Examination of Witnesses (Questions 200
WEDNESDAY 11 JUNE 2008
Professor Peter Furness, Dr John Crolla, Dr Imran
Rafi and Professor Stephen O'Rahilly
I hope it does not have the same fate as Modernising Medical Careers!
Professor Furness: It is worryingly similar.
Dr Crolla: It is worryingly similar and it is
some of the same personnel but, anyway, I will not divert there.
The current plan, although none of this is actually written in
stone, is that the pre-registration scientist training will be
dismantled and be replaced by a generic training, so this will
have a flexible workforce, so one day you will be in a biochemist
lab, the next day you will be in a haematology lab, then histopathology,
and then you can rotate into genetics. All the professional groups
have reported back that this does not make sense. It only seems
to make sense to the Modernising Science Careers framework. The
bit that we do 100 per cent agree with is the post-registration
part of the training which will be tenure-tracked, it will be
linked to the number of available consultant posts, and will be
fully funded right through the FRC Path through the examination
system of the Royal College of Pathologists, fully accredited,
et cetera. Thus we have a real problem there. What is actually
happening of course is that emerging technologies have blurred
completely the distinction between the cytology, the chromosome
and DNA. The primary analate will soon be DNA and it is effectively
already DNA, so if we are going to do a genome-wide test, we will
do it on a DNA-based test, be it an array looking for copy number
change or an array looking for expression profiling. This has
really happened so fast that I do not think we have got the structures
in place in order to accommodate this, but I think the professions
have really got to get together. The Clinical Molecular Genetics
Society and the ACC in collaboration with the RCPath must get
this sorted and soon. The other important point I really want
to make to this Committee is that we have now regulated ourselves
to a stand-still. This is so important and it was brought home
to me when my PhD student graduated this year. She had done three
years work in my laboratory on array CGH applied to constitutional
chromosome abnormalities. She was highly skilled, she was trained
at the Sanger as well as in my laboratory, and she decided to
go to Italy and she has now got a job setting up a diagnostic
array laboratory in Italy. I could not have offered her a job
in the UK except as a technologist.
Dr Crolla: Because in order to undergo the regulatory
training she would have to go back to supernumerary grade A training,
do four years, learn how to read chromosomes, and then go into
higher specialist training, so she would enter into a training
programme which actually was going to not make her fit for purpose.
I see why you say "worryingly similar"; there is no
flexibility at all.
Dr Crolla: That is an important point.
Q203 Baroness O'Neill of Bengarve:
I wanted to put in a small and slightly provocative supplementary
at this point. How far should the NHS take responsibilityand
I think this is particularly a question for Dr Rafifor
interpreting tests that people have got over the Internet? Might
it not be better to say, "You got this; you paid for it;
I am not even going to advise you on whether your money was well-spent.
You come to me with a symptom or no symptom."
Dr Rafi: I think most GPs would want to try
and support their patients and if
They are not your patients at that stage, they might be on your
Dr Rafi: They are still your patients and it
is in the same way that they come to you with the results of a
general health screening that they might have had done privately.
I think most general practitioners would be happy to help but
it is really having the competency to be able to deal with the
information that you have got. In terms of competency, the RCGP
are providing a competency framework for practitioners who want
to specialise in genetics within their PCTs and that has largely
arisen out of the work that the National Genetics Education Centre
did with the GPs with the specialist interest. The whole point
of education is to be competent to be able to deal with the questions
that you have framed.
Professor O'Rahilly: When you mentioned the
patient it struck me that there was an anomaly there with the
fact that patients with cancer, for example, who go "off-piste"
and buy drugs essentially are banned from having continuing NHS
care. It is a disgrace but clearly we are treating therapeutics
in the NHS rather different from diagnostics; in other words,
you can go out into the private sector and get diagnostics and
generate cost and income for the NHS. I agree that it is an anomaly
and we certainly do not treat the two the same.
Dr Rafi: The likelihood is that a lot of these
print-outs are going to generate low relative risks for these
diseases and there are going to be one or two that are going to
be particularly high risk, and it is being able to identify those
and being able to help the patient through the NHS system if they
need that help.
Professor Furness: Could I make a lateral link
at that point because part of the problem with direct-to-the-consumer
testing is that the consumer has no idea what he or she is buying.
If we had an established system whereby all the tests that the
NHS regards as being clinically justified, and maybe a few that
it does not, are on a database where information about them is
available to general practitioners and to the public, then you
would have a resource first of all for the GPs to answer those
questions relatively straightforwardly: "Here's what the
official answer is about that test." You would have also
have a resource for members of the public to look up and decide
whether or not this is worth doing or not. That is even without
going down the route of legislation to control what may or may
not be offered to the public, which I think in some circumstances
there is quite a strong case for, but even without that if you
provided the information, if the information is there and accessible
you have removed a lot of the problem.
Q205 Lord Colwyn:
I think in the States taking a test included an interpretation
of the test. They have got it down out there to a $1,000 a test
now but the quality of the interpretation varied dramatically
between the different laboratories.
Professor Furness: And the commercial pressure
to avoid saying, "Here's something we have picked up but
it does not make any difference."
Q206 Baroness Perry of Southwark:
I want to turn to the question of commissioning but just before
I do that could I press Dr Crolla, the Joint Committee says in
their evidence that more genetic counsellors are needed. Are people
not coming forward in sufficient numbers to do the training or
is there not enough training being provided?
Dr Crolla: I think for genetic counsellors it
has been a great success story, and certainly under the White
Paper initiative there were a numberI cannot remember the
exact numberof new genetic counsellors that were recruited
and trained, but I think the demand for genetic counsellors is
growing at the rate of the number of tests and scenarios which
require interpretation of diagnostic tests, so I think that was
part of the submission. Yes, it has been a great success but we
need to expand that.
Q207 Baroness Perry of Southwark:
We need to expand the provision; people are coming forward willing
to be trained? That is my question.
Dr Crolla: Absolutely, yes.
Q208 Baroness Perry of Southwark:
Turning to the question of commissioning, the Joint Committee
on Medical Genetics say in their submission that commissioning
is not structured to react to the rapid change that is happening
and as a result the introduction of genetic testing is patchy
and inconsistent across the country. Given that commissioners
have very limited understanding of the details about genetic testing,
does the existing policy framework for example enshrined in the
NHS Genetics Team, need strengthening and if so how?
Dr Crolla: I think the take-home message here,
as Peter has alluded to, is that the UKGTN as currently structured
is very focused on dealing with disorders with DNA as a primary
analate and developing gene dossiers which have the correct level
of clinical utility and interpretation built into them, so that
is a recommendation that then goes to GENCAG for acceptance, and
specialist commissioners sit on GENCAG (I sit on GENCAG) and that
mechanism works very well. What does not work is that there is
then very patchy uptake by PCTs etc of the recommendations of
GENCAG for the uptake of tests which have been approved by the
UKGTN Gene Dossier Committee. There are all sorts of different
reasons for this. It is very low on the list of priorities of
Health Service managers; the PCTs are too low a level for the
commissioning of this service; there is not the knowledge base
there to really understand it; and real health need is defined
as the "ability to benefit", which in turn is a function
of the prevalence and the effectiveness of interventions. These
are complex issues which cascade down to PCTs who are under enormous
pressure for a whole range of other interventions. I think the
recommendation from the Joint Committee would be that this specialist
commissioning should go back to a national level so that when
agreed nationally there should be provision for the rolling out
of these tests. Again, I will stress I am really talking about
rare Mendelian disorders. Also where the UKGTN needs to be strengthened
is that it needs to expand its gene dossier remit and it is beginning
to do that by taking on board cytogenetics and looking at ways
in which it could develop gene dossier-type recommendations to
commissioners for the commissioning of new technologies like comparative
array CGH, both dosage and genomics.
Q209 Baroness Perry of Southwark:
And you would want that ring-fenced funding to remain national?
Dr Crolla: That would be the ideal because what
is happening now when it cascades to the PCTs is it is very patchy
so it is a postcode lottery. I think it needs to be ring-fenced
Professor O'Rahilly: Let me just give an example
by Jenny Taylor who has drafted the document that the Academy
will send you in response to these questions. She was involved
in Oxford in the development of a service whereby people who died
young and suddenly of sudden cardiac death, of which there are
a number of genetic causes, would have their post mortem DNA analysed
and family members would be screened and then those individuals
who carried the risk factors were given implantable defibrillators,
et cetera, to prevent sudden cardiac death. That was accepted
pretty much everywhere in the UK apart from the Oxford region
and it could not be implemented there because of financial pressures
on the PCT, so there you had an example of the very place that
was developing and leading internationally in the area of development
was unable to find funding. There are numerous such anomalies
within the Health Service but that is a good example of how patchy
commissioning can lead to strange anomalies.
Professor Furness: I think in terms of Health
Service commissioning it is important to recognise that what we
are talking about here at the moment is small numbers, which makes
the model of the National Commissioning Group for Specialist Services
an appropriate one. But if, as we anticipate, this sort of methodology
and approach becomes relevant to more and more and more people,
then it will rapidly move out of the realm in which that group
has been accustomed to operating. It will not necessarily get
that much simpler for primary care trusts to understand, even
if a lot of people within one PCT might justify participating
in a particular investigation. The issues will still be very complex
and unfamiliar and the number of questions that are being asked
will become very large, so there really is a conflict here with
the policy of local decision-making in the Government, because
it will inevitably drive postcode diagnostics.
Q210 Earl of Northesk:
At the risk of stating the obvious, genomic data makes huge demands
on information technology and it is acknowledged that existing
IT systems, not least with respect to band width, the speed of
transfer and expertise, are likely to be insufficient for future
needs. In your view, where is the investment needed and where
should the balance be struck between investment in IT in reference
labs, secondary care and primary care? Who is best placed to develop
and establish the increased infrastructure and expertise?
Professor Furness: Could I suggest an answer
from the specialist labs and from general practice.
Dr Crolla: I have had some inter-activity with
Connecting for Health when we were first establishing array CGH
in our reference laboratory. The problem was that through the
NHS N3, the band width out to the Internet was 250 kb/sec speed
width for the whole of the NHS, for all 1.2 million users. I am
not sure it was anything to do with the pressure that we put on
but they were in negotiation with BT to increase that band width
to one megabit per second, which it currently is as I understand.
It urgently needs upgrading to much faster, 10 or 20, or as the
French are now installing in Paris 100 megabits per second as
a standard broadband band width. That is one point. The second
point is that in terms of data storage and data analysis there
is a real training issue as well where clinical scientists are
beginning to wake up to the real bioinformatic resources that
are required in order to do interpretation, be it genome-wide
association studies, sequence analysis or array analysis. I am
not sure that I would be able to give the Committee a particular
steer on where the resources need to be put but where we are finding,
and I find other colleagues are finding it particularly frustrating
is that we constantly are battling with our local trust IT departments
who work under very strict clinical governance rules, appropriately,
and also they have huge political drivers like Choose and Book
and the implementation of PIMS and other initiatives, and so our
requirements are relatively small-scale. I do not think that this
technology infrastructure improvement should only be in the reference
laboratories. I think it should be in all laboratories which are
accessing genomic information. I think the final point I would
like to make is that we have in the UK at the Sanger Centre one
of the leading genome campuses in the world, and within the Sanger
Centre there is the Ensembl Genome Browser with all the bioinformatic
resources, and they have a suite of servers there which is the
size literally of an aircraft hanger, and they have bioinformaticians
who I think would be approachable and who should really be collaborating
with NHS Connecting for Health. There are other initiatives that
are on-going where genetics laboratories are currently developing
new laboratory information management systems on a consortium
basis, again highlighting the rather clunky nature of trust IT
systems against what we need, which is fast, unrestricted access
to the Internet and genome analysis.
Q211 Lord Winston:
I am surprised you mention the band width and do not mention security.
Is not security rather more important?
Dr Crolla: The conflict is security and so clearly
we have to conform to all the acceptable use of IT policies, but
there has got to be a way of having parallel access which does
not impinge on or affect security. Clearly if you are going to
be sending large data sets across the Internet it has to be anonymised
before you can do that, but what I am talking about here is not
transferring data but access to bioinformatic tools, and the current
structure is really too clunky to do that in an effective way.
I do not think it has any issues in terms of patient confidentiality.
We would not want to in any way impinge or breach acceptable use
policies under the NHS clinical governance.
Do you get the feeling that the current IT development in the
NHS as it progresses is going to be adequate for genomics information?
Professor Furness: I think if you asked that
question omitting the words "for genomics information"
you would probably still get a no; it will not be adequate.
Professor O'Rahilly: I would be pleased if it
could get me an Electronic Patient Record in the next 10 years!
It will have to be bespoke developments, as Peter and John say,
because the nature of the questions that you are asking with information
is totally different. You need detailed bioinformatic information
about the likely effect, for example, of a particular mutation
on a protein. You need to be able to assess whether something
is likely to be disease-causing or not. There is a whole range
of information that is needed that is different from what we usually
think of regarding a simple transfer of patient records or simple
numbers with a normal range. This information is very dense and
deep and I think there will need to be bespoke and focused investment,
particularly on bioinformatics within the NHS, if we are going
to take genetics to the next level. I think that has to be not
just the regional labs but the specialist labs too.
Dr Rafi: In terms of primary care, generally
IT within primary care has been pretty good and probably in some
places a lot better than secondary care. There are certain needs,
though, that I think we are going to need for the future including
how we record family histories and pedigrees. There are no such
systems up and running at the moment within primary care, although
there are people looking at computer-aided packages that can give
family pedigree information and also risk assessments. That is
one thing that I think would be useful. The way that we record
data in primary care, particularly applicable to genetics, is
going to be important and as we get new classifications of diseases,
disease coding is going to be particularly important. I asked
Simon de Lusignan who is Director of Bioinformatics at St George's
about this. He is a GP with a real expertise in bioinformatics,
and his feeling was that what we need is a informatician in primary
care who can deal with psuedo-anonymised databases where you look
at the coding that goes into these databases, which may change
over to time. So trying to look at archived data from the past
may be difficult in the future when your queries change, when
your coding systems change. I think this is really important.
The other aspect of care is obviously surveillance, and with people
for example an increased risk of colorectal cancer or breast cancer
based on their family history, who need to have regular surveillance
over time, we need to have good methods in primary care to be
able to ensure that these people get the relevant procedure that
they need at the time, and so we need good registers in primary
care to be able to deal with that situation.
Q213 Earl of Northesk:
Back to your observation, Dr Crolla, that the existing NHS IT
systems are clunky. This may be a little provocative but it does
provoke the thought to what extent therefore are existing NHS
IT systems a potential barrier to active transfer of genomic data?
Dr Crolla: I used the word clunkyit is
a relative termand obviously people who are accessing the
data warehouse for PIMS applications would find the download speeds
perfectly acceptable. Where colleagues are trying to download
gigabytes of sequence information in order to blast sequences,
that is a very different scenario, and that is where it is clunky,
it is too slow. What often happens is that colleagues will go
home and use their own broadband connectivity and do the work
and then email it back to work, which is an unacceptable situation.
There are ways around it, and Lord Winston highlighted the security
issues, and you can have parallel broadband access within hospitals
providing it is within the acceptable use provisions but which
does not impinge on the security issues for patient confidentiality.
Where I think the parallel should be drawn is between what is
available within university IT systems compared to the NHS. I
wear two hats, I have a university position and an NHS position,
and because of our physical separation from the university I cannot
have access to the library resources in the university via the
NHS. The systems do not talk so I cannot instantly access data
or on-line journals and things like that on my desktop when I
need to see them, so there is a delay. Really if we are going
to develop and utilise these resources we need to have those resources
available at the desktop through the NHS IT system.
Q214 Earl of Northesk:
So another part of your wish-list is a much more federated IT
Dr Crolla: Yes.
Professor O'Rahilly: The problem is not unique
to separate sites. Even on one site at Addenbrooke's in Cambridge
our NHS colleagues cannot access journals, et cetera, through
their IT system, so it is not just a physical separation, it happens
in most institutions even when the university and the hospital
are in the same place.
But have we not had a recent directive about information about
individuals and how it can be taken out and there are strict rules
now that have to be applied? If you are familiar with that is
that not going to make difficult what you suggest where people
are taking information away from the NHS environment so that they
can work on it?
Dr Crolla: That is a very good point, Lord Chairman,
yes, indeed. Again, this would be accessing pseudo-anonymised
data or genomic data which is not personalised, so that was the
example I was giving. There are initiatives to encrypt all NHS
computers which I think will make life impossible.
That is the direction given just now?
Dr Crolla: Yes and our local trust is looking
at ways of implementing this encryption, but most of what we are
doing is in international collaboration and so we do need to have
access and be able to share anonymised data across international
boundaries. If it goes outside the NHS it is gobbledegook unless
they have the encryption software. It is a big problem.
Including your Blackberry emails?
Dr Crolla: Yes, especially your Blackberry emails!
Q218 Baroness O'Neill of Bengarve:
In the last section of the Royal College of Pathologists' report
you suggest a new policy framework and in particular the establishment
of a new UK-wide body, or the reorganisation of one of the existing
bodies to provide a single comprehensive function of horizon scanning
and for taking and evaluating new laboratory investigations. I
have several questions about that proposal of which the sharpest
is: which bodies would you abolish? There is a constant tendency
to recommend multiplication of bodies and then to complain that
there are overlaps and gaps and things that are difficult to deal
with. If you were being really quite surgical about this, what
would you abolish? Could, for example, the UK Genetic Testing
Network be enough or what else would you do? This body has got
to do a number of different things but just adding a body does
not strike me as perhaps the whole or best remedy.
Professor Furness: Those are my words so if
I could at least start the answer to that one. This is a new task
which could be taken as a justification for a new body. I have
listed the various bodies that do something similar on a small
selective scale. Obviously one would not abolish the whole of
NICE but that is an organisation that could potentially be reformed
and expanded to cover this function if it was able to alter its
working practices radically in this field. At the moment I gather
it has a charter which means that it could not do that. I do not
know enough about the other bodies to say whether they could actually
be abolished or not, but I suggest that their role in evaluating
diagnostic investigations could be abolished. The trouble is they
all have other roles as well, so one could readily extract those
roles from those various bodies and thereby reduce them a little
in size and either expand one of them to cover what is necessary
or set up another body. I would not wish to prejudge which would
necessarily be the best way of doing that as long as there is
an organisation that can speak with a single authoritative voice
on all these topics.
Baroness O'Neill of Bengarve: Thank you
Chairman: Does anybody else have any
comments about that? Lord Colwyn?
Q219 Lord Colwyn:
At our Genomic Seminar, which I know you came to and I am not
sure if any of the others came as well, it was suggested in evidence
from the Academy of Medical Sciences that in future pathology
services might be reorganised with a molecular pathology lab acting
as the hub of molecular tests in clinical genetics, cytology,
haematology and pathology. What is your view on the benefits of
organising pathology services in this way? What do you think the
costs might be and the desired timescale?
Professor Furness: That was an Academy point
and could I answer it, please, because it has been discussed a
good deal within the Royal College of Pathologists, and indeed
it is a topic that arises from Lord Carter's work on the provision
of pathology services. There are aspects of providing molecular
biology systems that are very expensive and nowadays rely on very
large expensive machines where you only have look at the economics
and it is absolutely obvious that it is more efficiently done
with a small number of those machines analysing samples from all
over the country, if that is the sort of thing that you are talking
about. However, on the other hand, the people who actually interact
with patients who need to decide what investigations are needed
for individual patients and to help interpret those tests for
the benefit of individual patients have to be where the patients
are. To that extent you are potentially talking about a hub and
spoke arrangement to make it most efficient. How many hubs you
have around the country is a difficult question and will probably
depend on the tests that you are talking about. If it is an investigation
that needs some massive high throughput machine and the country
really only needs one, then you are talking about one hub for
that investigation; for others you will more efficiently have
lots of hubs with spokes. I think the idea of the different specialities
of pathology collaborating to have one hub almost per hospital
to cover microbiology, clinical chemistry, histopathologythe
existing specialtiesis one which is being discussed repeatedly
and is of value in terms of generating local hubs to make things
more efficient. The barrier to that is, first of all, the need
for capital investment to do it and, secondly, the current structure
of NHS funding. We have "silo" funding where this amount
of money goes to this service to keep doing what it has been doing
year in, year out, irrespective largely of new demands and new
developments, and it is very difficult to get agreement to change
that pattern. The expense of that sort of reorganisation would
be there, I personally suspect it would not be enormous and I
think the savings could be greater than the expense if it is done
logically; but we have this hump, this barrier of organisational
inertia to get over to make it happen. I am very keen to see that
sort of development happening but it is the politics that get
in the way at the moment, I think. John, you have had particular
discussions within the UK Genetic Testing Network.
Dr Crolla: We have not actually discussed this
specific issue within the Joint Committee but when I saw the question
I canvassed the opinion of the Joint Committee on this. My personal
view was "if it ain't bust don't fix it" in the sense
that regional genetics laboratories and collaborations with clinical
geneticists have really served the population pretty well since
the inception of the regional genetics system, but most of my
colleagues responded by saying, "Hang on, there is a critical
mass issue here," and that really as the technologies are
evolvinghigh throughput sequencing, various new generation
sequencing technologies, the arraysas Peter was saying,
the capital investment is so large and the capacity of that equipment
is also very large, it really does make sense to centralise some
of these testing scenarios. What we must not lose is this link
between the interpretation and the clinical use of the data that
is generated by these hub and spoke models, but I totally agree,
as did most of the respondents from the Joint Committee, that
we cannot be ostriches and stick our heads in the sand and say
things will remain the same because the technology will drive
Professor Furness: It has training implications
as well of course. If you are splitting the doing of the tests
from the interpreting of the tests, how you train the people and
how you make the jobs attractive could be difficult, but I think
those are issues we will have to address.