APPENDIX 7
Memorandum by Ealing, Hammersmith &
Hounslow Health Authority (NC 35)
1. INTRODUCTION
1.1.1 The Health Select Committee has requested
submissions from four Health Authorities (HAs) to assist its Inquiry
into the National Institute for Clinical Excellence (NICE). This
document is the submission from Ealing Hammersmith & Hounslow
Health Authority. The principal author is Dr Michael Soljak, Director
of Public Health & Health Strategy. Although the comments
of some Health Authority and other local NHS staff have been incorporated,
the short notice of the request for evidence has prevented its
being approved formally by the Authority itself.
1.1.2 At this stage oral evidence could
be provided to the inquiry but it has not been requested.
1.1.3 The submission is in three parts:
— comment on the current functioning
of NICE with respect to health topics and interventions it has
reviewed or supported[3];
— comment on health topics and interventions
which have been accepted as part of NICE's work programme but
have not yet been completed; and
— comment on health topics and interventions
which are not currently part of NICE's work programme.
1.1.4 The Health Authority has had considerable
experience in priority setting and technology appraisal prior
to the establishment of NICE. As the host purchaser for Hammersmith
Hospitals Trust, it has developed a close working relationship
with Trust clinicians who are at the forefront of research and
development on new health care interventions. As a large London
Health Authority with a diverse population, it has also needed
to review in depth the health care needs of this population and
the current services provided to them.
2. BACKGROUND
2.1 Current National Policy
2.1.1 The Government stated in A First
Class Service that local variations in resource allocation
("postcode rationing") are undesirable in a single NHS,
and it announced the establishment of the National Institute for
Clinical Effectiveness (NICE) to issue national guidance on health
care technologies. Subsequent NICE guidance has to some extent
supplanted the previous ad hoc methods of funding new technologies,
such as the direction issued in 1995 in the form of a Health Service
Circular (HSC), that HAs should consider funding beta interferon
for multiple sclerosis.
2.1.2 Previous legal advice was that NICE
guidance is guidance, and is not a direction by the Department
of Health. This advice was based on HAs' responsibility, as defined
in the National Health Service Act 1977 and other legislation,
to allocate NHS resources for their populations[4],[5].
A Health Authority which determined not to fund beta interferon
for multiple sclerosis lost a judicial review not because it had
failed to follow the advice or direction in the relevant HSC,
but because it had failed to consider it adequately[6].
2.1.3 Some Health Authorities have not fully
funded NICE guidance in 2001-02. However on 5 December 2001 Lord
Hunt announced on behalf of the Government that directions would
be issued to place an obligation on HAs and PCTs to provide appropriate
funding within three months for treatments recommended in NICE's
Technology Appraisal Guidance (TAGs). These generic directions
have taken effect from 1 January 2002.
2.1.4 Local trusts have asked for clarification
of the situation where local expert opinion is at variance with
the TAG. Unfortunately there is as yet no legal determination
of the status of NICE guidance with respect to individual clinicians,
and whether it will supplant the Bolam test for medical negligence.
2.2 Characteristics of a Good Priority-Setting
Policy
2.2.1 A list "desirable" characteristics
of a fair and equitable priority setting policy has been proposed[7].
These characteristics or conditions are:
— relevance condition: the rationale
rests on evidence, reasons and principles that the various stakeholders
can agree are relevant to deciding how to meet the diverse needs
of a covered population and the necessary resource constraints;
— publicity condition: decisions regarding
coverage for new technologies and their rationale must be publicly
accessible;
— appeals condition: there is a mechanism
for challenge and dispute resolution regarding decisions, including
the opportunity for revising decisions in the light of further
evidence or arguments;
— enforcement condition: there is either
of voluntary or public regulation of the process to ensure that
the first three conditions are met.
2.2.2 The discussion of the process used
by NICE so far in reviewing or supporting health topics and interventions
is based on a consideration of these conditions.
3. HEALTH TOPICS
AND INTERVENTIONS
NICE HAS REVIEWED
OR SUPPORTED
3.1.1 At the outset should be noted that
the wide variations and funding and implementation of health interventions
and technologies outlined above was widely recognised as a problem
prior to the establishment of NICE. This Health Authority welcomed
the establishment of NICE, and considers that there has been a
substantial improvement, led by NICE, in the way in which technologies
are assessed.
3.1.2 The relevance condition has to a large
extent been met by NICE's current process. NICE considers both
evidence of clinical effectiveness and cost effectiveness/utility.
However the current technology appraisal process takes insufficient
explicit account of other values which may underpin priority setting
in the NHS.
3.1.3 For example, the "right of rescue"
in the case of life-threatening or severely disabling illnesses
is a philosophical principle which has been widely discussed in
terms of health care provision, and has greatly influenced resource
allocation in the past. Hence the NHS has made considerable investment
in life-saving services for established cancer and coronary heart
disease, but a relatively small investment in smoking cessation
services, which may have prevented these diseases in many cases.
The general public clearly gives considerable weight to the right
of rescue. At present NICE's process takes no explicit account
of the severity of a condition and whether it is life-threatening,
although it could be argued that ring-fenced funding for such
conditions through National Service Frameworks (NSFs) implicitly
recognises this right (see section 5).
3.1.4 Another factor is the relative difficulty
in accumulating evidence of (cost) effectiveness for interventions
for rare or even uncommon conditions. Sufferers from rare conditions
may be disadvantaged by the emphasis given in the TAG process
to the strength of evidence. Other values are mentioned later
in this evidence.
3.1.5 The 2001 Guide to the Technology Appraisal
Process says very little about how the Appraisal Committee should
come to its decision. Issues such as these could be given more
prominence by making the process of weighing the strength of evidence,
relative (cost) effectiveness, and other factors, more explicit.
Ealing Hammersmith & Hounslow Health Authority's Priority-Setting
Subcommittee uses a matrix originally developed in the Southwest
NHS Region which specifically trades off the strength of evidence
against cost effectiveness and some related philosophical values
(see Annex 1). The matrix is not used mechanically as a decision-making
tool, but provides a framework for reasoned discussion.
3.1.6 It should also be noted that the use
of cost utility data ie cost per quality-adjusted life year (QALY)
as currently employed by NICE is based on an assumption that a
technology has the same benefit for all patients, when clearly
this is not the case. A consequence is that some TAGs state that
interventions should either be funded for all patients or no patients.
In some cases where NICE has issued TAG not to fund a technology,
negotiations have been commenced by the Department of Health to
put in place "risk-sharing" arrangements to fund those
patients who show benefit and not those who do not. There is a
need to acknowledge this issue in the process of developing TAG.
3.1.7 There is also concern amongst clinicians
and patient groups about the methods used for the valuation of
quality of life in QALY development. There is a need for more
discussion at the national level about improvements in this methodology
to ensure that it does not unfairly discriminate against disabled
or older people.
3.1.8 In the longer term, TAG could be developed
within a population model which took account of variations in
response to the impact of health technologies. As better information
about prevalence and actual outcomes was added to the model, it
would feed back into clearer predictions for clinicians and individual
patients about the likely outcomes of care. This model would also
support the more accurate costing of NICE guidance.
3.1.9 The national Cardiac Surgery Register
already provides this form of feedback using Bayesian statistical
techniques. The development of more sophisticated computer technology
such as neural networks could provide further support. This would
entail better linkage between the TAG and relevant data collection
than currently occurs, and the adoption of a common currency for
health outcomes.
3.1.10 Moreover, one national size does
not fit all. Local populations, and their health care needs, differ
significantly. Some HAs have used the World Bank disability-adjusted
life years (DALY) model as a default source of data on disability
and changes in it as a result of health care interventions. But
the NHS has the capacity to develop a model of health care and
the cost and outcome data to populate it so that each health community
can demonstrate that it is responding to local needs effectively.
Local trusts have expressed concern that commitments to funding
NICE TAG can distort local priorities. A population model which
allowed local priorities to be quantified and costed alongside
NICE TAG would help to highlight genuine funding issues.
3.1.11 The publicity condition has been
met to a significant extent by NICE with regard to its own work
programme. There are public meetings of the NICE Board, and the
appraisal process guidance and decisions regarding NICE's work
programme are publicly accessible through its website and publications.
It carried out a consultation to enhance this process during 2000,
it has a Partners Council, and is establishing a public forum.
The issue of closer Ministerial involvement in public discussion
of the wider aspects of priority-setting has been highlighted
below.
3.1.12 The appeals condition has been thoroughly
addressed by NICE with regard to the technology appraisal process.
There is a clear mechanism for challenge and dispute resolution
of TAP decisions, including the opportunity for revising decisions
in the light of further evidence or arguments.
3.1.13 However it is unclear how individual
patients could appeal against NICE guidance. Presumably such appeals
would take place at the local level, but there is no guidance
as to how such appeals should be heard. In Ealing Hammersmith
& Hounslow Health Authority, anonymised individual appeals
are heard in the private Part II deliberations of the Priority-Setting
Subcommittee. It is clear that in future such appeals should initially
be dealt with by PCTs. Again a model of good practice whose principles
were shared both by NICE and the local NHS would be very helpful.
3.1.14 The enforcement condition has to
a large extent been met by issuing TAG as a direction rather than
guidance. The regulation of NICE guidance should be an explicit
responsibility of the Commission for Health Improvement. However,
as noted above, there is no mechanism for dealing with technologies
which have not yet been considered by NICE or which are not on
its work programme (see below).
4. HEALTH TECHNOLOGIES
ON NICE'S
FUTURE WORK
PROGRAMME
4.1.1 Despite the creation of NICE and these
new directions, a need to assess technologies and set priorities
locally has remained because of the inherent delays in the NICE
appraisal process. As mentioned above, judicial reviews of resource
allocation generally turn upon Health Authorities' responsibility
for resource allocation as defined in legislation. It would be
difficult for a Health Authority to justify in law failure to
consider allocating resources (as opposed to actually allocating
resources) to an intervention purely because it was to be considered
by NICE at sometime in the future. The legal situation was recently
confirmed in a letter to Health Authorities from the Chief Operating
Officer of the NHS[8].
4.1.2 However in the current financial year,
in response to a questionnaire about the funding of a new technology,
a significant number of health authorities gave its appearance
on NICE's work programme as the reason for not funding the intervention.
This may be in part explained by the fact that the drugs concerned
are for treatment of a chronic rather than a life-threatening
disease. Nevertheless at the local level it is apparent that there
has been no improvement in the extent (or quality) of "postcode
rationing" outside the NICE TAG process.
4.1.3 There would appear to be two solutions
to this problem. The first would be to add appraisal by NICE to
the licensing requirements for new drugs and perhaps to establish
a formal licensing procedure for other forms of health technologies—the
so-called "fourth hurdle". The second solution would
be to develop guidance and a framework for rapid technology appraisals
at the local level, which incorporated to some extent the good
practice inherent in the NICE model (see section 3).
4.1.4 As it is unlikely that any PCT would
be sufficiently well-resourced to undertake even a rapid appraisal
alone, such appraisals would need to be co-ordinated at the Strategic
Health Authority (StHA) or Regional level, possibly with input
from NICE Collaborating Centres. By seeking to develop good practice
at the local level, and drawing upon this for national TAGs, it
might be possible to significantly accelerate NICE's work programme.
5. HEALTH TOPICS
AND INTERVENTIONS
WHICH ARE
NOT CURRENTLY
ON NICE'S
WORK PROGRAMME
5.1 TAG Selection Criteria
5.1.1 Referrals to NICE so far have tended
to be for new drugs (and occasionally new devices eg stents) rather
than existing interventions involving human resources eg nursing
care or health visiting[9].
While some central resources are ring-fenced for such purposes,
there is a need to ensure that such interventions are considered
for funding, or alternatively, resources are redeployed from them
to fund other types of health care interventions.
5.1.2 This raises the issue of selection
criteria for TAG and guidelines topics. At present the topics
are proposed by NICE's stakeholders, especially the Department
of Health, using the selection criteria from the 2001 Guide to
the Technology Appraisal Process (see Annex 2). However it appears
that the criteria, which quite correctly include health and financial
impact, and possible variation in practice, are only applied once
the topics have been proposed. A more systematic review of the
impact of health care interventions would have resulted in TAG
being developed for common conditions such as femoral neck fracture,
stroke, depression, hypertension and dyslipidaemia. NSFs are tackling
some of these issues, but better coordination would ensure consistency
of approach and avoid duplication.
5.2 The Need For Better Integrated Priority-Setting
Processes
5.2.1 There appears to have been relatively
little in-depth discussion at the national level about developing
a more transparent and rational resource allocation process at
the highest level. The Treasury has rejected the concept of an
hypothecated tax for the NHS, largely on the basis that it would
complicate the setting of priorities between Government sectors.
Other branches of central Government—for example the Department
of Transport, Local Government and the Regions—use a similar
approach (usually cost per life year gained) to prioritise health
and safety investment. Hence There is a need to extend the approach
developed by NICE to other public sector interventions, in the
first instance perhaps to social care eg continuing care as it
is often delivered in conjunction with health care.
5.2.2 Within the NHS, Ministers have made
it clear that they expect the targets or "must dos"
of the NHS Plan, along with NICE guidance, to be a first call
upon resources, although the detailed rationale for this is sometimes
unclear. For example, a large proportion of new NHS resources
have been allocated to reducing waiting times for elective care
because of the high level of public concern about these in recent
years. However, while public concern is one important criterion
for resource allocation, it is not the only criterion, as NICE's
work illustrates.
5.2.3 Ministers could take a more prominent
role in a discussion about the extent to which values such as
public concern could be balanced against other criteria such as
clinical and cost-effectiveness. This may be what the Secretary
of State means by affordability[10].
However Ministers could be more active in leading a wider discussion
about affordability and priority-setting in the NHS, and the values
which underpin it.
5.2.4 For example, while many elective surgical
procedures are relatively cost-effective, universal waiting time
targets imply that they are all equally so. This is clearly not
the case. Because of differences in supply and expressed need
for services, waiting times do not have a direct relationship
with underlying need or the loss of benefit while waiting. They
have been chosen using the criterion of public concern.
5.2.5 If NICE were to develop more specific
TAG on the provision of an intervention such as joint replacement,
its evidence-based process should lead to the development of intervention
thresholds based on likely capacity to benefit rather than waiting
time. NICE has just released Referral Advice on 11 conditions.
But the advice for osteoarthritis, for example, does not attempt
to define a threshold for routine referral or intervention but
states only that: "referral should be based on an explicit
scoring system. . . developed locally. . . (which) should take
into account. . . pain, disability, (and) reduced functional capacity.
. ." [11].
But what should the threshold score be?
5.2.6 However in the longer term, there
is not necessarily a conflict between public concern about waiting
lists and an evidence-based approach to likely capacity to benefit
from healthcare interventions. As better information about the
outcomes of interventions becomes available, and it is actively
shared in a partnership between patients and clinicians, the focus
of public and political concern about access to care will naturally
shift from waiting times to the funded clinical thresholds for
intervention. This shift should be actively led and encouraged
by Ministers and the Department of Health.
5.2.7 In summary, currently there appear
to be at least two approaches to priority-setting in the NHS-
one based on political and public concern, and another based on
NICE's evidence-based approach to technology appraisal. To date,
NICE has not appraised a number of healthcare interventions funded
through the political/public concern stream. However in the longer
term the two approaches are not necessarily in conflict. There
is a need for public discussion and political leadership to begin
to integrate these values. In addition, NICE needs to further
develop its current technology appraisal process to consider other
healthcare interventions in addition to drugs and devices.
6. IMPLEMENTATION
OF NICE GUIDANCE
6.1.1 There is a large body of evidence
demonstrating the need for multifaceted strategies to promote
the implementation of effective practice[12].
NICE has a role in supporting implementation beyond the provision
of guidelines, an example being its stewardship of PRODIGY GP
prescribing software. In addition, NHS Trusts are directly responsible
for implementation, and the Commission for Health Improvement
also has a role in regulating and inspecting their progress.
6.1.2 There is a significant risk that TAG
will not be implemented not because of insufficient resources
to pay for the technology itself, but because of organisational
overload, lack of human resource, and lack of skills and support.
It is imperative that these issues be considered when TAG is being
produced. In addition there is a great need for more central support
and co-ordination of NICE guidance, for example through the definition
of national information requirements to audit implementation.
7. RECOMMENDATIONS
7.1.1 The Committee is asked to consider
including the following recommendations in its Report:
1. That NICE be commended for the substantial
progress it has made in a short period in ensuring fair and equitable
access to NHS care.
2. That a legal determination of the status
of NICE guidance with respect to individual clinicians be produced.
3. That NICE continues to develop its Guide
to the Technology Appraisal Process to explicitly consider the
decision-making process and the values considered in it, and to
improve the methodology for valuing quality of life in health
economic analyses.
4. That a national population model be developed
to quantify health problems (both local and national) and cost
NICE TAG at the level of each PCT, and for use at the local level
as a planning tool.
5. That a model of good practice for hearing
individual patient appeals about NICE TAG and local funding policies
be provided for PCTs.
6. That either appraisal by NICE be added
to the licensing requirements for new drugs and licensing requirements
be established for other forms of health technologies; or a model
of good practice whose principles were shared both by NICE and
the local NHS is produced. This model would assure the quality
of rapid technology appraisals at the local level, and ensure
that they were more widely usable and could be drawn upon by NICE
in its national TAG.
7. That NICE further develops its technology
appraisal process (and if necessary more research is commissioned)
to consider other common healthcare interventions in addition
to drugs and devices, in particular for common conditions and
for other NHS Plan targets (especially interventions subject to
waiting list targets) and National Service Framework standards,
so that priority-setting in the NHS becomes more integrated over
time.
8. That the process also be extended to
related public sector interventions, initially social care, so
that priority-setting in the NHS and other publicly-funded services
becomes more integrated over time.
9. That Health Ministers lead a public discussion
on the other values which should underpin priority-setting in
the NHS in addition to clinical and cost-effectiveness.
8. ANNEX 1: EVIDENCE/COST
EFFECTIVENESS MATRIX
Strength of Evidence
| Category | Description
|
| I | Strong evidence obtained from at least one properly designed, randomised controlled trial of appropriate size
|
| II | Evidence from well designed controlled trials with randomisation.
|
| III | Evidence from well designed, cohort or case controlled analytical studies, preferably from more than one centre or research group
Evidence obtained from multiple time series, or from dramatic results in uncontrolled experiments.
|
| IV | Opinions of respected authorities based on clinical evidence, descriptive studies or reports of expert committees
Evidence inadequate owing to problems of methodology (sample size, length or comprehensiveness of follow-up) or conflicts of evidence
|
| |
Evidence of Cost Effectiveness/Cost Utility
| Category | Description
|
| A | Cost utility of the intervention is less than circa £3,000 per life year gained (or quality equivalent)
|
| B | Cost utility of the intervention is between £3,000 and £20,000 per life year gained
|
| C | Cost utility of the intervention is between £20,000 and £50,000 per life year gained
|
| D | Cost utility of over £30,000 per life gained
|
| E | No life years gained or the treatment is detrimental to patients
|
| |
Matrix
| Quality of the Evidence | Cost Utility
| | | |
|
| A | B
| C | D
| E |
| I | ++++ |
+++ | ++
| + | -
|
| II | +++ |
++ | +
| + | x
|
| III | ++ |
+ | x
| x | x
|
| IV | + | x
| o | o
| o |
| | |
| | |
Key
++++ = strongly recommended
+++ = highly recommended
++ = recommended
+ = beneficial but high cost; may be recommended when other
factors are considered/ or in light of further evidence
x = not recommended
o = not proven
Stage 2
Additional factors which should be considered for treatments
that score +
Is the disease rare eg annual incidence 1/100,000 population
or less?
Are there ethical reasons why randomised controlled trials
could not be carried out?
Are there any other difficult/ unusual circumstances, which
restrict further research into this disease?
Are there any other criteria, which should be taken into,
account eg benefits to carers, efficiency gains etc?
9. ANNEX 2: SELECTION
CRITERIA FOR
TECHNOLOGY APPRAISAL
GUIDANCE[13]
— Is the technology likely to result in a significant
health benefit, taken across the NHS as a whole, if given to all
patients for whom it is indicated?
— Is the technology likely to result in a significant
impact on other health-related government policies (eg reduction
in health inequalities)?
— Is the technology likely to have a significant
impact on NHS resources (financial or other) if given to all patients
for whom it is indicated?
— Is the Institute likely to be able to add value
by issuing national guidance? For instance, in the absence of
such guidance is there likely to be significant controversy over
the interpretation or significance of the available evidence on
clinical and cost effectiveness?
3
Support here includes the provision of guidelines or referral
advice such as that just released for 11 common health problems. Back
4
". . .in establishing priorities- comparing the respective
needs of patients suffering from different illnesses and determining
the respective strengths of their claims to treatment. . . it
is vital for an authority accurately to assess the nature and
seriousness of each type of illness, to determine the effectiveness
of various forms of treatment for it, and to give proper effect
to that assessment and that determination in the application of
its policy." NW Lancashire HA v A, D & G (2000). Back
5
This function will presumably be transferred to PCTs in 2002
as part of legislative changes consequential on Shifting the Balance
of Power (StBoP). Back
6
The HSC itself was not a clear direction. Subsequent judicial
interpretation was that: "the (HA) had to have regard to
that national policy. They were not obliged to follow the policy
but if they decided to depart from it, they had to give clear
reasons for doing so..." R v N Derbyshire HA, ex parte Fisher
(1997). Back
7
McIver S and Ham C. Five cases, four actors and a moral: lessons
from studies of contested treatment decisions. Health Expectations
3; 1999:114-124. Back
8
"NICE are appraising (imatinib) and their guidance is expected
to be issued next August. The appraisal was originally timed to
coincide with the expected licensing process but the latter has
gone more quickly than expected. Health Authorities should
bear in mind that delaying a decision until NICE has issued its
advice is a refusal to fund the treatment and would be regarded
as a negative judgement." (sic) Neil McKay, 24 October
2001. Back
9
Raftery J. NICE: faster access to modern treatments? Analysis
of guidance on health technologies. BMJ 2001; 323:1300-3 Back
10
"There are two quite separate distinctions. . . which is
about assessing clinical and cost effectiveness and a quite separate
set of decisions which are around affordability issues. In the
end you would want. . . affordability decisions to be located
with an accountable politician who has to answer to the House
of Commons and to Parliament. It just so happens that the accountable
politician is me." Hon Alan Milburn, Evidence to House of
Commons Health Committee, 8 November 2000, paragraph 336. Back
11
Referral Advice, NICE December 2001. Back
12
NHS Centre for Reviews & Dissemination. Getting evidence
into practice. Effective Health Care Bulletin V5 No 1, York University
February 1999. Back
13
Guide to the Technology Appraisal Process. NICE, June 2001. Back
| 
