Tip:
Highlight text to annotate it
X
Laura Rodriguez: Okay. So, being the policy breakout group
-- is that better? Okay. Being the policy breakout group, I'm a little bit hesitant
to get up here because, of course, we don't have a lot of concrete things, whereas some
of the things that have come out of the discussions are very concrete. And I'm always jealous
of that when we talk about policy, because there's a lot of really important discussions,
but how we get at them sometimes can be a bit challenging.
So, like Howard, I'd like to thank the group. We had actually nine different countries represented,
and so we had a lot of experience, both in terms of what the individuals had done, but
also the cultures within which they are working, which is, of course, also very important to
the policy discussions. And also, we had about a third -- pardon? Not yet. [laughs] We had
about a third of our members that were responsible for healthcare delivery systems and for thinking
about how to integrate research into health. And they weren't all from the genomics field.
And so I think that also gave us a broad perspective.
And I wanted to be able to say that before I put up the first slide, which we can now
show, because one of our early discussions was to come out and make a clear statement
with very broad agreement that genomics was not unique in some of the questions that we
were thinking about and in terms of the policy solutions that we would need, and that it
would be very helpful if we were to think about it and think about the kinds of solutions
or pathways to solutions that we need to get to by thinking about genomics as another tool
for health, and that it's one element that needs to be considered in moving towards personalized
medicine, that we can look to past examples of how other new technologies have moved into
health care and learn from them, and think about them as we're trying to put together
our own next steps.
Also, I think one of the strong recommendations from the group is that we needed to think
about what genomic medicine has to offer with regard to the comparator of what our current
care pathways are, and where do we see the differences to make -- to enhance the current
care and move from that standpoint in what we're doing.
So, in terms of the priorities that we did talk about, we were -- talked about in the
beginning all of the various issues that the many speakers raised through their talks and
the challenges. But I think this list, while it might be a bit simplified and certainly
doesn't cover everything that came up or that -- everything that we discussed, I think these
were the key things that we acknowledged as being major issues. Several of them are addressed
or have components that are addressed from other working groups, and so we didn't spend
time on them. But, of course, engaging that stakeholders, the issues around funding, who's
going to support the work, the research that's needed, as well as the technology assessments,
et cetera, to move the work forward, who are the right decision-makers for the health care
in different countries. It varies. It's different in every case. That, of course, complicates
trying to think about moving things forward in one nation, let alone as we're trying to
look internationally.
And then, of course, also the patients are the stakeholders, and that as we can engage
patients -- and this has already come up in the breakout reports -- then they will also
push for this -- the benefits from genomic medicine as part of personalized care, personalized
medicine to come forward.
Data sharing, obviously it never fails to come up as a key issue. Privacy, informed
consent, the morass of legal issues, again, within one country, and particularly then
when we start talking internationally, about how to harmonize or find common principles
around the privacy and informed consent so that we can promote better data aggregation
to enable the research and the science and to promote sharing, as well, from the ethics
standpoint.
The regulatory oversight that came up this morning, in particular around FDA and the
situation that we're facing, but also in others, and then the cost benefits -- the costs and
the benefits of adding genomics to the care systems, again thinking about it from a systematic
-- or a systems perspective.
I think among these, the engaging the stakeholders issue, we didn't really talk about in terms
of opportunities and next steps. I think some of that has been addressed by Bruce's group
and others that just didn't really make it to our strongest priority. Where we actually
spent most of the time and pretty quickly got to a focus on was this issue of how do
we determine the costs and the benefits of adding genomics to our care systems. And so
that's where most of our recommendations or detail of the opportunities will focus.
So, first with the data sharing and the regulatory issues, the primary point that our group really
wanted to bring forward in terms of what are -- what should our next steps be, is that
there are many different groups and different alliances that are out there working on this
in an international perspective. We've heard a lot about IRDiRC. In rare diseases, there
is a global alliance, there's a Genomic Medicine Alliance that was talked about this morning.
There are just a lot of people out there already working on these. And also some of the issues
around privacy and informed consent that enable data sharing for the research to go forward
are also focused on the research end of the spectrum and not the implementation end in
terms of integrating genomics into care. And so this might not be the best place for our
group to form -- to do a lot of additional work.
But where it could be very useful in the future for people from this group to think about
is trying to map what the different activities are from all of these different groups and
alliances, and what are the issues that they're working on so that we can both know who is
working on what and how they're approaching it so that we can all learn from it, but also
to do the gap analysis of -- and see is there a unique place that we can get to where we
can make a contribution that is not already being addressed by some other group with perhaps
better expertise than we might be able to bring to it.
And so there were also talked about whether or not we might encourage a network of networks
to form, again, around those -- you know, all of the different groups working on informed
consent, or all of the different groups looking at particularly engaging different sets of
stakeholders. And we put this up here, but acknowledging that network of networks is
a great policy phrase that sounds good, but is really questionable as to what it might
actually mean. So, but we did think that it could be a way to share information and to
try and articulate more clearly what the responsibilities are and who has ownership of trying to push
different issues forward because we thought that would be helpful, again, just so that
we don't all try to duplicate in the same problem area.
So moving back over, then, to the costs and the benefits, and trying to think about that
in terms of what will genomics add to care in a delivery system. And we talked a lot
about trying to define what is it that we would need to have in order to go forward.
And so there is, of course, the technology assessment, the evidence, but there was a
working group that we've already heard from talking a lot about that, so we didn't dwell
on how do you determine what the evidence is. But, of course, in addition to the scientific
evidence, we need to have demonstrated clinical utility, and we need to be able to articulate
what the costs are. They could be small costs or they could be large costs, but those costs
need to be known so that they can be weighed and balanced to the clinical advantages, et
cetera.
So, again, this is a place where one thing that could be done would be to look at and
do some analyses on different successes in the past, those that had a great deal of evidence
to support them, such as PET scans was an example that we talked about, as well as those
new techniques or methods that are put into care very rapidly, sometimes without a lot
of evidence. And what is it about those that have them move in. And then, of course, the
difficulties of getting them out of care when there is evidence to say that they're not
helpful.
Another potential opportunity where a group could come together to look at this, and this
was something that we thought we needed to take a look at the literature and see what
might already have been done in this area, but would be taking the perspective of the
healthcare delivery system, trying to walk through that in a pipeline fashion and think
about where would genomics make a difference. Again, to help us get to the point of being
able to define where the economic costs would be and where the clinical advantages would
be. We thought about this along the lines of different disease models that could be
done. There's also talk about really, if we wanted to make a difference for health care
in a large scale, we'd be looking at what is genomics going to do for chronic diseases.
And they're seeing that maybe because that science may not quite be there yet, there
hasn't been as much work in looking at or having models or pilots to think about how
genomics might be able to improve care for hypertension, or diabetes, or mental health,
et cetera. And but -- that what we really needed to get to was where is genomics going
to make the biggest impact on care, and that's how we're going to engage the decision-makers
and provide them with information and evidence to get them to be willing to try and integrate
this into care.
There we go. So, I think this is about my last slide. So, again, thinking about what
will we need, the economic issues, and the need for more economic analyses, better economic
analyses, inputs to have into the system for helping us to determine. Evidence came up.
We talked a lot about who else is already doing this, again, thinking about the fact
that genomics is just one tool. This has been done in other places, and we should look to
places like the health technology assessment agencies in various countries, pharmacoeconomic
societies, that do this all the time to see what we could learn about how would we structure
some of these analyses for genomic methodologies in particular. They talked about the fact
that in Canada, as one example, they're requiring, in some cases, a health economist now be part
of the research team so that those questions are being brought to the table at the time
that the research is being designed and all along the way. And then also needing to recognize
that it's not just about technology assessment in terms of making the decision, but that
that -- the technology assessment then becomes an input to those who are making the payment
decisions around care for insurance or other health systems.
And, of course, again, as we've heard over and over again throughout the meeting and
in these breakout reports, needing to engage those who are making the payment decisions.
One recommendation for something that might be worth trying to pursue would be to work
in a system other than the U.S. We didn't come up with the ideal system, but where there
is one or maybe just a few centralized payers, so that, again, we could have conversations
with them about what is it that they needed to see where we have hope of trying to get
something that's definable, again, unlike in the U.S. where we get lots of different
answers, so that we could try and look at a particular case and carry that through for
a pilot or demonstration, and use that then to move on to other examples.
So that, I think, was -- in this area was our top recommendation for what the opportunity
would be in terms of what we might do as a next step. And then while this -- the other
issue about trying to integrate economics and economists into our research teams is
something that I think there was also a lot of agreement with, and that in terms of who
can act on that would be funders, but I think anyone, really, in how they're designing the
research projects and putting together their research teams, could look at it.
And I'll stop and take any questions, and also invite the group, please. The synthesis
of this was all very different from the actual flow of our conversation, so if I misaligned
anything, please say it.
Geoffrey Ginsburg: This was terrific. I guess I would just, following
up on your last point and the discussion we had earlier about evidence generation, it
would seem that if there's already a project -- pilot project going on in a environment
that lends itself, like a single payer environment that might lend itself to economic analysis
and it's not happening, that would be something to really try to embrace, that -- a way to
find the resources to make sure that that analysis is done. Or secondly, to consider
an existing pilot project from any of our global communities and making sure that it
-- or selecting an environment where that could be done, and what the appropriate economic
analyses as a demonstration project, as you say.
And the other, it seems that the -- this whole area should be of significant interest to
industry because it's -- their ability to launch a product, get it on the market, get
it reimbursed and adopted, is going to be predicated on having the economics work in
their favor. So, it would be an opportunity, I think, to partner not just with some of
the agencies you mentioned, but with industry as well.
Female Speaker: That's a really astute approach, I think.
And this community is in a place such a project or several projects would be easiest to do
if there wasn't a lot of population variation in the disease, for want of a better word.
Forgive me, I'm not a geneticist. And also, if there were a straightforward genetic test
that could be applied to identify the target population because that makes the Markov model,
or whatever economic tool you're going to use to do the economic analysis, more straightforward
to use. So, choosing the project wisely would be of high value. And this group's probably
an ideal group to make that decision from a research point of view.
Marc Williams: Yeah, I would agree that a focus on economics
is important. The only point of contention that I might make is that there are actually
different ways to look at this that can take into account variability, which is to model
so that you look at a threshold of effectiveness models so that you can manage to say, you
know, if the prevalence of this particular genomic variant is this level -- if it's above
this level in the population, it is cost-effective to do this. If it's below this level, it is
not cost-effective. And so in terms of how you define the problem and also define the
model, you can come up with very useful things that can be applied to an individual country
or group once they know what their -- once the prevalence of that is in their particular
population.
Female Speaker: Those things are also related to sensitivity
and specificity, and the cost that you would purchase test of.
Marc Williams: Right.
Female Speaker: There also is a threshold that you can give
for an individual [spelled phonetically].
Marc Williams: Yeah, so the point that was being made is
that there's lots of variables, and the difference between an economist and a statistician is
that statisticians require data.
Female Speaker: Correct.
[laughter]
Marc Williams: Okay. Right. So, all of these are assumption-based,
and so doing sensitivity analysis on the models -- and that allows identification of what
are the most important pieces of data to really understand because they most profoundly affect
the performance of the model. That is also extremely important information because that
can drive the research agenda to say, "You know, we don't know the answer to this, but
it doesn't matter because if you veer it between 1 and 100, the model performs exactly the
same. But if this -- if we're off by a factor of two in this one, the model completely works
differently. So...
Female Speaker: And that's why the question of what evidence
you accumulate and evaluate with the economist at the end of the R&D phase is helpful.
Marc Williams: Right. Right. So just the point she was making
was that defines the prioritization of the evidence collection.
Geoffrey Ginsburg: Thank you, Laura.
[applause]