Examination of Witnesses (Questions 200-219)|
TUESDAY 3RD DECEMBER 2002
200. Just a comment about surveillance on viral
disease. In, for example, gastroenteritis surveillance may not
be important for treatment but it is crucial in vaccination strategy.
Are there sensitive and accurate near-patient tests for viral
diagnoses available to GPs, for example?
(Dr Brown) It is very variable. There are at least
four virus groups that cause gastroenteritis and give you identical
symptoms so it is impossible to make the diagnosis clinically.
There are widespread diagnostic tests for the most important of
those, the rota virus and you will find all the local laboratories
are able to provide that. For the other causes there are not diagnostic
tests, except for electronmicroscopy, where that occurs, and that
is a very inadequate technique for a number of these infections.
So it is very variable.
201. How long does it take a newly developed
near-patient test to go through the validation process and get
into clinical use?
(Dr Brown) One probably needs to go back a step further,
and that is to say: is there a commercial company that is interested
in making the test? Do they believe they will make sufficient
money from this to go through the very extensive developmental
programme and validation and FDA approval if they are going to
sell it and make lots of money? In the case of viral gastroenteritis,
with the exception of the rota virus, which is the major cause
of hospitalisation of infants, that has not proved to be the case.
In terms of once the test is available, clearly different parts
of the world handle the validation of that in different ways.
It is still slightly informal within the European Community, although
this is being tightened up over the next few years. I think the
best way of validating these tests would be to use the evidence-based
medicine approach, which is to say you have to do a study which
indicates how that test result contributes to a change in what
the clinician does with that patient, and those are expensive
and difficult studies to do because they, by their very nature,
require clinicians and laboratories, and the funding structure
to do that is not in place. It is very much an ad hoc thing. You
see much more rapid progress, for example, where there is an intervention,
such as the influenza antivirals that were recently introduced.
The company that made the antiviral also invested in producing
near-patient tests that would accompany that to help focus the
use of it. However, for the vast majority of viruses that situation
does not pertain.
202. On a point of interest for Dr Pillay, we
heard about antimicrobial resistance to bacteria and about the
use of microchips that could identify the major problem of resistance.
Do the same exist in viruses against the antivirals and resistance
developing in the antiviral field?
(Dr Pillay) Yes. As we have heard, many of the viruses
which are amenable to treatment cannot be grown. They are highly
variable in terms of their genetic structure and, increasingly,
it is the sequencing of large bits of the viral genome that has
rapidly entered routine, clinical practice. This is an enormous
advance. We are not just talking about PCR-based molecular tests,
it is actually moving the sort of human genome sequencing project
into diagnostic laboratories, and all the laboratories represented
here would be undertaking those sort of tests. There are clearly
commercial companies that can support that sort of provision but,
equally, it is because of the costthe enormous cost(and
here we are talking about orders of magnitude of cost over and
above our existing tests), that there is a pressure for development
of our own methodologies that can then be circulated around and
tried within different laboratories in the network, as David has
mentioned with some other tests. That is a major advance in that
whole area. If you do not mind me very briefly skipping to a later
question which pertains to this, the key is how we can utilise
IT and so-called bio-informatic advances to utilise that sort
of data in public health. That is the challenge at the moment.
203. This may be too far in the future but to
what extent is there any possibility of recognising viruses, not
via the clearly very expensive and advanced techniques that you
are talking of at the moment but going one step further and actually
looking at secondary changes in body chemistry which may be indicative
of the presence of a particular virus that may indeed be much
more rapidly detectable once you know what they are?
(Professor Griffiths) There are some theoretical possibilities.
For example, you might look for interferon production as an area
of generic response to indicate the presence of a virus as opposed
to bacterial infection. In practice those have not turned out
to be very useful when looked at but it is an interesting idea.
We tend to focus on the individual infectious agent and you are
encouraging us to look at more generic classifications and dichotomy
for these infections.
(Dr Zambon) I believe there are a number of big pharmaceutical
companies who are interested in the concept of a dipstick test
which will allow discrimination of bacterial and viral infection
based on an innate immune response. However, I think that that
will require substantial investment and is unlikely to be available
from the public sector. If it is successful and being brought
into the health centre, it will require some form of public-private
relationship to bring it forward. It is a huge project.
(Dr Pillay) I would just add that although it may
be useful to identify a more generic method for diagnosing a virus,
there is a huge range of different viruses but not only that,
one of the peculiarities about virology is that a clinically important
virus is identified every few years and has been for the last
20 or 30 years, especially with the onset of more molecular methodologies,
and we want to be in a situation to identify and diagnose specific
viruses for the purpose of intervention, for instance antivirals.
204. To a degree you have covered some of the
points I wanted to raise in terms of surveillance, public health
and policy making, but it might help to clarify one or two points
you have made. My question concerns surveillance systems related
to viral infection and how effective they are in informing public
health policy decisions. If they are not, how could you improve
it? If you suggest a method of improving it, what evidence base
is there for that?
(Dr Brown) Perhaps if I start with that question.
I think the picture is quite variable. Probably the key issue
for distinguishing good and effective surveillance programmes
from other ones in my experience has been where you have an intervention
that you are applying such as vaccination. You can then focus
your surveillance on generating the data that will inform the
policy you need for your vaccination programme. With MMR that
is exactly the case. We have a surveillance programme based on
measuring the coverage and the proportion of people who should
be receiving the vaccine who do. We have a second programme which
is aimed at identifying measles, mumps and rubella cases in the
community and how well is our community protected from those infections.
We have a third programme which is based on measuring the serological
profile to those different infections in a representative population
to find out whether there are groups within the population that
are susceptible. We also have some adverse events monitoring and
we monitor the cohorts, that is stratified immune profiling, which
together with what we are seeing in the community and the predictions
which come in the coverage enable us to make very precise, forward-looking
models of what is likely to happen. Those have been used very
effectively to steer the vaccination programme, so I think that
is a good example. We have already discussed the problem of what
is the burden of disease due to different enteric pathogens and
how can you prioritise your efforts unless you understand that
basic principle. We discussed that in primary care where many
of these cases occur and we have very little idea as to how important
each of these infections is. We are collecting some information
on that from laboratory reports but it tells us almost nothing
about that. Probably the difference between those two is that
with a vaccination programme you have a very precise question
so you can actually design the surveillance need and fulfil the
circle in that way so that the two feed back into each other.
If you look across the range they are very variable.
205. Can I have a supplementary because I thought
in the evidence that Dr Zambon has produced you were a bit sceptical,
maybe that is too harsh a word, of the surveillance. What should
drive the surveillance? Should we ask the question or should it
be a passive collection of data and an aggregation of that data
that drives it?
(Dr Zambon) I think Dr Brown has answered this question
very much as I would have done. The only thing I can add to that
is there are examples where we collect data nationally but we
do not use it. There are a good number of examples there. There
are also examples where we do not collect data which could be
useful to us, for example, on travel associated infections, where
we do not really have any qualitative data on malaria in this
country in the sense there is not a national reporting system,
there is not the application of standardised reporting or data
collection, and there is not a systematic surveillance system,
and the reason, if you will, for the scepticism is that we can
identify the problems that we know exist already but we are very
unclear about how they may be improved and how things are to be
improved under the new agency due to start on 1 April. Indeed,
we are concerned about how things that are already good like measles,
mumps and rubella surveillance and some aspects of, for example,
the influenza surveillance programme are to be continued without
a clear mandate for a strong central lead. It is my belief that
the surveillance systems which we can identify as being good,
indeed excellent, not only on a national scale but internationally
as well, have two additional components, a strong R&D element
which informs the application and use of technologies, often cutting
edge and, secondly, a strong central lead. These elements are
not clear in the papers with respect to the introduction of the
HPA as to how the strong central lead for national surveillance
programmes is to be maintained. There is a great deal of talk
about regional data collection, regionalisation, but how is that
to be integrated in the national picture? So I think I would summarise
it by saying there are some excellent surveillance systems, some
systems where the information collected is partial and does not
always address the question, some and systems we do not collect
the right data in order to inform the policy, and I am not at
all clear any of this is going to be improved in the HPA.
Baroness Warwick of Undercliffe
206. I wonder if I can look at another aspect
of public health campaigns and ask you whether they rely too much
on attempting to modify behaviour given the evidence that we have
that these campaigns often fail. In your evidence you express
concern that strategies, and you particularly refer to the sexual
health and food preparation strategies, are based on attempts
to modify behaviour and yet surveillance does not take those strategies
into account at all in order to monitor effectiveness. I think
you go on to question whether substantial amounts of public health
money may be spent in ways that do not provide evidence of their
usefulness. Could you comment on how the links between public
health campaigns and surveillance might be improved?
(Dr Brown) I do have some reservations about behaviour
modification, not solely based on my own experience of my inability
to change when I might wish to. I think both the sexual health
programme and the recent Food Standards Agency initiative have
good features of them, namely they have good process evaluation
and at the end you will ask people has your behaviour changed,
and I have always found I have never been completely convinced
that that is a very good and robust method of establishing behaviour
change. I think there are two separate questions, the first is
that the best way to spend a limited budget without a validated
method for making a difference and I think that still could apply
to behaviour change. I think the two things that could make them
more justifiable would be that there had been a proper trial to
demonstrate, firstly, that you could make a difference and that
you had a marker that you could use to indicate that change. For
example, in the Sexual Health Strategy you might identify the
HIV incidence or perhaps even better as a marker of behaviour
change the herpes simplex virus because its acquisition is directly
linked to sexual behaviour so you have an objective marker. You
have to do a trial first because obviously you do not go straight
to population interventions, likewise with the handwashing programme
that is being broadcast at the moment. There are process measures
in place but is that a good way of measuring that you have made
a difference? I am not sure that it is. It is a question of prioritising
the fairly major expenditure that follows on from advertising
programmes against the other interventions that are available.
207. Can I follow up with one brief question.
There is also the question of whether there is value in itself
of raising awareness. We have had evidence from Dr Nick Beeching
that in the recent chlamydia campaign there were measures in place
to demonstrate that awareness had been raised but the incidence
of cases rose. I wonder whether again you could comment on that
because I think it really is quite difficult to look for those
sorts of objective ways of measuring whether, even with increased
awareness, it has resulted in changing patterns of behaviour?
(Dr Brown) Clearly there may be increased awareness.
Quite how you measure that, again, is a good question. In a way
you have answered your own point, if the awareness was there but
the acquisition was greater it seems to me that was not necessarily
a valuable addition to raise awareness and catch more of the disease.
208. Any other comments, gentlemen? Dr Pillay,
I missed your eye on the previous question. Did you have any final
comment to make on question six before we move on?
(Dr Pillay) The only thing I would add is to agree
with both Maria and David about the patchiness of surveillance
systems so where there have been historically very strong links
between epidemiologists and sentinel surveillance then I think
there is no doubt that that has been very successful indeed. But
as we in virology are able to do more and more things to not only
diagnose infections but ask quantitative questions about the nature
of infections, the strain, the differences between individuals,
and the amount of the virus and so forth, then it allows far more
sophisticated surveillance to be done. We are in a situation that
on 1 April we do not know what the structure of national laboratory
surveillance will be within England and Wales. I think given that
uncertainty and the huge demands there are on the potential where,
through the networks, we are developing the ability to generate
data link constructively and be forward looking with epidemiologists,
there is a danger that will be compromised.
- Just on Lady Warwick's question, question seven,
would you like to comment on whether campaigns to modify public
behaviour on the grounds of health actually fail. I was thinking
about smoking. It seems to me that smoking has been a campaign
which has been quite successful.
(Dr Brown) I am sure you should speak to your social
scientist representatives. The time-frame for that success is
perhaps quite a long one.
Baroness Finlay of Llandaff
210. I am concerned in terms of the relationship
between public health medicine and the virology services. With
all the changes occurring, who is going to be in charge of epidemic
(Dr Brown) Not me!
211. You were making comments about the situation
in England and Wales. We know that other parts of the British
Isles differ but what about Europe, are they better organised
on the continent of Europe than we are?
(Dr Zambon) Recently, it has often seemed ironic to
a number of colleagues and myself that we are being asked to speak
at international meetings and for us to elaborate on how good
the UK surveillance systems are, when we are in the process, apparently,
of dismantling them. In respect of influenza, respiratory disease
and respiratory viruses it is clear that the UK, along with France
and the Netherlands, have the most advanced systems but what is
also becoming clear over the co-ordination of Europe activities
is that there is going to be much more investment into the development
of sentinel primary care systems in Europe for the development
of disease and syndrome specific information linked to clinical
diagnosis, so a great strengthening of the primary care infrastructure.
212. Who would do that co-ordination within
the European Union?
(Dr Zambon) As far as influenza and respiratory viruses
are concerned, there is representation at every level from every
country by both epidemiologists and virologists, so there is usually
a pair per country who are responsible for the aggregation of
national data. These are being brought together with EU funding
to develop the regional networks, regional in this case being
regional European networks. Influenza is a particularly well-developed
model but I think there are other examples so there is a great
deal more co-ordination work in Europe. I think it is also fair
to say that some European countries where there have been very
decentralised approaches to epidemiology have recognised there
is a greater need for centralisation. Good examples there would
be Spain and Germany.
Chairman: Any further points? If not, can we move
on to question eight.
213. We have talked about IT and touched on
it a number of different times during today's session. Clearly
none of you were actually satisfied with the IT arrangements in
place at the moment. Would you like to say what you think they
should be? The NHS has recently announced it is going to spend
very large amounts of money on IT and there is a new IT director.
What do you think they should be doing from your point of view?
As a corollary, one of the new pieces of information for surveillance
which is now regularly available is the response pattern to NHS
Direct. I just wonder if any of you have used the monthly patterns
that emerge from that?
(Dr Zambon) Perhaps I will start because we are likely
to all have comments. It is important that whatever IT systems
are developed they do talk to each other. One of the greatest
difficulties has been trying to work with a data set from one
system and translate it into another system. Based on the remarks
with respect to primary care, there is a need to link diagnosis
in primary care with prescribing in primary care. It is quite
interesting that at the moment the information capture of prescription
in primary care involves one system whereas with sentinel practices
we have been involved in trying to acquire information on microbiological
diagnosis and that has involved a separate system. What one would
like to see in terms of developing data on disease burden and
also on antimicrobial resistance is a system which allows linkage
of information not only of consultations but also sampling, results
and prescriptions, so we have a fully integrated IT system in
primary care which then can link to what goes on in local hospitals.
That is very much a comment with respect to service delivery but
then there are also important IT elements required for information
exchange at national and regional level and there the emphasis
needs to be on systems talking to each other. I am sure Dr Pillay
will say things about the requirement for bio-informatics and
the requirement for a national data set there.
(Dr Pillay) Before I get to that I would say in the
UK we have a unique opportunity because there is the National
Health Service framework into which both PCTs and trusts fall.
Private work and private laboratories make up only a very, very
small proportion of the work undertaken within the UK so we have
with this new NHS initiative an ideal opportunity to capture truly
national data. There are not many countries in the world who can
do that. So really the challenge is to optimise that. It seems
to me there are two areas of IT demand, the first is to be able
to integrate clinical data/denominater data in order to give a
truer perspective on positive diagnoses which currently we do
not have, and Maria has discussed that, so I will not go over
it. The second area in terms of virologyand I have alluded
to this beforeis the complex nature of the data that we
are now generating in the laboratory. Unfortunately, many of the
existing surveillance IT systems allow for a diagnosis yes or
no. Really just talking from my own area of expertise in terms
of HIV drug resistance, there is more HIV in the UK than ever
before, there are more diagnoses, there are more imported infections,
there are more treated infections, there is more resistance developing.
In order to capture these parameters we have the opportunity to
get data from all virology laboratories so we have the potential
to have a handle on every HIV resistant virus within the United
Kingdom. That is a challenge. The ability is there for us to merge
data. The software requirements needed to do that are challenging
but certainly should not be insurmountable, and I would argue
that what is now the case with HIV and will soon be the case with
hepatitis C, and I am sure David will talk about other areas of
molecular epidemiology where they are looking at the gene sequence
of various organisms to look at transmission events and circulation
within the country. So I think there are major demands which I
would hope could be fulfilled.
(Dr Brown) I think the question of the interfaces
between different IT systems being bridgeable is one critical
issue so that information can be shared. To respond briefly to
the question of how one is able to use the sequence data that
we can now generate, clearly one issue is that you need to use
the same technique across the country and active management of
that is critical. The second is clearly web-based reporting and
web-based analytical tools are now starting to be used much more
widely. I am involved with one of those and there is the potential
to enable much more rapid exchange of this much more complex information.
(Professor Griffiths) To follow up a point Dr Pillay
made about the HIV side, and this links back to your question
6 really, it would be quite crucial to know which sub-types or
clades of virus are circulating in our country and to start trying
to anticipate potential antivirals and HIV vaccines because some
of the those will protect against some types of HIV and not against
others and it is quite important to have that in hand now as you
look forward a few years to those policy decisions.
214. None of you commented on NHS Direct and
whether you found any useful patterns emerging from that.
(Dr Zambon) I believe NHS Direct has received some
funding for research and development for evaluation for its use
as a surveillance system to assist in recognition of influenza.
I think the data so far are preliminary and I am not sure that
a clear conclusion can be drawn as yet. One of the things that
I believe is being explored is whether or not there can be some
form of sampling linked to NHS calls because one of the problems
is specificity in the system and how to actually use it in a very
specific way. My experience of NHS Direct is limited to that related
to respiratory diagnosis.
(Dr Brown) I think the point is that it is still early
and perhaps unproven and you can see the problems you may have
with syndromes caused by 25 different infections. How do you interpret
what that means, particularly for things that are widely prevalent
in the community anyway? How do we use it to detect specific epidemics
which are due to one individual cause when there are 20 others?
I am sure it will have a useful role but we are not quite there
yet from my perspective.
215. All of you have been asking for more information.
Have any of you had any experience of being overwhelmed by the
amount of information you have?
(Dr Zambon) One could say every winter!
216. Because the management of this information
if one gets good IT systems in place becomes critical and you
can become overwhelmed by it to the extent it is no longer useful.
(Dr Pillay) As someone who runs a national reference
unit, a major challenge for me and my team is to try and identify
patterns within the data and to get some useful outputs out of
all the information that comes in.
217. How are you succeeding in getting this
(Dr Pillay) I resort to having to write grant applications
to the R&D funding streams within the UK despite being within
a PHLS structure, to deal with that precise area of data assimilation.
We can generate the data; get the data from other laboratories,
what we cannot do is look at it in a coherent way.
Chairman: Can we move on to the final question. Lady
Baroness Finlay of Llandaff
218. Some of the evidence that has been received
expresses caution about rushing to use new diagnostic technologies
because of a lack of robust evidence about their effectiveness.
Do you concur with that opinion or do you think we should be exploiting
and pushing forward the frontiers and using some of the new technologies
(Dr Brown) I think we should certainly be moving forward
and exploiting it. "Rush in" is not a phrase I would
use. I think "orderly forward progress" would be a more
appropriate way of describing it. Clearly we touched earlier on
the issue of how you validate those tests and it has to be in
the context of evaluation of patient management which are complex
evaluations which will take some time to do for a range of these
(Professor Griffiths) I am not so sure what you mean
by the use of the word "new". Do you mean PCR and molecular
tests related to it?
219. I think one also has to look ahead to other
things which will emerge in the future. We have had a bit of a
discussion now just in terms of the general principle of should
there be some form of control mechansim in the new techniques
rolled out in the context of a trial which is fully resourced
and funded as a research programme to validate it before it is
more widely rolled out into clinical use?
(Professor Griffiths) That is something we are very
keen on in the network side, that there are enthusiasts of assays
for their own particular markets and within target populations.
What we would like to do is to leave behind the ones that did
not work and roll out the ones that have been fully validated.
There is no point in reinventing the wheel. We could use that
resource for everyone in the UK.