Evidence submitted by the American Pharmaceutical
Group (NICE 89)[31]
EXECUTIVE SUMMARY
1. The American Pharmaceutical Group has
a number of concerns about the effectiveness of NICE, including
NICE's evaluation process, the speed of publishing guidance, the
appeal system, and implementation. In our submission we illustrate
these concerns through case studies. These specific instances
give cause for concern that NICE would be ill-equipped to manage
an extended remit as suggested by both the recent Office of Fair
Trading report into the workings of the PPRS and the Cooksey Report.
INTRODUCTION
2. The American Pharmaceutical Group (APG)
represents the 10 leading research based US-owned pharmaceutical
companies who invest in the UK. APG member companies currently
supply over a third of all branded medicines to the NHS. We welcome
this inquiry and believe it is a crucial time for the Committee
to conduct a review of NICE to address whether the organisation
is fit for purpose.
3. The APG supports the objectives of NICE
and we value its independence and stakeholder engagement. However
we have a number of concerns about how NICE works in practice.
The recent Office of Fair Trading report into the pharmaceutical
pricing regulation scheme (PPRS) recommends a wider role for NICE.
However, the industry is not satisfied about NICE's ability to
deliver increased access, accelerated uptake of new medicines,
and proper assessment of value. The newer NICE Single Technology
Appraisal (STA) process raises particular concerns over NICE's
ability to conduct equitable appraisals close after the launch
of medicines.
4. The APG works closely with the Association
of British Pharmaceutical Industry (ABPI) and we endorse their
submission to the Committee's inquiry. Rather than replicate the
points of principle made in the ABPI's memorandum, we have used
this submission to illustrate some of these points by way of case
studies.
NICE EVALUATION PROCESS
5. A decision from NICE is becoming increasingly
dependent on whether the technology in question demonstrates cost-effectiveness
against an implicit cost per quality-adjusted life-year (QALY)
threshold. This can be a useful pointer, but does not capture
all dimensions of the value of a particular medicine or therapy,
both to patients, carers and society at large. Although other
measures of value can technically be included in HTA appraisals,
almost all of the key decisions seem to be determined by the cost
per QALY. The current implicit thresholds (£20,000-£30,000)
are set by the HTA bodies themselves, are not subject to public
scrutiny and do not appear to be linked to any kind of inflation
adjustment over time. They have never been the subject of public
debate or Parliamentary approval. This growing arbitrary dependence
on a cost per QALY threshold approach appears to be a back-door
means of NHS cost containment. While cost per QALY analysis has
a part to play in HTA, the APG does not believe it should be pivotal
to whether or not a product is recommended for use. The Cost per
Life Year, for example, may be a more appropriate measure in certain
circumstances. Affordability, innovation, burden on carers, value
to society and contribution to economic productivity are all important
factors that should be taken into account.
Case Study A: Cost effectiveness: the treatment
of mesothelioma with alimta (pemetrexed)
6. Exposure to asbestos is the main cause
of mesothelioma (a cancer of the lining of the lungs). Although
relatively rare, the number of mesothelioma cases in the UK is
steadily rising and is expected to peak between 2011 and 2015
and decline thereafter. Anticipated patient numbers will be between
1,950 and 2,450.
7. Alimta (pemetrexed), in combination with
another drug called cisplatin, is the first and only treatment
(chemotherapy) to be licensed in the UK for patients with mesothelioma.
Some patients have been given the treatment since its launch in
November 2004 but many other hospitals have been waiting for the
final guidance from NICE before using it, resulting in widespread
geographical inequalities in its availability.
8. Alimta (pemetrexed) entered the NICE
process in August 2005. Released on the 26 June 2006 the first
draft of the NICE guidance stated that the use of Alimta (pemetrexed)
for the treatment of mesothelioma is not cost effective. The ruling
was the subject of an appeal hearing in October 2006. The manufacturer,
Lilly, and the Royal College of Nursing (RCN) were both appellants.
9. A number of Lilly's appeal points have
now been upheld. The most significant was where Lilly successfully
argued that the failure to consider the benefits of Alimta (pemetrexed)
by reference to the Cost per Life Year Gained was discriminatory.
The Appeal Panel upheld the appeal on this point and requested
that the Appraisal Committee reconsider the use of life years
gained.
10. Other appeal points upheld were:
— NICE failed to disclose a written
perspective prepared by a clinical specialist who attended the
meetings of the Appraisal Committee.
— The initial Appraisal Committee hearing
on 7 March 2006 was inquorate. The Appraisal Committee should
review all matters considered on 7 March.
— There was potential for misunderstanding
the references to MSO1. The Appeal Panel requested that references
to it be removed.
11. The Appraisal Committee met again to
consider the issues upheld on appeal at a hearing on 8 March 2007.
Having had four appeal points upheld against them, we question
how the same committee can objectively reconsider the same issues
again.
12. An Appraisal Committee Document is now
available for comment by consultees. This document states that
the preliminary view of the Appraisal Committee is not to recommend
Alimta (pemetrexed) for use in the NHS. The appraisal committee
will meet again on 8 May 2007 to consider the responses and issue
the Final Appraisal Document. There is no further mechanism within
the current NICE process to challenge these findings which will
be presented to the NICE Board for ratification. The final Guidance
is expected to be published in August or September 2007, fully
two years after the start of the process. In the last year alone
Lilly estimates that 300-350 patients who could potentially have
benefited from treatment with Alimta have effectively been denied
it.
13. If the current guidance is confirmed
it will prevent patients receiving the medicine on the NHS. This
would be the first time that NICE had denied patients access to
the only licensed treatment for a condition on the NHS.
14. The Scottish Medical Consortium (SMC),
the Cochrane Centre, the London Cancer New Drugs Group and the
Drugs and Therapeutics Bulletin have all already positively appraised
the use of Alimta (pemetrexed). However, the SMC's advice is usually
superseded by final NICE Guidance, meaning that if the preliminary
views of NICE are finalised, patients in Scotland could also be
denied the only licensed treatment for mesothelioma from August-September
2007. Ironically, if NICE had appraised Alimta for mesothelioma
as a Single Technology Appraisal this would not be the case.
15. The approach followed in considering
Alimta (pemetrexed) is inconsistent with other similar appraisals.
The draft guidance did not reflect the benefits of Alimta (pemetrexed)
in a difficult to treat tumour in a small sub-set of the population
for which no other therapy is licensed or, more importantly has
shown comparable effects in medical practice to those shown in
phase III clinical trials. NICE's position is a clear disincentive
to future innovative research, commercial or academic, in difficult
to treat diseases. The overall budget impact submitted to NICE
estimates at most £3.2 million in 2006-07, increasing to
approximately £5.2 million in 2009-10. The current UK price
of Alimta (pemetrexed) is also the lowest in Europe.
16. The Government and Parliament have voiced
sympathy for people with mesothelioma—mostly victims of
workplace exposure to asbestos—through the passage of the
Compensation Act 2006. It would disappointing if ultimately the
only way to gain access to Alimta is to use compensation payments
to fund private treatment.
Case study B: Verifying the research, data and
models used in the NICE assessment: the treatment of Alzheimer's
disease with anti-cholinesterase drugs
17. In March 2005 NICE issued initial guidance
that three anti-cholinesterase drugs (Aricept, Exelon, Reminyl)
should not be funded by the NHS for patients with Alzheimer's
disease. In early 2006 NICE amended this decision so that the
three drugs would only be funded for people in the moderate stages
of Alzheimer's disease. NICE also recommended that Ebixa, the
only treatment available for distressing behavioural symptoms
in late dementia, should not be funded.
18. The initial decision and subsequent
revision, limiting use to moderate patients, has repeatedly ignored
the views of patient and professional groups about the value of
these medicines in early disease. The most recent guidance (2006
FAD) requires patients to experience irreversible and debilitating
decline in health status before treatment is initiated.
19. Residential care costs borne by patients
and their families were excluded in the Alzheimer's appraisal,
which has the effect of systematically under-estimating the value
of Alzheimer's disease medicines to society.
20. The use of QALYs in healthcare decision
making remains controversial. NICE rejected using the QALY approach
in its original Alzheimer's disease assessment as it was not considered
sufficiently reliable. NICE has used QALYs in subsequent appraisals
despite there being no evidence since the original appraisal to
demonstrate that this is a robust approach.
21. The Alzheimer's disease economic model
used by NICE to assess the three drugs contains numerous assumptions
and parameter estimates that do not reflect current practice,
resource use and costs. The economic model can generate cost-effectiveness
estimates above and below the notional £30,000 per QALY threshold
used by NICE, depending on which parameter estimates are used.
22. Repeated refusal by NICE to provide
a working version of the Alzheimer's economic model has prevented
the manufacturers from properly appraising NICE analyses. Given
concerns about methodology and assumptions, and errors in the
Technical Assessment Report, the industry is not confident that
this model does not also contain additional errors. Furthermore,
the manufacturers were given just two days to review and comment
on some of the later analyses, further hindering ability to engage
properly in the appraisal.
Case Study C: Evidence used in the NICE evaluation
process: NICE appraisal of erythropoetin (alpha and beta) and
darbepoetin for the treatment of cancer-treatment induced anaemia
23. Erythropoetin (alpha and beta) and darbepoetin
for the treatment of cancer-treatment induced anaemia was referred
in the 9th wave, and commenced with the scoping in June 2004.
The latest estimate for Guidance publication is November 2007.
24. The European Organisation for Research
and Treatment of Cancer (EORTC), the American Society of Clinical
Oncology (ASCO) and the US National Comprehensive Cancer Network
(NCCN) all recommend Erythropoietin for cancer-treatment induced
anaemia within the licensed indications for the products. NICE's
preliminary decision was not to recommend the use of EPO for patients
with anaemia induced by cancer treatment, although this is currently
the basis for an appeal. Withdrawing the option of treatment with
this medicine would be a major blow for people with cancer at
a time when their health is already under serious test.
25. There are a number of concerns about
NICE's evaluation process:
Government priorities on blood use
The UK has relatively low blood stocks and the
cost of producing safe blood is increasing. Despite being a requirement
for all NICE appraisals, NICE did not take into account the recommendations
from other Government bodies on preserving blood supplies. This
includes the 2002 Health Service Circular that instructed trusts
to take action to avoid unnecessary use of donor blood and to
consider effective alternatives.
Substantial errors in survival analysis undertaken
by NICE
NICE was instructed to perform a meta-analysis
of EPO trials consistent with the licensed indication to see if
a survival advantage existed. NICE concluded none of the EPO trials
was consistent with the current marketing authorisation and therefore
rejected all EPO clinical trials. NICE failed to distinguish between
indications for the use of EPOs and the other information
included in the Summary of Product Characteristics (SmPC) which
evolved post-marketing and after the trials programme was completed.
This was inconsistent with the remit. The approach NICE used for
this analysis was unscientific using a method we have never seen
before. If you include all clinical trials with a population,
dose and Hb level in line with the 2005 SmPC there is good evidence
of a survival benefit.
Iron data
NICE did not consider the new data emerging
that indicates a higher response rate when EPOs are administered
with iron.
Patient preference
NICE did not properly consider patient preference
and convenience. Evidence shows patients prefer EPO treatment
due to improved convenience of home treatment and concerns over
the increased risks associated with transfusion.
THE SPEED
OF PUBLISHING
GUIDANCE; THE
APPEALS PROCESS
26. There are often delays in uptake of
innovative medicines due to the time taken to publish guidance
and the length of the NICE appraisal process.
Case Study D: Length of NICE process: the treatment
of rheumatoid arthritis and of severe ankylosing spondylitis with
adalimumab
27. Adalimumab was granted its European
licence in September 2003 for the treatment of rheumatoid arthritis.
Abbott submitted evidence on 8 June 2005 for the multiple technology
appraisal of adalimumab, etanercept and infliximab for the treatment
of rheumatoid arthritis but NICE has yet to publish its guidance—some
way over the expected timeframe from submission of evidence to
final guidance.
28. The use of adalimumab for the treatment
of severe ankylosing spondylitis has followed a similar process.
Abbott originally submitted evidence for this indication in January
2005 and the expected date for release of final guidance is currently
unknown, as the third appraisal committee meeting requested additional
analysis to be conducted to inform the cost-effectiveness estimates.
The ongoing delays to NICE guidance for this indication are leading
to increased variation in access to adalimumab in the UK, with
the Scottish Medicines Consortium guidance recommending adalimumab
for treatment of these patients on 10 November 2006.
29. One of the problems with the appraisal
process as currently constructed is that any appeal lodged has
a major impact on extending the delay until final guidance is
released. There is currently a lengthy delay between submission
of a notice of appeal and the date of an appeal hearing (four
months in the case of adalimumab, etanercept and infliximab for
the treatment of rheumatoid arthritis). This appears to be due
to resource constraints in terms of the number of panels available
to hear appeals—NICE's guide to the appeals process indicates
it will endeavour to hear appeals within 10 weeks of the appeal
being lodged. There is a further delay in the time taken from
the appeal hearing to final guidance. For MTA appraisals where
only sections of the Final Appraisal Recommendations are being
appealed against, we feel it would make sense to release final
guidance on those recommendations which are not being contested
to reduce delays.
30. NICE's STA process is aimed at producing
guidance around the time of marketing approval for the medicine.
However, the availability of time slots on NICE's appraisal committee
meetings appears to be delaying the scheduling of topics until
well after launch. For example, adalimumab for the treatment of
psoriasis is expected to receive its marketing authorisation in
the 4th quarter of 2007. However, NICE provisionally indicated
to Abbott that no appraisal committee time slots would be available
to review this product until February 2008.
Case Study E: Length of NICE process and change
of remit: the treatment of primary and secondary osteoporosis
Technology appraisals for the treatment of primary
and secondary osteoporosis have been ongoing for five years and
are still not completed.
32. NICE is currently consulting on Appraisal
Consultation Documents (ACDs) for both primary and secondary osteoporosis.
The original technology appraisal for the prevention and treatment
of osteoporosis commenced in 2002 and was divided into two appraisals
during its development; one for primary prevention and one for
secondary prevention. The appraisal for secondary prevention was
published in January 2005, whilst the appraisal for primary prevention
is still in development.
33. NICE Guidance for the secondary prevention
of osteoporosis included recommendations for the initiation of
treatment and recommendations for those patients withdrawn from
initial treatments. Two years later, following the inclusion of
another agent and additional data analysis, the latest ACD for
the update of this guidance provides recommendations only for
the initiation of treatment. No guidance at all is offered for
patients who are withdrawn from initial treatments (ie unable
to tolerate or respond to alendronate). The ACD for primary prevention
produced in parallel also contains guidance only for the initiation
of treatment. There has been a clear change in the scope of the
review without consultation.
34. NICE has spent five years developing
this guidance already. The large number of drafts published for
this ongoing appraisal has led to increased confusion as to best
practice and unfortunately the current proposal does nothing to
resolve this. This is unhelpful to healthcare professionals and
patients and the current recommendations are likely to deny patients
access to useful treatments.
THE IMPLEMENTATION
OF NICE GUIDANCE
35. Different categories of NICE guidance
carry different weight. Technical Guidance is covered by
a mandate from the Secretary of State for Health in England. Typically
the stated timeline for implementation is three months from publication
in both England and Wales, 18 months in Northern Ireland, and
at the discretion of Health Boards in Scotland. NHS bodies have
a duty to make plans to implement Guidance issued by NICE
and to fund the treatments involved. However, the additional tier
of local bureaucracy whereby each local trust and board decides
whether to place a nationally approved medicine onto a local formulary
serves to exacerbate the "postcode lottery". The process
of potentially making three separate national HTA submissions
and then providing additional data for local drug and therapeutic
committees also places a significant burden on companies, which
goes against the principles of the Hampton and Arculus reviews.
Even when all these hurdles have been cleared, implementation
of NICE guidance is extremely variable. To some extent the Healthcare
Commission monitors NHS adherence to this through the Core Standards
in its "Annual Health Check", but the process is currently
weak.
36. Clinical Guidelines on the other
hand are even weaker, since they do not carry the same mandate.
The Healthcare Commission does not measure implementation of Clinical
Guidelines per se; instead they encourage trusts to
consider implementation as a developmental standard—one
which they should be aspiring to achieve in the future.
37. NICE has indicated that it intends to
begin incorporating Guidance into Clinical Guidelines
at the review which usually occurs three years after guidance
is issued. Feedback from NHS organisations suggests that, while
they would like to prioritise the implementation of Clinical
Guidelines, they do not have the funding or resources required
to do so. Consequently implementation is patchy and dependent
on available funds.
Case Study F: Guidance versus guidelines: the
use of atypical anti-psychotics in the treatment of schizophrenia
38. NICE issued guidance on the use of atypical
anti-psychotics in the treatment of schizophrenia in May 2002.
As is usual, this guidance was published with the intention that
it would be reviewed and re-issued after three years (May 2005).
Instead, in May 2005, NICE issued proposals to incorporate this
guidance into the Clinical Guidelines.
39. Even with mandatory status, the implementation
of Guidance issued by NICE varies. Regional variations
in use of atypical anti-psychotics and even variations between
local mental health trusts of up to 70% exist. A Healthcare Commission
report in January 2007 found a quarter of mental health trusts
have yet to get the funding from Primary Care Trusts to fully
implement the 2002 Guidance—despite the direction to implement
within three months.
40. By incorporating this Guidance
into a Clinical Guideline, the NHS is no longer obliged
to fund the use of atypical anti-psychotics. Some NHS representatives
have indicated to us they will be re-directing funds currently
allocated to atypicals towards other technologies which do have
mandatory status. This is not a deliberate attempt to withhold
treatment from those who need it, but the only way they can balance
the books. However, in an area such as severe mental illness this
approach could directly compromise patient care, and patient and
public safety.
41. At present the Guidance on atypicals
is the only technology appraisal that is being incorporated into
a Clinical Guideline on review. NICE has indicated it intends
to extend this method to other technology appraisals when they
reach their review date. Potentially, many other therapy areas
outside mental illness will find that funding for medication is
no longer statutory. That cannot be the intention of NICE which
was established to set clinical standards of excellence and achieve
parity of access to quality treatment.
CONCLUSION
42. These specific instances give legitimate
cause for concern that NICE would be ill-equipped to manage an
expanded remit as suggested by the OFT report into the workings
of the PPRS. The current NICE process seems unnecessarily adversarial.
For example, the process would be improved if stakeholders could
have on-going dialogue on technical questions with both the external
review groups and the relevant NICE appraisal committee. It would
also be beneficial if issues of substance, as opposed to process,
could be revisited on appeal. The appeal process itself raises
the wider issue of whether NICE should sit as judge and jury in
its own court—and then send any "successful" case
back to the same jury.
43. Ultimately the interests of patients
should be at the heart of the NICE process. We have documented
evidence of how it is falling short in delivering its aims in
terms of restricted access to proven medicines and accelerated
uptake of new medicines.
44. Under the recommendations of the OFT
for therapeutic reference pricing, HTA assessment of value would
still be pivotal in determining whether patients would receive
medicines. Given that the OFT proposes a new collaborative approach
to HTA between NICE, the Scottish Medicines Consortium and the
All-Wales Medicines Strategy Group, we urge the Committee to take
evidence from all of these bodies. Given the magnitude of the
issues at stake we also urge the Government to conduct a wide-ranging
independent review of HTA processes. Before a meaningful discussion
on the reform of the PPRS can take place, the APG wants to have
full confidence in the metrics used to determine the value of
a medicine within the total healthcare system. These should take
into account innovation, affordability, benefit to patients, carers,
society, the science base and UK plc.
The American Pharmaceutical Group
March 2007
31 The APG consists of the following companies: Abbott,
Amgen, Bristol-Myers Squibb, Janssen-Cilag Ltd, Lilly, Merck Sharp
& Dohme, Pfizer, Procter & Gamble Pharmaceuticals. Schering
Plough and Wyeth. Back
| 
