UNCORRECTED TRANSCRIPT OF ORAL EVIDENCE To be published as HC 503-iv

House of COMMONS

MINUTES OF EVIDENCE

TAKEN BEFORE

HEALTH COMMITTEE

NATIONAL INSTITUTE FOR HEALTH AND CLINICAL EXCELLENCE

Thursday 11 OCTOBER 2007

MR SIMEON THORNTON, PROFESSOR KARL CLAXTON,

PROFESSOR ADRIAN TOWSE and PROFESSOR JON NICHOLL

Evidence heard in Public Questions 434 - 524

USE OF THE TRANSCRIPT

Oral Evidence

Taken before the Health Committee

on Thursday 11 October 2007

Members present

Rt Hon Kevin Barron, in the Chair

Charlotte Atkins

Jim Dowd

Sandra Gidley

Dr Doug Naysmith

Mr Lee Scott

Dr Howard Stoate

Dr Richard Taylor

________________

Witnesses: Mr Simeon Thornton, formerly of the OFT, Professor Karl Claxton, University of York, Professor Adrian Towse, Office of Health Economics, and Professor John Nicholl, University of Sheffield, gave evidence.

Q434 Chairman: Gentlemen, good morning. I welcome you to the fourth session of our inquiry into NICE. Perhaps for the record you would introduce yourselves and give your backgrounds.

Mr Thornton: I am Simeon Thornton. Until recently I worked at the OFT where I led the study into the pharmaceutical price regulation scheme. That called for reform of the scheme towards what we call value-based pricing. I no longer work at the OFT and so am not formally representing the office today, but obviously I am happy to talk about the analysis we did and the conclusions we reached.

Professor Nicholl: I am Jon Nicholl, professor of health service research and director of the medical care research unit at the University of Sheffield. By training I am not a doctor; I am a statistician/epidemiologist. I was involved in the early stages of the multiple sclerosis risk-sharing scheme.

Professor Claxton: I am Karl Claxton, professor of economics at the University of York. I am also a senior researcher in the Centre for Health Economics. I have been sitting on the NICE appraisal committee since it started work in 1999. I represented the committee on appeal and most recently at judicial review. I have been somewhat involved in the debate about the OFT report in a number of different ways.

Professor Towse: I am Adrian Towse, director of the Office of Health Economics which is a research and consulting organisation funded by the UK-based pharmaceutical industry, but we operate at arm's length from it with an independent policy and editorial board. I also have a visiting professorship at the University of York and a visiting research post at the University of Oxford. I am also a non-executive director of the Oxford Radcliffe NHS Trust.

Q435 Chairman: Whilst we have just one session this morning there are four witnesses. If we have a question that is open to all I will not be too upset if a particular witness does not answer it because what he intends to say has already been said by another. Obviously, we have some specific questions to individual witnesses. If the witnesses believe at any time that something has not been commented on they should feel free to do so. First, let me ask all of the witnesses a question on the PPRS. What are the costs and benefits of PPRS? Do you agree that "if it ain't broke" we should not endeavour to fix it? If it is "broke", why has the reform taken so long? This mechanism has been around for some 50 years.

Mr Thornton: The scheme has been around for about 50 years and it has evolved incrementally over that time. There tend to be five-yearly renegotiations. Until we launched our study there had not been an independent review of the PPRS. That is perhaps one of the answers. It is also a fairly complex mechanism and so from the outside to look at the scheme and its implications is probably a daunting prospect. Perhaps I may give some background to the conclusions we have reached because I think that is germane to the question whether or not it is "broke". The NHS spends about ₤8 billion a year on branded prescription pharmaceuticals and the PPRS is one of the main means that the UK Department of Health uses to constrain that expenditure, to deliver value for money to the NHS and also give good incentives to companies to invest. We launched our study a couple of years ago. The scheme itself comprises two main instruments: there is a profit control which constrains the amount of profits that a company can earn on its sales to the NHS and there is also a series of price controls. Unusually in the world, under the PPRS one can set one's own price as a manufacturer for a new active substance when one launches a product but there are constraints subsequently on increasing the price. Another main dimension of the scheme is the price cuts that are imposed periodically. Most recently, in 2005 there was a 7% price cut across the board. To come to our concerns about the scheme, primarily neither of those two instruments - the profit and price controls - takes any account of the value of the drugs that a company produces. For example, if a company has produced a lot of cost-effective products it will be subject to a 7% price cut just as much as a company that has not been so successful in innovating. The same sorts of arguments apply to the profit control. To the profit control one can add a lot of difficult and practical details about how one assesses profitability. I will not go into those unless there is great interest but I am happy to go into them if there is. In doing the study we became quickly aware that we could not rely on those principal arguments alone and needed to look at the extent to which under prevailing pricing arrangements prices did or did not reflect the value of drugs to patients. That was what we did. We gathered some experts across the NHS who helped us assess the clinical and cost-effectiveness of a handful of large-selling drugs. We found that in some cases a drug cost almost 10 times as much as an available substitute without offering appreciable additional benefits to patients. For 2005, the most recent year for which we have data, we calculated that in total over ₤600 million of expenditure on those drugs could have been used more cost-effectively. That means giving patients better access to the treatments they need and allowing companies better returns for investing in truly beneficial and cost-effective products. Is it "broke"? I think it is no longer fit for purpose and there is a compelling case for reform.

Professor Nicholl: I have nothing to say about PPRS. I do not know nearly enough about it. All I can say is that I am wholly in favour of the idea of value-based pricing. Conceptually, it is clearly the right way to go and the question is whether it is practical.

Professor Claxton: I suppose that the costs to the PPRS are in the long run and short run. In the short run what we have are prices that do not reflect value to the NHS. By "value" I mean the possibility of improving health outcomes overall for the population served by the NHS. But I think that the greater costs of PPRS are the long-term consequences in terms of dynamic efficiency that fail to align incentives for investment and innovation with value to the NHS. There are significant costs to PPRS and an opportunity to put it right using prices that reflect value to the NHS.

Professor Towse: I disagree. We need prices that reflect value, and we can discuss how we achieve that. The main challenge that Simeon and the OFT put to the PPRS was not one that it was meant to meet. The PPRS was never meant to be relating individual prices to value which has to be done somewhere else in the NHS. That is why NICE was set up amongst other things, which is something we have to discuss. We need to differentiate between the practical proposals made by the OFT to better align prices to the value of individual medicines and what the PPRS is supposed to do - what it says on the can - which is to provide an overall arrangement that keeps the medicines bill under reasonable control within the UK. As far as one can see, over the years it has been extremely successful in that. The UK is not a high-spender on pharmaceuticals. We come back to the uptake of individual medicines which has been a matter of concern. Not only is overall spending reasonable within the PPRS but average prices in the UK appear to be reasonable. If you look at that bit of the OFT report it shows that of 10 or 11 European countries the UK is fourth; in other words, it is in the top half but it is by no means at the top of European prices. If one is thinking about what the PPRS is supposed to do - what it says on the tin - in terms of giving the Government some overall ability to manage the drug budget and have a reasonable set of average prices in the UK, it seems to me that it does that. The other thing PPRS is meant to do is encourage investment. Simeon has not raised the point, though it is very clear in the OFT report, that the office does not accept that is the case. If one looks at the success of the UK in attracting investment, we can go into the extent to which that is due to having good scientists and the extent to which it is due to a more general market environment. I think that both are important. I am very clear that we do not necessarily have prices which reflect the value of individual products. NICE has been doing things about it, but that is not what the PPRS is about. The PPRS is an overall agreement between the industry and the department which is designed to provide a general framework to ensure that the UK medicines bill is manageable, and I think it has been very successful in doing that.

Q436 Chairman: If it had not existed could it or would it have been introduced at the beginning of the 21^st century?

Mr Thornton: I think it is very unlikely that we would design something like that.

Professor Towse: The implication of that is that because we could not invent it today for a whole variety of reasons therefore it is not worth having. The reason we could not invent it today is because it has a 50-year history. It is equivalent to asking whether if we started again with no healthcare system we would create the NHS. The fact is that we have an institution nearly 60 years old that has huge benefits, but if we started again tomorrow we would lose a lot of what the NHS has built up over the years.

Q437 Chairman: I recognise that it has been around for 50 years and so it has some merit. Not many things have been around for 50 years.

Professor Towse: It is about the fact that it has underpinned a constructive engagement between industry and government.

Q438 Chairman: Professor Towse, do you agree that the PPRS rewards innovation for innovation's sake rather than for the benefit to patients? Should the NHS pay for innovation or should it be rewarded through another means?

Professor Towse: There are two or three points packaged up in that question. When the OFT makes that challenge it is saying that essentially the PPRS looks at total revenues and prices in the round in relation to profit and the capital that a company employs. In a sense, the implication is that it does not matter how good that product is: if the NHS is buying it then the companies will get returns on it, and that is literally true. The point is that the NHS should not be buying it if it does not represent good value, but that is not what the PPRS is about; that is what other bits of the NHS are about. The PPRS is about providing a backstop so that if these are products that the NHS does want they represent value for money. They have been through NICE and the NHS is using them. The PPRS provides a framework which says there is still a backstop guarantee to the Government that companies are not making unreasonable profits or that in aggregate prices will be too high compared with other European countries, for example. Literally, the challenge is correct but that is not what the PPRS is supposed to do. The PPRS is not there to solve all the issues that the NHS faces in ensuring that it is using its medicines efficiently. Other bits of the NHS, notably NICE, are very much in the vanguard of trying to achieve that.

Mr Thornton: I find that a slightly circular argument. It is being said that the PPRS does not currently look at relative prices and value and therefore it should not do so in future. As Professor Towse says, the PPRS in its stated objectives is about ensuring reasonable prices. I do not see how for a medicine one can consider in abstract what is a reasonable price for what effect that medicine has on a patient. Looking forward, the question is: can we do better with the pricing scheme? Currently, the main instrument of the scheme is the periodic price cuts. Those do not take account of the value of products.

Q439 Chairman: What about innovation which is one of the crucial issues, is it not? Is it innovation for innovation's sake or is it innovation for patients? Does it do the right thing?

Mr Thornton: I think it could be targeted much better. I think we have made a clear case that if the prices of drugs are looked at across the books in the UK they do not reflect value and do not give manufacturers a very strong incentive to invest in the most useful products for patients in future. One point we made in the report is that although the UK constitutes just under 4% of global demand other countries reference their prices to those in the UK. Those countries constitute about 25% of global demand, so the UK can make a difference in terms of global incentives to invest in useful products. I believe that the PPRS could be reformed to make more positive impact.

Professor Claxton: It is extremely important that we do not reward innovation for innovation's sake; it means that we pay twice for innovation and that we have insufficient incentives for the future. What we should reward are the benefits that flow from innovative products and value those benefits in terms of what the NHS stands to gain from overall health outcomes. Nobody pays for innovation for innovation's sake in any other market. It is rather like saying that everybody should be compelled to buy the first generation iPod on the basis that if they do maybe Apple will produce a product that people will buy without being compelled to do so. It over-incentivises innovation in things that are not of value to the healthcare system. We need a system that disincentivises those investments and innovations that are unlikely to be valuable, and that is good. We should disincentivise innovations that are of no value and reward those that are. To do that the private sector needs to anticipate the kind of prices that its products will be able to command if it invests and produces an innovative product that is beneficial. For that it needs prices that reflect value. To my mind, that is the real cost of the current PPRS and the benefit of value-based pricing.

Q440 Dr Naysmith: How does the analysis that you have just made apply to things like "me too" drugs where companies are anxious to use a drug similar to one that is already produced by another company because they know they will be able to market it? How does that fit in with what you have just said?

Professor Claxton: Under value-based pricing what we have is a situation where those who do research in a particular area and happen to be the first to market will get a premium price if it is beneficial. Those who come shortly afterwards will also get a similar premium price if it is equally beneficial. Those that come later will have a shorter patent protection to recoup their investment, but that incentivises the private sector to be quick to get it to market, which is what we want, and reduce the cost of development. For those that are real "me toos" where there is research around the patent and they come in some time after the originator, under value-based pricing they will have a short period to recoup any investment and their prices will then fall towards the generic; in other words, it will appropriately disincentivise research around the patent and "me toos" that do not offer incremental benefit. For those that do offer incremental benefits the price will reflect those benefits if they can be demonstrated with evidence. To me, that seems to be wholly appropriate.

Professor Towse: I should like to make a couple of quick points. The first is that it is important to distinguish, as Professor Claxton did, the initial "me toos" which are not really "me toos". What is happening is that one has some kind of scientific breakthrough and a whole series of companies competes against one another. Some will fail and two or three, maybe more, will get to the market within a very short space of time. As Professor Claxton said, if there is a big advantage in being first that is good because it adds to the pressure on companies to get there. Products that come later will face much more difficulty because the market is established. When the lead product comes off patent there is a very good generics market and the price of the leading product will fall quite dramatically. At that point you do not force down the prices of all the other drugs; it is simply that where it is right to do so the NHS switches to give patients the cheaper generic so they can be treated appropriately. You do not need to force down prices; you just need doctors to change their prescribing.

Dr Naysmith: I am not sure that clarifies the point.

Q441 Dr Stoate: As a cynical doctor, if only that were true! Probably as I speak there are legions of drug reps trawling the country trying to buck that particular piece of information. I want to move to something else which is rather more fundamental to what we are inquiring into. I should like to ask Professor Towse: on what do drug companies base their pricing in the UK? How do they decide on the price?

Professor Towse: The short answer is that you have to ask them, but what I hope they do is look at what they think the market will bear and what returns they can get. They link that into the analysis that Professor Claxton has done, for example, which is a good way of thinking about it. For what sort of indications and patient groups is the product likely to be cost-effective and at what price is it likely to be cost-effective? In other words, if you think that you have a drug that you hope will be successful there is a very high likelihood of NICE reviewing it at some point, or the SMC looking at it immediately. In terms of developing your price, because we are now beginning to have within the NHS good institutions that look for value and make strong recommendations about whether or not that value exists the hope is that companies essentially anticipate whether they have an evidence base that will support a particular price when they make a submission to the SMC or NICE.

Q442 Dr Stoate: What you said, which was interesting, was that they were priced on what the market would bear; in other words, we are talking about drug companies charging the highest price they can get away with.

Professor Towse: We started off the discussion about how companies got a return on their investment.

Q443 Dr Stoate: I did not ask that.

Professor Towse: I know you did not.

Q444 Dr Stoate: I asked how they decided on the price of their drugs.

Professor Towse: Where companies have good products we want them to get high returns because that encourages them to invest in more good products. Where they produce products that are not very advantageous we want them to get lower returns so they are not encouraged to invest more in those areas. How do we do that? At the moment in the UK we have a two-stage process. Companies come up with their prices and NICE, SMC and other parts of the NHS essentially ask: is that good value or not? If it is not good value the drug is not used. You have a straightforward choice: reduce the price or face the fact that your market share will be extremely limited and you will not get a return.

Q445 Dr Stoate: How do you justify the fact that there are such widely different prices in different countries?

Professor Towse: Then one goes back to international comparisons of pricing and reimbursement.

Q446 Dr Stoate: It is not that difficult to do. If you look at the price of a CD in any country of the world and you know the cost you can work out whether one is charging far more for it than another. The OFT has looked at that in the past. Why do we not do the same sort of thing with drugs?

Professor Towse: There are two stages. First, can we do the calculations? The answer is that we can. There are certain methodological issues and there is a whole annex in the OFT report that deals with them. The second stage is: whose price is right? Why are companies charging different prices in different markets?

Q447 Dr Stoate: How do we get to the price they charge? You said that it was because they got away with the maximum they could charge.

Professor Towse: We are getting into a series of complex issues. There are two obvious points to make. The first is that different governments have different mechanisms for setting prices, so to an extent it is not in the companies' hands. In many countries they are negotiating with governments about what sort of prices they can get. Therefore, they are administered prices and they do not necessarily reflect value. Second, let us suppose that every government effectively has a UK-type system, or a variant of the OFT system, whereby companies set prices and NICE or equivalent institutions look at value. Leaving aside parallel trade, one will still see different prices in different countries for two fundamental reasons. One is that there are different patterns of medical care; the other is that there are different costs associated with them. If one has a treatment that keeps someone out of hospital and the cost of the hospital day is quite different in Estonia from the cost in the UK one will not expect the same drug price to make that product cost-effective in those two countries.

Q448 Dr Stoate: Do you think that is the way they work? Do you believe that they try to bring in clinical effectiveness when they come to set the value of their products?

Professor Towse: You will have to ask them how they work, but the way I hope they work is that they know the NHS is looking for value and that it and NICE try to use cost per QALY thresholds, likewise the SMC; in other words, there is some understanding of what the NHS is willing to pay. The companies therefore have to look at the strength of the evidence they have and the benefits that they will provide to patients. They know what NICE is willing to pay, as it were, and they can set their price in response to that. My understanding is that what the OFT proposes is exactly the same process but in reverse, in that NICE looks at value, hands that dossier to the department and a price is then imposed which reflects that value. I do not think there is much difference between us as to the importance of identifying value. The difference between us is whether the company sets the price and the NHS decides whether that is value or whether the NHS decides value and passes a dossier to the department and the latter fixes the price.

Mr Thornton: I agree absolutely with what Professor Towse says in that our approach is fundamentally consistent. I agree with him that companies as profit-maximising entities will price according to what the market will bear. What we have tried to do in the report and through the reform of the PPRS is ensure that the market will bear only prices that reflect value, and that is not what happens at the moment. Therefore, one gets some very large selling drugs where the price is out of line with value, and that is the essence of our reform.

Q449 Mr Scott: Professor Towse, I am much more simplistic than probably most and new to the Committee, but if I understand you aright what you are saying is that if countries negotiate better perhaps they will get a cheaper price. I think that the particular bed cost in a country is irrelevant. The cost of producing a drug will be the same whether you sell it in Estonia or the United Kingdom. You are saying that they will get as much as they can for the drug in whatever country it is and whoever negotiates the hardest gets the best deal?

Professor Towse: If one goes back to where the OFT comes from, essentially we want to reward innovation. If successful negotiation comes down to forcing down the price to the cost of manufacturing a tablet then that is extremely low, but the company will not get a return on its research investment that is already sunk. The $800 million or ₤500 million that has gone into the cost of developing a product has gone. The issue is: how will the company get a return on that product? Essentially, in the NHS what we are trying to create - we differ on how to get there - is a situation where the company is entitled to charge higher prices if the drug is delivering more health benefit to patients. If the NHS took the view that it should just halve the prices of medicines in the UK the effect of that, as previously discussed, would be that the UK prices would impact the prices of drugs in other parts of the world. We would reduce the returns to innovation and signal to companies not to invest in these products. We are saying that we want them but we are not prepared to give industry a return on them. It is about rewarding innovation. The debate we are having here is: how does the NHS identify and achieve prices that reflect innovation? The starting point must be that if a company develops something that gives a lot of benefit to NHS patients it is entitled to a decent return for it.

Q450 Jim Dowd: Clearly, the idea of recovering research costs is critical, but a pharmaceutical company is essentially a research entity. It produces a product as a result of its research which it then sells commercially and that sustains the whole organisation. But the research activity is a permanent and significant part of its activities. How can it simply compartmentalise different parts of that research and say that these are its costs associated solely with one particular product?

Professor Towse: I do not think it can. We are not talking about a price that reflects a bit of the cost of research and a bit of the cost of manufacture because the research programmes, depending on how far back one goes to the basic science, are in particular areas. They will have lead compounds which may fail in phase I, II or III, so there will be products at various stages of development. It is very hard to do. One has to make assumptions as to how many failures there will be for every successful product. How many programmes does one have in the lab before one gets any lead candidates? The $800 million and ₤500 million to which I referred are based on a whole series of assumptions as to what has gone before in order to achieve a successful product. They are plausible estimates but nonetheless estimates.

Q451 Dr Taylor: I am struggling tremendously. I agree that somehow prices should reflect value. What bothers me is how you calculate it. To ask Professor Claxton specifically, how long it would take to make a judgment of value.

Professor Claxton: There are two questions: first, how we define and measure value; second, at what point that judgment is made in a robust way. As to the first part, if we accept that the primary purpose of the NHS is to improve health outcomes for the entire population it serves then the question of value is a simple one: do the health outcomes that we expect to gain from using a new technology exceed the health outcomes that will be foregone elsewhere in the NHS due to its additional costs? At lower prices one will find that the outcomes one expects to gain will be greater than the health displaced elsewhere as other NHS activities and treatments are displaced. If that price is higher, however, the additional costs and the health foregone because of those costs will exceed the health that we gain and innovation at that price will damage the NHS and public health. That is an appropriate definition of value.

Q452 Dr Taylor: This is your diagram?

Professor Claxton: I suppose that it was a means of trying to communicate the principles that underpin a lot of NICE decision-making. That principle underpins much of the decision-making of NICE. Do the health gains exceed the health that is likely to be foregone elsewhere? How that leads to price is that at very high prices the health foregone will exceed the health gain and at lower prices the health gain will exceed the health foregone, and there will be a price at which the health one gains is just offset elsewhere. If you like, that is the maximum possible value of the innovation to the NHS. If we pay that price the NHS will neither gain nor be damaged by the innovation.

Q453 Dr Taylor: Will it ever be practicable to do that at the time of the launch of a drug?

Professor Claxton: In a sense, we are already doing this at NICE. Increasingly, we have been asked to look at things at launch. We do that at launch under the STA process. What is different about trying to make these assessments at launch? Clearly, that is where the evidence base is least mature, so we are most uncertain about costs, effects and the question of value. Although I thought the OFT report was excellent, one critical omission was a full discussion about the consequences of uncertainty. At launch we have a choice; we can accept the new innovative technology and immediately get the benefits for patients who receive them, but if we are uncertain we will forego the evidence we need to support its use. What we need to do is consider both the value of the technology and the value of acquiring more evidence. The value of acquiring more evidence essentially uses the same framework. Evidence is valuable because it allows better decisions to be made for patients in future and that leads to better health outcomes. We need to account both for the value of its use now and the value of acquiring the evidence that is needed to support its use in future and that needs to be reflected in prices. For example, a technology which appears to be cost-effective based on what we know is very uncertain and we need more evidence. There is high value in getting more evidence. If we cannot acquire that evidence whilst that product is being used then we should say that whilst it looks cost-effective we will withhold approval until we are provided with the evidence. That provides incentives either to the private sector to come to launch with the evidence or to cut the price. I do not believe I am explaining this very well. There is a relationship between price and uncertainty. The more uncertain one is at launch other things being equal one will have to lower the price.

Professor Towse: I agree with that. If one looks at NICE and SMC - it is harder in the case of SMC because there is less in the public domain - not only the cost per QALY threshold but the quality of the evidence base influence NICE's willingness to pay, so to that extent the quality of the evidence base - Professor Claxton is arguing that that trade-off should be much more explicit - already features in NICE's decisions. I am not sure I differ with Professor Claxton in saying, in regard to one of the OFT proposals for risk-sharing arrangements and doing some things after launch, that where the company and NICE agree that more data is required then more evidence of a certain type is required to reduce the uncertainty so NICE can be clear whether or not it is making the right decision. One has some sort of arrangement that allows the company to run with the price that NICE thinks is acceptable if that evidence base is there, and if the evidence shows something different there must be some kind of reconciliation. There needs to be a recognition that the quality of the evidence is important; price has to be related to what we know about it, but I do not think the bun fight must be at launch and nothing happens after that. We have to be a lot smarter than that. For some things we can reach provisional views at launch but additional evidence leads to revised views after launch.

Q454 Dr Taylor: What bothers me is how one acquires more evidence without the drug being widely used.

Professor Claxton: At launch we need to ask: do we need more evidence to support use? If so, what kind of evidence do we need? Can we acquire that type of evidence whilst it is being used widely in the NHS? Some types of evidence we can acquire from observational studies and registries, but it will be very difficult to acquire some types of evidence, for example evidence that needs experimental design and the random allocation of patients to two arms of a clinical trial to estimate relative effects. The reason it will be difficult to acquire that evidence if we approve it at that point is that essentially NHS patients have access to that technology for that indication. The clinical community will inevitably regard it as unethical to allocate patients so that some have the technology and some do not. Even if it was regarded as ethical I suspect that very few NHS patients would sign up to enter such a trial. There are some kinds of evidence that we will not be able to acquire if we approve the product. For that reason we need either to withhold approval until we acquire the evidence or the manufacturer must cut the price to the point at which the benefits of saying yes now exceed the value of the evidence that we forego; in other words, you make the product so cost-effective that we are no longer too worried about the uncertainty. That is the relationship between price, uncertainty and evidence. Professor Towse has painted a picture to show that NICE already does this very well. I have been involved with the institute for a very long time. I am not sure that we are really reflecting the notion of trading off the benefits of issuing positive guidance on a technology with the evidence that we may forego. If there is anything that we do badly at the moment I think it is that.

Q455 Dr Naysmith: Surely, there is another difficulty in your analysis. You have to forego something if it is an expensive medicine even if it is good value for money. How do you know what will be foregone because some other part of the NHS does that and you have no control over it?

Professor Claxton: Indeed. We need some assessment or what we forego for an additional ₤10,000 of NHS budget.

Q456 Dr Naysmith: But that does not exist at the moment.

Professor Claxton: Whatever we do and whenever a decision is made, those things are foregone. In principle, what should we look for to provide that estimate? Basically, we need to have a feel for the productivity of things that the NHS currently does. Nobody can know, not even a practising physician, who foregoes health benefits because of a decision that is made at the bedside, and NICE can never know that. We need to have an assessment of the productivity of existing activities and what may be foregone. I think that in the past those estimates have not been particularly secure. Increasingly, we see research and empirical estimates. I believe that Peter Smith from the University of York who gave evidence to this Committee before the Recess presented some recent work on programme budgeting and marginal analysis. That gives more secure estimates of what the cost-effectiveness threshold might be. Other work has been conducted by NICE. That evidence begins to suggest that maybe our current range of ₤20,000 to ₤30,000 per QALY is a little too high, but at heart it is a scientific, explicit, empirical question that we can investigate.

Q457 Dr Taylor: You used the expression "consequences of uncertainty". Can that take into account the disasters that have happened over drugs which have developed side-effects long after launch? I am thinking particularly of some of the antidepressants and anti-inflammatory drugs. Is there any way that drug companies should have to pay for the damage that these drugs have caused?

Professor Claxton: That question has two parts. As to the first, if the original analysis of costs and effects is adequate it should reflect all the possibilities at that time that people think may be important. It would not just reflect improvement in a primary endpoint in a phase III clinical trial but should include a model - that is why modelling is important - which reflects all those possibilities and the uncertainty in those possibilities. We can ask the question: which of these uncertainties will be most important in future? In principle, if we do robust analysis which reflects all those possibilities we should in principle capture those issues. That is not to say that things which were unanticipated at the time will not happen; they will. The critical thing at the point at which you have to make a decision is that you have fully reflected everything that you currently know about the product. As to whether we should get some form of payback, perhaps you will give me a minute to think about it.

Professor Towse: Perhaps I may make a couple of quick points whilst Professor Claxton is doing that. One is to do with the threshold. I am aware that you have an evidence session with Nancy Devlin and her colleagues next week. The work of Peter Smith is very important but let us be clear that it is a cross-sectional analysis. To take the area of cancer, if the PCTs that are less successful at getting QALY gains for their patients than the better PCTs that is the extra money they spend and the extra QALY gains that we get. I think that what NICE is trying to do is look at what the smart PCTs are doing. Should they adopt the new technology that has come out? In other words, we are trying to answer different but related questions. The other point, on which I know Professor Nicholl has strong views, is that Professor Claxton has raised the crucial question: what can we collect once the product is launched? It is unrealistic to assume that everything must be on the table before the drug is launched. We should not allow ourselves to get the NHS into a position where that is the case. We must be able to collect good quality evidence once the product is launched. If we take the point that Professor Claxton raises we have to think practically about how the NHS collects good evidence post-launch.

Chairman: In terms of questions and answers we seem to have strayed from the script. Perhaps Sandra Gidley can take us back to the OFT report and value-based pricing.

Q458 Sandra Gidley: Value-based pricing sounds like a good idea on the face of it. Would it work in practice?

Mr Thornton: You will not be surprised to hear that I think it will. In 2005 we looked at a handful of drugs for the purpose of estimating what could be saved in terms of the more cost-effective use of resources. The data we used to reach our conclusions is already in the public domain. Clearly, there is a lot we can do with the available data in terms of the existing stock of drugs. To unpack your question a little, one dimension is: what would be the cost of doing these sorts of assessments? We estimate that there will probably be a need for 20 appraisals to accommodate the new active substances the come onto the market each year. There will probably be a need for a further 70 appraisals to look on an ex post basis at existing drugs on the British National Formulary, so we need probably about 90 NICE-type reviews a year. We estimate that the additional cost of that will be about ₤9 million. When that is compared with the potential benefits and some of the numbers that we have been talking about the cost benefit analysis stacks up pretty favourably. Inevitably, there are uncertainties, some of which we have discussed, but the question is not whether we can reach a position where we have perfect knowledge - clearly, we cannot - but whether we can do better than we are doing currently. I think the answer is a resounding yes.

Professor Claxton: As to the principles of value-based pricing, NICE is already conducting the kind of analysis that would underpin such negotiation. The role of NICE as I see it is, if you like, to flesh out the NHS demand curve for those products. It would be that on which the menu of pricing guidance would be offered to the private sector, so in a sense we are already doing it. I believe that existing processes would have to be strengthened. The principles that underpin the STA are correct and provide a suitable basis, but at the moment all the appraisal committee has to do is decide whether or not it believes that at a given price the product is cost-effective. Is it above or below a certain threshold? Here the demands will be a little greater. We would have to specify exactly what our preferred analysis was and what we believed its cost-effectiveness was at certain prices. There will be deliberations about that with stakeholders, including the manufacturer. I have detailed in the briefing document the key things that I believe would need to be strengthened. If it is strengthened and rolled out I do not see the necessity at some point in the immediate future of conducting that kind of analysis for all new products and everything in that backlog. I believe that a much more practical way to roll it out would be to do it for all the new products coming through the door. When indication by indication comes up for an ex post assessment that can be rolled into the value-based price approach. It does not have to be all at once. Other options include the fact that manufacturers have a choice as to whether or not at launch they go to NICE, or a NICE-like body, to get guidance and associated price for the NHS. They can choose not to do so but they will not get positive NICE guidance. They do not have to go through that process; they can set any price they want but they will not get NICE guidance until they go through the process.

Q459 Sandra Gidley: Do you recommend a hybrid system?

Professor Claxton: I do not recommend either; I say that it is an option. My personal preference is that everything at launch goes through a strengthened STA process where there is a possibility of offering a menu of prices and associated NICE guidance. I do not really see it as necessary to do that all at once including the backlog; it can be done as new products come through the system. It will take time to build up entire coverage across NHS activities, but to do these kinds of analyses that is probably more practicable in terms of capacity.

Professor Towse: It seems to me that this assumes a degree of certainty about each component that we just do not have. The starting point is the assessment of value. There are even issues about definition. Do we include the productivity effects on carers or just patients, or just NHS and PSS costs? There is a whole debate about what is meant by "value". There is then a question of how good is the evidence base at any point in time. Then we assume that the Department of Health can translate NICE's value assessment using some clear rigid cost per QALY thresholds into a price. It is not as simple as that. In reality NICE does not just have a threshold - there is a lot of debate about what that is - but takes into account a lot of other factors: the nature of the disease, the impact on patients and so on. At each stage of the translation you do not have an automatic formula by which everyone, including the company, can tell what it will end up with. It seems to me that what we need is a better version of what we currently have. The company sets the price knowing that there will be a rigorous assessment of value by NICE, SMC or whoever. It knows that if it gets its calculations wrong either it will not be used by the NHS at all or it will be used for only a narrow sub-group of patients where it is cost-effective. If it wants to be used in a wider group it will have to go back and change its price or collect better evidence. Yes, I want value-based pricing, but it seems to me that the formula that has been put forward is not necessary to accomplish it. We have to get smarter at using the institutions we currently have. I agree with a lot that Professor Claxton puts forward, but it is about making NICE, SMC and other related institutions more effective in what they do. You used the word "hybrid". As you are probably aware from my paper, what makes sense to me is some kind of hybrid scheme. Initially, you keep the PPRS in some form as an overall backstop control, but it may well be that for certain purposes you exempt from the PPRS products that have been treated in particular ways by NICE, for example price cuts. There may be some sort of hybrid scheme that enables the NHS to focus upon and reward value better but does not involve having to review every medicine. It is not just every medicine at launch. What the OFT proposes is that every time there is new evidence or a product goes off patent in a therapy class there is another review. We are talking about lots of reviews and I do not believe that is a sensible use of NHS resources.

Q460 Sandra Gidley: Does Professor Nicholl want to say anything? At the beginning he was very positive in thinking it would work but he has not said much.

Professor Nicholl: I do not think I did say that. If you recall, I said that the key issue was the practicality of it. I have said nothing because I have nothing to add. The main lines have been very well expressed to you. The key lines are to do with uncertainty. If we do not know the value then we do not know the price. Let us be clear that whatever line we pursue the scheme must be acceptable to industry. It is quite difficult to see how industry will be happy with a process of price setting that is based on what I think are nearly always large uncertainties. The question then becomes: can we resolve the uncertainties? I have doubts whether we can do that generally speaking. In some cases we shall be able to do so; in most cases unfortunately I do not believe we can. Although I believe that the principle of value-based pricing is absolutely right and we should do it whenever we can, practically we shall run up against some problems.

Q461 Sandra Gidley: Will there be a problem inasmuch as at the moment there is quite a lot of argument or discussion - call it what you will - about QALYs and whether or not they take into account everything? That is now a little more of a precise science than it once was. When it was first introduced it appeared to be more a black art than science. But if one then introduces the highly vague concept of value how does one get round the problem of transparency and the ability to challenge any of the decisions that are made? Should there be an appeal process? I can see that some drug companies will not be happy about it at all. Some may say it is good, but a patient who wants a particular therapy may be unhappy with the overall outcome.

Professor Claxton: I do not think that the definition of value is vague; it is around the issue of whether or not it is cost-effective on the basis of the evidence available. Do we think that the health gain from using this technology is greater than the health that we believe may be foregone? Those matters are uncertain, but the definition and the means and methods to measure it are not vague. There are some arguments about what the appropriate methods are, but those arguments can and have been resolved through the NICE methods guidance. That becomes the explicit rule of the game and it will be revised through a process that is predictable and based on advances in scientific method. All that is very predictable and not vague. That kind of analysis will flesh out the demand curve for a particular product and indication and there will be lots of arguments about what the scientific evidence implies as to demand, cost-effectiveness and, therefore, the prices they can charge. That deliberation and the potential conflicts between stakeholders exist now under the NICE process and under value-based pricing they will still exist, but those deliberations will be about what the science implies; they will not be political negotiations but scientific questions.

Q462 Sandra Gidley: You mentioned a particular indication, which is quite an important point. There are some therapies used for a variety of indications and one can attribute a different value. If they are all looked at in isolation one can have a different value for each. How does a final price emerge from all of that?

Mr Thornton: Within the constraints of the information we have currently there is a principled answer and also a practical one. Ideally, when we get better IT systems rolled out we will have information on the sort of indication against which a drug is prescribed. One can then explicitly price on a per indication basis. We have the information about the value for a particular indication; what we do not have in terms of aggregate prescribing over the course of a year is information about indications for which a particular drug is prescribed. In the absence of that what we have called for is essentially a more flexible pricing structure. One may have an hypothesis about the volumes of prescribing that are likely to relate to particular patient sub-groups or an indication for which a drug is particularly cost-effective. What we suggest is that one should be able to negotiate an agreement under which one pays a premium price up to a certain volume of drugs prescribed and a lower price thereafter. Currently, we have enforced a linear price over volumes prescribed. If one removed that constraint one could try to map imperfectly, but to a better extent than at present, the value that the NHS gets in prescribing a drug where, say, a disease may have very different levels of effectiveness for different patient sub-groups and different indications.

Professor Nicholl: It is true that the effectiveness of any treatment, not just drugs - incidentally, what we are talking about here is a very big question that needs to be thought through - varies with indication; indeed, it varies between individuals, so there is a ludicrous endpoint at which there is a different price for treating every patient. Clearly, nobody can go down that route. Basically, it sets a price for the average value for whatever indications are agreed to be included in that treatment. The argument is also a little theoretical because the process of trials and studies of cost-effectiveness by which value is estimated is not only inherently uncertain; it is very poor at distinguishing between cost-effectiveness in different sub-groups and different indications. The moment one starts to make those sorts of distinctions one gets into areas where one does not have sufficient information anyway. I think that to go down that route is a bit of a red herring.

Professor Claxton: I do not agree with Professor Nicholl or Mr Thornton. That is another point on which I part company with the OFT report. What Mr Thornton suggests essentially is that the private sector captures the entire value of the innovation for the NHS. It is called perfect price discrimination. You charge the high prices for those sub-groups where the value of the product is very high and then you get a lower price for the next sub-group and a lower price still for the least valuable sub-group. Essentially, the private sector captures the entire value within patent protection, which means that the NHS does not benefit from innovation during the patent protection period and the company takes the entire value of the product. I think that is inappropriate for a variety of reasons that we can discuss. I believe that a much more appropriate approach to whether we have different sub-groups or different indications is to offer a menu of prices and associated guidance to the manufacturer and it chooses one price that is common across sub-groups and indications. It is the manufacturer's choice and presumably it will choose the price that maximises revenue, but there will be a single price. What it means is that it will get some share of the value of that product but the NHS will also share some of the benefits. There will be positive net health benefits for those more valuable sub-groups. I am not sure I have explained that very well. I usually have the assistance of a little diagram which is in the briefing document.

Professor Towse: Professor Claxton makes a valid point but it seems to me that that can be dealt with. In principle, what the OFT says makes a lot of sense. We do not yet have computers systems that are good enough to know for which types of patient products are being used. Then you have some sort of price/volume relationship that recognises the most targeted sub-group where you get greatest value, but beyond that prescribing is going into sub‑groups for which there is less benefit and the price should reflect that. Professor Claxton is saying one must be careful that essentially one is not over-paying through the process and there are other means to do that. This is something that is done in Australia which is not noted for being amazingly generous towards the pharmaceutical industry. That arises precisely because in Australia there are no GP budgets or any of the controls and in a sense doctors have an incentive to prescribe more because that is the way they get their patients. It is therefore important for the national authorities to have some degree of comfort that when they have assessed whether or not something is cost-effective they have boundaries to ensure it is not being used for other indications where they think it is less cost-effective. There is a lot of important innovative stuff in the OFT report, but I am opposed to saying (a) that the existing PPRS is a waste of space, which it is not, and (b) that the solution is for the department to set the price of every product in the country. I do not believe that that is necessary or sensible.

Q463 Sandra Gidley: It appears to be almost taken for granted that NICE will have a role in this; likewise, presumably the SMC in Scotland, which has been mentioned, and maybe the All Wales Medicines Strategy Group. Would they have to work together? In Wales and Scotland they are devolved bodies. Could we end up with different prices in different parts of the NHS, or would it all come together very nicely?

Mr Thornton: There are three sorts of considerations that need to be taken into account in thinking about how those institutions work together. The first observation is that currently we have a single UK list price and we want to maintain that in future. If one does not have that one has parallel trade across the UK and supply side instability, so that is not a good thing. Second, under current primary legislation it is the Secretary of State for Health who has power to set prices; it cannot be devolved to another body. Therefore, in lieu of any additional legislation that would need to be the case. The third observation is that in thinking of institutional design we wanted to make the best use we could of available expertise in each of those three bodies, SMC, AWMSG and NICE. I believe that the design we came up with meets those three criteria. Essentially, we are talking about an appraisal system that will be co‑ordinated between the three bodies, so each will be doing different appraisals, thus avoiding the overlap that exists in the current system. That would need to be co‑ordinated by the UK health departments. They would provide analysis on the cost-effectiveness of drugs at prices proposed by the manufacturer. There would then be an input into the pricing decision of the Secretary of State and in practice that would be a pricing unit within the Department of Health. That has to happen within current legislative constraints because, even if one wanted it to be the case, NICE could not set the price without primary legislation. That is how the system would work. I believe that it is a good mix of recognition of the entirely legitimate desire to have certain functions devolved but also maintenance of the UK price.

Professor Towse: I do not believe that is true. Essentially, this is killing devolution. What is being said is that one is setting a price and there can be only one. I agree with that. Therefore, what one must do is get all these bodies working together. What is being said is that there must be common agreement about which types of patients will get access to the product. Under the existing system the SMC can reach an opinion which is different from NICE's. They do not do so very often but it may be they take a different view of the evidence - they do not appear to do so very often - or that they have different health priorities, but the point is that they can say that as far as they are concerned at the UK price they want the drug to be available for certain sub-groups of patients and NICE on behalf of England may take a different view.

Professor Nicholl: I think that this slightly academic argument really resolves the question of the PPRS versus value-based pricing. Under devolution and PPRS each of the devolved administrations can have its own price. That is the place where one has potential price differentials. If each of the devolved administrations conducted its own PPRS it could have its own price. If however there is value-based pricing the value really does not change when you cross the Border. Welsh people respond to drugs in the same way as English people. Under value-based pricing one gets the same price and hence gets rid of the problems.

Professor Claxton: This does not have to undermine devolution. As Professor Nicholl quite rightly said, by and large people who live in Wales probably respond in similar ways, so we can have a single price based on a common analysis of cost and effect, and that can be negotiated, and it can also be devolved. Given the common price that has been agreed, Scotland and Wales may wish to issue different guidance about the use of the product in those countries. I do not see that that necessarily undermines devolution.

Professor Towse: Professor Claxton cannot have it both ways. You have a national combined effort where everybody reaches an assessment of value - for which groups of patients it work or does not work - and then the department sets the price, and then you say that it goes back and makes a different calculations. Either there is a central view or there is not.

Q464 Dr Naysmith: If someone who sits on the fringes of this discussion can come in for a moment, it comes back to what we talked about earlier. The value can be different in different countries depending on what is displaced by the agreed price.

Professor Towse: Exactly.

Q465 Dr Naysmith: For instance, in Wales medicines to do with respiratory disease such as lung cancer and so on will be of much more importance than in, say, most parts of England.

Professor Towse: Yes; the cost-effectiveness threshold might be different.

Q466 Dr Naysmith: Is it not the case that they might decide to treat it differently?

Professor Claxton: Absolutely. That was why I suggested that we could have a single price and still allow Wales and Scotland to issue guidance that is different from NICE's. The alternative and other extreme is to say that devolution means we are independent countries which have different pricing and guidance, but essentially that is a political decision. I do not think that value-based pricing means an end to devolution.

Mr Thornton: Just to be clear about the status quo, currently we do not have different list prices across the UK. Hospitals pay different transaction prices but the list prices are the same.

Q467 Jim Dowd: That is because we have a national health service across the whole UK?

Mr Thornton: Yes.

Professor Towse: Perhaps I may make a quick point about the ₤500 million savings which Mr Thornton increased to ₤600 million.

Chairman: That question is quite a bit further down. I am not sure whether we are meant to cover all of them now. We have about two hours to go in the morning session. I should like to get through these questions and have quick answers to them, if they have not already been covered. Obviously, there will be a transcript of these proceedings and you can always come back if you believe that you have not had an opportunity to answer a question asked late in the session or it has been misrepresented. I ask my colleague Dr Naysmith to move to Q6.

Q468 Dr Naysmith: This question has been answered in part already. If value-based pricing were to come in obviously it would have an influence on NICE and the way it currently works in producing appraisals and guidance. Would it be necessary to recast the whole thing, or could it be grafted on without too much trouble?

Professor Claxton: As I think I indicated earlier, I do not believe that it needs to be changed fundamentally. There are particular parts of the process that need to be strengthened and those are set out in the briefing document. In my view STA is not yet a suitable vehicle for value-based price negotiation and it needs to be strengthened. NICE is the kind of assessment authority that one would envisage doing this work.

Q469 Dr Naysmith: Mr Thornton, do you think it can be grafted on fairly easily?

Mr Thornton: In terms of the institutional design and who does what I would refer you to the answer I gave a moment ago. Each of the existing bodies would contribute in terms of the cost‑effectiveness analysis. There is an issue of consistency of approach and co‑ordination. Everyone recognises that clearly NICE has a much more in-depth assessment process compared with something like SMC, but if one is looking across the board at all current drugs there will be some areas that are much less contentious than others and there the lighter touch and more pragmatic approach of SMC will fit in well.

Q470 Charlotte Atkins: As I understand it, at the moment prices are negotiated on the basis of a company's whole portfolio of drugs. Is there any reason why individual products should not have their prices negotiated with the NHS? Why does it have to be a complete portfolio?

Professor Claxton: I entirely agree that the price of a particular drug should reflect its benefits to the NHS. It is by that that we align incentives for future investment with value. If you think about it, it does not happen in any other market. Even in markets with patent protection prices reflect value to customers. Here in the NHS it is not individual customers but patients and the value to patients is signalled on behalf of a central authority that assesses overall benefits and cost to the health service.

Q471 Charlotte Atkins: Are all the witnesses agreed on that?

Professor Nicholl: I am.

Professor Towse: I am not. It assumes that we have a definition of value and the evidence and all of that works. It also assumes that we have a rigid cost-effectiveness of the threshold on which everyone is agreed - that is the bit which is done by NICE - and that effectively the department has the equivalent of Professor Claxton's demand curve and these straightforward choices can be made. It is much more blurred than that. It does not mean we should not try to reflect value but we should do it the other way round. Related to that is the uncertainty that is generated if government starts to set prices. Governments should look at prices and see whether or not they want to buy things. That is the appropriate way to do it. Companies already have huge volume uncertainty when they launch or research medicines. Will they be able to sell them? At what sort of prices will they be able to sell these medicines? If there is uncertainty about what your competitors are doing, uncertainty about how much value you will generate when the things are in clinical trials and in addition government then takes its own view on what your price should be you introduce an element of additional uncertainty that has not existed in the UK system before. I do not think that that adds anything.

Q472 Charlotte Atkins: Surely, this is just the way in which the pharmaceutical companies ensure there is no transparency?

Professor Towse: No transparency about what?

Q473 Charlotte Atkins: In terms of how they arrive at what the appropriate price is.

Professor Towse: What we are arguing about is which way round it happens. At the moment the company comes up with a price. If the Secretary of State refers it to NICE or the SMC looks at everything there is an assessment based on evidence as to whether that product provides health benefits that justify the price. If it does not it does not get used. The fundamental difference between me and Mr Thornton and perhaps Professor Claxton is that I believe it is volume that should take the strain. If the price does not reflect value the company does not sell the product; volume is not there. Likewise, if the value is reflected in the price the volume should be there and the NHS should be using the product. We have the same issue on lack of uptake in the NHS as here. If the rest of the system does not respond to recommendations that something is not good value and its use should be stopped or something is of good value and it should be used then to mess about with the price will not change what happens in the NHS, which is what we are trying to accomplish.

Professor Claxton: Professor Towse really cannot have it both ways. On the one hand, his response to how in principle prices are set by the pharmaceutical industry is that it anticipates how the NHS perceives the value of its products and guesses at the price that will be acceptable at volumes that will give it a return on investment. Therefore, the industry anticipates all this process that we have agreed is complex and has lots of uncertainties and methodological issues which sometimes it gets it right and sometimes wrong but on average it gets it right. On the one hand the pharmaceutical industry anticipates the whole implicit process; on the other the notion that all we should worry about is volume and we should forget about price basically says that it is volume that should adjust rather than price. Basically, he is saying that all parts of the NHS including the practising GPs have the time to assess cost and effect and what the threshold is for the NHS and make appropriate decisions to adjust their volume of prescribing. First, that does not happen; second, even if it could happen it would be grossly inefficient and an inappropriate use of GPs' time because they should be developing relationships with their patients. All that value-based pricing suggests is that it makes things more certain and provides clearer signals; in other words, instead of the private sector having to second guess how NICE will regard its product and its value we imbed that in the assessment of value and the offer of a menu of price and guidance. Therefore, we make it much less uncertain for the private sector in communicating and signalling value. That is really the job of NICE and the NHS. It is their duty to signal clearly and predictably value and then the private sector can make decisions.

Professor Towse: I am sure that you will talk to companies, but my sense is that not all of them are desperate to have governments fixing prices for them. Professor Claxton is right that there is a lot of uncertainty, but the issue is: is the department the best body to absorb all those risks and make those decisions, or is it ultimately a commercial matter for companies and, if they get it wrong, they suffer the consequences? One cannot escape from the question of volume. It is not about whether GPs understand the price of everything. If Estatin comes off patent which means that it is possible to put most patients on a cheaper statin, or NICE recommends the uptake of a new cancer treatment, something must happen in the rest of the NHS to translate it. If one accepts the idea that one solves all of these problems by a price mechanism, unless the NHS responds positively and negatively to what NICE does one will not change the treatments that patients get. If we can solve that we do not need to start fixing prices.

Professor Nicholl: This is a great spat between economists which some of us find enjoyable. I think that we are slightly straying from the point. No one is saying either that the NHS should set the price and that is it or that the companies should set it; rather, the value should be identified and there should be a negotiation based on the value.

Q474 Charlotte Atkins: Do you want to add to that?

Mr Thornton: I agree. I also agree that to say GPs should respond and therefore there is no need for value-based prices is a false dichotomy. I think we need both. That is an empirical question. Is it the case currently that when for example a close substitute product goes off patent and the price falls dramatically people switch as much as they should do? The evidence suggests that they do not.

Q475 Jim Dowd: The difficulty with that approach is that there needs to be a common definition of value before you can negotiate round it. I am sure that the pharmaceutical companies' attitude towards the value to them and the attitude of the heath service to the value to patients will be two entirely different things.

Mr Thornton: Obviously, we have been talking about the methodological issues and probably presenting things as more complex in essence than they need to be. Fundamentally, the notion of value is about improving someone's quality or length of life. It is also a question about whether one includes quality of life benefits for carers, for example. Those are real, substantive issues, but fundamentally that is what we mean when we say "value", and I think there would be a broad consensus across NHS and the industry that we want to target those two factors. That is meant to be encapsulated in a QALY. Clearly, there are methodological issues about how one assesses quality of life, and we probably do not have time to go into those. In essence, I think we are all agreed that that is what is meant by "value".

Q476 Jim Dowd: I suspect that with three professors and an economist before us our progress would have been somewhat studious. I just want to look at the issue of price at drug launch. If it was agreed with NICE prior to launch would it not end the issue of NICE "blight" of which the pharmaceutical companies complain and also make medicines much more speedily available to the public?

Professor Claxton: I believe that is absolutely right as long as the pricing issue and value also account for the need for evidence to support its use in the NHS. If we do not do that there is a danger that we will accept product at launch when we are very uncertain and we close down the possibilities of doing research to find out whether it really does work and is as cost-effective as it is claimed to be. With that proviso, I think that is right. If we incorporate into the negotiations the value of evidence not only do we get products earlier, which is what
Cooksey envisages, but we also provide incentives for them to do better and good valid research earlier, because those that come to launch with better and more relevant research which reduce uncertainty will achieve higher prices than those who do not. We also have an opportunity to incentivise that critical path for drug development envisaged in the Cooksey report.

Q477 Jim Dowd: If I understand you correctly, is your specification prior approval to launch with an agreed review process?

Professor Claxton: Yes. At launch the appraisal would begin and the demand curve based on cost-effectiveness for the NHS would be displayed to industry. There would be a debate, as with all NICE guidance, and deliberations about methodological issues and what the evidence implied. There would then be an agreed set of prices and guidance which would also reflect how much evidence the NHS needed to supports its use.

Q478 Jim Dowd: Mr Thornton, I see you nodding, so I presume you agree?

Mr Thornton: Yes, absolutely. In the study we looked at certain examples of drugs. There was some evidence about NICE blight which showed that when there was a positive NICE decision primarily on hospital cancer drugs uptake was improved considerably. The corollary of that is that if a drug is cost-effective it will need an assessment sooner rather than later so that patients get access to that beneficial drug.

Q479 Jim Dowd: Can you suggest any mechanisms beyond that which might improve the speed with which new medicines are taken up by the NHS?

Mr Thornton: Obviously, the report was about NICE's role in relation to pricing et cetera and that was why we focused on the benefits of an early view on cost-effectiveness. Clearly, that is not the end of the story; there is a much broader piece about clinicians interpreting and presenting guidance in a more user-friendly way, but that is probably outside my specific competence.

Q480 Jim Dowd: Presumably, what you mean is the speeding up of the process by which new effective medicines are taken up. I am surprised by the record of innovation uptake in the NHS. Sometimes we are derided for not taking up all these innovations, but the question is whether we take up effective innovations and medicines. If the ones that we are not taking up are not effective then we are doing ourselves some good. The problem is one of sorting out the issue of effectiveness in order to speed up the process. Essentially, Professor Claxton argues for a process which will try to get the robust evidence about cost and effectiveness earlier in the process and create incentives for that to happen.

Professor Towse: I agree with that, but, first, we must still recognise - it is all part of the Cooksey process - that we may end up needing things like risk-sharing arrangements. We will not have the evidence we need to resolve all the uncertainty at the point of launch. We have to tackle the issue of how to ensure we are capable of collecting evidence after launch. Second, there is nothing in the OFT report that will speed the uptake of medicines, and that is the whole problem. We keep coming back to volume. When something goes off patent you can solve the problem by ratcheting down all the prices so you no longer care which medicine the GP prescribes; in other words, the OFT has a pricing solution to the problem of volume in one direction but it will not work. If NICE says that it is cost-effective the issue is: how do we ensure that it is still used in the NHS? We have to tackle that issue.

Professor Claxton: I suspect that my discussion about perfect price discrimination and the importance of having one price and sub-groups sounds a little esoteric, but it is very important to the notion of early uptake of a new and innovative product. We need to make sure that we price so that the NHS shares some of the net health benefits of that innovation. If we allow perfect price discrimination on balance the NHS neither gains nor loses. There is no incentive to increase uptake and we may as well sit back and wait until it goes generic. The NHS does not benefit. That is why it is so important that the pricing schemes reflect the market, the NHS gets some benefit and there is real incentive on the NHS for early uptake. Incentives are really about implementation. One of the proposals in my briefing document is that because there are uncertainties about what will happen to revenue for both the private sector and Treasury we can use value-based pricing to get a predicted spend on a particular drug or branded drugs across the NHS. It is possible to think of a way of monitoring that and having a rebate system where basically if the NHS does not deliver on the guidance that is part of the pricing negotiation it owes the industry some money. Equally, if industry sells too much of the product and markets beyond guidance industry will then owe the NHS some money. That will provide appropriate incentives both ways: there will be incentives on the NHS to make sure that the guidance it agrees is implemented and appropriate incentives on industry that it does not continue to market beyond that guidance.

Professor Towse: I have not seen Professor Claxton's note. To be clear, what I believe he is saying in effect is that the NHS is committing to buying a certain quantity and to pay the company anyway, irrespective of whether or not it does, at which point the incentives for local prescribers change because in effect the PCT says that it is paying for it anyway so it should be used. Is that how Professor Claxton envisages it will work?

Professor Claxton: I think there are lots of different ways in which it could work. I do not suggest that that goes down to PCT level. It is an agreement between the Treasury and industry.

Professor Towse: But one still must get the NHS to use the medicine.

Q481 Chairman: The OFT report strongly hinted that there could be savings in this move of some ₤500 million per annum. Professor Claxton, in your evidence we are told that "it is perfectly possible that the overall NHS [spending] on branded drugs may increase". Are those statements compatible?

Mr Thornton: Perhaps I may go first. I agree with one aspect of Professor Claxton's view which is that value-based pricing is about getting relative prices correct. As a concept value-based pricing is consistent with doubling one's budget or halving the total budget; in other words, it is about making best use of available resources. Therefore, there is nothing in the principle of value-based pricing which says the NHS should spend ₤8 billion. One could easily design the scheme so that the NHS spent ₤16 billion, but that is an entirely separate question. Value-based pricing is about making the best use of available expenditure. In the study our estimates essentially looked at that question. We looked at a few large expenditure drugs and asked whether on the grounds of cost-effectiveness there was a case for, say, one of the statins costing a large multiple of the price of an available substitute, Simvastatin. We looked at the clinical and cost-effectiveness evidence and found that the answer was no. The sum you quoted, which Professor Towse questioned, is ₤640 million. In the report we said that it was over ₤500 million. For on-patent branded products it is about ₤575 million and ₤65 million for off-patent brands. That was just a snapshot and was not an attempt to look at all drugs on the books; it was simply a look at some drugs about which concerns as to their cost-effectiveness had been expressed. We are not talking here necessarily about reducing NHS expenditure or tax savings but making best use of that expenditure, so that ₤640 million could be used to give access to drugs for patients who do not currently have access to them on the grounds of cost-effectiveness; or it could be used to give access to patients for hip replacement operations or any of the other uses of those resources in the NHS. I want to distinguish between cutting the budget, which is an entirely separate question, and making best use of our resources which was what the report was about.

Q482 Dr Naysmith: Did the manufacturer of the statin that was much more expensive have a good reason for charging a higher price?

Mr Thornton: Obviously, I cannot speak for that manufacturer. There is a debate about the clinical evidence. Our conclusion was pretty clear on that front: the price difference was not justified.

Q483 Sandra Gidley: Was that based on the brand as against the generic?

Mr Thornton: Exactly. This is an important point that brings up one of the problems of the hybrid approach that we discussed earlier. I think it is important that we pick up on the drugs that are currently in the system. Under the current arrangements when a substitute goes off patent it is not the case that the volume of the existing on-patent substitutes changes dramatically and that has huge expenditure implications. Professor Claxton talked about our recommendations being over-generous to industry because of the way in which we want to remunerate it for benefits for particular indications, but we are talking here of expenditure that is clearly not cost-effectiveness against any of NICE's current measures.

Professor Claxton: I believe that the two are entirely consistent and both are possible. What we shall see with value-based pricing are savings. Those are not cash savings that the Treasury can take away from the NHS, hopefully, but savings which can be devoted to other uses from those products that are not valuable whose price must fall. On the other hand, it is hoped that we will see products in future which are valuable because of the incentives set up which will be able to command premium prices during patent protection. Therefore, on balance overall expenditure could well rise, but what we know for sure is that we will get more health improvement from that expenditure than we otherwise would without prices reflecting value. We have just reallocated the rewards and expenditure; we have reduced the rewards to the less valuable and increased those to the more valuable. Therefore, the two figures are entirely consistent one with the other.

Professor Towse: We are talking here about two points. The first is the ₤500 million headline which was clearly defined. The director general appeared next morning on the Today programme; there were headlines in every newspaper and it was very successful. But the small print gives the warning that value-based prices can go up as well as down. If we have value-based pricing in the NHS we do not know how much it will be spending. As Mr Thornton says, maybe the drugs bill will be ₤16 billion rather than ₤8 billion if we have value-based pricing. Almost certainly there are drugs in the existing system which, using the existing NICE cost per QALY threshold of ₤20,000 to ₤30,000, mean you will end up increasing the prices of those drugs. What we are talking about is changing relative prices and relating them to value. The point that is made very clear in the OFT report, as Mr Thornton has said, is that this is not about the overall drugs bill. If one is talking about potentially increasing the drugs bill by having more value-based pricing then one may want to have a hybrid system that provides a backstop - a PPRS control or other means - to enable the Department of Health to have some overall sense that the medicines bill is under control. The next point is the crux of the statin debate. We have statin A and B which are effectively at the same price. Let us say that the clinical evidence for one is that for some groups of patients it is more cost-effective than for another. The main product comes off patent and there is a big reduction in price because of generics. How does the NHS respond to that? The OFT response is very clear: one just pulls down the price of the drug. What I think should happen is that the NHS, as it is trying to do, should switch patients who do not need to be on the more expensive product to the cheaper product and those who do need the more expensive one should stay on it. The company then makes a commercial decision. Either it recognises that its market share will shrink because its product is cost-effective for only a small group of patients or, if it wants to try to compete for the whole market, it brings down its price. Mr Thornton correctly says that the NHS does not respond as well as it should to those situations by switching patients and does not save as much money as it should - hence the report of the National Audit Office and some of the initiatives announced by the department. The question is: do we think that is what the NHS has to get smarter at doing, or do we give up and just reduce all the prices so we do not care which product the GP gives out?

Q484 Chairman: I am tempted to say Professor Towse appears to argue that the best way to spread best practice in terms of statin prescribing is to change the price of the drugs as opposed to the habits of those who prescribe. Is that a crude analysis of what I have before me at this stage?

Mr Thornton: I believe that is a false choice and it is deliberately set up as such. Of course we want doctors to prescribe cost-effectively and efficiently and in the best interests of their patients. As Professor Claxton said, we cannot expect them to do these sorts of cost-effectiveness assessments in real time for all the drugs they use. That is why we need an instrument such as NICE to take a view on the cost-effectiveness of a particular product and see that reflected in price. The report is very clear: price is not the only thing to get right; we need to give the right guidance to prescribers. Price is however a very important thing to get right and currently we are not doing that.

Professor Claxton: I entirely agree. On the one hand, if we take Professor Towse at his word potentially we can just get rid of NICE and do not need it any more; we can rely on other parts of the NHS to anticipate value and adjust volumes. There are two parts to this question. I believe that the Department of Health quite rightly has decided that it is very useful for a national health service to have a central assessment of cost-effectiveness and to issue guidance to the NHS. All the value-based pricing proposals do is say that in that process we should also be able to offer to industry - because this is not about setting or negotiating prices - to reveal the combination of price and guidance that will be available to it. In my view industry should be able to make a free choice from that menu and, in signalling that from a central assessment authority such as NICE and the Department of Health, it will reduce the uncertainties faced by the private sector in making its assessment of value and the kinds of returns that it will get from its investment decisions. It is not that extreme and is not either or; it is a combination of price and volume and signalling value very effectively to the private sector.

Professor Towse: I am a strong supporter of NICE and believe that it should have a bigger role. I have never at any point suggested or implied that NICE should be abolished. When it comes to the crunch, NICE's health technology assessment of statins is: "When a decision has been made to prescribe a statin therapy should usually be initiated with a drug with a low acquisition cost." Likewise, we have NICE recommendations for the uptake of new drugs. We have to create an NHS that is in a position to respond to good central analysis of where value is. To me, that is the crunch issue and we cannot solve all of the issues by changing prices. Someone has to change behaviour.

Q485 Chairman: The pharmaceutical industry asserts that value-based pricing may lead to a reduction in its research and development investment in the United Kingdom. Why should we risk our third largest industry by changing the system for the pricing of drugs?

Mr Thornton: There are a few myths out there and it is as well to deal with them to begin with. The first myth is that the scheme provides explicit incentives to invest in the UK. That was a view we encountered particularly at the beginning of the study. That is not true; the scheme does not provide financial incentives for investment in the UK. If the profit cap were binding it would if anything have a dissuading effect on investment in the UK because of the way transfer pricing rules work, but it is unambiguously not the case that the scheme provides incentives to invest in the UK. Further, one could not provide any of those incentives under EU legislation. I believe everyone now agrees with that. As a second order argument, people say that maybe there is an implicit link between prices and where investment takes place because possibly companies can threaten to leave the UK if they do not get a favourable price for a particular product and so there is a loose bargaining relationship between prices and where investment takes place. As to whether that is credible, there are lots of explicit drivers for where investment takes place, such as having good clinical trial networks, financial incentives, tax incentives and a well skilled labour force. There is a question as to whether that sort of threat is really credible. The third point I make is that even if that threat is credible and there is a loose relationship between prices and where investment takes place, as our discussion just now shows there is nothing in value-based pricing that is necessarily associated with lower aggregate levels of expenditure on pharmaceuticals. What we are talking about is making best use of expenditure. Therefore, if there is a loose link between the returns one gets and where one does one's investment the UK is just as likely to attract the most innovative companies as it is to dissuade those that are less successful. I believe that we need to deal with this string of red herrings that is imbedded in the argument.

Q486 Dr Naysmith: Why do you think that the drug firms are opposed to the change, because they are not stupid?

Mr Thornton: I shall deal with public utterances that I have seen in evidence to this Committee. The general sense of it - others can intervene if I am wrong - is that it sounds good in principle but the question is whether or not it will work in practice. People tend to keep their cards fairly close to their chest, which is entirely consistent with the current situation in which we find ourselves. As you know, there is currently renegotiation taking place with the PPRS. Therefore, I do not expect anyone to reveal his cards at this stage.

Professor Towse: I am concerned that essentially the OFT is setting up another straw man. Nobody in the world thought there was a clause in the PPRS that said if one built a factory in Kent one could charge the NHS twice as much as one would otherwise be able to. Not surprisingly, the OFT has discovered that no such clause exists. The question is whether that creates a climate in which companies are more likely to invest in the UK. The second point is: how might that mechanism work? Essentially, Mr Thornton says that there is no such mechanism and he does not understand how it can work. There is no clause 22 which says one can get higher prices. The reality is that the most important thing if one is investing in, say, research is whether one has good scientists. If one is investing in clinical research the most important thing is: does one have good investigators? Can one recruit patients? But there happens to be more than one place in the world where one can do that. The UK does have good scientists, but so do other parts of the world. If one has a choice whereby one has available similar skill sets what determines whether one invests in the UK or somewhere else? Other things might come into play. Whether or not the particular healthcare system will reward one's innovation if one is successful is bound to be one of the factors that comes into play. In other words, if any government adopts the position that it wants the research investment but it will be damned if it is going to pay for industry's successful products, where companies have choices they will not invest there. It has an impact. As the OFT report says, prices in the UK have knock-on implications not just for UK users but for prices elsewhere in the world. The final point Mr Thornton made was that the system he proposed was even better than the one it would replace, ie companies with innovative products would get super-prices and so would make the UK even more attractive. The problem is not the linking of price to value; it is the fact that under his proposal the department would be setting the price. I would strongly support the idea that where companies set prices there should be an assessment of value. If you do not deliver value no one buys your product which is what commercial reality should be. What I believe creates unnecessary uncertainty is the reverse situation where the department fixes the price. As far as I can see, that is what causes the companies concern.

Q487 Chairman: Does not the department in using PPRS in this way affect the decision-making and setting of prices by companies on the basis that it potentially limits their profits?

Professor Towse: It does. The companies can set prices but in various ways they are constrained in changing them and whether they have to make pay-backs if they exceed the profit threshold. That is one part of the constraint. At the moment the most effective constraint which we are all agreed should be made more effective is the need to demonstrate that one is delivering value; otherwise, when industry sets its price it will not be able to sell any product to the NHS, or it will be extremely restricted. I do not think there is any disagreement among us that the NHS is getting better at identifying value through NICE and other institutions, but we need to enhance that process and get prices closer to value. I think the issue is which way round we do it.

Professor Nicholl: One question is whether companies will continue to invest in R&D in the UK. It is driven by the scientific and clinical research environment. I think that the scientific research environment remains excellent and is almost incomparable with anything available round the world. The clinical research environment is set to develop very rapidly and improve. There is every incentive for the companies to continue to conduct R&D here. As to why they are against the scheme, a rather sinister voice whispers to me that perhaps the products are not quite as effective as they would have us believe.

Professor Claxton: I agree with everything Mr Thornton said about incentives for the location of R&D. It is about infrastructure, public investment in fundamental research, translation of research and, very importantly, cost. The notion that this could lead to a relocation of R&D is a simple threat. First, is it credible? Second, even if it is credible is it appropriate for public policy to respond anyway? As to whether it is credible, they have located here. Do they have any strong financial incentives to relocate if their bluff is called? No. I cannot believe that that threat, if that is what it is, is credible. Even if it was credible, should we respond? That is really a political question and not really one for an economist. I strongly agree with Mr Thornton as to why we hear that now. Anybody worth his salt will not show his cards right now. We also need to recognise that value-based pricing will necessarily have winners and losers and that is the way it provides the incentives for dynamic efficiency. Those who believe that they may be losers and will not be the innovators producing valuable products for the NHS will not like it and they will lose. Some of the current incumbents and also new entrants into the market will be the new innovators of the future. Those are the ones who will gain. They may not have a voice right now and some may not even exist, but what we should do is provide an environment that rewards innovation, which is what value-based pricing does. Professor Towse appears to believe that in the current situation the private sector can anticipate value and set its prices accordingly. If that were true it would be great. All that value-based pricing does is say that instead of having to guess the value to the NHS industry will be told and it will be included in a process of acquiring evidence, analysing that evidence and there will be a deliberation about the relationship between price and guidance. If Professor Towse is right why not cut out the middle man and all the guesswork in trying to pitch a price that will provide NICE guidance and sit down in the same room with the same evidence, go through it and agree that industry can have one price with narrow guidance or it can charge a little less and will be given unrestricted coverage? There are huge benefits to the NHS in terms of getting it early and reducing the costs of picking the wrong price; there are also huge benefits to the private sector, making it much less uncertain. Therefore, my pitch to Professor Towse is that if that is what is happening why not cut out the middle man? Let us all be in the same room and reduce the uncertainty. When it comes to investment we should do everything we can to reduce unnecessary uncertainties. Some uncertainties remain and that is what motivates good investment decisions, but those that can be removed should be removed. We can remove that uncertainty.

Professor Towse: First, it seems to me that what Professor Claxton is looking for can be achieved by better dialogue within NICE. At the moment NICE and the companies are not allowed to talk price in any shape or form. Essentially, there is no ability for NICE to ask a company if it realises where it will end up if it sticks to a certain price. We have no means of circumventing that. I do not believe that the solution to it is to say that NICE will still have nothing to say about price and the department will then have a dialogue with the company based on NICE's results and essentially it will impose a price on the company. Second, I should like to deal with the idea that there are some nasty, sleepy companies that have not invented anything of any use and a whole bunch of companies we do not know about with lots of innovation which would love this new system. It is already the case that companies that believe they have good products lobby to have references to NICE. Companies that are innovative already try to use the existing system to get their value recognised. They are then frustrated because the use of the product is not as they would wish. I do not believe the big problem is that there is a bunch of useless, sleepy companies.

Q488 Jim Dowd: But they do that for all products, not just the good ones.

Professor Towse: If I was running a company and had a product that I did not think had very much chance of meeting NICE's cost-effectiveness threshold I would not be lobbying for it to be referred to NICE; I would hope to get sales quietly without having to do that.

Jim Dowd: Maybe you would but they do not.

Q489 Dr Naysmith: We have the current system until probably 2010. If a decision was made to adopt the new value-based pricing system as recommended by the OFT how long would it take for that to happen, and could there be a smooth transition from one to the other?

Mr Thornton: We envisaged that the new system would be phased in from 2010 which fits in with the timetable for scheme renegotiation. We have spoken about the costs of the new scheme. We estimate that it will cost ₤9 million extra a year. We have also spoken about how the institutions will work together. According to the estimates of costs that we undertook we believe that it is perfectly practical to implement the scheme we envisage from 2010. We have never claimed that the study is the final word in all the analysis of this. If further research emerges to suggest that we should not be looking at all of the drugs in the back catalogue but at 80% or 90% of the most valuable, fine; we will take that forward. In this discussion sometimes the assumption is made that if we cannot do everything perfectly we must do nothing and have across-the-board price cuts that take no account of value. That is certainly not our vision. We intend it to be a pragmatic process that is informed by evidence.

Q490 Dr Naysmith: When would the decision to go ahead with it have to be taken to meet the target of 2010?

Mr Thornton: Normally, the PPRS negotiations start about two years before the end of the process. Clearly, there is a renegotiation with the PPRS and surely that is the beginning of the discussion, so from now on there would have to be talk about the modalities and how things would be implemented.

Professor Towse: It would be do-able but complicated. Using formal government parlance, the OFT is not required to do a regulatory impact assessment. That is something with which the Government is currently struggling - hence its interim announcement setting out principles but saying that it is still trying to do more work on understanding the implications of how it will work if it is implemented. The crunch question is that at least from the point of view of the use of NHS resources what we would be doing is concentrating our health technology assessment resources on medicines. We would start off doing all new medicines. The SMC does that but with a process that is close to STA. I believe what Professor Claxton is saying is that that may not be enough. But the crucial point is also to make reviews and assessments post-launch when there is new clinical evidence or something comes off patent and the status of the therapy has to be reassessed. That is something which NICE does on occasion when the matter is referred by the Secretary of State. Here we are talking about that happening potentially routinely in therapeutic categories. How often would one be doing those reviews? My strong preference is to say that we should expand the investment in NICE and what we do on HTA. How much of that effort goes into drugs depends on whether we think that is where we get the best value. Is that where we will save most money or identify most health gains? Maybe it arises in some surgical procedures or diagnostic tests. We cannot assume that we should pre-empt a huge increase in HTA resources and spend it all on the review of medicines.

Dr Naysmith: On every occasion over the past six to eight years that we have had NICE before this Committee we have asked them to look some old treatments, tell us they are useless and get rid of them. It says that it will do that but when it comes before us it says it is under such pressure from new drugs that it never gets round to it. I suspect that is already one aspect of the system that is not really doing as well as it should be.

Q491 Charlotte Atkins: Is the quality and quantity of the trial data presently submitted by companies good enough or should it be improved? What can NICE do to make sure that it encourages higher quality trial data?

Professor Claxton: Perhaps I may answer that since I look at two or three appraisals every month. There is constant frustration about the relevance and quality of the clinical trial evidence. As to its relevance, there are inappropriate comparators, primary endpoints that are not easily linked to ultimate health outcomes and a shorter follow-up than we would like to see. There are inclusion and exclusion criteria that do not match the NHS. There are huge frustrations. Only yesterday we ended up with a set of trials in the particular area we are looking at which were conducted on a group of patients who did not have any particular problems about control of blood glucose, which meant that all of those trials were entirely irrelevant to the scope of the appraisal we were considering. In the end we had to base that on uncontrolled before and after evidence. That is some of the weakest forms of evidence one can find in clinical epidemiology. That is one of the reasons I see value-based pricing, if we can reflect the need for evidence, as a real opportunity to provide incentives to produce relevant data. The other frustration that is often encountered is that even when we have relevant trials of decent quality more often than not no research has been done to link the primary endpoint in the phase II and phase III trials to quality of life. I am absolutely astonished by that. It is quick, easy and cheap; it can be done in a few months and it will cost peanuts to do it, and yet often it is not there. One can be cynical and suggest that despite reams of text claiming a strong link between the endpoint and quality of life maybe there is no such link, because if there is why has that study not been done? I will stop moaning about the appraisals, but there is frustration that there appears to be no incentive to provide and generate the kind of evidence that is needed for decision at the point of launch or even beyond.

Q492 Charlotte Atkins: Do you think that the OFT's recommended changes would affect that and better research data would be provided?

Professor Claxton: They would if - it is an important "if" - there was explicit incorporation into the pricing and value decision of the need for evidence. If basically we ignore the uncertainty and just say that price can reflect the expected costs and effects and not take account of the need for evidence it can have the opposite effect. There will be no incentives on the private sector to produce that evidence.

Professor Nicholl: Current data is poor.

Mr Thornton: I endorse what Professor Claxton has said. I think that value-based pricing can give better incentives to companies to produce the sort of information that we need. When we look particularly at the effects on sub-groups and particular indications that is precisely where we believe we should give full value to companies for demonstrating benefits. I believe that is where I take issue with Professor Claxton's particular model of non-linear pricing. That aside, value-based pricing can give good incentives to provide useful information. I also believe that Cooksey's recommendations for earlier stage engagement between NICE and the companies is extremely important so that companies are clear as to what question they will be asked when the time comes to do a cost-effectiveness assessment.

Professor Nicholl: I think it is about setting up a framework and a completely different understanding of what is required in terms of evidence from the position at the moment. The discussions that we have had all morning, including this particular one, assume that the evidence is about effectiveness, but it is about cost just the same. It needs to be clearly understood that cost is not the same as price. When data does emerge almost without exception not only are there no long-term outcomes or quality-of-life evidence; there is no cost evidence in the broad sense of that word, that is, evidence not only about price but the cost consequences of the treatment to the NHS, which are very important. At the moment the quality of data on which to base pricing decisions is inadequate, but it could become adequate. I do not think that will happen unless we set up a framework in which the companies recognise that in order to achieve the price they desire they will have to produce the necessary robust evidence on both effectiveness and cost.

Professor Towse: I agree with everything that has been said on this matter. The question is: what does one do about it? It seems to me there are three parts to it. The first is the dialogue as in the Cooksey report. I believe that at the moment NICE has one pilot with the HMRA and one company. Can the concerns described by Professor Claxton that are likely to come up at the appraisal committee be signalled at an earlier stage so that the companies have an opportunity to do something about it? Second, let us suppose that they are. Here I agree with Professor Claxton that at the moment the quality of evidence has an impact on NICE decisions, but he says that it needs to be much more explicit. The third point is that when one looks at NICE's recommendations for further research one can analyse them and see what types of research he thinks should be done. Typically, a lot of it is stuff that one can do post-launch; it is head-to-head comparisons where there are competing therapies. One comes back to the question of when one can collect that evidence. A good source of understanding NICE's frustrations is the technology appraisal decisions to see what extra evidence NICE says ought to be collected in some form. A lot of that is stuff that one cannot do pre-launch. Obviously, the crucial matter raised by Professor Claxton - whether in clinical studies one can have the bio-marker or endpoint translated into something that is meaningful in terms of patient quality of life - that should be do-able.

Q493 Charlotte Atkins: The other important issue is when trial results are ambiguous or poor the results are often not published. Should that not be available to both NICE and researchers? What can be done to ensure that that happens and all companies have to register and publicly report all their trials, not just select the good ones?

Professor Nicholl: The key first step is registration and that is being addressed at the moment. All trials that are started which involve NHS patients or premises need to be registered. Quite rightly, the results of all those trials need to be made publicly available.

Q494 Charlotte Atkins: But there is a voluntary code, is there not, which is being ignored?

Professor Nicholl: Yes, so I understand.

Professor Claxton: Ever since NICE started it has been a source of frustration and concern that there is no power to demand to know all the trials that have been conducted or see that evidence in the way that a licensing authority would. Since the decisions that are made by NICE have a profound effect on use in the NHS that just does not seem to be appropriate. The decision that is made can have a profound effect on cost and outcomes in the NHS and yet we do not have the power to demand to know what evidence is out there, or to see it; nor do we have the power to demand access to it, or that independent analysts have oversight of any analysis conducted on commercial and in confidence patient level data. That is certainly true in the STA process. We can ask industry to conduct a certain analysis but we cannot check by way of an independent review group whether or not it has been well conducted. All these matters are of concern. My personal view is that the decisions that NICE make are just as important as the licensing authority and it should have similar powers. There have been a number of examples where trials that have been conducted have been withheld from an appraisal. Those trials are always ones that show the product in a poor light.

Q495 Charlotte Atkins: Should NICE have power to ask for this data and the voluntary code be swept away because it is not working?

Professor Claxton: Yes.

Professor Nicholl: Many studies have been made of the influence of these sorts of issues on published results. I think the biggest one was conducted a few years ago and was published in the Journal of the American Medical Association. It included a review of more than 1,100 original trials and concluded there was incontrovertible evidence that published industry-sponsored trials had pro-industry conclusions at more than three times the rate of non‑industry-sponsored trials. There are two possible ways in which it can arise. One is because trials that do not have pro-industry results are not published; the other is because the conduct of the trial somehow leads to a pro-industry bias. I do not think anybody knows which of those two answers is right, but the fact that it could be the first one leads automatically to the conclusion that if there is to be value pricing based on trials conducted by industry all the trial evidence must be available at the point at which the pricing decision is made.

Professor Towse: As far as concerns the voluntary registration scheme, it seems to me that it is the last chance saloon. The industry must get its act together and disclose what trials are going on; otherwise, sooner or later somebody will impose some requirements on them. That is separate from the question as to what requirements NICE should impose for appropriate evidence. It seems to me that NICE should have some powers to ask for appropriate evidence. I make a distinction between understanding what trials are being conducted and what the results of those trials are. Professor Claxton went further and said there should be access to patient level data. There is an issue as to how deep NICE can mine, as it were, but to have some understanding as to what work is going on and whether or not it is relevant to NICE's decision-making seems to me to be pretty fundamental. One point that has not been mentioned but is implicit in the question is whether the evidence should be made available just to the appraisal committee or assessment group or should be in the public domain. I believe that the only issue of substance must be around publication rights. Are there issues around access to journals et cetera if information is in the public domain? The fundamental point is whether NICE has the information it needs in order to make a judgment of value.

Professor Claxton: To raise a linked point, another frustration is that NICE can issue guidance on the use of products only in their licensed indication. No guidance can be issued on the use of a product beyond that indication, quite rightly. Nevertheless, in practice it means that sometimes a product is routinely used in clinical practice outside its licence and we are in a situation where we cannot consider that as a comparator or in the guidance, so our guidance seems somewhat irrelevant to NHS practice because we cannot do that. We also have situations where comparators could be looked at and might get a licence if the company sought one, which would mean that the drug being appraised in terms of dosage and regime would not be cost-effective. An example is Aseptin in early stage breast cancer where we could not look at the lower dose and shorter duration regime because there was no licence. Both these circumstances are quite frustrating. The committee has had a lot of debate on whether it should go ahead and try to issue some indication beyond licence. Quite rightly, there has been a very firm view that it cannot because ultimately it is the licensing authority that is the backstop. It has the power to review evidence and assess safety; NICE does not. As soon as NICE starts making statements about unlicensed use of the drug then it is undermining the safety of the licensing authority. We cannot do that. But there must be a middle way. We do not want to undermine licensing which is essential for safety. At the same time, we do want to issue guidance that is relevant to the NHS and includes all comparators. At the moment licensing is in the hands of the manufacturers and they will not ask for a licence for something that is not in their financial interests. It would be really nice if the NHS could approach the licensing authority and say that clinicians are using the drug in this or that context, ask it to take a look at it and give it the equivalent of an NHS licence for that product. That would be absolutely marvellous, but it is beyond the remit of today's meeting.

Professor Towse: This is very interesting. Does not NIH do this all the time, ie run clinical trials to see whether, for example, products in one cancer might work for patients in another? This seems to me to be an area where the NHS should in principle be doing work. It is not just about whether the company may not regard it as being in its interest; it is simply that the cost of doing a trial for regulatory purposes when there is already a situation in which clinicians know perfectly well that it works in that patient group may not make sense.

Professor Nicholl: One of the hats that I wear is as deputy director of the HTA programme. I also chair the various HTA research commissioning boards. We are very conscious of this and try hard to commission head-to-head trials which companies would not invest in themselves so that some of these questions are answered. The trouble is that there are rather a lot of questions.

Professor Claxton: That is absolutely right. There are many circumstances in which the public sector should invest in the research about the comparisons that we know the commercial sector does not do. At the moment the problem is that even if we invest in the research to look at something that is unlicensed we do not get the licence and so NICE cannot issue guidance on it. That is a real frustration and a mismatch between investment in research and putting guidance into the NHS that is appropriate. That is something that needs to be looked at.

Professor Towse: The situation in which NICE finds itself is quite understandable but it is a straitjacket that does not make complete sense.

Q496 Chairman: Do you think that NICE should have power to look at the guidelines on the use of Avastin for wet AMD?

Professor Claxton: Yes, absolutely.

Q497 Chairman: Professor Claxton, you talked about the submissions made and whether or you were happy with them. Wearing your appraisal committee hat, if you get one that you are unhappy with do you ask for a resubmission?

Professor Claxton: It depends on the process. Under the NCA or old process that has happened many times. Some of the most controversial appraisals have been through that kind of process. If we were unhappy we would ask the assessment group whether it would be able to do a re-analysis. If it was unwilling or unable to do it we would ask the national decision support unit to undertake that work. If we needed additional information from the sponsors we would involve them and ask them. Alzheimer's is an excellent example of a whole series of those kinds of re-analyses, including sub-group analyses, that have been conducted by the sponsors. Beta interferon was another long-running example. Under STA it is a little more difficult. Ideally, issues as to what needs to be done in the appraisal should have been flagged up between the scope and the meeting. If the independent assessment or evaluation group had any particular problems or questions in looking at that submission it should have been in contact with the sponsor and the appraisal committee should be able to see those responses. But at the appraisal committee meeting we have a choice. We can say we believe that the submission is of sufficient quality that we can base a decision on it and it is cost-effectiveness or we can say no. We can say no for two reasons. One is that we believe the submissions are of sufficient quality for us to make a decision and it is not cost-effective, or we can say that the submission provides an insufficient basis on which to make a decision.

Q498 Chairman: At that stage can you ask for more information from the sponsor?

Professor Claxton: We can between the provisional decision and the FAD.

Q499 Chairman: But not when it is a yes or no decision?

Professor Claxton: No. It is a yes or no decision.

Professor Towse: But the NICE guidance before it is issued can say it is clear that the appraisal committee has been struggling in that case and before it issues it in final form it will have one more go at asking for the information that it believes is appropriate. In principle it can do that.

Professor Claxton: In principles, yes. I cannot recall an example under STA where that has happened. One point that needs to be made very clear about STA - as a member of the appraisal committee it is a frustration - is that basically the submission is in the hands of the sponsor with a review by an independent group. That means it is very important we have the option of saying no to a product because despite the dialogue that has gone on the submission has not come up to scratch and we are not confident with it; in other words, the sponsor has failed to demonstrate cost-effectiveness. It is very important that that decision is sustained at appeal. Essentially, the burden of proof in