SUMMARY AND RECOMMENDATIONS

  1.  Lilly recognises that NICE offers a valuable service to the NHS in providing it with guidance on which medicines are clinically and cost-effective. We appreciate the aim of the Government is to provide a publicly funded health service with evidence on the value of our medicines and we were supportive of the Institute at its inception.

  2.  Two recent reports—The Office of Fair Trading report on the Pharmaceutical Pricing Regulation Scheme and the Cooksey Review of Health Research Funding—have both advocated an expanded role for NICE in the future. In light of these recommendations, and the ongoing public debate about the Institute's decisions, we believe this inquiry by the Health Select Committee to be timely and welcome the opportunity to contribute.

  3.  We urge the Committee to review NICE in the context of the current processes of the Scottish Medicines Consortium (SMC) and the All-Wales Medicines Strategy Group (AWMSG). We believe the Government should conduct an independent review of health technology assessment (HTA) before any considerations for reform of the Pharmaceutical Price Regulation Scheme can be made.

  4.  Lilly believes that over the eight years that NICE has been operating, a number of serious flaws have emerged in its processes, which are undermining confidence in the Institute's recommendations. NICE is tasked with making very tough decisions and it is therefore essential that all stakeholders in the process can be confident that those decisions are made in a constructive, fair and transparent manner with rigorous consideration of the evidence.

  5.  In order to build more confidence in the NICE process and improve the transparency and speed of decisions, there are a number of improvements we would recommend:

—  The current reliance on the incremental cost per Quality Adjusted Life Year (QALY) as pivotal to the decision of the Appraisal Committee (AC) should end. There are important technical and policy reasons why this is the case. NICE itself has listed a number of other factors that should be considered—these should be given greater weight, and other measures of cost-effectiveness used where the QALY may not be appropriate.

—  The advice given to NICE by its external advisers (Technology Assessment Groups and Evidence Review Groups) should be subject to quality review either within or without the Institute, to ensure that all advice is of an equally high standard.

—  A mechanism should be found whereby NICE does not completely reject a drug that is being appraised early in its lifecycle on the basis of insufficient evidence of cost-effectiveness. More creative solutions to confirming value after launch could be considered other than "use only within clinical trials".

—  The Institute should follow the example of the SMC in using a flexible, constructive and open process of engagement with stakeholders rather than the generally adversarial system it currently employs. This should include the opportunity for manufacturers to have dialogue with AGs/ERGs and the ACs throughout the process.

—  The appeal system should be made independent of NICE and the grounds for appeal widened to allow substantive review of the scientific merits of the AC's decision.

—  A cost benefit analysis of performing an assessment for a given drug should be a consideration in topic selection. This would help to avoid lengthy and time consuming appraisals of medicines that may have minimal budget impact. Together with central funding and commissioning for rare diseases, this may also lead to more pragmatic decision making for expensive but rarely needed treatments.

—  All of these are of little merit however, unless the fundamental issue of implementation of NICE guidance is addressed. To hasten access to approved medicines by patients, and to reduce the burden on business in line with the principles of Hampton and Arculus, all medicines approved for use by the NHS should automatically be placed on local formularies without further debate or delay.

  6.  We have set out below our views on some of the issues that the Committee specifically wants to address.

Why NICE's Decisions Are Increasingly Being Challenged

  7.  The simple explanation is that more and more decisions are negative. Prior to 2006, approximately 80% of appraisals recommended use or restricted use of the drug in question by the NHS. In 2006 NICE introduced the Single Technology Appraisal (STA) process, which aimed to issue faster guidance on life-saving drugs closer to launch to avoid the incidence of "NICE blight".

  8.  80% of first draft recommendations (ACDs) for the 16 single technology appraisals currently ongoing have not recommended the drug for use by the NHS. These documents are all in the public domain so it is inevitable that negative decisions are questioned by patients, clinicians, industry and the media both formally and informally.

  9.  The Institute's own processes and behaviour contribute to the increasing number of challenges. It is very difficult for industry to have constructive dialogue with the Institute, which compares very unfavourably with the SMC and the AWMSG in this respect. For example, there is no opportunity for manufacturers to meet with ACs to explain their data, clarify areas of confusion or misunderstanding or challenge the analysis provided by NICE's external advisers. While there may be some commercial justification for this where many medicines are being appraised simultaneously, there is no justification for it in the STA process.

  10.  In the event of a negative decision, the only option for manufacturers or other stakeholders is a formal appeal. The SMC in contrast gives reasons for why it has not recommended a drug for use in Scotland, and allows companies to resubmit their case—perhaps to present new evidence or make the case for use of the drug in a more rigorously defined sub-group of patients. NICE does not allow resubmission of evidence in this way, which consequently leads to a greater number of formal appeals.

The Evaluation Process

  11.  We have a number of concerns about the Institute's evaluation process for Health Technology Appraisals.

Stakeholder engagement with the process

  12.  It is difficult for many stakeholders to engage effectively with the process. Although NICE encourages and accepts evidence from a range of stakeholders, in reality it is hard for patients and healthcare professionals to play a meaningful role in a lengthy and economically complex evaluation. NICE appoints a "clinical expert" for each appraisal, who in effect should be representing wider medical opinion. It is not unusual for these "experts" to come from an unrelated medical field or to have no actual experience of using the drug in question. Patient groups cannot be expected to provide evidence on cost-effectiveness, but their views on clinical benefit, societal benefit and patient choice are important and should be heard. The increasing reliance of ACs on the incremental cost per QALY calculation (see below) above all other considerations disenfranchises patients and risks their participation becoming a "tick box" exercise.

Technology Assessment Groups/External Review Groups

  13.  NICE commissions independent academic centres (AGs/ERGs) to review the manufacturer's submission and the published evidence on a medicine. The group prepares a summary report for the AC either using its own economic models, or making alterations to the manufacturer's models.

  14.  The quality of reviews varies considerably between the groups, with reports sometimes lacking fairness and balance. For example, in the ongoing appraisal of Alimta (pemetrexed) for Mesothelioma, the ERG questioned its efficacy, in direct contradiction of the findings of the medicines regulator.

  15.  The reports produced by external centres show a poor understanding of the pharmaceutical development process. The AGs/ERGs would benefit from training and better ongoing dialogue with the industry. Their reports should be subject to quality review.

Realistic expectations of evidence

  16.  NICE has increasingly begun to suggest manufacturers should carry out new clinical trials to provide further evidence about a medicine and in some cases it has recommended use "only in the context of research". Unfortunately, by the time this decision is made it can be too late to design a clinical trial to meet the needs of patients in the UK, which reflects the Institute's poor understanding of the realities of commercial pressures driven by licensing and patent law. The research suggestions are sometimes unrealistic from an ethical perspective and would not necessarily answer the decision problem under consideration. The Institute's fixed timelines for re-reviewing guidance do not allow sufficient time for trials to be set up from scratch and data analysed in between reviews.

  17.  The high number of negative recommendations at the first stage of the STA process shows ACs are clearly uncomfortable making positive recommendations based on the range of data typically available early in the lifecycle of a medicine, a case of "guilty until proven innocent". Studies prior to launch focus on efficacy and safety to meet the requirements of the licensing regulators, not the different requirements of HTA bodies round the world. In our view, NICE needs to be realistic about what data can be provided at this stage, and to accommodate a degree of uncertainty. A negative decision means patients may be denied a medicine on the basis of uncertain rather than incontrovertible evidence and that "real life" data can then only realistically be collected from general use in patients overseas, which may have limited applicability here.

Measures of cost-effectiveness

  18.  NICE uses a measure known as cost per Quality Adjusted Life Year (QALY) to determine whether the cost of a medicine can be justified by the impact it has on health. There is a value of £20,000 per incremental QALY below which a medicine would be considered more cost-effective than the comparator treatment. When a medicine exceeds this NICE will only recommend it for use if there are other important factors.

  19.  No measure of health outcomes is perfect, and while the QALY is a plausible and attractive concept, there are a number of technical limitations in practice:

—  The QALY is essentially the arithmetic product of length of life (LY) and health-related quality of life (QA). One of the key assumptions is that these things are independent of each other and the amount of time a person is willing to trade for an improvement in quality of life is independent of how long they have left to live. In terms of health, it is easy to see why this will not hold—patients' quality of life is likely to be influenced by the amount of time they have left to live, especially if they have a terminal condition. This is the reason why many feel the QALY biases against people with shorter life expectancy eg terminal stage cancer, the elderly.

—  The health-related quality adjustment (QA) that is needed to obtain the QALY can be calculated in a number of different ways, each producing a different answer. This introduces a high degree of uncertainty around the cost per QALY estimate.

Cost-effectiveness versus other considerations

  20.  Of fundamental concern to us is the reliance of NICE on a QALY threshold as pivotal to its decision making. We believe that while it is appropriate to have a threshold as a guide, particularly for comparing one medicine against another, the QALY should be one of the factors considered in reaching a decision, and not effectively the only one.

  21.  While the academic jury is still out on the QALY, we believe NICE should adopt a more pragmatic approach to the assessment of cost-effectiveness, and should consider taking into account" "cost of life year gained" analyses, particularly in the appraisal of end of life therapies. Additionally consideration should be given to the broader impacts of a medicine, as valued by society. These might include innovation, affordability, burden on carers, societal benefit and contribution to the UK economy.

  22.  When a drug exceeds the QALY threshold, the Institute has listed the other factors it will consider. A number of recent decisions indicate to us that these are in fact given little, if any, weight in comparison with the QALY calculation:

(a)  Alimta (Pemetrexed)—not recommended for use in non small cell lung cancer. There was a considerable degree of uncertainty surrounding the calculation of cost/QALY, which ranged from £9,010 to £1.8 million.

(b)  Alimta (Pemetrexed) for Mesothelioma. Lilly is appealing negative guidance on the use of Alimta for mesothelioma. Mesothelioma is a rare but devastating industrial disease caused by exposure to asbestos. The government has committed to compensating victims, but at the same time, NICE is denying them access to the only licensed medicine for their condition, giving little weight to the particular features of this population.

(c)  Evista (Raloxifene). In the appraisal of Evista for osteoporosis, NICE refused to include in its evaluation of cost-effectiveness full consideration of the wider costs and benefits associated with the demonstrated reduction in breast cancer risk in people taking Evista.

Clinical guidelines development process

  23.  The Institute has declared its intention in the future to focus more and more on developing clinical guidelines, which incorporate technology appraisals. Since the guidelines do not have the same statutory backing as the technology appraisals, this will result in less dedicated funding, less uniform implementation, and more variability in local practice, contrary to the founding objectives.

The Speed of Publishing Evidence

  24.  NICE does not evaluate all medicines at the time of launch and has not evaluated many of the medicines currently in use. When there is no guidance from NICE, decisions are made at a local level based on financial budget, clinical benefit, value for money and affordability. NHS trusts and PCTs frequently cite "waiting for NICE" as a reason to prohibit prescribing, which can lead to wide variations in use across the country. The Government and NICE have taken steps to tackle this problem, by issuing advice and introducing the Single Technology Appraisal (STA) process, respectively.

  25.  While we appreciate the need for NICE to be rigorous in its analysis of each drug, the processes—both STA and MTA (multiple technology appraisal)—for publishing guidance are far too slow and the timelines lack flexibility. Appendix 1 includes a case study which illustrate this.

  26.  Numerous efforts have been made to address the issue of NICE blight, but it remains a serious problem that causes immense emotional and physical cost to patients and limits the return on investment to pharmaceutical companies during the period of patent protection, which has a knock-on effect on investment in the next cycle of medical discovery. The speed of issuing guidance must be addressed through improvements to the appraisal process and more appropriate resourcing.

  27.  As the review process for NICE STA is similar to that of the SMC—consideration should be given to adopting SMC advice across England, Wales and Northern Ireland. This would result in NICE avoiding duplication of effort as well as reducing resource costs and more importantly it would enable patients to have quicker access to medicines that are considered to be clinically and cost effective.

The Appeal System

  28.  We believe there are a number of deficiencies in the current appeal procedure which, if addressed, would improve the fairness of the process and ultimately the credibility of guidance issued by the Institute.

  29.  NICE's constitution requires the appeal panel to be chaired by the Chairman or Vice Chairman of NICE, and for at least one other member to be on the NICE Board. While we do not in any way question the integrity of current members of the Board, a conflict of interest nevertheless exists. The appeal panel should be independent of NICE.

  30.  There is no opportunity to appeal other than on process issues or perversity of recommendation, so no chance to question whether the scientific interpretation of the data is fair. The possible grounds for appeal should be widened, particularly given the absence of dialogue between ACs, their external advisers and the manufacturer and the subsequent potential for misinterpretation of evidence.

  31.  The practice of disclosing new evidence in the Final Appraisal Determination (FAD) must be discontinued—there should be a requirement that if there is significant change following the ACD, then a second ACD should be issued for consideration, in order to allow for comment and to reduce the likelihood of appeals. The danger is of introducing further delays in the process but this could be avoided by better dialogue with manufacturers throughout the process.

  32.  When an appeal point is upheld there is no subsequent opportunity to appeal the appraisal committee's reconsideration of that point.

IMPLEMENTATION

  33.  The NHS has been directed to implement NICE guidance within three months in England and Wales and within 18 months in Northern Ireland. Implementation does not consistently happen within the required timescale, and sometimes not at all.

  34.  While we recognise that manufacturers have a role in providing evidence on the value of our medicines, we are now forced to do so many times at different levels in the health service. When a national HTA body such as NICE has appraised a drug, there should be no need for local duplication. NICE has long recognised that implementation is an issue and has recently created a department with a £2.5 million budget to address it.

  35.  Fundamentally, issues around implementation stem from its affordability to the NHS and the priority placed on it by the Health Care Commission. Cost-effectiveness and budget impact are different things and both affect decisions by the local NHS on when, if and how to implement recommendations. For this reason, we believe implementation would only be assured if it were accompanied by ring-fenced funding and assigned a higher priority by the HCC than currently. It is not acceptable that the money which should be allocated to fund medicines recommended by NICE, is being used elsewhere in the health economy to meet budget deficits.

  36.  There is no obligation on the NHS to implement NICE clinical guidelines (which offer a whole system approach to the treatment of a disease), which are a developmental standard as opposed to a core one. For this, and other reasons, it is unlikely the health service will attach much priority to guidelines—a worrying situation as NICE is increasingly moving the re-reviews of technology guidance into the guideline process, and thereby removing any obligation to implement.

Helen Llewellyn

Lilly UK

March 2007

  In this appendix we set out two case studies to illustrate why we feel the processes used by NICE in publishing guidance are too slow and lack flexibility.

1.  ALIMTA (PEMETREXED) FOR MESOTHELIOMA

Timescale

November 2004	Alimta receives licence
July 2005	Scottish Medicines Consortium advice published
August 2005	Submission of evidence to NICE
June 2006	Final appraisal determination published
October 2006	Appeal hearing
December 2006	Appeal board asks committee to reconsider several points
March 2007	Original appraisal committee reconvenes
May 2007	Final appraisal committee meeting
August/September 2007	Final Guidance anticipated
February 2009	Scheduled re-review begins
February 2010	Re-review decision

Impact

  It will be three years from licence to guidance, during which time NICE blight has widely impacted use of Alimta. Should the appraisal committee reaffirm its negative guidance after considering the appeal points, it could be 2010, if ever, before NHS patients have the opportunity to use this drug, which has been recommended for use in Scotland by the SMC since July 2005. We estimate that approximately 300-350 patients in England and Wales in the last year could have benefited from treatment with Alimta but have effectively been denied it. Those people would, on average, have gained an additional 4.8 months of life. The estimated cost of this medicine to the NHS is £3.2 million this year, rising to £5 million by the end of the decade.