultra-orphan process. One suggestion in the review was that the final decision responsibility would no longer rest with the SMC – a move that could be seen as diametrically opposite to NICE and NHSE’s increased focus on cost-per-QALY as the measure of value for these drugs. We await developments in Scotland with interest.
Would NICE’s decisions be different under the new process?
We estimated the cost per QALY for three of the HSTs, where data were available in the public domain; whilst our estimates are highly approximate, we are reassured that they are at least consistent with those published for these drugs by SMC. In one case, the new technology was ‘dominant’ – a term used in health economics to describe a technology with increased health gain at lower cost compared to current treatments. For three other cases, we found costs above £400 000/QALY. Introducing the new QALY gain weighting, we would find a maximum weighting of 2 – meaning that none of the three would have met the new £100 000/QALY limit for automatic funding. Hence taken in isolation, the new limit – even with QALY weighting – appears to reflect a less permissive environment for patient access. In these cases, presumably the deliberative process would remain as important as ever; the revised process retains the broader HST framework, which the Committee would debate in order to reach a funding decision.
Are the limits and weightings consistent with societal values?
Two questions are worth addressing. Firstly, is there a societal preference for funding treatments for rare conditions? An important study in the UK (click here), replicated recently in Australia (click here), found no such preference, and similar results were found in Norway (click here). Whilst we have not performed an exhaustive review of the literature, we conclude that there is no consistent evidence for a preference to prioritise specifically on the basis of rarity. What we do see, though, is a suggestion from a range of studies that society values equity – in the sense of ‘something for everyone’ – and hence that it is ‘unfair’ for some conditions to have no affordable treatments available. We also see a consistent preference for treatment of severe illnesses (click here); it is not unreasonable that rare metabolic diseases affecting children, for example, would be considered in this way.
Secondly, is there a societal preference for preferentially funding treatments with higher health gain? A significant UK study (click here) found that the public do indeed favour treatments with higher health gain, although at a diminishing rate (each additional unit of health gain adds slightly less value than the one before). The increasing multiplier in the new HST process does not appear to reflect fully this subtlety of the preference.
And finally: note NICE’s consistent use of language. This is not a discussion about adjusting the threshold, but about weighting QALYs – ie allowing some known QALYs to be worth more than others. This contrasts with threshold adjustment, which effectively says some other NHS QALYs – unknown – are worth less. Perhaps a subtle distinction, but reflecting NICE’s remit to be fair to all users of the NHS, not just the ones whose treatment is being discussed on a given day.
-Dr Liz Morrell