Tip:
Highlight text to annotate it
X
Male Speaker: So actually we'll segue now with Lisa Parker
playing the role of Geoff Ginsberg.
[laughter]
Lisa Parker: All right, thank you. So, since my two fellow
panelists both mapped out many of the issues, and actually Wendy had given us such a good
introduction to the terrain of issues raised by return of results, I will focus on the
summary just to highlight a couple of particular opportunities that I think eMERGE presents
in this regard.
Our morning discussion certainly highlighted that eMERGE is engaged in and poised to continue
the research on the clinical implementation, this middle ground before we actually get
to clinical implementation broadly and fully. That eMERGE has the opportunity to ask the
questions about what we should be doing with regard to all of the normative issues and
preferences and educational needs that were recently discussed. I think that with the
opportunity to continue doing site specific as well as network worldwide projects, we
have an opportunity for innovative pilot studies around return of research results as well
as comparative studies, indeed capitalizing on the non-uniform health care system that
we have and differences across institutions and their electronic health record systems;
and indeed the fact that some of our eMERGE sites, the health care system is also involved
in insuring lives, and so that we do have an opportunity for some of the economic and
broader policy examination that's been raised. That we can do this across time, in two senses,
both the longitudinal study, querying deep diverse phenotypes, and also across the life
span that we do have, you know, pediatric patients, we have much older patients. And
that I would suggest that at least some say, despite the ACMG recommendations, that the
age of the person in whom the result is identified may indeed matter, whether this is an elderly
person for whom that disease risk may not be likely to be manifesting, or you're talking
about a newborn or a pediatric patient.
And so I think, very briefly in the summary so we do have time for discussion, both with
Gail and Susan and then everyone, I just want to emphasize that what I have found eMERGE
to have most particularly in contrast to any other research setting or research network
is that there's an integrated infrastructure for conducting the empirical ELSI research,
and that return of research results is a key topic for this integration. There are certainly
practical logistical issues, but that those are so intimately intertwined with the ethical,
legal, the economic, the social aspects, the social risks, if you will, of returning results,
and that eMERGE has proved, through two phases thus far, that really more than elsewhere,
and perhaps more than the ELSI CEERs to some extent, there's a real integration of empirical
ethics researchers, policy researchers, into the science. And therefore, there's a real
opportunity here, I think, to examine both what I'll call the individual participant
level issues as well as now system-wide issues of economics, the impact on health care costs,
quality insurance issues, data security, and participant privacy issues at a systemic level
as well as looking at the clinician preparedness and needs of the decision support and participant
responders, including as we go forward and consider the implications for families whom
may be outside of the given health care system or beyond the electronic health record, implications
for them, liability issues, and just sheer family member responses.
So, I think eMERGE is well-positioned to continue to try to address the issues raised by return
of results. And I will stop.
Male Speaker: Would you like to complete the polling of
the --
Lisa Parker: Oh, I will, yes, absolutely. I had intended,
in fact, to turn on -- now, Gail, if you wanted to weigh in on this discussion, and then we'll
have Susan, and then we'll put it up.
Gail Jarvik: Yes, so, I think, you know, one of the things
we talked a lot -- we had several calls, and our panelists did a great job of, you know,
being prepared and, you know, [unintelligible] prepared. We talked a lot about how, you know,
what kind of balance of discovery and implementation works that where you're getting results that
you can return to people that have value to them. And -- but also, if there are good ELSI
questions, there are good discovery questions, and that's actually sort of how we kind of
honed in on this, you know, more penetrant gene idea. And I'm very much in favor of that,
and I like pharmacogenetics. There are a lot of people working on that space. I mean, we
have some good pharmacogenetic data, but I think, in part, returning things to actual
people and learning how to do this, that the highly-penetrant stuff is more, you know,
critically demanded, and that there is a really lot we could learn, not just about which variants
are pathogenic, but about which variants aren't pathogenic, and issues like penetrance and
other things; there's a lot of ELSI questions in that space. So if we kind of went over,
well, what do we want to learn, what are our unique strengths, I think that idea of more
highly-penetrant conditions jumps out more and more. That's all I have to say.
Lisa Parker: Susan, did you want to chime in?
Female Speaker: Sure. I want to go back to the question of
what unique opportunities eMERGE may have in the return of results phase. And I've got
a couple thoughts on that. First of all, I think eMERGE probably needs to, as well as
can, cast and go beyond the ACMG statement on return of results, return of incidental
findings, actually, in clinical sequencing. Because, of course, that was about clinical
sequencing, and [unintelligible] is agnostic on its implications, at least in the text,
for research. But eMERGE is already facing these issues, as we know, in the research
sphere.
So -- and ACMG, in that statement, was modest, appropriately, in saying this is a first pass,
we need a better evidence base. So, I think eMERGE can really play a leadership role there,
including on the specific question of the role of participant and patient preference
on what to look for and what to return. Those -- as everyone knows, they're contested questions.
Another domain I think where eMERGE has unique capability is digging in on some of the consent
questions, and Iftikhar had on his slides some of the cutting-edge ones, like consent
to include genomic data in the EHR. Others are consent to share with kin, consent for
posthumous use, so there are really some emerging important issues.
Third, I think I'd like to put a proposal out there to be innovative and take a hard
look at how we're defining actionability. Because I'm very interested in Heidi's slide
about developing a scoring scale on actionability, but overwhelmingly, the dialogue in actionability
has been driven by investigator and clinician views of actionability, and has largely ignored,
I think, participant and patient views on what do they see as actionable in their lives.
What about reproductive views? I understand that would open a Pandora's box, but what
are their priorities? What about life significance even if the clinicians can't fix the disease
or the risk factor? So, I think actionability needs a lot of work.
And the last thing I would throw in is the description of the rich diversity in the eMERGE
population, including diversity by ancestry, opened a really important opportunity. So
much work on return of results and ELSI work more broadly suffers from a lack of diverse
populations in which to ask the question, are there differences in view on what people
want to have happen? I think that's a really important domain that is sounds like eMERGE
can uniquely contribute to.
Lisa Parker: Thank you.
Gail Jarvik: I want to clarify -- this is Gail -- that
I am not endorsing any of the opinions of the ACMG statement with regard to mandatory
return or return for minors when I say that the list, the actual list of genes, is a good
starting place for us to look for what is -- you know, agreeable to many people is actionable.
I think there are more actionable genes; there are a couple that I am dicey about on the
list, but [unintelligible] thrown off.
[laughter]
But the list of genes, I think people have generally a positive review of what to do
with that, those variants. There actually is wonderful research space from eMERGE. What
do people actually want? Not just what does the ACMG say should happen, but what do people,
what do patients want, what do providers want? And I think that's --
Marc Williams: This is Marc. I think this is a really high
profile issue. As one of the authors on the paper, first of all I'll say, Gail, you only
get one gene off, all right, that you're --
Female Speaker: I don't need the other two, I'm okay with
them.
Marc Williams: Okay, good. Glad to hear it. But I think,
you know, one of the things we faced in creating the statement on the list, is that as everyone
here knows, there's been almost no research in this area at all. And so while we gave
it our best effort based on what we, you know, thought was a reasonable way to proceed, I
think the whole process would dramatically welcome, you know, actual formative research
about, you know, how do people feel about this, what's really the best way to do it,
and develop a research agenda around these types of questions. Now, I don't know if Heidi's
still on the call or not, but I know she's leading the ongoing process that's being developed
relating at least to the list, and so you may have some insights in terms of your perspective
on that type of research.
Lisa Parker: Heidi, are you --
Female Speaker: I think Heidi got off.
Lisa Parker: Okay.
Female Speaker: I saw her step off. But I just want to acknowledge
that the paper was very clear that an evidence base needed to be developed.
Female Speaker: And I think, you know, the paper that Wylie
Burke led on opportunistic screening, and the need for real evidence around the fact
that these people are coming in and largely asymptomatic as to the 56, so, what do we
know about -- do we have enough evidence in that context? That's a really important set
of questions.
Female Speaker: That we can address at eMERGE, absolutely.
Male Speaker: So, as a non-specialist in this area, but
having the benefit that when you're a grandparent, you get to see digital natives, the millennials,
and I think, well, what would they think of this entire discussion? And they would think,
this was laced with nothing but degrees of paternalism. There was no option at one end
that sounded like no paternalism, which is [unintelligible]. Why are you withholding
them from me? So, I guess it seems to me, at least, some recognition that this space
has moved a lot in just the amount of time since eMERGE was created. The public expectation
of, well, you accept the risk of getting it, you know, just telling it like it is, but
I want everything about me. It doesn't seem to have been in this discussion of incidental
findings.
Male Speaker: I slid over that slide.
[laughter]
Lisa Parker: I actually think that has been in the debate.
I mean, I think that was part of the original impetus for even a decade ago raising the
idea that participants in research and patients might want their incidental findings or even
their research results. And, I mean, I see the trend moving towards providing choice
so that some people might indeed say, "I want a ton." I mean, in fact, I define action ability
much more broadly than you do. Some people might even say, actionable or not, "Give it
to me now." Whether you understand it or not, "Give it to me now because it's mine." And
some people might be much more cautious. I don't think we know yet about what frequency
are attached to those attitudinal conditions. We need to know, but ultimately, I suspect,
we need to have a range of options that people can avail themselves of.
Male Speaker: And is it any more moral to ignore that view,
"Give me all my data now, it's my data," then it is to say, "No, I don't want that data,
don't return that for me," or, "I don't want that result." We all agree we need to do that.
But what about the person on the other side? And we just saw data that for the pharmacogenetics
update, the error rate is down there 1 percent or less. Well, you know, why is this really
something that we shouldn't return? Are we hiding behind these constructs that we created?
Research, you know, was designed because you didn't know the answer, and now we're getting
findings that are true, that's data, and should not -- even if we can't interpret the data,
can we give the data to the participants?
Female Speaker: I think we're back to the CLIA.
[laughter]
Male Speaker: Maybe, or not. I mean, we could write a research
protocol that, "And we will take the genetic information and give it to the participants."
Could we not?
Male Speaker: Sure. Actually, there is one like that, it's
called the multiplex [spelled phonetically] study. Yeah, with specific candidate genes,
but, yeah.
Male Speaker: But we could do a GWAS, or what is close to
GWAS.
Male Speaker: Yes. [unintelligible] implementation conundrum,
that perfect is the enemy of useful.
Female Speaker: I'm all for giving people their data, and
I'm probably at one of the few places that would let us give back non-CLIA data, but
my impression from my peers is that many places would not be excited by --
Male Speaker: Well, but yeah, so -- yes, this is to extend
that, right? So I think in the 21st century, that saying "Knowledge is power, and the people
that have the power want to keep it," right? I mean that's what the millennia would be.
Male Speaker: [unintelligible] liability.
Male Speaker: Liability is a fallacy, it's additional power.
Female Speaker: It also suggests that eMERGE is in a good
position to study the next step, and not just the next day or next month or even six months
out, response of individuals to having received their results, but in fact, over a certain
amount of more time, what they actually do with and what the effects on demand for health
care services is, how that links up to clinical practice guidelines for actually responding
to the level of, call it, you know, disease risk or preventative interventions or whatever,
that are recommended. Do people come in and ask for more because they're now more concerned?
Does it actually inspire them to follow a clinical practice guideline that's recommended?
And eMERGE is especially well-suited, structured, to be able to ask those questions and address
them.
Female Speaker: eMERGE is also really importantly positioned
to deal with the fact that more and more people have access directly to their electronic health
record or some version of it. So, if it's in the electronic health record, you would
think, sooner or later, it's going to be available to the individual. And what are the implications
of that? I mean, for example, ACMG talked about a dialogue ensuing between the clinician
who'd gotten the 56 back from the lab, or whatever came up in the analysis of 56, and
the patient. But if the patient has full access to the electronic health record, that's not
much of a filter, not much of a screen. It can be a source of education. So, I think
the trend, including among millennials, towards increasing access to the EHR has to be dealt
with, and eMERGE can really play a leadership role there.
Male Speaker: Yeah, I think the other point that's worthwhile
to note relating to that access issue, that is the, Justin Addal [spelled phonetically]
pointed out in our "Crossing the Omics Chasm" viewpoint article is it's not clear at this
point what all is going to be represented in the electronic health record. In other
words, would it just be the results that are deemed by someone to be actionable and we
don't give any of the other information? I think most of us believe it's not going to
be an entire variant file, but that's another type of research question that could be asked,
is what do our patients think would be a reasonable amount of information to have in there, and
how would they like to access it within portal or tethered personal health record environment.
There's a lot of real opportunity here.
Male Speaker: I think something to remember is that we are
having this discussion about return of results in part because of the behavior of generations
of previous researchers, some of whom basically said, "Here's this information, and if you
have questions, talk to your doctor." And, so, now we have clinical organizations that
are conditioned that when a researcher says, "I want to give a research subject information,"
their answer is, "No, because my doctors are going to suffer as a result of it." So, if
we are going to return the information, we have to scope it appropriately that we are
in a position to answer and analyze the questions that are asked.
Female Speaker: So, just to follow up on that, I did a study
years ago where we were measuring LPA, and we were required to ask every provider if
their patient could get their LPA back along with the rest of their lipid panel. And most
of the providers said no, their patient couldn't, because they didn't want it, they didn't want
to know about it, they didn't want to explain it to the patient, they didn't want to incorporate
it into care. And I think, to me, that emphasizes how important it is to pick something to return
that is important in clinical care that the provider will want to know about and will
need to learn about.
Male Speaker: That often happens, the patients are asking
the doctors for it, instead of the doctors responding to what the patients want, rather
than imposing on the patient, you know, what he thinks they need. It's just a version of
health care being mostly for doctors as opposed to for patients.
Female Speaker: Those were the days.
[laughter]
Male Speaker: Well spoken. We still have some more time,
if there --
Male Speaker: I find it curious, you say the data wouldn't
be deposited in the EMR. Well, it may not be deposited in EPIC because it can't be,
or in CPRS, but isn't it, in fact, part of the health care record regardless? If it's
a lab test that comes back, even if it's faxed in and it's not incorporated, it's part of
the record.
Male Speaker: Well, not necessarily, Larry. The example
that we used is the ancillary system, which is, if there's a system that's tied to the
electronic health record, but we don't have all of the compiled data for the images in
the electronic health record. We have a way that you can view the image through the electronic
health record, but the data does not reside in the EHR, and there are specific performance
issues where we think the huge volumes of genomic data, if stored within what we consider
the EHR environment, would degrade performance to the point that it would become unusable.
Male Speaker: Granted, that it's in a separate platform
and in a separate system the way that, for example, radiology is in many, if not most,
health care systems. But if the patient signs a consent that I want my MRI [unintelligible]
to my other doctor, it's part of the record insofar as it is transmitted to the other
system. Or if a patient now under MU-II [spelled phonetically] says, "I want that data, they
can get the data."
Male Speaker: I think a good example is if you look at lab
tests, okay, in many places when you order a serum sodium, they do a 20, because it's
cheaper to just run it through the multiplexer. And they throw the other 19 away and you can't
have them. Also, they run QA samples with lots of tests, and if you want to see a disaster,
see what happens when the QA results start bleeding into the clinical results and everybody
gets confused as heck. So, there are a lot of results that are collected by ancillary
systems which never come across the wire.
Male Speaker: Yes, but if you run a panel of 56 genes, and
those are returned to the clinician with the result file in a PDF, and you're in the position
where you're going to reinterpret DUSes [spelled phonetically], they better be in some ancillary
system that's considered legally part of the EMR.
Male Speaker: I think it's a terminological difference.
When we say EMR, we mean EPIC or its market partners as opposed to the entire clinical
digital record.
Mike Gaziano: Right, I think -- Mike Gaziano -- I think
there is a distinction, even if it's behind the firewall, between a research database
and a clinical database. If you're doing a randomized trial and it happens to be, you
know, the conservers have to be -- happen to be on your premises, those data are secured
data and not accessible to clinicians for routine care in any way. Certainly there can
be a dialogue between a participant in a trial who wants to get access to the data, but only
with the investigators that are involved in that trial. It's not as if he can go into
the medical records office, sign a form, and say, "I want my randomized trial -- secure
randomized trial data," and the medical records office will be required to give it to him.
They are really separate enclaves even though they may be contained behind the firewall
of a health system.
Male Speaker: Right. And I think our talking about only
CLIA results ordered for clinical care.
Mike Gaziano: Right, and that's a different, but now we're
starting to blur the line between CLIA results for research purposes that will eventually
find their way back to a medical record. Some which might find their way back, some which
might be retained within an enclave, a research enclave because they're not ready for primetime
consumption. So, the lines are beginning to blur, but not everything that's contained
behind a health system is all treated the same, and a patient doesn't have the exact
same access to all the data at all points in time.
Male Speaker: This is an excellent preamble to the next
topic, which will occur after a 15-minute break. Now having shown that we can actually
do a 15-minute break, you'll see 20 on your schedule, but we're actually going to restart
in 15 minutes, and we can almost pick this up in mid-sentence.
[laughter]
So, we'll see you at 15 minutes past the hour.