Memorandum by Professor Alan Maynard (NC
1. Alan Maynard is Professor of Health Economics
and Director of the York Health Policy Group in the Department
of Health Sciences at the University of York. He was Founding
Director of the Graduate Programme in Health Economics at York
from 1976-83. From 1983 until 1995 he was Founding Director of
the Centre for Health Economics at the University of York. He
has worked as a consultant for the World Bank, the UK Department
for International Development, the World Health Organisation and
the European Community in countries such as South Africa, Latvia,
Cyprus, Chile, Mexico, China, Thailand, Brazil and Botswana.
2. In addition to this academic and international
advisory work, he has been involved in NHS management since 1983,
being appointed by Conservative and Labour administrations to
a variety of roles. He has been Chair of the York NHS Trust since
1997 and, in this role he is involved in the provision of secondary
acute and mental health care to the local population of 300,000
from a budget of £130 million.
3. In May 1997 the then Parliamentary Secretary
of State for Health, Baroness Jay asked for advice on how better
to manage the pharmaceutical budget. After some meetings, the
advice, which had been offered to Baroness Jay, was published
in the British Medical Journal (Maynard and Bloor (1997)). One
of the proposals in this paper was the creation of a new body
with a remit similar to that subsequently given to the National
Institute for Clinical Excellence (NICE). The proposals put forward
in the British Medical Journal were further developed during my
subsequent involvement in the Comprehensive Spending Review in
4. I have advocated a mechanism to evaluate
the cost effectiveness of new and existing health care technologies
for nearly 20 years. Whilst Government and policy makers such
as Michael Rawlins, the Chair of NICE, deny the existence of rationing
in the NHS, such prioritisation is unavoidable. The issue is not
WHETHER health care resources are rationed but HOW ie what principles
should determine which patients should get access to health care?
5. Rationing involves denying patients of
care which of benefit to them and which they want. It is ubiquitous
in all health care systems; the only difference is the way in
which rationing takes place internationally and nationally. The
proposal to create NICE in 1997 was to ensure that the rationing
principles were explicit and that the best possible advice was
offered to managers and clinicians so that scarce NHS resources
were targeted at those patients who could benefit most at least
6. Thus I was an originator and an architect
of the idea of NICE. However I have also been one of its most
persistent critics, as I believe that the organisation I helped
to create is dealing with a range of issues either by ignoring
them or acting inappropriately. Some of these issues were dealt
with in another paper in the British Medical Journal. (Cookson,
McDaid and Maynard (2001)). The purpose of this paper is to develop
and add to this work and pose a series of questions, which I hope
the Select Committee will find worthy of fuller consideration
in both their oral proceedings and their final report.
7. The demand for health care is driven
by at least three powerful forces. Firstly the population is ageing.
Whilst there is debate about whether subsequent cohorts of the
elderly will be healthier and less demanding in terms of social
and health care, the impact of demography on the demand for NHS
services cannot be denied.
8. Another powerful force increasing NHS
demand is new technology. The ingenuity of pharmaceutical and
equipment manufacturers is considerable. However many of these
products create marginal rather than substantial benefits for
patients. Consequently it is of great importance that the rhetoric
of producers of new products to treat patients is evaluated carefully
to determine the technology's clinical effectiveness (does the
product improve the health of the patient more than existing treatments?)
and cost effectiveness (does the product improve the health of
the patient more than existing products at the same or a lower
9. The adoption of a product by the NHS
should be determined by its relative cost effectiveness. A new
technology may be clinically effective but not cost effective
eg beta interferon for the treatment of multiple sclerosis is
clinically effective but not cost effective, in part because of
the high price set by the manufacturer. However any product, which
is cost effective, is clinically effective (Maynard (1997)).
10. The quality of evaluative evidence provided
by manufacturers is uneven. During the last decade clinical scientists
who conduct trials and health economists who do economic evaluations
have been scrutinised better as data bases and critical evaluation
of practitioner's performance has improved. (eg the Cochrane Collaboration).
However the incentives to corrupt the evidence base remain considerable,
especially when it is claimed that it costs $800 million to bring
a new drug to market. With investments of that magnitude, the
industry does not want to fail clinical and cost effectiveness
tests just before marketing!
11. A third cause of inflation in health
care demand is the expectations of the population. Most of us
are reluctant to accept that we are suffering from a terminal,
sexually transmitted disease called life! In decline we struggle
for small increases in survival even when the quality of life
is poor and the cost is high. To treat, for instance terminal
cancer patients, with relatively ineffective technologies is to
deny an older person of sight restored by a cataract removal or
middle-aged patients of the eradication of crippling pain and
disability caused by angina.
12. The producers of new technologies supported
by patient lobbies and doctors practising in the treatment area,
make a formidable combination when advocating the adoption of
new treatments. Their efforts are re-inforced by politicians and
the media who tend to present new treatments as being costless
(with no opportunity cost). Such inflation of expectations when
not substantiated by evidence of clinical and cost effectiveness
is improper, not in the best interests of the population but commonplace.
13. The combined effects of these three
powerful influences on demand when set against the finite and
limited resources if the NHS inevitably creates conflict. When
demand is increasing due to technological ingenuity, demography
and often ill-informed expectations whilst funding is parsimonious,
finite and deployed with considerable local variations in efficiency
and equity, political dissent about the outcomes of national and
local rationing processes is inevitable. Guidance through this
jungle of advocacy and rhetoric has to be informed by well researched
evidence. That is the function of NICE in England and Wales and
the Scottish Inter-collegiate Guidelines Network (SIGN) and the
Health Technology Board of Scotland (HTBS) in Scotland.
14. Enforcement of these guidelines is the
function of the Commission for Health Improvement (CHImp) in England
and the Clinical Standards Agency (CSA) in Scotland.
15. Whilst the principles of the technology
assessment are clear and the need for policing of compliance with
guidelines is obvious, these are novel ideas both to the medical
profession and NHS managers. The much-abused concept of "clinical
freedom" can never be the freedom to use society's resources
inefficiently and without accountability. Unfortunately the medical
profession has failed to set evidence based practice guidelines
and remains reluctant to be accountable for its performance.
16. However as guidelines are established
and policed it is essential that the science supporting NICE and
other regulatory bodies be of the highest quality. Poor science
and bad management can discredit the whole process of establishing
which of the many competing treatments should be made available
to patients in need of care.
17. NICE was established to produce national
guidance on individual technologies (appraisals), the management
of specific conditions and clinical audit. A series of issues
have emerged from the work of NICE since it started work in April
(i) How should the work of NICE be prioritised?
NICE is funded parsimoniously and charged with a large work programme.
It has allocated most effort to the area of appraisals of new
technologies, in particular new drugs. This work involves "horizon
scanning" to determine which new drugs are likely to emerge.
It appears to have led to relatively little resource being used
to evaluate existing technologies. What criteria were used to
determine these choices?
(ii) The Scottish SIGN allocates all its
funding to evaluate specific conditions (eg stroke or asthma)
rather than the individual technologies used in combination to
treat particular conditions. What is the appropriate balance of
investment in appraisals of new and existing individual technologies,
and the evaluation of specific conditions?
(iii) The NICE focus on new technologies
may have a particular opportunity cost in that it may direct NHS
funding to them rather than to proven older technologies where
there are waiting lists (eg hip replacements, cardiac surgery
and cataract removal) and which are more cost effective. Thus
it may be better value for money (or more efficient) for NHS trusts
to invest in these older technologies than in drug treatments
for cancer (eg taxanes), obesity and motor neurone disease. The
Government has decided to make NICE "advice" mandatory.
This decision may distort the allocation of resources in the NHS
and lead to the funding of relatively inefficient treatments.
(iv) The chair of NICE indicated in his annual
report (July, 2001) that the cut-off for the approval of technologies
appeared to be £30,000 per QALY (a QALY is a quality adjusted
life year and represents one year of life with full physical,
social and psychological well being). The Australian equivalent
of NICE, the Pharmaceutical Benefits Committee, uses a cut off
of less than half that adopted by NICE. How was the NICE cut off
decided and how defensible is it?
(v) NICE has approved all the drugs it has
appraised, with the exception of beta interferon. These approvals
have been conditional in that they are made for patient sub-groups
with particular characteristics. However has NICE (because of
its de facto high cut level of £30,000 per QALY) been
too generous? Some medical practitioners have argued that they
have eg general practitioners in North Devon rejected NICE's advice
on the influenza treatment Relenza.This drug appears to reduce
the duration of a bout of `flu by one to one and a half days if
it is used in a timely manner and the doctors decided that the
benefits were not worth the high cost of the drug.
(vi) The North Devon decision about Relenza
emphasises the continuing discretion which practitioners have
in regard to their use of treatments. The decision to make NICE
decisions mandatory will challenge this discretion. However the
primary problem arising from discretion is that once approved
by NICE it is difficult to control the drugs use. A "nice"
combination of patients' lobbies and provider groups suitably
"oiled" by pharmaceutical marketing funds can lead to
excessive and inefficient use of the new technology beyond the
recommendations of NICE. Thus the diabetes treatment Rosiglitazone
was approved by NICE and their estimate of the annual expenditure
cost was £15 million. It seems that currently expenditure
on this drug in the NHS is nearing £50 million.
(vii) How can such expenditure inflation
be controlled? Local management and the Commission for Health
Improvement may influence such trends when information about practice
in the NHS is improved by electronic prescribing and better investment
in information technology and the management of its results. Another
method of curtailing NICE induced NHS inflation is to give NICE
a budget to fund NHS purchasers when it approves a technology.
If these could be made "hard" budgets and if expenditure
exceeded allocations, purchasers could demand recompense from
NICE. Such a proposal would not be popular with NICE. However
expenditure inflation cannot be ignored by NICE as being someone
else business: its opportunity costs affects significantly the
delivery of local services.
(viii) Decisions about technologies are informed
by data from clinical trials, some of which have economic components.
Pharmaceutical companies to meet the requirement of the 1968 Medicines
Act carry out these trails. This legislation, introduced in the
wake of the thalidomide tragedy, obliges the companies to provide
the Commission on the Safety of Medicines (CSM) with evidence
of safety, efficacy and quality. If this evidence is adequate
the CSM gives the company marketing approval and the product can
be sold in the NHS. The evidence provided to the CSM is a limited
and imperfect basis for making decisions about the clinical and
cost effectiveness of products. As a consequence pharmaceutical
companies are investing in economic evaluations, which now have
to be conducted in compliance with NICE guidelines. Also clinical
trail data are modelled.
(ix) Any decision made by NICE must, given
the quality of the data, be tentative. Long-term use of a product
may reveal benefits and consequences (eg side effects), which
are not detected by the clinical trials required by the CSM. It
would seem sensible to collect post marketing data and for NICE
to review its decisions over time as more becomes known about
expensive products with which can be dangerous if not targeted
efficiently to appropriate patients. It would also seem sensible
to merge NICE and the CSM as they are complementary organisations.
In future marketing approvals will be valueless unless NICE sanctions
NHS reimbursement of technologies. Whilst such a merger would
require legislation, it would be efficient and is timely.
(x) The quality of the science on which NICE
decisions are made should be evaluated independently. The timeframe
in which NICE makes decisions is tight and much advice is bought
from academics for little or no reimbursement. The novelty of
NICE and the collection of "cv brownie points" by academics
are unlikely to ensure a continued supply of high quality analytical
advice. Errors in decision-making will be used unscrupulously
by companies to undermine NICE .The scope of NICE's work is large
and its potential importance for the NHS cannot be underestimated.
Poor funding should not undermine the quality of its work.
(xi) The Government believed that NICE would
eradicate postcode rationing. At present over three dozen patients
in North Yorkshire with multiple sclerosis are given the drug
beta interferon. In Midlothian no patients can get this treatment.
Experts in the two places disagree about the clinical and cost
effectiveness of this drug and it is, as a consequence, rationed
by postcode. NICE's decision about this drug, if it is confirmed
and if the Government had not circumscribed the decision by allocating
funds which undermine NICE, would have led to no new patients
getting the drug. Whilst rejections of drugs, if enforced, will
lead to equality over time, approvals may not. The decision to
approve a drug, even if based on excellent science and adopted
appropriately by practitioners, would lead to equal treatment
for those patients across jurisdictions. However to fund such
a drug other treatments would have to be given up and without
national guidance about what to abandon, a new form of postcode
rationing would be created.
(xii) The current systems operated by NICE
do not take account of equity issues. For instance should technologies,
which benefit the poor more than the rich be given a higher weighting
in terms of benefit (ie equity weighted QALYs)? Such issues are
very contentious. Current QALY measures value an additional QALY
received by the poor as being of equal value to a QALY received
by a middle class patient. Weighting issues should be explicitly
discussed in the policy debate about the future of NICE.
18. The Committee is reviewing NICE in the
light of key goals set out in A First Class Service. Each of the
four questions will now be reviewed in the light of the preceding
19. Is it providing clear and credible guidance?
The quality of decision-making has been questioned in some cases
(eg the case of Relenza where Dr Joe Collier in particular was
very critical in the Drugs and Therapeutics Bulletin). Given the
quality of the clinical trial data which is used by the CSM to
determine marketing decisions, and which is used by NICE in its
decision-making, it is inevitable that some advice will be contentious.
This makes it important that initial decisions by NICE are reviewed
as more evidence becomes available. This review process should
be formalised so that industry and the regulator agree what new
data should be collected and when reviews should be undertaken.
20. The guidance to the NHS is generally
clear and, in a period of increased funding, the advice, particularly
in the secondary sector, seems to be followed sometimes slavishly
and sometimes not too well ie with the usual variation in practice
endemic in all health care systems! The usual view appears to
be that it has to be "owned" by local practitioners
and where this involves an expansion in the clinical portfolio,
it does not seem too difficult to achieve, given time. The Commission
for Health Improvement appears to be attempting to police adherence
to NICE advice.
21. Has it ended confusion by providing
a single national focus? Doctors are easily confused as they are
showered with advice about how to treat patients from the net
and from pharmaceutical companies wishing to sell their products.
There are over a quarter of a million health care web sites on
the net and patients can log in to American and continental sites
and carry their print out to their NHS consultations. The challenge
for the doctor in declining such material and using NICE guidelines
remains considerable. Given NICE has acted like a "girl who
can't say no" (as in the Oklahoma musical!), the nice test
will be when it rejects more drugs for use in the NHS and doctors
have to take negative rather than positive messages to patient
22. Is it providing guidance that is locally
owned and acted on in the right way? In the absence of systematic
appraisal of take up of NICE advice, casual empiricism indicates
variation in local take up and ownership. The question begs a
definition of the "right way". As argued above, the
current relatively high cost per QALY cut off, the propensity
of NICE to reject very few technologies and the difficulties of
controlling the precise use of drugs once they are approved, means
that such interventions may be used inappropriately and that NHS
funding so consumed is not available for other more cost effective
23. Is actively promoting interventions
with good evidence of clinical and cost effectiveness so that
patients have faster access to treatments known to work? The evidence
to answer this question is scant. Undoubtedly for some patients
NICE guidance has led to the provision of treatments but the relative
cost effectiveness of some of these approvals is relatively low
(eg the drugs for obesity, motor neurone disease, cancer and influenza).
Providing faster access for relatively cost ineffective interventions,
has an opportunity cost as in deprives other patients of access
to more cost effective treatments.
24. Is NICE independent? Failure to get
NICE approval is a major threat to the cash flow of pharmaceutical
and medical equipment manufacturers. Their capacity to fund research
is very large and there is evidence of a general nature and not
specifically linked to NICE, that the evidence base can be corrupted.
The potential rewards for clinical and economic scientists to
"manipulate" data in a biased way are considerable.
Separately I have argued that research groups should be licensed
and independently audited so that bad practice is better policed.
As NICE progresses and becomes an even more significant impediment
to the marketing of products of marginal cost effectiveness, pressure
on the independence of researchers and NICE staff is likely to
increase. It is important that these people are seen to be independent
and transparent in their activities with public processes for
reporting of conflicts of interest.
25. The National Institute for Clinical
Excellence, like the Australian Pharmaceutical Benefits Committee
which has operated since 1993, challenges the inventors of drugs,
medical equipment and other new technologies to demonstrate that
these products of their ingenuity offer significant health benefits
at affordable cost to patients in the public and private health
care systems. This task is enormous and challenging, not the least
because the demand to demonstrate value for money has the potential
to damage strong commercial interests. A Government influenced
by those very same commercial lobbies has in the last year undermined
the Australian Pharmaceutical Benefits Committee.
26. Whilst the work of NICE is essential
to inform purchasers of health care and ensure that they deploy
the NHS budget of £50 billion so that it produces the greatest
level of population health improvement possible, there are areas
where its performance can be improved. The following issues require
more careful resolution:
(i) How was the cost-QALY cut off decided,
and is it too high? Is NICE accepting new technologies too easily
and, as a consequence, absorbing funding which would have been
available for older technologies of proven and superior cost effectiveness
such as hip replacements and cataract removals, where waiting
times remain excessive?
(ii) How can NICE contribute to the avoidance
of expenditure inflation induced physician discretion and failure
to follow its advice on the targeting of treatments? Would it
be efficient to discipline NICE by giving it part of the NHS budget
to fund the technologies it approves, with it being responsible
to purchasers for overspending on these treatments?
(iii) What can NICE learn from SIGN which
focuses its evaluations on patient episodes of care rather than
individual technologies? What should be the balance of investment
by NICE in appraisals of individual technologies as opposed to
treatment packages? Who is responsible and how should the collaboration
of NICE and SIGN/HTBS be best implemented to ensure that their
work is complementary and duplication is avoided?
(iv) Who should audit the quality of the
science of the NICE appraisers? How can the supply of independent
high quality research activity be sustained when some are paid
inadequately for their work and industrial pressure to "bend"
data is increasing as the commercial consequences of a failure
to get endorsement by NICE is increasing. Is it time to have national
and international accreditation of research quality offered by
academic and consultancy groups so that corruption is minimised
and its consequences made more significant?
(v) Is postcode rationing removed or merely
altered by NICE given there is no advice available to determine
the technologies, which are dropped to fund the products it, approves?
Should health care benefits from technologies, which benefit the
poor more than the rich be given an additional weight in terms
of benefit assessment (equity weighted Quays)?
Much current political and commercial practice
in the UK is based on the teaching of Marx. Thus Groucho Marx
argued: "the secret of life is honesty and fair play. If
you can fake that, you have made it!" The yawning gaps between
the claims made about new treatments and the evidence to sustain
these claims is enormous. Regulation to ensure that this Marxist
faking is minimised will always be complex and expensive. However
this is not an argument for giving up, it is a reason to reform
and improve essential bodies such as NICE so that cost effective
health care is delivered to patients in a timely and humane manner.
Maynard, A, Evidence based medicine: an incomplete
method for informing treatment choices, Lancet, 1997; 349:126-8.
Maynard, A and Bloor, K, Regulating the pharmaceutical
industry, British Medical Journal, 1997; 315:200-1.
Cookson, R, McDaid, D and Maynard, A, Wrong
SIGH NICE mess: is national guidance distorting the allocation
of resources? British Medical Journal, 2001; 323:743-45.