How costly or cheap, relative to its benefits, does a healthcare technology have to be to justify its rejection or acceptance by the National Institute for Health and Clinical Excellence?

NICE performs an essential task in a healthcare system where the consumers (patients) do not directly bear the costs of their healthcare consumption. It provides a Which?- type guide for the NHS on best buys. It has taken on the burden of deciding whether the benefits of a treatment justify the costs. This requires judgements about the value of health and life. Since its inception in 1999, NICE has used an upper value range of£20,000-£30,000 for a quality adjusted life year, equivalent to£1.4m-£2.1m per quality adjusted life.

This value could be used to determine the budget of the NHS: just keep spending up to the point where the only things to be bought cost more than, say,£30,000 per QALY. It could also determine the level of health outcome the NHS should produce given a predetermined budget; for a budget of£100bn we should expect the NHS to generate at least around 3.3 million QALYs a year (if the value of a QALY is£30,000). Aside, there is the question of who should enjoy these additional QALYs: is, for example, 100,000 patients gaining 30 QALYs each equivalent to 10m gaining 0.3 QALYs?

So, our opening question has implications for our view about the value of life, the limits to healthcare spending and NHS productivity.

While this valuation issue is most obviously associated with NICE, it underlies every NHS decision that commits scarce resources - from a surgeon deciding to admit a patient for an operation to ministers deciding to reduce maximum waiting times to 18 weeks. Both decisions imply a cost per QALY and both imply that the health benefits justified the cost of achieving them (but does anyone know not just how many QALYs have been generated by reducing waiting times, but even how much money has been spent on this?).

The big problem with the question is that while NICE has an answer, there is little theoretical or empirical evidence to justify its figures. An even bigger problem is how a credible and acceptable answer could be derived, or even whether there is just one answer or many depending on individuals' characteristic, such as age, felt to be important in terms of fairness). One approach is simply to ask a population sample their 'willingness to pay' for a QALY - but this will be influenced by their income. Another is to calculate the cost per QALY of NHS decisions - but what validity would such figures have?

Topics