House of Commons - Health

Select Committee on Health First Report

4 Affordability and rationing

202. NICE aims to promote good health and prevent or treat ill health through the use of treatments which are both cost-effective and clinically effective. If NICE provides guidance that NHS organisations cannot afford, it has failed to achieve this aim. The evidence we received revealed that many PCTs find it difficult to adopt NICE technology appraisals and clinical guidelines because of their cost implications. Witnesses had the following concerns:

NICE's threshold is not evidence based;
The threshold is higher than that of other NHS organisations;
As a result, there is a risk that PCTs give other possibly cost-effective therapies, which have not been assessed by NICE, a low priority;
Although NICE and PCTs have taken some steps to address the situation, there has to be a re-examination of the NICE threshold; and
If the threshold is wrong, NICE's role in rationing is compromised.

Criticisms of NICE

NICE'S THRESHOLD IS NOT CLEAR AND IS NOT EVIDENCE-BASED

203. The incremental cost-effectiveness ratio (ICER), or cost per QALY, which NICE uses to determine cost-effectiveness in its technology appraisals was described earlier. The cost per QALY is one indication of whether a treatment or procedure will be approved by NICE.

204. As we have seen, Professor Rawlins told us that NICE employs a threshold of £20-30,000. Some research has indicated that it could be even higher.[179] Professor John Appleby from the Kings Fund told the Committee:

Some fudging is going on as far as NICE and others are concerned, not just about what the threshold is but how it is applied.[180]

205. There is clearly confusion about the cost per QALY threshold. Witnesses questioned whether there was any evidence to support the level that appears to be used. Professor Devlin told us that, "the threshold has no explicit basis or location in evidence".[181] Others agreed that it was "arbitrary".[182] Professor Smith confirmed:

[NICE] has had to undertake its work in the absence of secure information about the most appropriate cost-effectiveness 'threshold' at which to approve new technologies.[183]

206. Professor Rawlins admitted that the threshold was not based on "empirical research" as no such research existed anywhere in the world. He told us instead that the threshold was:

...really based on the collective judgment of the health economists we have approached across the country. There is no known piece of work which tells you what the threshold should be.[184]

207. No public discussion has ever taken place of the suitability of the threshold used. The American Pharmaceutical Group pointed out that the threshold has "never been the subject of public debate or Parliamentary approval".[185] Cancer Research UK also argued that the threshold should be discussed openly and the reasons for its level should be determined in consultation with interested organisations:

The public should know where the £30,000 figure came from and why it is set at that level…Government [should] enter a debate with stakeholders about what threshold is appropriate for a country such as the UK and for the NHS in the future.

The pharmaceutical company Bristol Myers Squibb agreed that, "it should be much clearer how the threshold is determined".[186]

208. The cost per QALY threshold used by NICE does not appear to have changed over time.The range it uses now of £20,000 to £30,000 is the same as it was in 1999. Witnesses claimed that this was further evidence of its arbitrary nature.[187] If it had kept pace with NHS-specific inflation, then in 2007 it would be £28,000 to £42,000.[188] Professors Devlin, Appleby and Parkin claimed:

By whatever means the threshold is determined, it should be adjusted over time. NICE appears to have operated the same threshold or threshold range since 1999.[189]

209. The stability of the cost per QALY threshold over a period when NHS spending has increased significantly indicates that there is little or no relationship between it and the NHS budget. Witnesses claimed that the two should be directly related. Professor Bryan told us that:

The issue of affordability and the overall size of the budget cannot be unrelated to the appropriate level of threshold.[190]

NICE'S THRESHOLD COMPARED TO THE THRESHOLD USED BY OTHER NHS ORGANISATIONS

210. Inevitably, in view of the lack of any firm justification for it, witnesses expressed concern about the current threshold range. A number of witnesses claimed that the threshold should be higher than its current level, to ensure that patients received the best treatments.[191] For instance, Breakthrough Breast Cancer stated:

If NICE does not review its current cost thresholds there is a danger that the NHS will not keep up-to-date with treatment advances and that patients in England and Wales will not be offered the most effective treatments for their conditions.[192]

211. Manufacturers and patient groups in particular argued that certain treatment areas should have higher associated cost per QALY thresholds. It was claimed that orphan drugs (drugs for rare conditions) should continue to benefit from a higher cost per QALY threshold.[193]

212. NICE argued that the current cost per QALY threshold was appropriate. In particular, Professor Rawlins maintained that a lower threshold would have negative effects on the treatments available to patients:

The truth of the matter is if we halve the broad threshold, we would have declined as cost-ineffective most of the new drugs we have looked at and I do not think that is what people want.[194]

213. Other witnesses, in contrast, expressed a fear that the threshold used by NICE was too high, and that the efficiency of the NHS could be reduced as a result.[195] These witnesses questioned whether the threshold, or the threshold range, currently in use was in line with spending decisions made by PCTs about treatments not assessed by NICE.[196]^,[197]

214. Professor Smith told us that his research had compared the cost per QALY threshold in funding decisions made by PCTs to that used by NICE in its technology appraisals.[198] Using data from the Department of Health's programme budgeting system, Professor Smith showed that the cost to PCTs of "a life saved" was around £13,100 for cancer patients and £8,000 for heart disease patients. Adjustment for the quality of the year of life saved (to make the figures comparable to the ICER used by NICE) indicated a cost of around £19,100 for cancer and £12,000 for circulatory disease. Professor Smith stated:

Although these figures are rudimentary, they do suggest that the existing cost of a QALY secured in these programmes of care may be lower than many commentators have assumed….

A threshold closer to £20,000 is more appropriate than £30,000 on the basis of the evidence that we have been able to uncover.[199]

215. Further research by the same group evaluated three more areas of care: respiratory disease, gastro-intestinal conditions and diabetes. A comparison of the figures for all five areas, unadjusted for quality of life, indicated that the cost of a "life year saved" for respiratory problems was lower than cancer and circulatory disease and higher for gastro-intestinal conditions and diabetes.

216. Professor Rawlins claimed that this was evidence that NICE used an appropriate threshold. He cited the later work by Professor Smith stating, "His more recent and more extensive work indicates that we are in about the same sort of ballpark".[200] However, Professor Smith maintained that his results actually suggested PCTs used a cost per QALY threshold closer to £20,000 than £30,000.[201]

217. Although this research implies that PCTs make use of a cost per QALY threshold, other witnesses claimed that most decisions by PCTs were not made on the basis of a specific threshold. Professor Appleby told the Committee that "every decision seems to be separate".[202] Dr Anderson from the Royal College of Psychiatrists, the Hepatitis C Trust and others, questioned the basis of PCT decision-making more generally.[203]

Consequences

CROWDING OUT

218. Since guidance given in technology appraisals must be funded within three months, too high a threshold would cause significant financial problems for local health organisations. Professor Smith described the significance of the threshold [the cost per QALY above which a treatment is unlikely to be approved] for the NHS:

If NICE adopts too liberal a threshold, it may approve technologies that drive out more cost-effective conventional treatments…if NICE adopts a threshold that is too strict it may inhibit the development and adoption of cost-effective new treatments.

Professor Stirling Bryan told us:

A serious concern on the threshold topic is how the band of £20,000 to £30,000 has been arrived at and whether it has been set at too high a level. These worries are driven, in part, by the mandatory nature of NICE technology appraisals guidance …An important consequence of applying this threshold is, therefore, that local NHS organisations will find it even more difficult to remain within budget.

Moreover, while a lower threshold would deprive patients of treatments which were just marginally useful, it would thereby free up funding for more cost-effective interventions.[204]

219. Positive NICE technology appraisals must be funded within three months of publication. As NICE evaluates more and more medicines and procedures using the current threshold, there is a risk that treatments which NICE has not examined, or other areas of healthcare, will be 'crowded out' as PCTs are forced to prioritise NICE-evaluated approaches.[205] Patients with conditions not yet covered by guidance may therefore receive cheaper or less up-to-date therapies than patients who receive treatments which have been the subject of NICE appraisal. As the ADPH stated:

The disadvantaged group is the majority of patients who have conditions that are not covered by NICE, and in particular those who, unbeknown to them, are denied interventions because the funding has been diverted to implement NICE's recommendations for others.[206]

220. Many witnesses thought that areas of spending not evaluated by NICE, which were foregone as a result of NICE guidance, might represent better value for money, particularly in the long-term, than those evaluated. These could include older, established and often cheaper medicines, or public health measures. Professor Bryan told us:

[My research] suggests that people have to displace things that they perceive to be of greater value as a result of the NICE guidance. There is a perception, whether or not it is the reality, that that is the case.[207]

221. The NHS Confederation agreed. It stated that many of the treatments examined by NICE were only just within the defined limits of cost-effectiveness and that other areas of care suffered because of their implementation:

As a result the paradox arises that NHS funding is mandated for a marginally cost effective drug and local NHS organisations may have to achieve this by not spending on treatments which may be very much more effective and could benefit more people.[208]

222. An example, which highlights the difficulties caused by the mandatory funding of drugs over other types of therapy, concerns the new thrombolysing drug Actilyse (alteplase). Alteplase was the subject of an STA that was published in June 2007. Dr Nigel Dudley, a consultant in elderly and stroke medicine, argued that the appraisal should not have been published ahead of NICE's stroke guidelines or the National Stroke Strategy. Many areas that lacked far more basic stroke services had to purchase this expensive drug before addressing other, more pressing issues:

The priority given to Alteplase means that those who have shouted loudest…have gained in this particular case at the expense of other parts of an underfunded stroke service…

money that has to be spent by law on thrombolysis for patients aged 80 and less will not be available to spend on patients of all ages in rehabilitation units or early supported discharge services who would benefit.[209]

223. The problem is exacerbated by the exclusion of some particularly expensive treatments from the payment by results tariff. The increase in costs associated with NICE guidance is incorporated into the uplift to the tariff, determined every year by the Department of Health.[210] Exclusion of drugs from the tariff means they are paid for separately after specific approval by PCTs, which is usually automatic if following NICE guidance. The overall effect is that there is an incentive for hospital clinicians to use NICE-approved new technologies, since such an approach brings in extra income to the trust, rather than weigh up the relative value against other more established forms of treatment covered within the tariff price. While this encourages the uptake of new technologies, it also means higher costs for PCTs.

What should be done

MEASURES TO BE TAKEN BY NICE AND PCTS

224. Some steps have been taken to help PCTs afford NICE guidance and provide funds for treatments which NICE has not assessed and, we were told, others could be taken, including:

Better communication of NICE's programme of work to PCTs;

Inclusion of the cost of all drugs in the tariff; and

Better evidence about cost- and clinical-effectiveness in all areas to enable PCTs to prioritise treatments more effectively.

225. The Department and NICE have attempted to help PCTs plan for the expenses associated with NICE guidance by providing information on NICE's work programme and new treatments in general.[211] NICE has also started developing commissioning guides and templates.[212]

226. Some argued that NICE could do more. The Academy of Medical Sciences recommended that better communication between NICE and PCTs could improve the situation:

…close communication between NICE and PCTs so that Trusts are financially prepared for the provision of new treatments…Advance preparation of all PCTs would reduce inconsistencies between those that provide a treatment and those that do not.[213]

227. As we have seen, the fact that high-cost drugs are not included in the tariff means that there is no incentive for hospital doctors to consider whether they represent value for money. Inclusion of all NICE-approved therapies within the tariff would share the financial risk and encourage caution since the average uplift could be kept closer to what is affordable in general. The Minister agreed that this approach should be considered:

The final roll-out does not cover all specialties until 2007-08. Obviously that could include high cost drugs and it could give greater certainty to PCTs. Overall it would not increase the funding. This is something, bearing in mind the roll-out finally through 2007-08, we are prepared to keep under consideration.[214]

228. Other witnesses told us that the limited evidence available on unassessed areas of healthcare meant that PCTs had difficulties deciding where to reduce spending as a result of mandatory NICE guidance.[215] This has led some PCTs to come together to pool knowledge to determine how best to commission in the areas of care that have not been assessed by NICE. Professor Devlin stated that the lack of information available on the value for money of these areas had prompted some PCTs to attempt to provide this evidence themselves:

One of the aspects of PCT decision-making that we found was a desperate demand for information on what services were poor value for money, what should it be disinvesting from, what would be the appropriate responses to cost pressures and so on. We found instances of PCTs working together to try to create some sort of information and evidence base on which to do that.[216]

229. Dr Lise Llewellyn, Chief Executive of Berkshire East PCT, told the Committee that this approach had been taken in her area:

Collectively PCTs have got systems and processes where they work together. In my patch we have got a public health unit that works across PCTs so that where there are concerns or queries about treatments or types of drugs, et cetera, we actually try and take a collective decision.[217]

SETTING THE THRESHOLD

230. Some witnesses claimed that the current NICE threshold was simply unsustainable.[218] As PCTs are forced to fund more technologies, the resources left over for PCTs to use as they see fit, will shrink so much that they are unable to tailor healthcare delivery to the needs of their own areas. Dr Kiran Patel told us:

There needs to be a level of local PCT freedom to do what is appropriate for its population.[219]

231. NICE is clearly concerned about the threshold it uses; a feasibility study is currently being undertaken by Professors Appleby and Devlin jointly with the Institute to examine the cost per QALY of decisions taken by PCTs to invest or disinvest in treatments. This will allow comparison of the cost of PCT decisions with those recommended by NICE.[220]

232. Witnesses argued that the problems outlined above could be mitigated if the threshold of cost-effectiveness was set independently of NICE. The Institute was certainly not established, nor constituted, to make political decisions of this kind.[221] Some suggested that the threshold should be set by Parliament.[222] Professors Devlin and Appleby argued that the NHS should be given independence on this matter in a manner similar to that of the Bank of England on interest rates. NICE, PCTs and other purchasing bodies would then be required to adopt this threshold. Professor Devlin told us that this could lead to a fairer and more efficient system:

We are suggesting that NICE's threshold is not just a matter for NICE alone: it is not just NICE's business. If NICE makes a mandatory decision that PCTs must implement it completely alters the bundle of services which PCTs can afford to deliver. That affects the services that all patients can potentially consume or benefit from, so NICE's threshold should have an input from the sector and a much wider range of expertise.[223]

233. Professor Rawlins spoke out against this suggestion, however, stating that this would substitute one system lacking in evidence for another, potentially less informed, one:

My own view is an independent body would have exactly the same difficulties we have had. They would have to use judgment about it because there is no empirical basis.[224]

He added that NICE had commissioned research on the subject, which would be reported towards the end of 2007.

Rationing

234. The question of the threshold used by NICE is key to any discussion of healthcare rationing. Ensuring that guidance is affordable to purchasers is vital if NICE is successfully to perform its vital role of helping to ensure that the NHS's budget is spent as cost-effectively as possible. The resources of the NHS are limited, and the funding increases of the last few years will not be repeated in 2008. The cost of new health technologies, including medical treatments and diagnostic and surgical procedures, is rising. At the same time, the population of the UK is ageing and with age comes increasing healthcare needs. As Professor Rawlins told us, "every single developed country has got the same problem".[225] Demand outstrips resources; as a result rationing occurs in all healthcare systems.[226] Several witnesses highlighted the consequent need for a true public discussion of healthcare rationing. As Professor Stirling Bryan told us:

I also believe there is a need for a more open and frank public conversation about rationing in healthcare. If one is to get the public engaged and supportive of NICE one needs to be much more open and frank with them about resource limits, how decisions are made and the fact that there are very difficult judgments to be made in terms of healthcare priority-setting and rationing and that NICE is part of that.[227]

Dr Keith Syrett, from the University of Bristol, also stated:

There remains a need to engage in a comprehensive debate about rationing in the NHS in order to secure public acceptance of the authority of limit-setting bodies such as NICE.[228]

He stressed that, "it is the function of government, not of NICE, to initiate such a public debate".[229]

235. To date, however, politicians have shied away from such a discussion, perhaps because it is uncomfortable to talk about the 'cost' of a life (ie. the threshold that should be used), or the maximum that should be paid to improve a patient's quality of life. The subject of cost does not sit easily with the concept of a NHS that is "free at the point of need". Professor Nancy Devlin stated:

It is a very difficult debate to get going in the current climate. A big policy issue within the NHS has been improving patient choice. Patients should be able to choose what treatments they get, where and when. That sort of theme around responsiveness and individualisation of treatment alongside a debate on rationing in the health service is somewhat uncomfortable, is it not?[230]

236. The problem is exacerbated by the threat of legal action, which has been used to override PCT rationing decisions, resulting in patients receiving the treatment they desire. On rare occasions, as we have seen, Ministers have intervened to achieve the same aim. Such actions may fuel public perception that rationing is wrong and that denial of treatment is in a sense cheating the patient. Dr Daphne Austin, a public health consultant, claimed:

The perception that it is wrong to deny treatment is fuelled by the fact that PCTs frequently step down when there is a real threat of legal action. This is interpreted as an acknowledgement of the PCT being 'in the wrong'.[231]

237. Mr Dillon agreed that NICE should play a role in discussions about rationing, and that "the realities of decision making" should be communicated:

I would quite like NICE to contribute to a broader debate about why it is necessary sometimes for those controversial decisions to be taken, and I think that has to be a debate that involves government with its stewardship responsibilities for the health service.[232]

Dr Austin concluded that without such a discussion, achieving affordability—and therefore determining the right threshold—was unlikely:

Until we can accept that not all needs can be met we will continue to have a distortion in health service priorities and, ironically, fail to get value for money overall.[233]

238. The Minister agreed that rationing was necessary, and that NICE should play a part in public discussion of the subject:

in a cash limited system we clearly cannot pay for absolutely everything so there needs to be an approach which attempts to prioritise on evidence what is available. NICE has a role to play in that in starting to shift the debate on to a more robust footing that is about what the evidence tells us.[234]

She added that "transparency and engaging" were needed to communicate to the public that not every medical treatment or procedure is available for them through the NHS. She also told us that "managing expectations", through communication and better understanding of local needs, was necessary.[235]

Recommendations

239. The threshold or ceiling NICE employs (measured in pounds sterling per QALY) to decide whether a treatment is cost-effective, and so should be available in the NHS, is not based on empirical research. Nor is the threshold directly related to the NHS budget, since the threshold has remained constant while the budget has increased hugely since 1999.

240. The threshold used by NICE does not take into account the funding decisions made by PCTs generally. For interventions not assessed by NICE, PCTs appear to use thresholds which vary from treatment to treatment but for the most part seem to be lower than the NICE threshold.

241. Many PCTs struggle to afford to implement NICE technology appraisals, as well as clinical guidelines. As more interventions are evaluated it is feared that the position will become unsustainable. Funding is essentially ring-fenced for technology appraisals, leaving PCTs little room for manoeuvre in their budgets to reflect local needs and priorities.

242. A number of steps were proposed by witnesses to alleviate the situation. To improve coordination between NICE and PCTs, we support the wider use of implementation consultants, who would provide information both from NICE to the PCTs and from the PCTs to NICE.

243. There must be incentives for clinicians to be very careful about the use of expensive drugs. We recommend that current exclusion of high-cost drugs from the payment by results tariff be reviewed.

244. It is difficult for individual PCTs to decide which areas to prioritise and in which to reduce spending when their expenditure rises as a result of new NICE guidance. In the absence of NICE guidance on disinvestment, we recommend that groups of PCTs should work together to determine appropriate areas of spending in consultation with the public. Such groups should also examine existing treatments to determine which are not cost-effective.

245. While the measures listed above would mitigate the problems PCTs face, the fundamental problem which has to be addressed, according to several witnesses, is NICE's cost-effectiveness threshold. Given the uncertainties, for example about the thresholds used by PCTs, we are not in a position to decide authoritatively whether the current threshold, or threshold range, is appropriate. We recommend that more work similar to that undertaken by Professor Smith and colleagues at York University takes place on the thresholds used by NICE. We are encouraged that NICE has commissioned its own research in this area.

246. During the inquiry, doubt was cast on whether NICE alone should continue to determine the level of the threshold. We consider the present situation is unsatisfactory. We recommend that a separate body, with representation from NICE, the Department, PCTs and others should set the level, or range, to be used. NICE's threshold should be closely linked to that used by PCTs. The threshold should also relate to the size of the NHS budget. The new body should decide whether orphan drugs continue to be treated differently from other treatments.

247. Demand for NHS services will always exceed the ability to meet it. Not every treatment can be provided to every person. NICE has a vital role to play in the rationing arrangements and, working with Government, should make clear to the public how and why such decisions are made.

179 Q 540 Back

180 Q 531 Back