Event ID: 1492520
Event Started: 2/4/2010 8:25:07 AM ET
Please stand by for real-time captioning.
Captioner dialing audio line. > >
We will begin today with some preliminary planning on the June -- whole-genome sequencing -- comparative effectiveness task force. The rest of the morning will be devoted to the discussion of the committee's review -- draft recommendations, which we hope to have ready for release for public comment.
After lunch we will explore objectives, mechanisms, policies for genomic data sharing and share five models, discuss future directions, technology.
Tomorrow we will discuss the SACGHS gene patent and licensing report, and conclude the meeting with updates from our federal agencies.
So, lots has happened since our last meeting. In September we transmitted a letter to secretary Kathleen Sebelius that will support effective healthcare reform and in November received a reply from her, [indiscernible] to the office of the national coordinator of health technology, using integrated health records for the genomic and -- the letters are in tab nine of your briefing book and on the SACGHS website.
Last month we received a letter from Myriad Genetics Laboratories in response to comments at our last meeting. We advised them to provide specific information, notify their membership about how to do that. The response we got from Myriad to our letter is in your table folders along with comments from CMS, FDA and FTC on Myriad's comments.
On Friday, Alberta jut gutter yes will be talking to us -- developed laboratory tests.
On January 13, CMS, centers for Medicare and Medicaid services and office of national coordinator of health information technology published regulations to help implement the national health record incentive program under the American Recovery and Reinvestment Act of 2009. The -- incentive programs including defining the central concept of meaningful use under the EHR technology. [indiscernible] sets standards and certification criteria for EHR technology. Both these regs are open to public comment and we will discuss later in the meeting whether SACGHS should provide additional comments. Excerpts are found in tab 8 of the briefing book and you will hear a presentation from ONC tomorrow.
The paper on direct to consumer genetic testing approved in October was revised based on edits received from you in October. The -- moving forward to transmit, if you have additional edits, comments, please give them to Cathy To misby the end of the day, thanks to Sylvia on getting that paper completed.
In response to a suggestion by Dr. Francis Collins at our last meeting we formed a small group to draft a general commentary that highlights the committee's prior work on emerging issues in genomic medicine. The draft commentary is in tab 9 of your briefing book. I hope you have had a chance to read it. I would like to thank David Dale, Jim Evans, Andrea Gonzales and [indiscernible] for their work in developing the draft. It was an amazingly efficient process and with staff -- we appreciate your reading the document tonight, letting Cathy or me know if you have comments before we submit it. The target is to submit to JAMA for publication as a commentary.
I would also like to call to your attention the materials in tab 8 of the briefing book regarding the healthy book 2020. The goal of healthy people 2020 is to promote health and help make informed health decisions, a process conducted by the -- leveraging lessons learned from the past decade to set national health objectives for the next decade. For the first time the process includes a new set of objectives in the topic be area of genomics. Public input and staicialgd dialogue is insured to help healthy people 2020, and development of our work bagsed on previous recommendations, forwarded to the HP 2020 -- and the prevention of disease promotion, along with -- genetics and genomics as a necessary component.
The comments are in tab 8 along with a list of objective areas for healthy people 2020. We will provide more information tomorrow. We are making progress.
Finally, at our last meeting we received updates from the departments of labor, CMS and Treasury, IRS, the HHS, office civil rights and Equal Employment Opportunity Commission, EEOC, on the various regulatory developments underway to implement GINA. Prevents information from being used on a discriminatory basis. To help insurers and employers comply with the law.
The interim final regulations implementing title 1 of the law, the insurance provisions, took effect December 7, 2009. Three departments, Labor, HHS and IRS, joirntly issued the regs, will review the comments received before setting production schedule for final rule.
The office of civil rights received about 25 public comments on the proposed regulation issue on October 7, 2001 72009, most responded positively, and OR is considering the publication of public comment, to determine the feignal changes needed to the rule. Expect to file later this year after coordinating changes with the other agencies.
The final regulation implementing title 2 of GINA, the employment discrimination implications, through the OMB, and at the end of the clearance process. Final rule not yet issued, the statute became effective December 21 of last year and the E, OC began enforcing the protections against use, acquisition and disclosure of genetic information in the employment setting on that date.
Now, I would like to have, let all of you know we have two new members of the committee. They have been nominated. One is caris An. And the other is Janice Box. We are delighted to have you. Caris is the chair and founding director of the genomic medicine institute at -- and founding -- at the Institute's center for personalized genetic healthcare, priept -- Case Reserve School of Medicine and -- molecular School of Medicine at the Cleveland School of Medicine -- a full biois in your folders.
Janice is the state genetics coordinator, manager of genetics disorders section in the Michigan Department of Community health, worked in pediatric settings, Michigan State genetics plan, federal grants and cooperative agreements relating to birth defects. Newborn screening and genetic service delivery.
There's still a bit of paperwork left to finish before both can become voting members of the committee, we expect it will be completed by the next meeting employ on behalf of the committee I would welcome both of you, we put people to a lot of work, we look forward to engaging you fully and look forward to the contributions you can make.
Whenever we welcome new members, it also means we lose members, and we are going to be saying good-bye to two of them after this meeting and because it's really hard to let them go we are not saying good-bye until tomorrow. But tomorrow, Sylvia and Al, and Julia L, will be having their last meeting as formal members of the committee, but as they know, they never truly leave.
[indiscernible]
In addition to macadamia nuts, Eric green, the newly appointed director of NIGHI will be joining as E fishio member, and will be a great addition to our group as well.
Finally, I want to introduce a new member of the SACGHS staff, over there probably changing the airline res vaigzs, Allison Lee, a BS degree in psychology from George Mason university and MA in professional writing from Chatham, joined the staff in December, put immediately to work, instrumental in getting all of us here, getting the meeting organized.
Before we go any further -- Sheila, are you on by phone? She will be joining us, we will hear the beep shortly.
Now we come to the highlights of the morning session, the briefing on ethics from Sarah.
Good morning everybody. As you know, and I know you look forward to this little lecture. You have all been appointed as special government employees or will be soon, and that's how you serve on this committee. Because of that there are special rules that employees have to follow. I want to review a couple of them. First, about conflicts of interest. Before every meeting you provide information about your personal, professional and financial interests. Information that we use to determine whether you have any real, potential or apparent conflicts of interest that could compromise your ability to be objective in giving advice discouring committee meetings, we waive conflicts for general platters because we believe your -- we rely on a great degree to be attentive to the possibility that an issue could arise that could effect or appear to effect -- we are provided with a list of your fortunately fortunately financial interests and -- ask you to leave the room if -- since we are not too far from the capitol, government employees are prohibited from lobbying and we cannot lobby as individuals or a committee. If you lobby in your professional capacity or as a private citizen it's important you keep that activity separate from activities associated with this committee, and always keep come mind that our role is adviser to the Secretary of the department of health and human services, not the Congress. It's important that you be attentive to these rules. We appreciate your attentiveness.
Well, enough from me, now we need to -- [audio disturbance]
Is that Sheila? Anyway, our first topic is to talk about our plans for addressing the issues surrounding the affordable genome. This has come up repeatedly over the last few years. As we near the time when the affordable genome is likely to be a reality, thought it would be important to take it up as a topic in its own right. The next generation sequencing methods are bringing the clinical use of whole genome sequencing -- a variety of issues, seem to be surmounted, but an issue of affordable how the information can and should be impleed into clinical care. In tab 3 of your binders is not only some articles which hopefully you have had a chance peruse, but a set of questions. I would like to spend a few minutes this morning discussing what you see as the issues the committee should be taking up, so we can begin to formulate our plans for the future.
I will open the floor to thoughts about what are the issues we should be taking up.
Thank you.
First, I will ask a question. Have we received any specific guidance, questions, from the Secretary, the Secretary's office of high-priority issues, whether short-term or long-term, that the Secretary would like us to consider?
To my knowledge we have not. When I met with coul ins back in September, this was one of the items high on his priority list and thought a way to bring together many of the things we have been working on, genetic testing, utility, those sorts of things.
This meaning the implications of the affordable genome.
Yes.
What we are looking for here are your sense of what the -- technological issues we may want to talk about, may want to talk about issues around how it's incorporated into DTC or where it fits in with clinical testing, newborn screening, where it fits in with -- what are the down stream consequences. We have a 1,000-dollar genome, there are enormous human consequences, clinical down stream testing, all sorts of things to be done. We have a broad range of topics we could be taking on. My guess is we will end up forming a task force to help us with all of that, have informational sessions, but we would like to get your thoughts about where we might focus energies. Gwen?
I thought it might tie into some of the issues on the DTC task force, as it relates to affordable -- people do it outside a type of provider context.
Lar a?
[indiscernible]
I was going to echo what Gwen was saying. I don't think in reading the materials beforehand, that we should focus on the proximal issues, that is what are the challenges and closing the gap between the 10,000 and 1,000 dollar genome. I think that will happen without us and much more rapidly. I think we should focus on down stream issues and the kinds of things we always emphasized. Clinical utility is a big one, and the other, Gwen also alluded to. I suspect much if not most of this sequencing will be done outside the clinical arena and will only filter in to the clinical filter in round-about ways because people bring their genomes, providers, et cetera. I think we should focus on interpretation and trying to bear it out clinically until --
I will agree as well. We should assume that there is an affordable genome and define affordable at the beginning of the report. Some would say affordable genome at 1,000-dollars isn't truly affordable, but we get to that piece. I would be less inclined to focus on clinical utility issues, but take an assumption that if there are tests within there that have important clinical utility, and if indeed that is the case, similar to the early years with genetic testing, three areas. First, the health IT piece, what are the implications in terms of data that comes out of this, both from a magnitude of data and the issue around privacy of data and how that data, especially if it's done outside of the traditional system is shared or not shared.
Secondly, I think the issue of the payers, starting with the public payers is an issue, so if indeed someone who is on public payer system has information, how is that integrated or not into their care, what are the implications for reimbursement, or testing implications related to that. Lastly, with maybe education task force, what this means for physician education in the broader perspective as to if indeed this is available and everyone is bringing it -- lots of people are bringing it to physicians, what information do physicians need to be equipped with to choose to integrate or not integrate that information. Three core areas.
Let me push you on one thing. I would not focus on clinical -- actionable, reasonable health benefiting, huge amount of information we don't know what to do with, lead to additional testing that may be good ill. You don't think it's an issue to take up in this context? Not necessarily gene by gene --
I agree with your conclusions, the information, may flip-flop over time as we learn more. My concern is the amount of time and effort it takes to put together an assessment of the clinical U till, may be beyond what we can do in this committee in a reasonable amount of time employ it's not to say it's not important to be looked at. I see that less as our core competencies to do in the period of time I think this is relevant. I think it's, in a couple of areas, more important to have a core opinion on some of the issues than a lot of opinion on something else if it takes another year to get there. My issue is clinical utility is a bigger nut than we can crack short-term.
[indiscernible]?
Okay, well, I think this dialogue between you and Marla jogs my memory here that probably clinical utility is the most important thing this committee could focus on. The fact that it will take some real-time and effort and studies and money to establish the clinical utility of the personal genome should not discourage us from doing it. After all, we spent billions of dollars to get to where we are now, and I think it's very important to evaluate from a societal perspective, the balance of benefit and harm. I agree with you, Marla, there are actionable things in the genome, but many more non-actionable things, but people will take action on the basis of these. They might even remove their prostate or other more drastic surgeries as result of the knowledge of the genome. In addition to all what you said, I think the importance of the balance of benefits and harms has to be exploredded from a societal perspective. I wanted to refresh the committee's memory. Last year CDC and NIH held a workshop on personal genomes, the results of which are published in Jim Evan's -- journal, for which many people, including Francis Connors, Steve you were on that committee, made recommendations for actions. I think it's important to put that in the context of what we are trying to do here. If you think we are struggling with what to do with 1 million data points, you anticipate seen nothing yet. There will be six -- direct to consumer education, it touches on all areas this committee has been exploring, including clinical utility.
I think it's important, I urge the committee to keep the report as practical as possible. The whole genome sequencing there's so much public health issues. If we were doing this in newborn screening, the whole shift in [indiscernible] medicine is going to be given upon to families. If you have your whole genome from the time you are a newborn, what does that mean, we don't usually test minors, there are legal issues, patent issues. There's education issues, we don't have the workforce, an educated public. The practical issues, I think, what need to be highlighted to the Secretary, that these bring all those genetic discrimination issues, reimbursement issues, education, patent, like direct to consumer, brings back some of the prior reports the Committee has done and shows this makes all of that explode even faster.
I will make two points, probably both of them relatively less practical, but philosophically important. One is that the issue be of whole genome sequencing is really not going to be -- we can't look from the paradigm of what we have traditionally been doing relating to testing. This is going to be a huge problem of knowledge imagine am, not an issue of understanding all the different data points. It's really, we will have phenomenal amounts of knowledge, manage in a different way if we understand it is. For the session I think we would be well-served to hear from someone who has a content expertise around knowledge management, and the second area that is important to consider as we're -- I am just -- it slipped away for a second, hang on, let me get it back. Oh -- having some of the -- people that, person that comes to mind specifically is Zach Cayhoney -- we face some of the problems that [indiscernible] and Mara mentioned before, we find things we know what to do with, and others we don't. At least when looked at from the perspective of say whole body scanning there was very interesting concepts that emergedded from looking at that process, could potentially be relevant here as well. Someone that has done thinking about what we do with incidental findings, what's the response that people have to information that they don't know for sure what to do with, those are conceptual things that are going to be necessary to frame this.
Mike, Jim, then --
Jim, did you want to say something relevant to follow on to --
No, you go ahead.
All right. I want to bring to mind some of the practical issues that probably the Committee might want to consider, things like data quality. It's not -- [indiscernible] is not perfect yet, the issue be of that. Integration of the whole genome with electronic health record, because it's going to have to be -- you don't want to have these things separate because both are going to be important. Interoperability of the systems used to store the data, to manipulate the data, if all sorts of different companies make these systems independently, then they will never be able to talk to each other and be useful. Data security is absolutely critical, and data transmission, the issue be of just moving large amounts of genomic data from one place to another with perfect integrity is not simple, not trivial. Then, the most important thing is twep developing systems to connect to the -- genetic information is only important as it relates to the patient. There's really practical issues of how to do that. We've been talking to the National Library of Medicine on how to integrate the systems to standardize the way genotype is an owe annotated, any type of large scale clinical studies.
Jim?
Yes, I wanted to try to focus for a second on what our main role and our capabilities are as a committee. I think, like Mara points out, this is going to be an absolutely huge issue, right. There are going to be gigantic issues with utility, records, probably the best thing we can do is help the Secretary prioritize what the most important aspects are. I would again come back to the point that even though -- like Marc says, this is a qualitative game changer with all this information, but having said that, the rules haven't changed about the application of this kind of information to clinical medicine. We have to continually enforce to the Secretary that all of the won won dross information, has to -- as to the patients. We need to focus on perhaps a role of prioritizing and triaging for the Secretary, because we sure aren't going to be able to solve these problems ourselves.
I understand Sheila has joined, welcome.
Charmaine?
Mike be made one of the main points I wanted to make, integration with information about the person tested, and to piggyback on Sylvia's point about public education. I think that how people use the information, what happens when children get tested, how they handle that, the public education piece of it is major.
Andrea, then Eric?
I agree with every comment made, but I wanted to point out two issues. We have more than two different issues, but I want to point out issues that need to be brought to attention. One thing is that the $1000 or affordable genome will happen, there's a race to continuously decrease the cost -- data management are also being dealt with expeditiously, but will need help. From our Committee point of view we can look at some of these more -- issues that are practical to -- that we are going to foresee, going to be needed, to bring these to the testing or type of information into clinical electronic medical record. We know there are informatics needs, information about vocabulary, even [indiscernible] genetic information we don't have a genetic vocabulary. These are crucial issues that are important. The issues of analytics, quality control, it's crucial how we call these issues, also how we do proficiency testing. These are things we can start prioritizing or identifying for the Secretary maybe somebody else can work but we can do there. Connecting devices, interoperability into the different systems. These are practical issues that need to be solved, we can start to bring to attention.
The other component is how we practice having the whole genome sequencing, how we -- the information, education, so forth. Maybe we can start looking at these issues from the practical point of view that will effect how we practice, and the clinical utility is a huge issue to deal with. Just looking at different aspects, not just the clinical utility?
Eric?
The only point I would make, heard several speakers allude to it, jism Evans said it, I will hope the discussion doesn't try to focus subtle sees -- I can tell you in two months of being NIH director and about that, two year sequencing, the advancements are truly breathtaking, there's a wow. Truly, I have been involved in production of genomes in two years and what I see happening in sequence technology, even the past 12 months is truly spectacular. No matter what you think you are planning, issues you are dealing with, trying to get to it is impossible. It's happening faster than a committee like this can operate. I would think very ambitiously as to the amount of data potentially generated, and all the discussion about bottlenecks of information handling, is real, then multiply times five. There's no sign the pace at which these technology advances -- there's no sign it's slowing down. What I learned in the last six weeks, phone call after phone call, announcement after announcement, both vendors and scientists working on this. It is absolutely here, and the pace of acceleration will continue.
Paul, then --
I think following on that, a breathtaking review of the technology at some level, I would return to the first comment, which is affordability. That said, in the context of thousands of our fellow citizens not being -- going to free clinics because they can't the get any health care, can't afford any of it. I think we do have to deal with the notion in the critical sense of what affordability of this information, all members of society will present to whatever healthcare they are getting or not getting, with the genome sequence in hand. I am not so sure the pace of technology and our being able to provide that are equal.
Then the other aspects I would like to reecho, the medical and non-medical implications of broad based full genomic knowledge. Are there significant non-medical implications of this, I don't know if there are or not. Certain maybe genealogy, other things, I don't know. I think that needs to be considered. I really do agree with the knowledge management and the whole comments about the incidental, and ask Jim and others, there's also a patent issue in here.
AH.
There's another -- for Jim, we would like you to stay on for a few more years if you don't mind. How do we deal with the patent issue there.
Mara?
Well, too easy to tee up, but I am not taking on the patent issue. But just a broad comment, recommendation to the committee is Wayne greets ski had a great quote, the hockey player, somebody asked how he scores you all the goals, he said skate to where the puck will be, not to where the puck is. That has to be the overriding comments about the technology and movement going forward. We need to skate to where the puck is going to be. That alone will give the Secretary insight, given the thoughtfulness of this Committee, we can do in a thoughtful way.
Jim
In a spirit of camaraderie, with Mara, I am not going to talk about the patent issue either. I wanted to bring up one interesting thing. When you think about the whole issue of privacy, I think it behooves us to think about what drives that. To me what drives that, the reason people accord their DNA and genetic information, some increased level of protection, privilege, is that it can tell us something about the behavioral aspects of the person. Something br about our proclivities towards behavior, the non-medical issues. I think that's germane to a consideration by the group. It brings up the issue whether parts of the -- to be treated in the same, for example as psychiatric information is accorded special status in the genome. I think it might be worthwhile to not think of human genomic information as a monolithic entity, but the qualitative differences in the information that arises and should be accorded different treatments.
I like the greets ski analogy, and following the puck, the way I follow it is, not about technology, it's about health. And I think that's to the extent this information, like any other biomarker information can improve health, and can be affordable and can be used by all segments of the population, I think, we will have a winner, otherwise we will have a mess on our hands. I am hoping SACGHS will tackle a lot of these things.
Interesting discussion. I am glad we have taken this up. I agree with Eric the price tag shouldn't be the focus. Looks like we establish the price. The key thing in my mind, goes along with some of Jim's comments. Somehow to be in the position to help int great the scientific, technological development, with the physician's office based what you need to know and do; we need to help as much as we can with thinking about that process. Given the genome is going to be sequenced for somebody somewhere, somebody will need to know what to do with the information. That's not a very orderly process at all right now. If we can define this step, help define the steps, we will really do a service to our colleagues in the country.
I am hearing a lot of enthusiasm for a lot of issues. One thing, before we bring some of this together, on the affordability issue, whatever the price of this is going to be, that's the smallest part of the cost of the test. What's going to happen is consequences of it, it will be cost-prodousing and benefit producing. We need to understand what that's all about. But hearing the array of issues, this isn't about whether the technology will come, it's how we bring it to reality in a way that enhances the health of the population. My suggestion, and I think having heard from others prior to the meeting, that we use some of our time at the N meeting, come I believe is? June have informational session to get up to speed on various aspects of this, and then probably form a group to help us create a charg does that seem like ay reasonable plan?
We will need folks to help us pull that together, at least for June, presumably on -- Paul, you expressed interest in that. Carris is raising her hand. Perfect, someone who has been around the block, you can help, and caris, I am thinking we won't get this done so fast. You will be around. If you need more you can draw on others. In June we will expand the group to figure out how we will go from that information session on to a working group.
Great. Thankk you. That should be an exciting process, and an important one.
We will turn to the task force on clinical U till it and -- last discussed in June 2009, and established a task force Marc chairs to help create a charge to identify the issues that we should explore. So, Marc has been working diligently on that and will give us information about what he proposes we do that will be constructive in this -- actually pretty new and changing area. And one that is particularly challenging right now because we don't actually know what's happening with all the funding for comparative effectiveness in the healthcare reform bill.
Thank you.
Thank you. And I want to thank the task force members. Our charge is which issues if any should be explored -- our immediate focus was to try and access where things were at in terms of federal furpdzing in CER that effects genetic and genomessics. In 2009 there was a billion dollars, sorry $1.1 billion appropriated for comparative effectiveness for research divvied up to the NIH, HRQ and office of the Secretary of the Department of Health and human services to be targeted for comparative effectiveness research. The 400 million for the must be used to "conduct supporter synthesize, compairative effectiveness research or -- clinical data networks, other forms of electronic health data to be used to generate or obtain outcomes data."
The act also required the sk Secretary to task the funds appropriated to the Secretary and required the Secretary not only to consider the OIM recommendations, but also from the federal coordinating council for comparative effectiveness research, referring to as FC CRC for obvious reasons and spending $founders million appropriated to the office of the Secretary. Our strategy was to review the recommendation that emerge and identify those relating to genetics and genomics, to assess the degree to which these projects funded by NIH and HRQ with their CER funds satisfied recommendations and identify studies or projects not yet funded, in as much as we could. Then it led the opportunity that we could potentially recommend to the office of the Secretary directions for the funding that could support projects recommended by IOM or FCCR -- not funderred currently through an HRHQ. The FCCR is composed -- physicians with responsibilities for health related programs. Issued a report recognizing personalized medicine, effectiveness of intervention by patient subgroup. I will talk here about a synopsis we did of the report that focused on genetics, genomics and personalized medicine, the purview ive that, the written synopsis of this, others is behind tab 4.
Now, I also included a report by the Lewin Group, produced for the personalized medicine group that assessed, provided input to OIM and FCCR, how money could be used for comparative effectiveness research and the Lewin Group does a good job of how -- can complement one another.
The FCCR concluded the -- data infrastructure for CER. One example is patient registries. Secondarilily, recommended significant investments for dissemination, translation of CER, particularly CER studies on priority populations, types of interventions, defined as racial and ethnic minorities, persons with disabilities, multiple chronic conditions, elderly and children and priority types of interventions, comparing different home models, surgery versus medical management.
The full portfolio of -- investments, critical to consider CER and health IT wholist irkally" as such our committee may want to continue to encourage health IT policy that supports collection of genetic information useful for CER and [indiscernible] barriers, we will spend the afternoon obviously talking about some of these issues.
The IOM report issue the June 302009, generated 100 prioritized research topics and 10 recommendations of the 10 topics, two explicitly mentioned judge beings or genomics, the first is core priority looking at effectiveness of genetic and biomarker testing with usual care in treating breast, coal owe rectal and -- the third is compare effectiveness of biomarker information including genetic information with standard care in motivating behavior change and improving clinical outcomes. Eight other research topics that could conceivably -- in scope, but not explicitly mentioned.
The NIH reviewed all the recommended 100 topics, concluded most are already being studied through ongoing NIH research, the task force identified numerous funded projects in the genomics and personalized medicine space. There is good progress relating to this, particularly in that first core priority of cancer.
Of the 10 recommendations there were two we thought were of particular relevance to the committee. Seven, devote sufficient resources to -- research in the innovation of CER. We would potion posit beyond CER we need clinical utility, as we already heard in the discussion about affordable genome. The HHS should had twep large scale -- data networks for use in CER. Raises privacy, informed consent issues that will likely overlap with issues raised by genomic data sharing and ongoing efforts to create such networks. The recommendation implies we need to collect clinical level data. In some ways what we will be discussing around meaningful use will also relate to this issue, because if we are not representing some of this in meaningful use we are not going to be able to collect it.
Now, I did get a chance to play around, and thank you to Mike Lawer for helping with searches on this, to look at the NIH era fund CER grants, several projects are going to directly relate to genetics issues that the IOM recommended. 24 of these were specifically funded under the comparative effectiveness research monies, detailed under tab 4. I was exhausted but didn't do an exhaustive search, so if you want to parse it. There's probably at least 50 to possibly hundreds that address personalize medicine issues not directly related to the I OM top hundred. Some of these studies are using the methods ever comparative effectiveness research although not specifically funded by the CER-designated funds. Many of these projects will serve as investments in data infrastructure and dissemination in translation of CER findings consistent with the FCC, CER's recommendations.
We don't have much information yet on the HRQ CER-funded grants. [indiscernible] did provide some information that two of the announcements, the choice and I-adapt are closed and the rough estimate of applicants is [indiscernible], a small proportion of have a focus on genomics, but detailed reading may reveal others. The project and EDM announcements are still open. [indiscernible] estimated perhaps 10% of the these may have something to do with genomics, would be a substantial number. All of these grants will be reviewed, funding decisions and awards before close of 2010, September.
If we are to look at gaps in terms of what is actually happening, I think there were three that could reasonably be characterized as such. The first is evidence of adequate evidentiary applications and the third, cortile priority recommendation, healthcare delivery systems and the third the coordination of efforts. I will briefly talk about each.
I thank Steve for allowing me to borrow his slides. Some of you have seen these in another context. This slide overlays [indiscernible]'s T 1 to T 4 translational efforts against when do evidence-based guidelines actually come out. This sort of represents what might be considered sort of an ideal model with everything in balance where our evidence-based guide lines are occurring before we go into health practice. The problem, of exowrs, is we really don't know where the evidence bar should be. If we lower the threshold for translation into practice we may have things moving into practice that have little evidence on clinical validity, utility, may effect coverage, potential for increased harms and also benefits, moving things out that actually work. We are usually relying on expert opinion at this level, but this type of evidence bar would stimulate innovation.
In contrast, we move the evidence based weight to the other side, we [indiscernible] with good prospects for e imbursement, but lower innovation, the cost -- reducing the likelihood of harms and may disciple inish benefits, some treat wants never make it to the arena, where we can't generate sufficient evidence.
I am not going through this matrix, but something discussed at least super officially superficially -- saying in each of these bars what evidence do we have around efficacy, regulation, good evidence there, reasonable evidence, feasibility, no evidence on cost, these type of things. You can fill that out, use that in some type of decision-making process.
So, I think this is an area where we have heard about this before, this Committee, we have heard this morning about where that evidence bar has to be. This is something where the Committee could potentially play a role in helping determine this. I will also mention, not in tab 4, but another part of the packet there's a comment of the CMS med CAC recently surveyed on what type of evidence you need to make a coverage decision. There's interesting findings from that, I think support the same issue. We are really struggling to say what is the evidence bar we really need.
The second gap is this third cortile, the effective of biomarker -- standard care, motivating behavior change and improving clinical outcomes. There are very few of the funded projects I reviewed that specifically address these critical issues. Maybe more emerge in the HRQ projects. But something that being be a fair point of -- particularly related to the issue of behavioral changes. Both for are providers and patients.
Then the third thing is coordination. There's all of these different projects, all collecting information, creating a lot of registries, but are we really using standardized representation and storage. Will this impair our ability to share findings across projects. Could we learn something about genomes in one condition associated with risks for another condition, combine the information. I use psoriasis and coronary artery disease, something that came up, looking for a larger control group for psoriasis, do you have genome typed individuals -- but they are only to be used for cardiovascular -- did you know it's a huge independent risk predictor for coronary artery -- turns out none of our cardiologist knew that. Now they are very excited about working together. This is something where there could be a lot of synergy -- something we were thinking about as a possible role for the Secretary.
At present the Secretary's funding decisions are unknown. The Secretary is required to send operating disitionzs to Congress in July of 2009 concerning funding decisions but the report is not as yet publicly available. I almost took this slide out because I was depressed. There was a bill introduced into the Senate I believe that -- an independent bill indicating studies should take into account molecular and genetic subtypes, codifying this type of work. That bill was folded into the overall health care reform bill and was in fact represented in the house and Senate versions that were passed, but as we all know, the status of that right now is unclear. So whether this particular bill will be extracted from healthcare reform and brought up independently or not, I wanted you to know there are things at the legislative level that may also impact what it is we are going to do.
So here are potential next steps for the task force. One is to try and get a handle on these evidentiary standards for genomic tests. Outlines considering adjusting an evidentiary bar. For example, if we have something like a Warfarin for genomics, we are potentially going to be applying to hundreds of thousands of individuals a year. We probably need strong evidence this is going to work. On the other hand, if we have -- doing A or B, perhaps we don't need as much evidence to say we think there's genomic information that would distinguish between going with therapy A or B, it may be reasonable in that type of situation to move forward with a lower degree of evidence, since right now we are essentially equal.
There are other entities that have begun to address this issue. This was one of the major areas of focus at the gap meeting that took place last fall. It may be the Secretary could charge this entity with taking ownership of this particular issue, but it's one we thought was quite important. We could create an inventory or clearing house of genomics -- the CERA, agenda, could inform -- with the special attention to the healthcare delivery system end point.
We also thought about the possibility of having an informational workshop on this issue for the June meeting. We need to continue to monitor the health IT issues that continually arise. In particular reviewing the meaningful use rules, which we will be doing. By the same token, we could say our work is done, enough happening, maybe there isn't a role for the task force to move forward. That would be a potential next step. Some of you may come up with brilliant ideas I haven't thought of, in which case we could consider other options. With that, I will end and we could have discussion.
Wayne?
Thank you, Marc. I would not suggest to dissolve the task force. I think we are just beginning to do the work. I think CER, when it's all said and done, is sort of a good sort of median by which this committee and other groups can tackle the issue of clinical utility. It's a way to address the clinical utility in the real world. Whether CER will live or die in congressional language, the issues it raised are real and they are already on the table. Just by the way of clarification, additional information, I was looking at the 24 projects you identified from the NIH list. Many have nothing to do with genetics or -- but coded as such. I wonder if you have issues on that. But let me just finish my thoughts. As part of my other hat, I have two jobs, one at NIH and much time at NCI, from the NCI perspective we funded seven out of these 24, part of a network of CER and genomic and personalized medicine. We had our first meeting with the grantees in January and connected those groups with both gap net and E gap. They are going to, and hoping we can find across NIH other worthy projects that can actually join that network from a non-cancer perspective. I think cancer is sort of the dominant fields in CER right now, and the [indiscernible] priorities reflected that breast cancer, O var can cancer, ets cetera, but there are other worthy areas. I think if this committee keeps shining a light on CER from what its true meaning is, for clinical U tillity in the real world, have a discussion, work with the other groups, develop some kind of report to the Secretary with specific encouragement or recommendations, I think it's a good way of spending the time, because it's a window, it's an opportunity to shine the light on so-called clinical utility issues. Thank you.
Let me explain, on the issue of what we were talking about on clinical utility, a fairly defined thing, harms, benefits in health terms. But the decision factor matrix, how different people -- context is very important. FDA has a specific set of regulatory requirements how it makes decisions. Safety, efficacy, payers have other criteria. Patients have a set of criteria. You can think about all these things not just as clinical utility, but we can add value saying how we can help get the information necessary for decision-making, which the clinical is one, and I suggest that patients, clinicians think about these things differently than a regulatory agency or even a payer, but different people need different information. How people understand that, the information needed, where they get it, to make better decisions is one of the pieces I think should come out of that slide you showed. Jim?
I wanted to put a plug in for -- you highlight something in your synopsis early on that I think we should make a conscious effort to address and counter. That is the bizarre ac ackizations you hear, research is antithet biological to research. Steve's commentary beautifully articulates why that's not the case. But because you hear that a lot that should be high on our radar screen to counter. It's simply not -- [indiscernible]
This group is rarely at a loss for words.
Mara?
Just for fun I will say I very much agree with Jim. I think you continue to hear that about comparative effectiveness, looking more broadly than just against the standard of care today is the key change to that perspective, because there was misinformation at the beginning that it was only looking at the current standard, and that brought about some of the concerns that personalized medicine was not always in comparison to the current standard, and therefore by changing it would not be appropriately viewed. But in both the report and other work, the broader definition of compairative effectiveness has done that, but I think misinformation and perception is still very much out there.
Gwen?
I think it plays into a lot of erk moshings emotional fears, same as -- that it will lead to health racks rationing. I think it has to be very clearly articulated, taken out of an emotional context.
It's interesting you mentioned the R word since ARRA mentioned particularly you couldn't in the research, most of us said -- that's really tying ours hands to some degree, the whole idea of how we do the research is still up in the air as well. The FCCR report spent a lot of time talking about alternative methodologies, methods not traditionally accessed or scored well in NIH opportunities, a little less so in -- adaptive trials, things really new types, new ways of doing research, research off the clinical data we are beginning to accumulate is a critical piece of this. That emphasizes the need to capture the data that is important. Some of that will be genetic and genomic, we need the ability within the systems to pull that information out.
Andrea?
To ma Marc is saying, I find missing in this -- use of genomic and genetic information, the [indiscernible] research laboratories, we have to be concerned about the quality of the testing being performed. There are clear regulations that establish even for research purposes, information transmitted for decision making should be done in a [indiscernible] certified laboratory. I didn't see anything about that. The other issue is not only that the quality of the testing is how the result will be transmitted to healthcare providers or researchers. Being a practitioners, I know the challenge to convey specific information, what you can test, the specific limitations of the test. Something mitionzing is comparative methodology research. [indiscernible] I can give example, different technology to use to do detection, make changes of -- decisions on treatment. That research, I didn't see anything, but I think it's critical you add that part of the information.
To talk to Mike's reference -- normal way to do proficiency testing, no part of anything I have seen. I would like to maybe recommend the Secretary to create a clearinghouse for information similar to the clinical trial.gov, where the information is put, model there to use, recommend the Secretary to use to put some of the comparative research in publication. The last is biobanking. As we continue to work through all the issues we talked in the previous session, and the current session and session that will follow, the use, storage of specimens well-annotated and quality control is critical not only for continued research, but to go back, do new testing with new methodology, all need to be part of our discussions.
Marc, this is what you put us, for us to think about, something tells me you are not totally agnostic about which of these we should be pursuing and when. Do you want to lay out what a reasonable agenda would be?
I am not sure I can define a reasonable agenda.
An unreasonable agenda?
I am much better at that. I think from a practical perspective, the -- you know, some guidance on evidentiary standards is going to be critsically important. Whether this is something that could reasonably be expected to be completed by a task force of this committee or whether this is something we need to get an idea of who actually is in the game relating to this and say okay, here are the people taking ownership of this, this is something we need to support and hear back on, I really don't know on that. Again, I think it would be beyond the scope of a task force to be able to create an inventory or series of inventories. I think it's a critically important thing to do. One thing the task force might reasonably do is to say we need a clearing house of information and on these different issues and recommend that be created within some entity, that was something discussed at the initial gap Net Meeting, one thing gap net could do to have a clearing house of projects so people know what actually is going on in the space.
In terms of the informational workshop, we already know we will have a workshop on affordable genome, may not be reasonable in June to have another meeting or it may be people think we have heard enough from prior presentations we don't really need to go there again. Certainly that would be something the task force could reasonably take ownership of in terms of pulling that together. That doesn't really answer your question all that well, I don't think, but --
The good news is [indiscernible] is raising his hand and since he's -- in all aspects, he can tell us what's go on with Gap Net and other nets --
There's [indiscernible] fa bet soup out there. Here's what's going on. I suggest this committee can weigh in towards the end of the year, maybe after June. The reason I say that is for a couple reasons. One, the projects that are actually being funded now, and that 24 plus or minus 10, I think are doing the work, plus getting together, and trying to develop that number one, and the roadmap type issues, and they are going to have maybe joint meetings with an IOM roundtable on genomic transtranslation, in the background already, and the clearinghouse projects, the genomics -- all kinds of things this year, and [indiscernible] can tell you more. Waiting until the end of the calendar year and then having just another session to figure out, really, what's going on, could inform this committee as to what the next steps should be, just waiting and seeing what the other groups are doing. There is really no need to rush immediately, because the work is being done, and maybe if we put the place holder, maybe at the June or the October meeting for a quick update on the various efforts by NIH, CDC, G -- and the roundtable could give more information to play with. This is rapidly moving target this year.
Do you see any gaps at the moment that -- which others are not addressing?
I think our gaps in all of these things, obviously, whether or not the other groups are going to address them fully is not clear. I would suggest that we work with them somewhat, since many of us are involved in these things, and wait and see towards the latter part of the year what kind of recommendations this Committee wants to make to the Secretary. Remember, all these other entities are doing it from various vantage points. NIH is doing their thing, but this is committee that provides advice to the Secretary. We shouldn't wait too long, not suggesting to wait another year, but towards October meeting we will be in better shape information-wise.
Andrea, Marc?
I am agreeing the issues may have to wait until the fall, but wondering if we can do something in the meantime, the issue of [indiscernible] testing, genomics, genetics in research laboratories, information being used to trigger patient -- [indiscernible] need to be done in a -- certified laboratory, rigorous quality control. We need to bring that to attention of Secretary or somebody, in those areas.
Andrea, I am wondering if that falls under the general rubric of clinical utility, the data sharing, issues we heard --
These grants are already being granted, testing is being done, do we need to bring these issues up?
Yeah, I guess I would share the issue about whether that's something this task force will be primarily tasked with, because as I hear about this it really seems much more related to the work we had done in oversight. I am not saying we shouldn't, we probably as a committee some respond, but I am not sure the best wear to do that, would defer to Steve. I would certain the not disagree with what M said, there is some wisdom in that. There are two things we can probably do as a task force, even if we were relatively inactive, one is to monitor the Secretary's report, so when that emerges into the light of day we can review that, see what the priorities the Secretary identified will be. The second thing is when we do actually have the information on the aarc funded projects, look from the perspective of how genetics, genomics is representative and give a better idea of the overall scope of what's going on.
Two more quick comments and we will try to wrap this up.
I was going to comment, I think probably the space for us to be the second two words in our name, health and society. That is, the patient's question often is does this information matter to me? And the parent's question, is my child healthy? The piece we need, which really doesn't fit with the acute stimulus money, but is long-term. The information sets that provide clinical information to link to genetic analysis, and so we need to encourage the government and other sources to invest in -- people say registries, patient databases that allow for drawing good conclusions. Those are long-term investments. The huge value of the Frappingham -- long-term investment, looking for ways structurally to fund projects is very important.
Mara?
Well, maybe it's a good summary following up on Andre a's question, there time-sensitive issues that need to be addressed in the short term? I understand M's comment about from October on there are other issues, but in the light of this set of grants now are there comments, summaries on what's been put together to date that need to be useful and actionable, need to get to people before the October time frame. That to me is the key time-sensitive question. As I understand the health questions, I also focus on the relevance of this committee. I want to ensure we are doing something, people need the information.
I am hearing we should be monitoring, also hear we should defer until October to get a real presentation of what's going on with the other entities and we can make a decision about what's going forward. We can do -- ask staff to monitor, maybe provide information for June.
And are there any implications for which there are action items that can be impacted by the Secretary's office for which our view of it, even initial look at the data is relevant.
I could ask Andrea -- you have a good notion, I don't. Maybe you could coordinate with staff about what could be done in the interim and look to the fall to get an update, activities where we can add value.
If I understand this, the issue is, as I see it, in these funded research projects currently doing genomic testing there are concerns you have about how the testing is being done interest whether the results actually represent the quality that needs to be -- that we would need to have to draw conclusions.
Well, there is a federal reg eulogy that covers those type of testing. If you are making a clinical decision on how to treat a particular patient, even research, should be done in CLIA certified laboratory. Bringing to light the agencies the issues that need to be mindful of.
Is this really something, since right now the primary funding is through NIH, is this something that would need to go -- rather than go to the Secretary to NIH?
My issue was slightly different, really a question. Are there any decisions that are being made, less on the previously-granted grants, which Andrea mentioned, but more on those coming up, for which the analysis we have done, and you, Marc, in conjunction with other pieces, is useful to get in front of the Secretary or others. Is the work done so far useful to anyone in the granting of additional work between now and the end of the fiscal year?
I think I can answer that question, which is right now everything, I don't think there would be any way to insert anything into the ARC process, my guess. My understanding is the Secretary's report is done, just under consideration. I don't think for those things, the other two pots of ARRA money that haven't been distributed that we would have a opportunity to weigh in on that. I think that would be going beyond.
We need to wind up this session.
That would work well on other issues, that was the core of mine.
Just to clarify, the scope for this committee, task force, was the ARRA CER, but funding many that predate this. Some of the issues raised by Andrea, the validity of the test, performance of the test, we actually have a methods report I will talk about tomorrow, which discusses the quality issues and looking at the evidence. So there's also other grant projects like the work on pharmacogenomics, coming to a close, I will suggest that if we wait it might be useful to get a lay of the land, and there are other things that we are not discussing here that will also be part of the discussion. Also, it's a fast-moving fields, some have started using the term patient-centered -- I think if we stay true to what the overall goal of our project is, regardless of the label, we will have a more long-lasting impact.
[Captioner transition.
] > > >
That brings us to a break. I know we are running a little late. If we could start back 10 minutes from now.
Thank you, everyone. > >
Secretary's Advisory Committee on 10-minute break and will reconvene at 10:10 a.m. Eastern Time.
Captioner has been disconnected from the phone line, and will redial at this time. Thank you for your patience queue
Remind everybody on the race site Security -- right side of your table folder is this form to fill out. Thank you very much.
Let me turn this over to Barbara.
I will not delay too much did I will start off by thinking everybody for giving me the opportunity for presenting this report. Before we launch into it we will hear a presentation by Jana and she will share with us their findings and then when she's finished we will launch into our report, Jana?
Thank you, it is a pleasure to be here and see some faces from another committee. It is the pleasure to share with you our report as we both feel that we value the need for education and training in genetics and especially for us in newborn screening. Our committee, our subcommittee I should say is comprised of myself and Dr. Tracy Trotter who is much more colorful presenting as well as members from other organizations to include ACOG and the American family of pediatrics and the National newborn screening center and Genetics Research Center P. these are some other individuals. One of our initiatives is to come up with a newborn screening clean house and to help facilitate the discussion on that and we're happy to announce the genetic alliance and the National newborn screening and Genetics Research Center with HRSA will serve as the Greenhouse did their website is not active. The purpose of this is a place for people to go directly to gain information from a professional and public perspective. I will not read each slide for the purpose of time. These are some other updates of what is going on. We have the perinatal family health history to work with family practitioners to provide a family health history tool to help again, educate and learn what is behind these genetic issues and newborn screening and to really prepare families. The American College of Medical Genetics has a great program that is on the horizon and that is their summer scholars program. Their rationale is that statistics show that about 18,500 medical school graduates, only 163 enters the field of medical genetics. Currently there are five states that have one or fewer medical geneticists. Over the next ten years over three medical geneticists are expected to retire so we'd need more pits of the developed this program that will be launched in 2011. The purpose is to address this workforce issues and capture students' interest and involving the students by practicing Genetics in their work settings to include clinics, labs, government and regulatory agencies and hopefully Foster professional memberships and highlight the many the First employment opportunities -- diverse and limit opportunities that the medical field has and hopefully get more tinnitus out there in the field -- geneticists out there in the field and the collaboration will be strong in helping to move followed with education and training. This is another list of folks addressing the issue of education and training and working as partners. Another quote that supports are in need, advances in newborn screening service new challenges to the PCP. They require access to information, collaboration with local, state and national partners is essential to optimize the function of the newborn screening system because it will not be as productive as it needs to be people are not educated and trained. These are the various partners that will working with to help enhance this. The focus from all of these perspectives is to really address the response to the initial out of range result, what to the positions to? How did they do a? How do they handle it? To coordinate the complete evaluation, what are the next steps? Providing medical home in court -- and coordinate care because everybody plays a role in this important aspect of newborn screening Creek are education and training committee service in eight advisory capacity -- and eighth advisory capacity both to PCP and public presentation. It has been very worthwhile to help bring everyone together to help address this issue. And because we all value the fact that we just need to avoid duplication and enhance that collaboration and we will be more productive.
In regards to PCP Education we were able to participate in the National Institute for Health genetic Research Institute and their conference of developing a blueprint for primary-care education and we were able to house a roundtable session on the second day which included about 30 participants appeared to really talk and address the issue of water specific educational needs and barriers for them from each of their perspectives and what we can do to lift those barriers and enhance the education. A report for publication is being prepared by Alex Hemper. Some of the target areas are listed here as you can see. Again from each perspective and how those agencies and organizations can address these issues into the the resolve them and provide better education and training. We feel that each organization from the time, from the prenatal time right up until the family practitioner, everybody really does play a role. We also address some of the barriers to educating some of the primary care providers. These are some of the comments that are made that we have to address which is lack of time. Everybody only has so much time in their day to really get in depth with such an issue of genetics. The lack of geneticists to train the primary care providers including especially those that are already in practice. That is where we really value the fact of getting those medical students and educating them early on lack of a enthusiasm. But there is literacy out there that interests people. And lack of certainty and confidence. It is easy for people to say, that is not my area of expertise and the concerns about relevance to child health care and the fact is that Dr. Trotter likes to say everybody does genetic screening, a genetic testing if they take care of a newborn that day.
These are some educational interventions that are taking place that we feel will really help move things along and that is to keep up educational curriculum's. Again, it is taking steps backward and going to the very beginning of future positions. Insuring the board's certification exams to assess basic literacy and an oh medicine and have been CMEs and incorporating them into primary care as well as promoting the participation and related educational activities to the maintenance of these fortification processes and to create a web site that will be a tool for everyone. Genetics and the primary care training institute are working on a learning collaborative that will help prepare positions with busy primary care practices with experts in genetics and genomics medicine so they can work and provide a hands-on opportunity to be educated in genetics and newborn screening speed and at the end of the year to share the results and formally evaluate the project impact. Our next steps as the look on the horizon are residency training materials through the regional activities pick partnership with our organizations and the American Board of pediatrics and a development of the genetics in primary care Institute and to continue falling up with your committees educational taskforce. We strongly valued the need for education and training both on the professional level and the public level. And as technology advances and the awareness and the newborn screening programs continue to develop and progress, the need for this kind of education and training is far more important than ever has been pared we think it would be a hockey puck analogy. We have to look ahead to where it's going especially with the other disorders on the horizon that are being addressed and look at to add to our panel of disorders. So with that I thank you for the opportunity to be here again and share our initiatives and look forward to further working with you.
Thank you so much, Jana. Great minds think alike. I think you're finding spring much articulate with ours. It is nice to see that we come up with the same barriers as well as the same solutions so it makes logical sense that we would be able to continue to work together and we will talk more about how to be able to do that we have a lot to do in a short period of time. The most important part for me is to have the discretion and receive your comments on our recommendations. Before we launch into that I want to give a quick overview of this committee. I actually am familiar with the report as one might assume can last night on the airplane I read it from beginning to end and came away with a couple of impressions that have not necessarily had before and I wanted to hide it goes a little bit for you pick one is that the report makes the case that since the very earliest days when there was a secretary at's advisory committee genetics has always risen to the top. Any time, education is there and there is always a non and it has an influence on genetics' bid so it clearly has been on our landscape forever. Over the years much has been written about the challenge of translating findings from the human genome project into something that may be clinically useful more recent attention is paid towards looking towards a chronic and this is and what we can use -- how we can apply genetics as well and also the promise of personalized medicine is also on the horizon. A common image that I think all of us are carrying in our heads these days is this continuing. It may be something like genome Sciences and on -- and on the other side, different words but it is an incredibly thick line between the two. I think that reflects that is the challenge to do that translation from one to the other. So we are kind of looking at the right-side of that in this group looking at health care but I think that we also sort of know around here that is a loop and there are pushes and pulls back and forth. So we do not want to be thinking about health care and health professionals in isolation. There are a few things that are not controversial, I think and I think overall the whole report is not controversial two that are sort of slam dunks and one is that we all might agree that we are all best served if we have a knowledgeable work force that understands appropriate use of how to use genetic information. The other thing is that consumers are for disciplines in this as partners of these endeavors rather than simple recipients of services. So as I think are probably shared values for most of us. What might be in the less obvious and batted in this report is this notion of the translation of signed into clinical application. And also about the transportation of thinking. Perhaps even in the absence of anything of any on the ground applications that second idea is often called requiring a paradigm shift. And if we think about the original use of that word, paradigm shifts, we use that a lot. They are dramatic and often cause disruptions in the science when they happen. They are rare and we do not know if we are in the middle of one or not but they do cause a big change so I want to suggest that that may be some subtext in the report that is not necessarily openly stated principle. Thinking about a paradigm shift in scientific revolutions, you have seen these names before. They are an interesting group of people. It is a huge group of people and the richness of knowledge the so it is a big group. We have divided into 93 work groups and each of those have leadership, health care professionals and David is the terror of that group -- is the chair of that group and Joseph has stayed very involved and he is here to help us answer some of the questions and Vince is the chair of the consumer patient group and he has hung in there the whole time. The time line. We picked this up from the previous group that worked on it in 2004. We had an international round table. We were then passed with forming a task group did at that early meeting there was a decision about the boundaries and we came away with the setting in this report, separated into three groups for those discussions were long and hard and it seems like that actually, the boundaries are tighter than many people suggested. We narrowed it down to those three. Those three could also be three different reports and perhaps that is one way to approach it. What we attempted to do is to think about the notion that ideas and people moved to a systems. They do not just stay in those three silos. So the intention of combining into one report is to take a nap and appreciate that integration of services across -- a nod and P sheet that integration of services across the landscape to thousand eight and 2009 was for the bulk happen to it -- was with the bulk happened and there was a session in D.C. and then they were heavily the soft by staff after that and that is what we will be looking at today. The final report will have an Executive summary recommendations did the one that you see here does not pick it has the ordinary background and scope which is the literature and then the three working groups have their own sections of their literature as well as the data that they collected. We have a freestanding survey of federal activities that was intended to follow up on what has happened since the previous 2004 report and then conclusions and recommendations on our data a gathering activities including a review of all of the literature including the groups that we mentioned and then each workgroup conducted their own research and they paid and administered surveys. They also did some interviews could each of the work groups function. Thomas Gleason of their data collections -- each of the work groups worked autonomously. Within those Workgroups they were the ones that decided what data gathering activities were to be done. So they had a lot of autonomy. Before we talk about the discussions I want to highlight what we are trying to accomplish here today. We do have a couple of discussion questions that we're going to ask at the end of them. One is to the findings follow from the literature repute answer to a? To the draft recommendations target the issues identified in this report? Are these recommendations specific enough? We always talked about that we want them to be actionable. Do they rely on the appropriate degree on the public sector, private-sector and public-private partnership? Are we targeting it to the right places? And overall, is this report ready for primetime? When we go through it you will see that the recommendations are fairly dense and we will talk about whether perhaps the message gets lost or it is required to get our point request and there is a couple of decision points on how to freeze these. I have talked about this report in the past -- in the past as an unruly teenager. Not to kill a metaphor I will do it one more time but right now it feels like it's a young adult. It is feeling quite confident free and is ready to enter the real world. That it can handle any criticism that may come its way because how hard can that be? And perhaps sort of optimistic that good intentions to lead to good outcomes. So part of the question that we're asking here is is it really ready for prime time? So that will be at the end of the session.
So we are trying to turn it into a cynical, older adults? Is that the idea?
That is what I will not keep playing this support out.
Findings, generally. We came through both Datapoint, the review of the literature and the original data that we collected came up with the couple broad conclusions paid there is just not enough and we need to see more education. The need for clinical services has increased but the work force is insufficient. We need more numbers spit and health care professional organizations report about competing priorities. These are legitimate concerns that this is not a primary concern and where did they put it in this list of very important other tasks that they do.
The current public health force is not well-prepared to receive and simile genetic and to no information in to public health and there are a number of barriers to that because the work force is uniquely differs because it covers such a range of population health issues. And consumers prefer to obtain genetic information from the providers but they also turned to the media. A couple needs identified. The need to understand the concept of multiple risk factors. This is in contrast to a very deterministic view of genetics. Understand the role of the informant and the complexity of that a need for various tools that are understandable to evaluate the veracity of the information. And then Of course, concerns about direct to consumers genetic testing. Most consumers be the government as a trusted source to information. There are seven recommendations. So it is not 1 million. I went back and forth trying to decide what to do with this and I will read them in case some people cannot see the screen. I actually find them easier to follow on page 110, tab 5. So I will read through all of them first, rapidly. There are a few that you see require very concrete decisions and others you can see if you have comments about that. Each recommendation has the preamble or preface. So for number one a significant body of the chair from the nine states and about how leads the inadequate genetic education of health care professionals as a significant factor limiting the integration of genetics and to help care. It is often minimal and help care education programs and focuses primarily on a single gene disorders and is not assisting with long-term all detention for critical applications give innovative approaches that could make the efforts of entities will be required to remedy the situation did these entities include but are not limited to health professional organizations educational institutions, specialty certification boards and academic accrediting organizations. So there are two Options of recommendations that all of this preamble. We will need to choose between one of the two or combine them. The first one is that HHS should for remotes a dictionary public or private panel -- should form a multidisciplinary public or private panel. They should have established programs and genetics question gnomic educational education as well as help professional organizations engaged in genetics and genomics and continuing efforts. They will identify successful education and training guidelines and models that are outcomes based. Identify what works. Number two, and identify current funding streams for developing and promoting education as well as gaps in funding. Three, recommend mechanisms for expanding and in Canton the content needed to prepare health care professionals for personalized and oh health care. Before, recommend how full the standards of vacation accreditation, and continuing education activities might incorporate to gnomic content. That is about the whole world of certification. And five, published findings and recommendations and develop a plan to monitor the outcomes of the torque. Option B is that HHS should convene a workshop to identify and the rest is the same did something keywords is to "convene a workshop" the purposes are the scene and the traces are between forming a panel and convening a workshop at the end of that there is a recommendation connected to this to act on the recommendation from a previous reports. And this relates to the notion of decision-making tools -- I know there is a word missing there, clinical decision support we can decide whether they should be part of a recommendation or part of the preamble or whatever but that is the choice. Okay. Recommendation to, consistent findings in the literature indicate that health care professionals and public health provider serving under certain and and a representative groups and representations' face additional challenges and they have specific needs in their involvement and development of effective education models pits of the recommendation is that HHS to promote the development and implementation of innovative genetic anti gnomic education and training models for health care professionals and public health care providers serve in underserved and underrepresented groups and populations. Specifically HHS should target research funding to identify effective educational models for health care professionals and public health providers and understand committees. Some funding to identify models. Identify and support programs to increase the diversity of the health care work force in general and the genetic specific work force. And in sure that consumers and representatives of the rule, minority and disadvantaged communities the dissipate in the process to a mother to make sure that they are culturally and linguistically a prepared and tailored to the unique needs of these the diverse communities. Staff recommendation Number three pick in Harris diversity of the public health work force makes it difficult to target educational efforts to improve the genetic and knowledge across the work force. It's systematic ever that a one reads the composition of the public health work force with current job as possible these related to genetics and genomics and identify speech needs has not been done. This has to do about serving the public health work force. That group that is so diverse. Specifically tapping the expertise of its agencies with relevant missions and public-health HRSA, CDC, IHS, NIH and assess the work force to determine the number of public health providers to ascertain current trends. To sort of look forward to the public health and see where we are now and where we might need to go. I'm sorry, I missed a sentence, and future needs to identify training needs to promote leadership development in the field could based on this, HHS should support and encourage the incorporation of relevant competencies' and require genetic knowledge. The key here is the core competency should be based on those. And fund educational programs based on these capacities that promote genetics and the enormous talent and recognize the potential impact of affordable to no analysis and incorporate the concept of environmental interactions in risk assessment for population based genomics. They should be based on these trends that we're seeing.
Recommendation Number four. Consumers have consistently expressed the desire for genetic information that is comprehensive, accessible and trustworthy. And again, this is the second recommendation that we have to actions that we should decide on today. The first one is that HHS should indoors and insure sufficient funding for persisting government resources such as those developed by NIH and CBC to provide trustworthy genetic web based information for consumers and they should include scientific a delegated information and also links regarding to the topic such as genetic contribution to health and disease to reach a broad range of commodities they should also include links to information that are not web based such as television and radio programs and print materials and they should -- the availability of these resources should be promoted using a wide range of strategies from clapping with developers of Internet search engines to working with community leaders at local level to adopt new information. The key here is the notion of working with existing government resources. We might think about things like a home reference here as well as the rare diseases Web site might be models that we are talking about here. The other option, option B, is that HHS should endorse and insure sufficient funding for a web based information center and the rest is the same. The difference between these two choices is the first to work with existing resources and the second is recommending that the Secretary facilitate the development of a new freestanding web based information resource perhaps that builds and the gaps that the other ones don't bring the rest of the recommendation is the same.
Recommendation five. With a vast increase of scientific knowledge stemming from genetic into no research and new technologies and the increase in Teaneck to consumer genetic services, consumers of all literacy levels our talents to understand and use disinformation to make a proper health decisions. Review should support research the identify.
The defective for transmitting genetic engineer McMillan's into information that consumers and patients can use to make health decisions. HHS to also support research and then devise methods that patients communicate. Based on this research and to reach people in committees HHS to develop educational programs that use a wide awake of media, television, radio and mobile phones and provide for translation of materials into locally prominent languages. HHS should then support the dissemination of these programs. As part of this dissemination the Secretary of review should work with other relevant departments and agencies such as the Department of Education, National Science Foundation. This is recommending that there be research to identify models or the best methods for patient consumer education, patient and consumer education strategies and the best way to disseminate these programs.
Recommendation Number six. The background is about family health tools with the public as one means for individuals and families to gain help literacy and take a more active role in preventing and managing disease, particularly inherited conditions produce are a powerful asset for consumers and health professionals use this assessment and the they must be capable of accepting the information provided by the consumer oriented tools. Otherwise the value of family histories diminished as a factor. The recommendation is that people should support continue efforts to educate health care professionals, public health providers and consumers about the importance of family health history. Specifically for health professionals, HHS should support these family history and clinical cares to develop critical decision support tools and mechanisms to integrate pedigree's into electronic health records through the tools that we have been talking about and the EHRs bid for public health providers they should promote research identifying the will of women history and public health and how this family history of it into the population count for consumers, HHS should support research on how they use family history to make health-care decisions. For example, things like lifestyle changes. They should assess the effects of caffeine family histories with and diverse cultures and communities -- of getting family history. Expand public health awareness programs and materials on the importance of sharing family history and permission of primary-care providers. This is education again and promote the impending of educational materials and family history collection tools and ensure access for all by providing the schools in various formats using those as another educational venue for consumers. And the final recommendation, Number seven.
Given the reality that health care professionals and the professional societies representing them are likely to invest the resources and education and training and content areas for which services are only partially or not all reimbursable, a critical step of promoting increased knowledge of genetics in genomics among health care professionals is ensuring reimbursement for time spent in direct patient care that delivers genetic to gnomic services. They are calling attention to notion of time. And to increase incentives and to increase the willingness the secretary should insure reimbursements such as interpreting of tests, collecting family history. In sure their reimbursement for all members of of teams and distance consultation -- four members and for distance consultation. And act on the recommendations.
Good reading, huh?
The next steps are what we're doing right now to make a decision if this is ready for prime time. If it is it will go out for public comment. We will analyze those and report back in June with a final report. If it gets accepted at that point it will go to the Secretary in August. So, I know we need to talk about number one and number four. Since I have an urgency to settle that issue I have the microphone open so I will open that up first. Going back to recommendation one. Again, the issues. Two proposals presented by the task force are forming a multi disciplinary panel. Looking at cutting edge ways of thinking about education and translation and that panel would have with never a party secretary gives it. Another one is to form a workshop that is considered to be a single one time or couple daylong event that would come out with some things and we can open it up to any combination of that. I think that Maura was first.
I think he clarified it at the end. A panel is an ongoing even.
It tends to be, yes.
And the idea of potentially combining the two and starting with a workshop to kick off the issues to then better inform potential panelists going phone.
A risk would be if the workshop decides that getting it done in a day would not be enough but if the recommendation is simply for a workshop, that may be the risk of doing it that way. But the idea of blending the two, there is some -- good reason for that.
Does the committee have a recommendation? Did the Committee have a preference?
I think there was not 100% consensus pick the benefits of the panel is that it could be in greater depth of the benefit of the workshop is it might be something that the Secretary actually does but the panel may be, not one more panel.
I would go with the combined idea to start a workshop so it actually happens with the possibility of forming a panel thereafter and we get the best of both worlds.
Thank you.
I also wanted to endorse the notion of doing both. To assess and this may have already occurred but to assess the role the private sector plays in providing education. There has been a lot of focus on marketing and the negative aspects potentially of the private sector materials linked to marketing. But there is also an enormous amount of education material produced by the private-sector that is, in fact, a substantial part of educational activities now and needs to be thought about and In fact, I would strongly encourage it been a topic and representative of the activities being included in any ongoing review. The other point I wanted to make is one of personal experience which is at a community hospital that I am involved with we are trying to improve genetics' education for the medical providers at the hospital. And CME rules are interfering with our ability to get into the curriculum because of the rules of all parties, establishing priorities based on needs of the clientele. The fact is genetics is not viewed as a need at this point so some attention to those issues are also important.
Really quickly, yes, the whole notion of the perceived need is a definite barrier to education and should not be taken lightly. The idea of using new educational bottles as part of this number one recommendation and try to think about the new technologies.
This may be a naive question but if we want the combination of the workshop and a panel we cannot say that the workshop will be said that there needs to be a panel and then there is no reason to do the workshop. Is that correct?
I would imagine that we would get advice from staff on wording but it would be a hope that the workshop addressed the following issues and one of them would be the need for a panel or something, in multidisciplinary panel or something.
This is just to facilitate this be what I would recommend that we do given what I heard is to take option B and add an F to that that part of the charge would be to determine the need for and develop the charge for a panel to move forward with the issues identified by the workshops.
I agree.
And Joseph?
Just come to the table, Joseph.
Thank you very much. Actually, Dr. Williams beat me to the point that I was going to make. We had a discussion as part of our task force on this issue. We were pushing in the direction of the workshop calling for the charts to be developed. The challenge is that we want to look for something that is a low-cost opportunity.
Okay, I like our solution. I think we will go with it. Done a.
On a different notes one of the things that I worry about is that people that are not initiated in this will read it and see training and education all in terms of residency and medical school etc.. I know we see in the context of clinical care I just wonder if this is not wordsmithing too much to say, "and integrated with" clinical care because the only way is to integrate it with clinical care.
That is good to add that word.
Maura?
I completely agree with Jim's comments and what Paul had said. I was wondering if we -- It may be awkward at this point. I am intrigued by the area of domestic violence which has been a very important and key area for physicians to be the gatekeeper is to recognize domestic violence. Mike understanding is that after a workshop of sorts, it was a recommendation that it became a required piece CME education. It is probably premature but my understanding of the process, it happens in about five years and now bistate it differs somewhat but to Jim and Paul's point it integrates its and the affordable genome and putting a piece on Genetics and genomics as a required piece of CME. I know that that may be too much to put in the report as it stands now but I would ask the committee to think about it and for bringing up as a panel discussion I personally have written a couple of articles on this exact issue and in small groups of physician associations they were quite intrigued with that because it would put some rigor and National View so we would get in all committees a requirement not because one State Association position was interested that they get more information than others. There are some areas in the country with your academic centers potentially bad as one logic. To have less focused energy on this issue.
I think that domestic violence is a terrific example because it is not only as you mentioned freezing to the top in terms of CME and other health providers but is also making it into a required part of the medical chart. So it is translating from learning and that -- in your conference in Hawaii when you are many about continuing education to at your clinic have been added been similar to a vital sign. So it is that translation, you know that we're talking about of clinical education. It would be great if we could keep moving in that correction.
Okay. Number four is the other one where we just could not the side so we decided to let you all help us with this. This is the idea of community of consumer resources. The data from the survey, the literature and the entry fees have led the fact that consumers simply have too much information out there. They do not know what is credible. There are specific sites for one thing and if they need something must have to go to another site and pretty soon they're frustrated by it. A lot of those sites were developed a number of years ago when some of them are looking dated. Coupled with this is the very strong message that we heard is that consumers trust the government as a clearing house and gatekeeper peered one is to not throw the baby out with the bathwater. There are existing resources and the other is to develop or ask for the development of one that may be unique that may be in a little bit more forward-looking. So any thoughts on those? Again, this is going to the Secretary of HHS. That is very important to remember.
I think there is a real opportunity for a one-stop shop, if you will. That would be a novel resource. That is the more pragmatic perspective and one that philosophically we cannot try to do everything. It has to be cognizant of the of the resources out there and direct people to those resources as appropriate. But be a place where people can go to have a one-stop place with a can facilitate navigation and deal with some of the frustration. To someone interesting that the study results show that the public does, in fact, trust the government. There is not a lot of empirical evidence to support that point but be that as it may that is what they said.
Gwen?
If you look at it, the two places that people go is cancer.gov or cancer.org. I think there is a compromise here to develop a new portal in an existing system so you end up on this site and there is actually a portal that you can -- that has its own name that has its own URL so you get everything together and people are constantly trying to replicate what is out there and better at it without saying, while -- without integrating it into that place I do think there's actually a middle ground there.
I think I was seeing that you said in a much better. The idea of the portal. You can look at this as some of these newer search engines that are coming out with a are really trying to understand what is exactly that you're looking for. So rather than going to cancer.gov NC I cannot find what I need here they can go in and there could be some methodology with a sick, you know what based on what you told us here is the best resource. So the content does not have to be extensive but the thought process about how to interact with the consumer might be quite novel.
So part of the challenge here is, of course, communicating to a wide variety of offenses including the providers and consumers and additionally it has been bought even within the government all of these resources, I mean, NCI, cancer.gov and others the consumer is bombarded with a wider array of so-called information. It is hard to know what works and what does not work did so as a sentiment what we're doing right now is to develop this Jean -- genomic application to develop this sort of what you call an information resource that actually has the -- it is A. Burch will link that also has creeps that actually capture what we know. And for those of you who watch the Federal Register we just put out a RFA yesterday or the day before, cloning for the creation of reaching genomic knowledge center. It will work with EGAP, CAPNET. If it will try to distill a process of systematic reviews as well as click topic breves for particular applications and what we actually do not know and whether there are evidence based guidelines out there that can lead to the consumer decision-making process. So we have been thinking about these things for years and I welcome the opportunity to work with other agencies to see how to best implement an information resource that is both centralized and also virtually can link to other information resources so you can have one said that its the demand of everybody.
I would speak up in favor of trying to augment the existing resources. Kind of like remodeling an old house but is a good thing to do. And in particular there is so much material that has been developed that can be adapted for different audiences. I have been a participant in the past and helpers the issues we tried to look at how people learn and how to get to their level. I think adopting additional materials is a way to get there in a for a shorter time with far less work involved.
So I want to go back to the consumers and where they get the information. I think what we need to keep in mind is that this Wayne Gretzky analogy and and we need to educate people at the higher levels. What happened is my son goes to middle school, pulled years old are learning this stuff -- 12 years old are learning this stops so maybe at the workshop we should include some of these middle school -- not middle school kids, you know, but the education.
I think that is a really good point P1 of the other recommendations that we're currently looking at specifically indicates the need to connect with the Department of Education be you are right, if we begin it from too in education then we will have a genetically knowledgeable public and work force of 20 years from now.
You know, a genetics right now is hot. They are teaching them at the middle school level but who knows where it is going to go.
So I think this portal is like the congressionally mandated newborn screening clearing house that Jana talked about where it ranks thank you existing resources and I think that one of the things that we're doing is developing a system so that when people come in, they will say I am a parent and I had my baby in Hawaii so that filters the results so that Hawaii specific materials would come up first. So maybe I am a primary care physician and I'm looking for information about genome sequencing and then been able to have some of those results will turning like that.
I think that speaks to Marc's idea of the search engines that could be more specific.
I think to go on the issue of what cancer is getting in school I think is important to remember we did talk about chloroprene with the Department of education but has to go through -- about collaborating with the Department of Education but how many of us got As in algebra but cannot help teenagers to their elders -- do their algebra homework so it is important that it goes through a longer life span of education.
Right. Okay. What I hear is the notion of a portal that would have some of this decision-making abilities and help the person be more specific with the exception of David's comment of a recommendation to reifies with its existing. If we go with the portal method the idea and, of course, it would have links to those existing ones. Maybe it would improve those or update them or whatever. The way the recommendation is written, is an actionable to the Secretary of HHS? Can we picture what she might do in response to this? If we see that we would like a new portal developed that has all of these features.
Yes?
If I were the Secretariat I would immediately ask, what do we already have?
Okay. I agree.
Yeah. As usual I was going to say -- one of the things that we had our discussion around was to take a advantage of existing resources. What I heard was not a new portal but an add-on to the site. Modification of another site where there was one already existing and you could add. It just one more at on that will allow you to do this so it is not integration of a new one but just an add-on and use existing resources. That would be something of part of what we're trying to get at which would be actionable and you could use something that could be slightly modified if not use what already is in existence.
Marc?
If we look at the evidence that was generated I think you can make the case based on the studies that were done to say that yes, we know there are a lot of existing resources out there but they are clearly not meeting the needs because we're hearing from the public. So some of that is incumbent on what David is saying that we need to modify those existing resources and also argues for the fact that is not just those resources -- not that they're not designed as best as they could but people are having difficulty getting to them. I think the's idea is very compatible -- David's idea is a very compatible with the idea of having a one-stop shop. I think that those working together to improve the existing resources and then have, if you will, a surface layer on top of that that really helps get people to the right part. In the electronic health record environment this is the issue that we do with all the time. We have all of this information in the electronic data warehouse and people want to get at the information. If they are turned loose in there they will never find this so you say, what are you really looking for cracks so they can enter in plain language and go directly to with the need and it saves a lot of time. I think it is a very elegant approach.
Soap a one-stop shop to me means a unique portal. Okay.
Vince, I am going to ask if you have anything to add.
I echoed Dr. Williams' comment and some of the comments that we received from the interviews. We have a lot of resources that are great resources and great data but people do not know where to go out and meet identify some resource that can lead to other resources of that was the whole perspective about the development of a portal versus just enhancing the current resources. So my comments suggest's Dr. Williams -- just echo Dr. Williams.
I am feeling a consensus that suggests maybe because it's a little bolder or new thing is to suggest a development of this new portal. We risk it being dismissed as too ambitious. But I'm sort of feeling the tone in the room for that. Should I be corrected on that? We will get public comment as well and we can revisit this again. So let's go with the portal for now and see where we go for it. Those are my to pressing agendas but I would like to open it up if there are specific recommendations that we would like to talk about Premier doing alright could we have another half hour, I think.
And also if there are recommendations that should be included that are not.
I do not think you get lunch early just because we do not talk, though.
[ laughter ]
I will start. I thought it was a great report. I thought it was quite comprehensive and I thought that the recommendations as well as the report itself was actually remarkably easy to read and flows through it and did not feel terribly at -- sort of appropriate technical. Not quite sure it was the best page turner but it was good. It really got to the substance of the issues without for the most part getting too deep so I am happy with the recommendations as the stand.
It would have been a better page turner but Salinger died before we took advantage.
He did not rate for 30 years but he would -- write for 30 years, but he would for our report [ laughter ]
I'm sorry, are the recommendations in priority? We never voted.
Do you think that they should be?
I just do not know -- does the Secretary think recommendation one is most important? I am a logical person pick for me, when I get the report I think that recommendation one is the highest priority and recommendation ten would be the lowest priority so that is how I think but that is me.
That is a good point and certainly something that we could consider to see what is really resonating with the people. Part of the process would be to rethink the priorities of the recommendations.
Do you think that we should try today?
No.
Let the public weigh in.
Andrea?
I just wanted to move recommendation seven up, that is all.
Let me ask that question. There are none to places, reference to previous reports as recommendations. I do not know if it's a style or philosophical difference. I stomach one is to leave them freestanding or the other is to put the text in the preamble or go to some read that there are relevant reports that relate to this be what do we think is a better approach to take? Leave them as recommendations or take them out or put them in the text.
Kathy, can you move one slide back? That is the one that is that represented in the actual hand out. That is the oversight report. Those are the two reports. Okay.
Are people familiar with this one ? Okay. Kathy has the text, if you want it. Just pop up the hand. It looks like people are okay picketing of.
When I picked up the report again I looked for the recommendations and I had to turn back to page would ever to find them so I would put them in the front. I think the readers will like that and then they can see, why did you say that?
Yes, the executive summary is the very first page. It is not here in this draft. That is like a page of background and then the recommendations. Exactly the. What about keeping these references to previous reports as recommendations? What do we think? Sylvia is kind of nodding, okay. Done. So you can see that there is seven reports be we would try to be fairly equal on the ones that address the needs for the health care providers. Which are clinical providers, public health providers. Their educational needs be we tried to address the needs just of consumers. We tried to address the needs that education tries to help eliminate health disparities. That is one of the major missions of SACGHS and we are highlighting family history because that is an easy portal for Education predicted the cover what you would think, if you had to take away your big messages?
I think you have denigrate.
We do not need to beat the source to death, do we?
There is plenty of the team to be done and, please send your comments to Kathy -- of editing to be done and we have a couple weeks to make it is a little bit per year. And it will go up to public comment. And then we will revisit this in June.
So you will not see this again.
We will get your edits, any changes and I'm hearing the consensus.
Yeah.
So we will let the committee to the final adjustments and get it out and hopefully we will be in good shape to review in June and get it finalized. So I think the consensus is testimony to the fine work that you and your colleagues have done on this so many thanks. Great. Now we can move it forward.
All right. We will get a that a bit of a jump on our public comments which is a good thing to allow plenty of time to hear from the public. We do this at all of our meetings in BP sheep the input that we to get. and we appreciate all of the input that we do get. I do not know all of our speakers. Let's begin with Marc, are you here? He is speaking on behalf of the Association of pathology chairs I reminded the committee that the written testimony from all of the folks is in your table folder. Thank you for coming and I look forward to what you have to state.
Good morning I am representing the Association for molecular pathology which is a nonprofit medical association representing approximately 1800 positions, a doctoral scientists and medical technologist to perform a laboratory testing based on knowledge derived from biology, genetics and genomics. We have long been concerned that the U.S. Patent and trademark Office has granted broad patents on genomic discovery's including individual genes or mutations and untended consequences has been the patent holders and their exclusive licensees have frequently chosen to monopolize public attesting by restricting other health-care providers and facilities from developing and improving tests covered by these patents and licenses. We believe that this in many cases are sticks access to health care and in more extreme cases may even in danger patients. So we strongly endorsed the report on patents and licensing processes and their impact on patient access to a genetic test be we commend the committee for challenging DNA patents and extending in taking steps to limit or eliminate exclusive licensing practices. If implemented the committee's recommendations would be a significant step forward to reverse years of policy that has hindered innovation, restricted patient access to test and constrained the widespread clinical application of biomedical research. We urge the committee to penalize unchanged the recommendations presented last October and to encourage the secretary and a administration to act swiftly to implement them in their entirety. The committee reached these conclusions after more than three years of careful analysis, sufficient public comment and the colder engagement. And the report, even as released in draft last year, was written after the completion of the study initiated by the committee in 2006 to assess the positive and negative impact of licensing practices on patient access to genetic tests. We believe the research was the role we played by the full committee with many opportunity for public comment and has led to a well researched and documented reports. We agreed that attaching intellectual property rights to tracks of innovations such as New their critics, a diagnostics or technology platforms is essential to encourage investment and a word innovation. A single gene or a sequence of the genome is the only a product of nature but contains information that should be not be palatable. Threats of enforcement from a patent holder and the ensuing litigation costs lead to a chilling effect on the ability of genetic testing that could otherwise directly benefit patients bids and clinical laboratories are reluctant to develop new tests under the current restrictive environment we urge the committee to move expeditiously to finalize the report as presented last a October so that these much needed recommendations can be put into practice. Thank you.
Thank you. My apologies. I have you down as misrepresented with your affiliation so I apologize.
I also have comments for two other societies but AMP is the lead organization could would you like me to continue with those or continue later?
I do not know which organizations we have as well.
He is to steer as -- as support for AMP.
I have a joint statement.
So is that in addition to what you said on behalf of AMP?
Yes. They have separate comments in support of AMP's position.
Why don't you go ahead and tell us what they have to say.
I will accept, you have the written comments in your folder.
We represent the academic departments that are accredited in North America and represents 145 institutions and the American Society for investigative pathology representing 2000 members to promote the discovery advancement and dissemination of basic in transitional knowledge and experimental pathology and related disciplines. We support the AMP report and AMP's comments on the report and we support the extension of patient care givers from infringement liability stemming from patent claims on genes including anyone making, using, offering for sale or selling a test developed under the patent for patient care or in the pursuit of research. In addition, we call for enhance transparency in licensing activities, public access to the permission of a licensing actions and the adoption of efforts to promote a broad licensing practices P we view these recommendations as a call for actions to protect all patients from the detrimental effects of the gene patents and exclusive licensing practices but we support the following recommendations and will promote better access and quality of innovative molecular testing services. The patent to a single gene in biological station be discontinued either as an insult of traditionally spew or a act of congress. as a result of judicial review or an act of congress. To ensure that access to innovative molecular tests remains widely available and affordable to patients' financial terms should be reasonable, license agreements should also be free of any terms that limit the number of tests that can be performed by a laboratory. The percolating the technical performance or clinical uses of the test should not be lessened support professionals and technical performance and profit clinical use. License agreement should likewise be free of terms with in a corporate research related to testing or the public dissemination of the resulting research finding it.
Positions researchers, patient advocates, government officials, research funding agencies and other stakeholders should work corporate week to develop alternative models. These innovative model should increase patient access to health care and achieve benefits to the existing body of intellectual property.
At [ Captioners
Thank you, our next speaker -- the united cerebral palsy, look forward to what you have to say.
I am Maureen Fitzgerald. The collaboration is a partnership of the ARA of the United States and united cerebral palsy. Each represents persons with disabilities for over 60 years. Our comments today about the discrimination act. People with disabilities have experienced a long history of discrimination. Now they have something to look forward to, also something to be worried about. The wellness programs, term manifested, and filing a complaint under GINA. The disability organization was and is a strong supporter of Gina, through the public comments process we have commended the agencies who have written strong regulations governing the implementation of GINA. We are especially appreciative of the strong protections for wellness programs, and for health risk assessments. Wellness programs can be a real important part of an employment setting, as long as they don't discriminate against people because of a disability or risk of disability.
From the perspective of the disability community, I am not aware of any significant problems under GINA, yet, but I am aware of some confusion, and I think it has to do with what people perceive GINA actually does. The terms manifested and manifestation are clear to you, but are very difficult for a lay person to understand the subtle tees in GINA. Downs syndrome is -- based on genetic information, when a health insurer denies that family corchl or charges exorbitant rates because of the person with Downs syndrome, they feel they have experienced experience discrimination under Gina. The term [indiscernible] isn't a common noacialgz. Finally, underrifyinging a complaint, should be readily available, how the process works, the detail of trying to go online, figure out how to file a title 1 discrimination complaint under GINA. I spent quite a bit of time trying to figure it out and I couldn't. That should be something fairly available to people.
And [indiscernible] applauds GINA, we ask this advisory committee continue in its leadership, education and the issues that will challenge us all in the future. Thank you very much.
Thank you very much.
As you know this has been a topic we have taken up here in the committee.
Do you have --
For example, the explanation, the individual with do you Downs Syndrome, health insurance companies can't discriminate because of a genetic condition, when the family can't find coverage, theyy feel they have been discriminated, what's discrimination, what's a manifestation, where the breakdown comes, is this clear? The whole discussion about the recommendations, education, is so critical to people. A lot of the folks I am concerned about are not sophisticated, don't understand medical language, so being real clear about what GINA does and does not do would be part of the education process.
Obviously we are hoping some of the healthcare reform issues will deal with some of that. We will keep our fingers crossed.
Our next speakerrer is Joanne -- here representing American society of human genetics.
Thank you very much. I am the executive vice-president for the American society of human genetics, which is a very diverse genetics organization of over 6000 members. We represent communities that perform basic research all the way to clinicians that see patients. So in fact, achieving the consensus of statement that our board could in fact endorse heartly has been a challenge, but one that in fact, the process of developing this, I think, was an education to many of our members. Some of our members do not understand the patent and licensing process really at all and others are very immersed in it. At this point the leadership of ASHG applauds this group for the enormous amount of work expended to produce the report on gene patents. The recommendations made are in essence, consistent with the ASHG principles relative to intellectual property and genetics, specifically the [indiscernible] community supports the key principles quality, assurance, insurance accessibility in the genetic testing arena. In the past and continuing action, the board of the American society of human genetics has taken steps to support lawsuits and positions involving intellectual property and our position had has usually been manifest a party to amicus briefs rather than serving as plaintiffs. The board is of the view the genetics community must continue to make it clear that exclusivity may result in issues around cost, quality of testing and patient care. As noted by your committee, the current IP environment may play an important role in relationship to these issues.
Our scientists must comply with their own institutional regulations regarding all disclosure of findings or inventions that might be commercialized. However, and this is one of the areas we are trying to inform our members more fully on, they must also understand their obligations related to beyond disclosure and protection of intellectual property the responsibility, as well as the degree of authority they individually have in determining the terms of any licensing agreements made based on their own intellectual property as disclosed to their institutions. The recent recommendations of this committee suggest there are serious issues around access and the quality of testing. Both of these concerns are of primary importance to the genetic community and the board of direct ors of ASGHC strongly recommends the guidelines that address these issues. We all know there are incredible and continuing challenges associated with efforts to change patent legislation and policy, including the interpretation of exclusion clauses for research and testing protocols.
However, given the rapid evolution, El advance relevance of technology and [indiscernible] members of our organization, a consideration of such policy change seems absolutely essential. Indeed the technology in the field of genetics and the application of testing for human health are moving extremely rapidly as Dr. Dr. Green stated earlier, toward the trend of increasingly complex and complete genomic data driven by the efficiencies of the whole genome and whole [indiscernible] approach. The advent of comprehensive DNA and [indiscernible] sequence alters profoundly the implications of restricting to any particular locus or the [indiscernible] of that locus. We in the scientific community are striving to fully understand the impact of the current legal and regulatory framework while in fact we in our labs are forging ahead in the development and imploament plement implementation of full genomic -- discussing the important issues, following to response to activities that may change the landscape, comment further on policy implications when appropriate and I would add continue to inform and educate our own members.
Thank you very much. We appreciate that.
Our final speaker who signed up is Jeff Boyd. Mr. Boyd is with Medical Device Consultants of [indiscernible]
Thank you. I would like to thank the committee for the opportunity to speak today. I have spoken before you in public K0789 comments back in June as relates to the issue of clinical utility. It's good to see this moving forward. I suggest, however, clinical utility, one issue that's important, as well as evidence is another issue that needs to be dealt with. There are two separate issues that really need to be addressed and sometimes they get lumped together. I have have prepared comments and I will be as brief as possible. I had the opportunity to participate in the January 2010 MedCAC meeting on pharmacoE conommics, pharmacogenetic testing and the panel, what they thought the most important take-away was from the meeting. The vast majority stated clip clin cute for these tests were the most important, the understanding of clinical utility was very different to almost every person on the panel. They had totally different definitions of what it meant. It was all over the map, ranged from a change in patient management occurring from the result of a test to the benefits A A accrued to the patient, improved survival based on therapies providedded. As well, the panel started talking about another issue. They started talking about let's gather evidence, enough evidence to prove out clinical utility. They started talking about Cadillac evidence and about different end points that needed to be looked at, as well as the types of studies they needed to engage in, ranging from registry to perspective randomized trial. Those particular issues, obviously, have ramifications for people who do these tests or actually are involved in developing the tests, and can ratchet up the time, effort, money associated with proving out clinical utility and going down a path of developing evidence. So, it's unfortunate that the issue of clinical utility has really not been adequately addressed by policy makers, and since there is no clear definition of what clinical utility means, many policy makers are taking it upon themselves, especially those at payers, and I happen to work with a number of the prief payer and Medicare, they have taken it on themselves to does he fine clinical utility with many defaulting to the most conservative definition, which typically means improvedded patient outcomes in some form or another, as well as Cadillac evidence looking at perspective randomized trials. These can be very onerous to answer those end points, and sometimes may be really unfeasible to be able to answer the question.
Frankly, no one should blame them for coming up with that, definition because they don't have a definition. But, the problem is, it's a one size fits all mentality, which is really -- it's seen in the issue of evidence based medicine, come is characterized by the value of treatments which has also resulted, obviously in a one size fits all assessment. A one-size-fits-all perspective ignores the technology type, its applications, intended use and other practical factors involved in evidence development. This is further reflected in the criteria developed by Blue Cross Blue Shield for tech assessment, for useful for [indiscernible] diagnostic tests.
As mentioned, not only is the issue of clinical utility important, but the quality of evidence, the study design is also very important. Dr. T hit on that with the technical factors involved in putting evidence together, where and when the test is used, what happens besides health out comes, those things need to be considered and sometimes can be ignored by payers when they are looking for this type of Cadillac evidence. It's encouraging to see that the clinical utility task force is moving forward with a roadmap. I highly encourage that to be facilitated as quickly as possible. Without this definition of clinical utility in looking at the evidence, the concern is that payers will continue to fall back on the most conservative view of what it means for appropriate level of evidence.
Jeff, we have your comments which are great. Can you come to -- wrap up with a few of the other final thoughts?
Yes, I am going to do that. So it relates to -- the definition ultimately arrived at with evidence gathering and CMS has been at the forefront of this, put together such tools as coverage with evidence development, very important, but the process as defined right now can still be very onerous for people to really meet those particular criteria. In the way it's set up coverage development is a -- first you have go through the NCD, then CMS basically says okay, it's not covered but we will potentially go through coverage with [indiscernible] development. The problem is that I think a lot of companies are really reluctant to want to go through the process because the end result is you end up with a noncoverage determination, in turn picked up by private payers and kind of a roll of the dice, especially if they do not know whether or not coverage with evidence development is remotely available. A couple of suggestions for payers like CMS is for them to be more of an Olmstead buds man in the process and facilitate technologies they deem to be be clinically useful to them, help push those technologies through the process a bit faster, rather than having to wait a long time to engage with coverage development. Also, become more transparent in the process, especially with the public as they are going through this. I would also encourage private payers, they pay 160 million across the United States, half of what CMS covers, yet CMS is doing a lot of the heavy lifting. It would be extremely helpful if they were encouraged, especially by this group, to participate in the process of more flexible coverage policies, like coverage with evidence development.
Thank you, I appreciate that. Do we have any other individuals who would like to make public comment?
If not, then I think we have come to that point in the program where we get some lunch. I know some of you ordered sandwiches, there are also places out on Connecticut Avenue, why don't we plan to meet at 1:00, it's 15 minutes than it says on your schedule, but still gives you a little over an hour. We will plan to meet back here at 1:00 and take up the session on Janelle -- did you say something?
1:00 p.m. eastern time. See you back then.
[Recess until 1:00 p.m. eastern time] > >
All right, so folks listen carefully. Our agenda changes from moment to moment and I would like to do a quick can this be how many people on the committee have flights out of here or have to leave before 11:00 tomorrow?
Leave the meeting before 11:00. That is what I figured. So we will not have a forum midmorning. Here's the plan. We are going to extend the session this afternoon and listen to some of the presentations primarily from our federal colleagues. I do not know how long it will be but it looks like it will be and our beyond the scheduled time. We will start doing the patents report at 7:30 a.m. tomorrow. I am sorry, Maura. Fortunately you are an Eastern time.
I apologize we have got to did it down because after 930 we risk losing the forum. So we will start at 7:30 a.m. and we will have the patents report and first thing in the morning. We're trying to work on the public comments and are in the process of trying to reach those two individuals to figure out what the best plan is going forward so that we are the beneficiaries of their input so we will extend this at least one hour. I know that Sarah has been working with all of you to see who can stay. I appreciate everybody's flexibility. I know it is a problem. This looks like a few are not out of here by mid afternoon tomorrow you are here for the weekend. I know it is a lovely place. So the order of business for today -- Allison wants to know how many people are coming to dinner. How many people are planning to go to dinner tonight? It will be at 7:30 p.m. over at the Tea Fleur or somewhere. The main reason we are here is because they have been working very hard on genome data sharing and the challenges of the clinical data research interface are becoming blurred and we raced this to a high level of concern and we have heard other issues today about how one does this with confidentiality as well as human subjects so the materials are and have -- are in have. So Charmaine Royal will introduce in this question -- will introduce this session and we have a very tight schedule so I will turn over to you.
First I must say thanks to Sarah and Kathy because they have been working very hard on it.
I will give a brief overview because we have a wonderful lineup of speakers this afternoon. I want to remind us why we are even talking about this in sharing of genomic data has the potential to have all kinds of ethical implications that are associated with it. Another issue that is raised is the potential blurring. We are not saying that this is being caused by genomic data sharing, certainly the blurring of a line has been happening for a while and the question is whether genomic the Schering will increase this even more and then the questions about informed consent and will we need to think about new approaches as we move ahead with widespread genomic data sharing. So what have we been doing so far? In December 2008 it was decided by this group that genomic data sharing was an area of priority. In March of 2009 there were briefings on the IOM report wall. And in September last year they got a contract to draft a report on genomic data sharing and work along with SACGHS to do that. The project is a yearlong project. And in our meeting in October we discussed this and we formed eight -- formed a steering group and we met and talked about what we would be doing today. In our meeting in October we did the said that it would be great to have a session and on our conference call we fine-tuned that in terms of what shape it was cleaned to -- it was going to take and the Lewin Group has done some background work and what this will explore is whether there are some new issues, issues about consent, how can the process be improved? Whether the benefits and risks of population based registries and how can we address the issues related to indigenous groups and special populations that participate in this research so what will we do today? We will have a group of speakers that will talk to us about the models of genomic data sharing. We will gather some information from them and figure out the next thing that we're going to do is talk about the information that they give us speed what issues are raised and try to think about where we might go with this one. The presentations, we will start out with an overview by Laura Rodriquez and then Joyce Mitchell pitted than we have speakers that will talk from different sectors on consumer control policies and genomic data sharing. As we listen to those talks some things we want to think about it come as a listen to the models that are presented think about what are the implications for informed consent and what are the things that are common and what are the things that are different in terms of consent and the storage of data and issues regarding access. Privacy, confidentiality, protection in terms of reidentification and deidentification of to -- of genomic data. At the end of the talk we will have a discussion. The key things that we will be focusing on are what the elements that have worked well in these models and what are the ones that have not? And are there issues that could benefit from more policy discussion and development? And then we will try to think about what the next steps should be. Will there be a need for us after we hear all of this information? Will there be a need to identify a best practices. Could SACGHS contribute to this or should we wait until the report is done? The report is generating some information about what is happening out there. Should we wait until the report is complete before we decide what we should do? And in the interim should we plans and individual sessions to try to explore this a little bit more. We will go right ahead and move into the discussions because we will run a tight ship this afternoon so Laura, please come and give us a talk about the federal policy.
Okay. I would like to think the Committee for having me come to speak today on behalf of everyone and clarify quickly here that I'm just reporting on what is going on from the different people and the am happy to see them sitting around the table so they can answer questions that you might have as it goes through this conversation Charmaine has gone over some of the goals of the committee and some of the rationale on why they want to look at genomic data sharing was the potential for this kind of data sharing to facilitate important research and things going for a word at the moment in a very rapidly and in the way that raised questions that we wanted to be very deliberate about as we handle them as they moved. Additionally, too something that Steve mentioned already is the fact that the is kind of data but are blurring the line between Research findings and critical care so that something also we wanted to think carefully about and in doing so the number of ethical questions that are raised by not only the potential applications but also how we manage them and what protection to put into place for the individuals whose data we are looking at that is generated in large volumes around some different and new areas and the kinds of data that we are gathering. And also again something that Charmaine mentioned is the fact that genomic data by its nature is challenging the traditional paradigms and how was that going to change the way we need to think about managing the data in terms of providing appropriate balance between wanting the research ago Ford and also maintaining and protecting the interest of the participants from whom the state that are derived -- these data are derived. What he wanted to do in hearing from this area was largely because of the amount of money that the federal government is putting into this and also, in fact, in building resources for the sharing going forward. And in doing this the government is not only playing a role in the research as a founder but also providing some leadership to the community and thinking about how to go into these new domains and hopefully after today you have in a little bit better sense of what kinds of things that the different agencies are thinking about as the air during this period -- they are doing this. So the survey included ten questions that focused on the issues here. Trying to find out what research programs in any existed in the various ex officio agencies and if they did have research programs or did not did the seed genomic data sharing as chief leading to the agency mission in any way -- did they see genomic data sharing as leading the agency mission and how did they develop policies to try to implement these expectations and how did they incorporate elements into these policies. And then finally again going back to the process that these data are planning a line between research and clinical information. Was there any allowance or expectation with in the policies to provide in our connectivity between the research data and electronic health records? So this survey was sent to all of the ex officios as well as separately to USDA and NSF to find out what kinds of policies they were putting forward. We have responses from 12 of the ex officio plus NSF. So here are four of the groups that had node to no Data sharing activities and they did not see it relevant to their mission at all. I think from looking at who these different groups are it is not surprising in some regards how they were doing it [ laughter ] I am just reporting.
There were also several others that reported that they did not have any genomic data sharing activities but they did see it as being relevant to the agency mission in some way and again this is OHRP and they have some policies that overlap in the realm of gender research protections and then the specific considerations are around how genomic data sharing is done. OCR with their involvement in HIPAA and there were five others that did have genomic data sharing activities and research programs in the did see it as directly relevant and these were the ones that were more research based and will be the dominant activity for what they did. Looking at those five they are very different. Not surprisingly the largely deal with Plant genomic activities in that was not something that was close to this committee but even this looking at this you can see there were still very different states in terms of their thinking and activities in this regard as they were approaching it. So the VA at this point, they have a aspect with data sharing but as they are based, the sharing is the info on system. The CDC has several different programs and policies vary among those programs but they tend to be based on how they have set up their traditional sharing and how they have interpreted looking at the urging of McTeague and systems that they have. They tend to work -- looking at the genomic data and the systems that they had paid aid to research at the sites. And then, of course, the NIH has invested significant time and energy in Building database repositories and having the broad Chang take place in a way that is much more in direct as for center resources I have tried to provide links for many of these programs Killeen Ford and just to highlight one, the NHANES Program they did a genomic collection in the hosted several meetings and workshops to think about that going for and what would be appropriate with that particular structure for NHANES whether it is actually legislative language that structures how they can go forth and how they share their data. The VA has a large genetic medicine program that they have been moving forward and they formed and buys a committee in 2006 that meets an immediate basis to think about the different questions regarding how to appropriately share genomic data and how that relates to the participant population that they have at the VA and also asking how zero of those veterans felt about different aspects of data sharing and what they would want to get out of the in terms of return to results and those kinds of things. So the policies themselves at the VA four-cylinder development and they have been very active as serving as a forum -- are still under development and they have been very active ascertain as a former. NIH has been very active in putting together policies pick one that I will talk about this afternoon focused around Chino my dissertation studies and we also have specific policies for other -- genomone Association studies and we also have specific policies and what their expectations are a sharing of genomic data at the IC level. We have been very politic of putting together different ideas of how they apply to the specific programs and what our expectations are for genomic data sharing. We have worked very hard at doing that but, of course, we're still working on that. Coming back to the agencies with they had policies or areas that did not directly involve genomic data sharing we look at OHRP and coated specimens that pertain directly as to how the genomic data sharing in terms of it is a human subject research and the regulatory implications then of that type of determination, engagement in human subjects research. Clearly these are very relevant to what is going on in the research programs themselves. Several of the other ex officio agencies, the EEOC, OCR and OHRP noticed that they have overlap in this area. And then lastly I really just wanted to mention that NIH is now going forward and extending our existing policies to putting together a trans NIH policy. We see this as extending what we have done in the past for all of these different individual project by project policy development activities but Pope the will provide a way that will be consistent for investigators, institutions and those investigators that led to use the data as well as those in need the data as well as the public. So there is a common expectation as to what the NIH is to lean in been a steward of all of this data that comes into our resources.
The themes that came forward from looking at all of the information of the agencies that responded is clearly that genomic data coming from many individuals will include sensitive information. And also that brought sharing of the data does unable and a celebration, the potential for some acceleration of scientific research. So with those two principles accepted concept is that policies must insure privacy and confidentiality of research subjects did even with those agencies that said they had no activities they were still concerned about the ethics of this kind of activity going forward. If that protection mechanisms were needed against unauthorized access and there needs to be careful attention to the use of data. And, of course, that the issues regarding their management for their distribution and their collection needed to be very carefully considered and that they must remain relevant and timely to the technologies that are used as well as the public conversation taking place around this type of information and how we reducing it with an research and with inciting more broadly. Potential gaps that are identified in the survey. Informed consent was the most frequently mentioned place for more guidance was needed or Best practices were needed. Again, not surprising I do not think but it was again something that was mentioned even by those groups that did not have any activity in the area. Something that is really permeating the discussions. Additional considerations around what accessed participants may have to the deed itself either through the databases or to resell to prevent their own results -- or to results for their own results in the situation. It is a fairly hot the contested issues with the opinions on both sides. There is the recognition that Wyatt policies at the moment did not preclude incorporation, there really are not at the moment any clear structures that will make the inclusion of the data into EHRs something that is easy to see will happen so more attention to that area was another place identified as a gap. And with that I think I will come back to some of the same questions that Charmaine put up again in terms of questions for you all to consider whether there is the need for additional policies for genomic data sharing in this area and from a federal perspective does the committee have thoughts about whether an agency specific initiative or the way in Chicago or if it should be coordinated in some way. And also is there a need to try and deliberately waste public awareness around the importance of sharing genomic data and the inclusion of these kinds of the debt and their electronic help record and a targeted way. With that I will thank Symma, Cathy and Sarah for all of the work that they did and I will take any questions.
This is a question for you but probably more broadly to Charmaine in terms of the task force. How much time was spent not looking so much debt issues of privacy and confidentiality but actually physical aspects of sharing the data like standards across. Is that in scope, out of scope of the task force? Is that something that you just in your surveys to the various groups?
I can say that it came up minimally in the survey and the questions were not structured to drop it out but I think only one or so of the answers that I saw come back and mentioned the standards for that kind of thing with regard to the scope of the task force that is definitely a question for Charmaine.
I think we sort of put that on under the HIT umbrella.
Any other questions?
I will raise a question. In the data sharing area he talked about a shearing with in the government agencies but when the NIH or government agency sponsors a study then it is really governed by the IRB, usually the local IRB might issue is a huge differences between IRBs and different feelings about data sharing. No one has corralled the wild horses in a way and it is very confusing if you are a researcher in this area. You spend a huge amount of time trying to share data.
I will agree with all of the statements. I am not sure what to do in terms of solving as I will talk about more in my talk later about the policy that NIH developed while the decisions for whether or not they'd sharing is a profit still reside with the local institutions. And we have tried to put forward an infrastructure for some persistent protections to be put in place and mechanisms for how the data are shared and what the considerations are in making considerations about sharing data to try to bring in a little bit more easy to the process of doing it both for investigators again trying to access the data and those submitting the data. Ideally we have tried to provide some help to the community and thinking about these issues but obviously this is moving very quickly and we are not anywhere near a consensus on how to do it.
Can I pass along the same lines, are there standards or guidelines across to various agencies that each of the institutions at least NIH or NCI grantees have to use in any of these key issues such as informed consent?
Across NIH, that is -- there is a consistent threshold that is supposed to be used to speed we are a very large agency is sort of interpreting the policy is somewhat subjective. We have done a lot of and trying to develop rubrics and SOPs to try to bring them up to a consistent level but that is taking time as we are all learning to go about this and everybody has different ways of doing things. So ideally there is consistency with an NIH but I am sure it is not perfect. And in terms of other agencies I think every agency is trying to do this on their own and there are some general consistencies and principles and I think how each of us are deciding to do it is evolving.
Clarification and a comment. AHRP does not have any activity on genomic data sharing but we do think its relevant to our agency but going to the point that David had raised this issue has been discussed in terms, of course, mean the difference IRBs and we have a consortium of research in different centers and there are differing policies. How do we coordinate this? There has been some talk about coming up with new policies for multiple IRBs or blanket IRB but I do not know if that has led to any conclusion or activity within the NIH. We're certainly discussing that within AHRQ involved in this area.
We are discussing it, too but I would not see that we are at a conclusive point.
Thank you, Laura now we will have Joyce Mitchell for HIT.
Thank you. I am assuming that you will get it so it shows up here? Thank you PI am delighted to be here to talk about existing and emerging technologies involving genomic data sharing. It is a very large topic to cover in a short period of time and the tactic I have taken is to be more broad in terms of general areas and terms that I see emerging. This group clearly knows a lot about genomic but I felt it was. Useful to give a broad overview there has been huge progress made in the last decade or two decade certainly. And certainly the broad overview of the whole thing is here at the bottom there are 5,000 genomes available online in and public data repositories are routine and that is different than the situation would have been ten or 20 years ago. There are lots of genetic tests that are available today to anybody in the physician community who wishes to order them. There are almost 1900 routine studies expanding clinical care today and next generation sequencing has arrived and has taken all of our small-scale single gene experiments into something enormous as to try to do with the average. The final of the human sequence from three to 4.5 million HITs and you end up with 10% larger than that. So it is a fairly large data problem.
Yes?
G2P is what?
Dino type to phenotype -- genotype to phenotype wife.
Select the same time all of that information is out there then consumer demand for Genetics is exploding. I take you first to the genetics home reference. This is a site that I have a particular -- it is particularly dear to my heart. I was a senior scientific advisor on this site from 2001 to 2009. It was the first sight that actually targeted the public and said that the public would actually like to know how to bridge their consumer health questions with the buyout informatics data coming out of the genome experience -- bioinformatics data coming of the genome experiment. So here is the web site that you can go to explore at your leisure, it is part of the National Library of Medicine and currently has about 500 health conditions and about 700 curated gene summaries and it has 215 million hits per year but it is never advertised. That is 215 million hits per year from the public and clinicians to go there a lot with the public hates them with questions about diseases and disorders that they do not deal with on a routine basis.
And then of course, the interesting phenomenon of direct to consumer genetic testing. I note his controversial in that it understand the issues behind that but he's a huge force and it is their aunt -- happening daily and doing it in some interesting ways. First of all, it ends up in the fashion and style section instead of the scientific section in the middle of the New York Times. But there are three major companies and a lot of other companies that deal in direct to consumer testing and this one is 23&me. I am sure you have all seen it before. This is a clinical report and there are other research reports. Let me just show you very briefly, cystic fibrosis is the example of carrier trait and here is the example of eight drug response -- -- or a drug response. It tells you a lot and this is data sharing, for the person who paid for the test cane them complete information on the test. It has not only the 23&me name and has other names. The most common variants certainly you would be looking for an cystic fibrosis. Down here at the bottom it does not have any of 31 CFDR mutations. It tells you where you could find out more about the gene. It most likely has no disease, it may still be a carrier due to other mutations and I would see that is a pretty sophisticated kind of report and a direct sharing of the debt in that particular regard. And here's another one with the action until you some clinical information for pharmacogenetics. This particular person is a homozygous for that allele. Here is the results, increased sensitivity. So there are lots of folks who were dealing with patients who have questions about all of this burden, on the other hand, I have said before and I see again it is out there and it is hours to deal with as the profession and there is a lot of data sharing going on with this and a lot of curiosity and a willingness to get these results and investigate war. The most telling thing about the data sharing is that you can actually download your entire set of results and go investigate them yourselves if you're so inclined. Large files and you have to do some leaning in order to figure out how to deal with it. But that is what the public is doing at this point. Now let's talk about genetics/genomics and the EMR. These single gene tests for specifics syndromes and mutations better known to cause disease or be associated with diseases. This takes you to 2000 feet and if you go to a 2009 that is where you get the 1900 of these tests available and the purple is the laboratories. 600 laboratories around the world. In addition to those single gene tests there are a number of single gene expression tests that are growing rapidly. I give you two examples. The first one is the mammoprint that is used to do a gene expression profile on the tumor. It is for prognostic purposes. In 2007 the FDA cleared for marketing this test. It determines the likelihood of breast cancer returning within five to ten years and the first cleared a molecular test profiling genetic activity so too is aging expression microarray test and you can send off a sample and get the results back. It is used routinely in some places. Here is another example. It is called AlloMap. It is a molecular expression profiling for heart transplant patient management. This is any lack in gene profile -- 11 gene profile for that is used to alleviate were pretty associated with intracardiac biopsies. So you have to going regularly to get an evaluation to see whether or not you are rejecting your transplant or not and the standard we for doing that it is a cardiac biopsy with. It is someone in the city in my mind I have heard surgeons say that they are routine -- evasive in my mind but I heard surgeons say that they are routine did you have to go in once a week. What they're doing now with the AlloMap is to take a blood sample and run your leucocytes through a expression analysis, microarray and if you are not rejecting you do not need the biopsy but if you have signs of rejection then you need the biopsy and further tests. So those are numb to examples that are coming fast -- two examples that are coming fast. So you have tests been done in all of these laboratories throughout the world and private laboratories. You have a lot of test interpretation which are faxed back. Tests are not stored in a structured form. Not generally available for decision support. If your own laboratory does the test you have a much better chance of making that happen. The interpretation does not give too many details. A MammaPrint does that tell you the expression of the 70 genes, it gives you an interpretation overall and clinicians are struggling to explain these tests. I would suggest that the rest of us are trying to figure out how to interpret them as well and explain them to the patients. And the business models of most of the laboratories during the testing include the gene patents four of these expression profile test speed they do not have a business model that promotes data sharing. They make money on doing the test and not of ensuring that the debt and compare and contrast this with the consumer data sharing policy where we gave you the test, we gave you the rat data, we keep the interpretation and it is yours. So comparing that is a fairly major deal. For electronic medical records it has got implications for all of the component systems. Assertedly at the laboratory exams are the ones that impacted first and foremost but these ones as well. One of the big things in standards is messaging and vocabulary standards. There is HL7 standard and a genomics standard that has been approved that has started to be used for testing and that is a big step. Here is a screen shop of intermountain Healthcare. Dr. Williams was involved with that along the week, sharing genetic data using the HL7 chronicle of genomics standard. At the same time it is and all systems. It is certainly represented in some form in newborn screening and the Department of Defence requires DNA samples of all new recruits and the identification of World Trade Center victims was a hallmark in the tools methodology and techniques by which you could identify people based on small bits of tissue which you find after a bombing or Trade Center collapse and those techniques are used on a routine basis daily throughout the world with the suicide bombings and terrorist attacks that have been. Later on you will hear use of this to no data and looking at the affected each and identification and the origin and spread. This year H1N1 is big. The data is clearly there and it is not necessarily represented in a way which standardized yet. If there are definitely strategic information issues which have not been solved. If they are being discussed. How to represent this data in electronic medical records is a large question. There are some systems that to that already. I talked about the helix with molecular biology subsystem. How to send a structured genetic data between systems is still being worked out of the zero the HL7 clinical genomics standard has started to solve that problem. How do you make this understandable to both providers and patients? It keeps emerging as more and more information comes along appeared it is not a settled issue and we've learned more all the time so that is an ongoing issue to be dealt with. And then how did you keep all of this knowledge up-to-date is a problem for all of us. This is all emerging and what are the implications for health care and providers and patients and how is it that you notify people of the appropriate information and in all places in the world. He can have, once again, some examples. Here is an example of a genetic test that can be represented. These are examples in single cases and not generally available throughout the world what is coming, all of this stuff is coming but you cannot just settle in and think that you can deal with what is there today when tomorrow the world changes. Certainly next generation sequences is interesting here and now and the buyer mental variables have to be core needed in order to figure out what is the appropriate interpretation on many genetic tests. I think one that is interesting is the microbiome and I need to know what not only is my DNA but I need to know my critters DNA to learn how out fight metabolize food and react to different situations and all of these things are coming in and in some way or another will be part of our electronic medical record as we go forward. HIT standards, of course, are hot news today. There was a technology standard panel established in 2005. A public-private partnership and Maybelline the Bush first and now Obama's vision of electronic help record sharing -- help record sharing in things are going along rather rapidly. There is an interim final rule that was issued the last date of 2009 and goes into effect next week. It is a final rule so it goes into effect at the same time to still being discussed and altered as things go forward. If the things are obviously happening at a national level. I was safe and effective data sharing requires standards for data and transmission and all of that is emerging in the genomics world. There are standards that are being discussed and developed, the clinical genomics standard is one appeared you have a CDA clinical architecture that is part of the RIM that is being worked out. You have representation of how to share the data for the microarrays and you have MIAPE. The same thing for proteomics. You have a ways to represent the data and a week to exchange the data being worked out. You have vocabulary's with in the health-care system which are -- SnowMed has been named as one of the standards and you have other vocabulary's and representation between entities that are coming through these various oncology speed if you cannot represent the data and talk about it on a conceptual level and then you do not go to very far. All of these are emerging, promising for the future but not quite there yet. Thank you.
Any questions?
I have a question you talked about of 23&Me and is their differences with how they share their data with consumers?
I have not looked into the details of all of them. I do know that both of them will allow you to download your complete data set if you request it. There are some software packages that are available, and open source, that will allow you to accept that data and we -- and manipulate it. I think Navogenics says that if you would like to do that they want to talk to you first, do you know what you're getting? And to some pretty technical stuff but that is a complete data files and there is a community of people who are, of course, doing that. They have their when little facebook pages and 23&Me in particular suggests that you might wish to share your data with the research community and other entities, you know. It is very possible that 23&Me may be making money off of various contracts that they have with companies that would like to have this data on people. I am not sure of the details of that.
Yes?
A very helpful report appeared a couple of comments. I believe that several of the firms have said that they are not sharing any data with any companies in the midst of the genome. That is part of the agreement going forward and that when you do that they have no other commercial relationships in doing that. So one of --
Unless you agree to share.
Not with commercial entities pay they have a number of agreements that they have talked about with some of the patient groups that than to its. That may be another interesting piece about it is the patient groups.
Patients like me.
Exactly.
Looking for people like you, yes.
That was one of the aspects that was interesting that I think could be something that is increasing. The Alzheimer's patient family group has been very active in saying, share full genomes pick how do we as a patient group want to take this? And then bring that to researchers who ever agreed to deal with that. That is one additional model there. I think the only comment that I would have when you talk about the company piece for some of the companies that have products that are currently on the market, several of them to not have products currently on the market. The amount of information you are able to share really depends on whether you are CLIA approved or FDA approved and the ability to give that additional information is not possible because the regulatory authorities will not release the additional information if their approval is on a composite that you cannot do that the genomics companies, the patient genomics until very recently have been regulated differently and therefore have been able to give more information as they so choose.
Yes?
So maybe just a comment. I think the companies that have FDA approval, it depends on what kind of information you are talking about. Because if you have FDA approval you can say that your test is just for whatever you have delegated studies for and what you have approval or clearance for. You cannot say oh, there is all of this other stuff but I do not know what it means.
Or it is approve for this and it is the combination of these mutations produce some data that says this one mutation alone is relevant to this other disease in the are not allowed to give that out because it is not approved.
Right.
But just to clarify for instance if you have 70 genes, it is their prerogative to say which are there or not.
It definitely makes a difference if you are providing a test for a medical purpose as opposed to for the curiosity of the person who wishes to pay for it.
I'm sorry, I just another point. That is right. It's interesting because these companies that say we're doing this just for educational purposes but take your data to your medical provider and then it's a question of what can that medical provider to.
I have a story on that which is I have a colleague who is the emergency room physician that said that he actually had a patient comes in for a visit bringing his Navigenics report with him and been the anxious about the whole thing until my colleague said, you know, I could treat you for your anxiety and that would be inappropriate emergency room visit but wanted to talk to me about being anxious to this direct to customer test is not prepared and I will refer you to a geneticist. So it is true that you get folks who are starting with these test results out of curiosity and then it does make them anxious and what do you do about the whole thing? Does a phenomenon and it is impacting the care providers -- it is a phenomenon and it is impacting the care providers.
Thank you.
Thank you, Dr. Mitchell.
We will open up for a little discussion if we have any on the two talks. We will not take a break as we have in our program. We will move into talking about the different models. So if there are any other comments related to those to talks by Dr. Rodrigues and Dr. Mitchell, then we will move on to the next topic. Any comments? All right. Okay.
We will go ahead. Let me see, where is my program? The next talk is from Dr. Katherine -- Catherine Schaefer.
Thanks very much for inviting me here today to be part of this discussion for topics rather than one topic of genomic data sharing. It is a imported said of the issues -- important set of issues and the dippy she to be much been able to be part of this -- and I appreciated very much for being able to be part of this be you ask for my perspective of the health care delivery system. I should point out that being part of Kaiser Permanente in Northern, California that this is a health care delivery system with a very large and active research division that is creating a very large and comprehensive resource for research on genetic and environmental influences on help and therefore may not be typical of all health-care delivery systems or even those that do research. This is the perspective that I am going to be talking about today is the issues that arise in the integrated health-care delivery system with an electronic medical record that is preparing a very large resource to facilitate research on genetic and environmental influences on health. The resources that we are developing will link to gather data on 500,000 members of Kaiser Permanente including continuously updated clinical data from electronic medical records, data from a participant surveys and environmental exposures including social determinants based in the geographic Information System database and genetic by -- biomarker data. The purpose is for the proper scientific community to conduct research on genetic and environmental influences on disease susceptibility, course, prognosis, and out comes in response to treatment as an pharmacogenetics and also to conduct research and translate findings into improvements in medical care and public health appeared from the beginning we have also have the aim of conducting the research on the ethical, legal, and social implications of genetic research and the use of genomic information in care. I thought it would be helpful if I gave you a little bit of background about this resource. With initial funding that we received in 2005, 2006 we developed a lot of time and effort to engaging the membership of Kaiser Permanente in northern California and the proper organization providers staff and so forth through focus groups, internal communications and media about what we were planning. And to sample concerns and values and better understand the perspective of our organization, its membership about development of this sort of resource. We organized committee and by zero ethics advisory panels -- and bioethics advisory panels to facilitate research creeping over ten registry's creating over 100 diseases and conditions.
In 2007 we started with enrollment of a general cohort to a mail survey to 1.9 million people in Northern, California. That is virtually our entire adult membership that sought information about demographic and background factors not included in the electronic medical record. About 400,000 people completed the survey will for the course of about a year and then beginning in late 2008 we again contacted survey respondents and asked them to provide written and informed consent and the saliva sample. As of about last month 130,000 people have consented. Our current activities include continuing efforts to enroll the planned participants sample. We plan to enroll a total of 200,000 individuals by the end of this year and reaching the goal of 500,000 participants by the end of 2013. We are a beginning the collection of blood samples, easing out the collection of saliva using the clinical and the structure. And if we are continuing work on several fund did Gino wide Association studies including an ethics study of bipolar disorder and the steady -- study of prostate cancer. We have also developed a collaboration's portal and access review committee that will be able to receive applications later in 2010. And importantly we recently received a grant funded by the National institutes of Health that supports Catherine genome-wide genotypings. This study was designed to be a resource for longevity. The average age of this first 100,000 participants is 65 so we have a large number of aged individuals and a large number of people who are perfect for beginning to study the factors that affect. We will be genotyping 650,000 SNPs and the resulting data will be linked to the electronic medical record a survey to create this resource did it will be accessible through dbGaP and Direct collaborations with us and will require direct consent in dbGAP. So considerations in this environment, First of all, this is a very rich resources that would be difficult and extremely expensive to replicate in another environment. It is large and it is aid -- it is diverse and generally represented of the population. The comprehensive continuously updated EMR allowance characterization and follow up. Kaiser Permanente recognizes that they can make an important contribution and wants to insure that the best and brightest use is made of this research consistent with its members. So our perspective on data sharing issued by this commitment to our members. We are invested in them and they determine the future of this organization so all our situation is difficult bit different than makes this an academic models -- that may exist and academic models we clearly have a lot of skin in the game so to speak. So we are the very interested and focused on these issues and are extremely committed to the two data sharing and the answer is that we hope the Sully will bring. Over 50% of our first 100,000 participants have been members of this organization and receive health care for over 20 years. So this boat is a pretty rich -- both a very rich source of research since we have comprehensive data going back to 1995 in an electronic format. And then data going forward pass as well. Trust in Kaiser Permanente by our members enables us to do research. So we are a three -- it is. Important and we are pretty committed to maintaining that trust. In terms of factors that affect the sharing. Certainly informed consent and that nature is quite central. We use written informed consent that is brought it. And includes no restrictions on any kind of health problems that could be studied. Health information can be updated from the electronic medical record going Ford -- forward in time and all studies must be approved by an institutional review board. And data can be shared with scientists outside of Kaiser Permanente that agree to protect confidentiality and follow rules for use. The informed consent stipulates that using and sharing a genomic deal will be for research purposes only. Results will not be placed in the electronic medical record and participation is confidential. Genomic data will not be returned to individuals or their providers. Participants however may be contacted if information develops that has significance for their health. Participants may withdraw and ask that their sample be destroyed. One question that the rise is is how we ensure that these other commitments made in the informed consent armets when data are used for through a public the base where we have -- database where we have less information, less probable feedback from investigators who use information that way. Informed consent does not historic we really address issues of individual autonomy. But has less to say a word it is not historically used to address issues that can rise above social harms that may -- about social harms that may arise in the process of carrying out some kind of research. So in our environment concern has been expressed about data sharing three federal database such as dbGAP PR committee advised the panel, focus groups that we have conducted in some survey respondents very directly have been concerned about this issue and expressed the idea that the government may take or misuse data. The building of the other faculty and eight databases increases the perceived at vulnerability of this NIH database to reidentification or misuse at least in the individuals who are concerned and the use of the DNA or other forensic uses is also a prominent community concern. There is the concern that research may be done that could be used subsequently to stigmatize a vulnerable group. There is actually no recourse of the individual other than actually withdrawing from the resource. No choice is involved about the kind of research that is undertaken then with the resulting data. And then there is the perception on the part of our members that storage and control of data by Kaiser Permanente and the resource gives participants better recourse and control over events. I just want to mention. I think you will hear quite a bit from Daniel Masys about this later today but the obvious fact that in most research data it means sharing Peter -- sharing phenotypic data. And health plans with EMRs as is the case for our research has huge investments in the EMR data that is linked to the generic data. The quality of phenotypic data is critical to the best use of the two genomic the net and the resources and is challenging -- to the use of genomics data and the resources challenging because it is useful to all. The best use of the data depend on knowledge of the system that generated the data. This is also the meaning of this sort of clinical data even when standard diagnostic codes are used and efforts are made. It really takes an understanding of how that data is generated to make the best or most valid use. Partly in response to a variety of these concerns we have begun at this point to perform a series of stakeholder interviews led by Carol Lewis the head of our ethical, legal, and social implications board with the goal of informing our development of policies and procedures. We have been conducting these with a variety of stakeholders as listed here with the following sort of research questions. Such as, what are the specific data sharing and governance issues inherent in a biobank. That is really what you wanted me to tell you about today and I regret to say that we have just begun these interviews so I really do not have the death that I can present about the outcome -- data that I can present about the outcome of the results of some earlier focus groups and interviews that we conducted. I hope there is a chance actually to come back and tell you a little bit more about the outcomes of this intricate effort at a later date. Thanks very much for your time.
Thank you Dr. Schaefer. Any questions?
Your institution like many has policies all along subjects to withdraw and by that meaning having this samples destroyed. Are you aware of any cases where such a request has been made at your institution?
Yes, it happens really but it has happened and has already happened with this particular resource. But I know of only five instances out of 130,000 individuals participating with that has happened -- where that has happened.
Could you comment on how the providers in the Kaiser Permanente system understand what the heck you were doing and also I am curious about the uptake or the frequency of participation by your members. It seems quite high and I wonder whether -- what you have done to foster such types of situations.
I am delighted to hear used describe it that way. It has mostly been an effort, the traditional ways we know how to contact people which is essentially meaning people materials that are descriptive of the research program. The consent form is eight pages long if you include the HIPPA authorization. So most of the participants that we have garnered so far and perhaps this is one reason why we have a very good representation in older age groups are people with the time and patience to basically make their way through printed material that they have received through the mail. So our next efforts in enrollment are actually to carry out different sorts of records that do not involve only essential to reaching out to our members through a mailed, written material format but involve other ways of in teaching people. With respect to providers , are providers, we have had a research division since 1966. They are familiar with essentially having patience recorded for studies. The research division essentially operate side by side with the providers but we do not typically recruit providers. We do not ask physicians to talk to their patients and ask them to participate in studies. There are certain clinical trials that are exception for that but in this gender research we do not do that. What do they think about it? They are very hopeful that in the not too distant future they will be able to do a transitional studies that will fulfil the promise that is held out there that this kind of research will results and things that directly improve health care -- will result in things that directly improve health care.
You may not know the answer to this but I am wondering if there may be training to the health care providers with physicians or nurse practitioners who are the primary providers to address issues if their patients, may be having read something in the press about a genetic study or knowing more about the study. Who do your participants go to for sort of small questions? Not informed consent type of questions. Do you know what I mean?
About this study, you mean?
I would assume reading through this consent form would make the participants a little bit more alert and they might then go to their providers with general questions and is there any training for those providers within Kaiser Permanente?
I do not think as of yet there has been any provider training in how to respond to sort of general questions about this kind of data in particular. The providers have but we have provided to them. We do have a strong medical genetics department that is depicted across the region. And those providers themselves, for example have organized about the time that, for example, BRCA when testing became available. We anticipated that there would be a lot of general interest in this even though the test was not in the appropriate for women who by virtue of family history might have a low to moderate risk of inherited susceptibility. The genetic providers pioneered a class that in a woman come to that sort of explains what it is and the sort of pull of family history in risk of breast cancer and women could sell preferred then for genetics counseling if they thought that this was something that they really needed. So we have a of a bit of a model of how to handle the situation with -- where there is general interest in something but, in fact, there may be a case with the test is really only appropriate for a relatively small number of people.
A couple of brief questions. One is a follow up to Michael's questioned if they leave Kaiser to go to another payer how do you handle that and the second question represents a lack of information which is given the age distribution and the membership plan, how representative is your sample compared to the rest of the Kaiser membership in relation to the population of California in general.
Let's see. We actually do not have a good solution in terms of continuing participation if they leave Kaiser Permanente and go to another system. Essentially we have no way to really follow up in the same sort of point their medical history at the point at which they leave us. So the informed consent gives us permission to use the data that we do have and the samples even after the person has left the system or died for example but right now at least we do not really have agreements for continuation observation.
Can I answer his question about represented Ms.?
We were going to take one last one. And then we will move on to the next speaker.
The answer is that while we have good representation of different groups it is actually Kaiser Permanente which is generally represented in Northern, California now the way that we have been enrolling people for example, it is older and more female and more white and better educated.
Thank you, Dr. Schaefer.
We will hear from Daniel Masys who will talk about the academic model and then we're going to take a break. thank you, and since I am standing between you and a break I will move with all expediency.
Medical records combined with genome-wide scan, in my capacity as the principle investigator for the national coordination center for a consortium called electronic medical E records and genomics consortium.
Three topics in the next 15 minutes. First, what emerge is, lessons we are learning about data sharing, and then a focus in particular on wherey we're at with respect to this science, the emerging science of data identification and reidentification. This grew out of a request for applications from the genome research institute in 2007, the key element of which is highlighted here in red. In essence, support for affiliated groups, you had to have an existing biorepository and you had to have the ability to extract phenotypes from medical records. The consortium, members of the [indiscernible] are five, pretty wide distribution in the differences with which they acquire their biobanks and clinical data. Here's a map that shows the primary phenotypes that were parts of the original submissions, part of grant submission you had to propose what phenotype you were going to do a G was on We have since expanded that. You see here they range, and I will actually note as well there is a variety of sizes of biobanks, so in the upper left hand corner, Pacific Northwest, group health of Puget wound had sound had a -- bank of about 3000 samples. The cataract phenotype represented the most mature health system based biobank, one that many built from a [indiscernible] cohort of 22,000 individuals in northern Wisconsin. The Mayo Clinic was about 3000 samples in peripheral vascular disease. North western was looking at Type II diabetes with a general purpose biobank built from all comers into an environmental medicine prospective consented biobank participation.
Vanderbilt is looking at -- future cardiac events and that model is non-human subjects, deidentified biobank from discarded blood samples, the DNA is extracted unless the patient elect to oment out. Vanderbilt is the -- started in 2007, just a little North of 76,000 samples growing at 500 per week.
Again, the features of our network, each site having DNA linked to the concerned electronic medical record data, an important component and requirement of the RFA was community engagement, investigation into models of consent and re-consent. Two of the five members have to re-consent because of the last condition on the slides, submission to DB gap because of funding, part of the model of consent for others and so it was not the case that all groups had to do that.
The core was a 3000, roughly, subject, G-was subject with roughly 30,000 genome scans we could do not only the primary phenotype, but the associated data, I will show you more about that.
We have since received supplemental funding for additional new gene typing across -- that work is in progress now. This is an example of conditions that were not part of the original proposal, but do represent data that, because it's so commonly acquired, just in the natural course of people having routine testing and electronic medical records, gives the system an opportunity to test the samples, the range for most conditions in the thousands of samples for which the genotyping is already done. The basic platform is 600 K alumna geno scan, although African-Americans we have about a 1 million snip chip -- the ability to look, seize -- lipid levels, G-was studies on -- already published, dedicated research cohorts and [indiscernible] rate are emblematic that it is a data-rich environment, such as Cathy described, and allows us to begin with a geno mine scan and look at various aspects of the genotype.
What discovered along the way, and the inside joke is any fool can get a genome scan and many do; the really hard part is the -- the informatics we are engaging are essentially because we are going to pool, do meta analysis of phenotypes, how comparable are the populations who walk through the doors or sign up for the cohorts in these five different health systems. If biologically there is some inherent bias in the nature of the patient populations, pulling genomic data may mislead with respect to statistical populations.
We have discovered that genotypes are pretty easy to share by virtue of the NIH genotyping centers, so sending samples and receiving data sets, not unlike the ones you can download from 23 and me or Navigenics, is actually the easy part, but there are no standards that represent the genotype/phenotype package where you can send the whole thing in one envelope. We are developing standards for how clinical data can be put into a format that is useful for association studies.
Clinical data has a number of features that make it different than the classical cross-sectional research cohort that has been published in the G-was literature to the current time. One is we can't predict how many times some measures will be done. For example, if you have a diabetic with blood sugar measurements, there may be thousands. How do you decide which to include in a research data submission, as well as the feature of EMRs, clinicians are absolutely comfortable with the notions that they are -- clear, some people have definite diseases, others may have probably or possible, and that notion of uncertainty that lives comfortably in the clinic is not well suited to this research environment that looks more like a case report form with dichotomous report you have or you don't. One thing is the assertion of whether a condition is present or absent and our level of comfort that exists in EMR.
I would like to focus on the last thing we are making progress in the network on. That is the re-identification potential that arises particularly out of clinical data and the phenotypes associated with the genetic samples. With the general model we will of course like to maximize scientific value while complying with the federal privacy policies Lara Rodriguez mentioned. This is the screen shot from the dbGAP data submission policy. If you are submitter to dbGAP you have to send data without identifiable information, using a unique code and such. The question is when you have clinically derived phenotypes, how do you do that? Is it calls to mind, I think, an important set of vocabulary, because IRBs always get balled up with this, about the notion of well, is it anonymous or not or what you mean by de-identified. In this regard, at least in the computer science and informatics community we regards anonymous as things are not traceable to an individual and it was a concept prevalent from about 5000 BC, the type hip uke rat ease, to -- anonymous, or it wasn't, and the IRB was happy with that, assessment. What we know now, we had to replace that with something that looks like a slider bar. It's replaced anonymous, we recognize the biologic data is rich in attributes, reidentification potential essentially never goes to zero.
So it's a continuous variable whose properties can be calculated for some, but actually not all types of health data. The primmer on reidentification is a simple one. That is, if a data set has ostensibly been de-identified, then the way, the pathway to trying to find out the identity of the individual from who it is derived requires two conditions. The first is out of many records, getting a unique set of attributes, what people call the logical record with one individual. You have to get to uniqueness first. Then, that's necessary but not sufficient. The population, whose understanding of genetics is informed by the OJ trial and CSI believe DNA is inherently identifying as if you famed a pommy polly vial on the carpet you would know -- you need a naming source, be able to impt sect that with a person's -- basically, de-identification is preventing you from getting identified to a unique record, always more than one that carries a unique set of records, but you can block the linkage to a naming source. This is a graphical view of this from work, and I will present a couple of slides from brad Moll in, a data privacy guy. Your briefing materials have one of his recent publications. In essence it shows on the left-hand side, actually in all three of these conditions have to be satisfied, unique with respect to the de-identified data set. You have to be unique on the right side in terms of named data such as a voter list, vital statistics registry and a firm linkage, one to one between the models.
Let's look at uniqueness. If you take clinical data, our own cohort of the 2500 Vanderbilt paicialt patients, you can say how alike are they based on common measures in clinical data. One is ICD 9 disease coding. To cut to the chase, using the reference population now about 1.9 million records in our EMR, about 97% of people are out of the box unique. The combination of their age, gender and ICD 9 codes, on average we have about 12 codes per person. It only takes five codes and all of a sudden you are in a box where [indiscernible] it would seem we are in very good condition to be able to sending out the entire detailed set of rich phenotypic attributes representing even ICD 9 codes.
We are working in a world governed by HIPAA. The standards, more stringent is called safe harbor, allows you to release race, gender, only year of birth, not date, only state as the smallest geographic entity in most cases. There's the thing called a limitedded data set, which allows you in addition to those to increase the specificity, row lease the actual date of birth and go to a county level.
Then the question is what linkage potential does that give you for data sources. Here is Brad's work on the pooled U.S. census data from the year 2000 that shows the fraction of unique individuals under HIPAA safe harbor, releasing year of birth, Secretary Sebelius -- roughly a 10 to the minus 4 uniqueness, depends on the states, how sparse or densely populated the state is. There's a dramatic increase in the number of records that become unique in the actual data birth, sex, race and county. 30% to almost 100% of individuals can be uniquely isolated inside a clinically derived data set. How about the naming sources? If the issue is lots of things are unique in EMR, here the story is very highly variable across the landscape of information resources on the Internet, importantly, across states. A common re-identification case is vital records statistics or voter records. These are the state policies, items developable for the participants, Illinois, Minnesota, Tennessee Washington and Wisconsin. Authorized users in three states include anybody. In Minnesota you have to be a Minnesota voter. You have to be a political person in Illinois, isn't everybody? And so you can get this stuff on a disk and it ranges from $30 to $12, 500 and you get a variety of different data elements, including date of birth. You always get name and address. The question is what are the other things that map to the limited HIPAA dataset items. As a result of that availability, we have done this for all the states in eMerge. What happens to uniqueness when you merge with a naming source. You see on the right-hand side that the number drops, but it doesn't drop dramatically. In essence, K here is the cell size, you could say well, re-identifiablilty doesn't begin with a single record. Maybe it's when you have a pool of five records, close enough, where we could then use other methods to try to zero in. Even a K of one, where 30% are unique it drops to 15%, but that means 15% of that entire population you have a name, address, all -- successfully and fully re-identified the individual from the de-identified data. As a result of that ability to do a quantitative analysis, what we found in the network is the clinical data we are sharing with dbGAP will need to be a sub-set of the full clinical record, specifically removing uncommon codes that support elevated risk -- re-identification risk. Elevated, above the HIPAA standards so we can qualitatively say what HIPAA standards and meet by a number of -- we all have lawyer-approved data sharing agreements at the individual record level. That works fine among the consortium. Took us 18 months to get -- the way we did that, everybody made the agreement to share with the coordination center as opposed to having to do with four other agreements with four other institutions.
In the capacity as a data analysis center, providing data privatization, including quantitative reassessment of data sets before they go to dbGAP. You might management Alzheimer's population highly skewed to older individuals, the HIPAA standards for people 90 and above. The good news, we have a couple manuscripts in review and are just about to release tools that will be open-source useable by mere mortals for actually determinating the quantitative risk of the demographics of publicly submitted data sets and how you can in essence trade off for scientific purposes the grange granularity of one item, smudge the zip code if it's important to maintain age. That's an important variable in the analysis. You can do various permutations in order to meet the federal standards and you are losing some content, but if it's not important for the key scientific hypothesis it's still a kind of win. That's the work in progress, and we will be reporting in the literature, happy to report to you as we make progress on these issues.
Like all good networks, we have a URL with the unpretentious of G was.y net. All our publications, practices within the network are posted on that website. With that, I would be happy to answer any questions.
Thank you, Dr. Mason. Any questions? Mark?
One of the first slides where you talked about the RFA, I think there was a reference in there to the use of natural language processing. I am curious how is that working out for you?
What we discovered is that -- well, some people said -- the null hypothesis is that EMRs are so bad you couldn't use them for anything. Just a mess. What we discovered is in order to get a positive predictive value that works across multipeel ENRs, you need -- codes, IC 9, labs, and medications. In a sense, medication is the sincerest evidence a clinician thinks you have a disorders and that generally gets the RPVs in the range, roughly only about 65, 75%. We have to use natural language processing, that is teaching computers to identify concepts, diagnostic concepts and whether they are inserted or negated into the record to get PPVs in the 95% range. The good news is we use experts, generally takes five iterations to get it right. Then when one of our institutions gets it right we fowntd found we can actually transport that across the network and with relatively minor modifications, the PPVs fall only a few percent when reused in very heterogenous -- that's the unexpected big win here. If one group does the work of creating the phenotype creation logic, other institutions that want to use these to find codes of interest, either for administrative interest or research, can reuse that without having to re-do the wheel. Yes?
Question: Very impressive presentation. I was curious to see if you can elaborate ability about the process of informed consent of the patients. What I understood you were talking about, use procedural specimens that would be discarded otherwise. You mentioned unless the patients opt out. What is the process of informed consent, or blanket as they come to the Vanderbilt institution, they will be involved unless they -- how that process works.
Vanderbilt is the member of the process that identifies both record and biological samples are de-identified and we cannot construct identities and go back or contact individuals. The general model has been published, and I would be happy to provide it as the reference, but the short version, OHRP-approved. In this non-human subject space, the federal regulations would have allowed us to view this as existing tissue and data without notifying anybody. Our ethics board and IRB said it doesn't sound right. In the conceptualization, and [indiscernible] preceded by surveys of patient attitude and such; we added this component of very extensive public notification campaign, the fact that people re-signed a consent for treatment. Right above the signature line the only bold-faced type in the whole thing is a big box, says "I understand Vanderbilt extracts DNA from left over blood samples and I should check this box if I don't want my blood used for that research."
Having run 30 months, we had a predicted opt-out rate of 5% and that's about what we are seeing. Generally broad acceptance of Vanderbilt patients based on what Cathy said. Our patients may not trust the government, but basically they trust the institution they are getting healthcare from. They are willing to let Vanderbilt do the research. To not turn this into an hour-long discussion, we can come back, give you the full soup to nuts, we published it and I can send you the URL.
Okay, any other questions?
No, thank you, Dr. Mason. We will take a 10-minute break, come back at 10 after 3:00.
[Recess until 3:10 3 -- p.m., eastern time] > >
So I would like to thank the committee again for having me to talk; and I promise this will be the last time.
I -- that is, I am not sure I will say it is the model, but it's an NIH model we are seeing become increasingly consistent across the different institutes going forward. There were several questions the task force was asked to answer, I will try to get down to some of the nuts and bolts, the process of how this works, informed consent, responsibilities for protection along the way.
Data sharing is nothing new to the NIH, a long tradition of sharing resources and tools, and of having large policies for all of our extra mural -- around how they share data. Traditionally this has come forward in one of the more major statements in 2003, for any grant over $500,000 to have the data shared at the time or post completion of the study, the data was to to Shared broadly and made available. What's different, one thing that's different about GWAS, began to merge, making issue available prior to publication in the NIH-realm of the broad data sharing. The reasons we did this were largely based on of scientific opportunities coming forward, technology became accessible to do whole-genome scans, try to tease apart the genetic underpinnings of common diseases which have been so difficult to address through standard genetic mechanisms and strategies in the past. The tune the tune Opportunity to do this, and the breadth of the different institutes that were attempting were a strong force for leadership at NIH to say we needed something going across the institutes, consistency, expectations for the investigators andy again for the public in terms of what they would understand about data and what would be in place for protections about how these data would be shared. You are all well-aware of the power the genome-wide association data had in terms of the richness of genotype and phenotype, and ask many questions of the data, all supporting the reasons to have as many investigators have access to the data as possible to ask as many different questions as possible.
This brings us back to the guiding principle and really the foundation upon which everything came from as we constructed the policy and all of the different elements within the policy. That was to try to achieve maximum public benefit from the federal investment and the wealth of information and data generated through the different studies NIH was beginning to fund.
The policy itself is broken into three primary sections, the bulk of the language focuses around data management, speaks to the importance the NIH put on both standards and expectations for data submission and data access, also for the protection of the data. And of the interest of the individual's whose data was in the resource the NIH was creating through this policy.
The way the process works, of course, everything is built on a primary research study, taking place perhaps outside the realm of NIH study, investigator, structured around an informed consent discussion interest agreement, about how the datal be used, whether shared, et cetera. At the point in time an vast investigator they wanted to apply to the NIH for funding was the trigger point for the GWAS policy, the expectation the data will come to a repository house the at the NIH, data for genotypes and phenotypes, newly constructed at the time within the NCBI. These different screen shots portray the range of information about different studies available, from the protocol, survey instruments, arch averages of phenotype data, people can zero in on where they want to look.
The other advantage to having the central resource besides being able to make documents such as these examination procedures, would for many be in lab file cabinets, now available and searchable through open access pages on the web, people can find new collaborations, preview studies and find out if they are relevant to the kinds of questions they were able to ask before they attempted to ask questions of NIH.
All the data that come into dbGAP are deidentified and the way we define deidentified, Dan mentioned, this is a variable term, to look at HIPAA standards, the 18 identifiers in the privacy rule, use that as the basic rubric by which investigators are asked to deidentify information, hold the key to the code in their institution, not share with the NIH, so when information came we did not have any way to link back to the code for any individuals within the data sets.
The third and final phase is to make this data available to the secondary investigators. This would be through a controlled access process, which I will talk about in a moment, they are only ever getting information to coded information. They request the data for a specific research purpose and project.
Coming back to measures of proy tection, one of the things that NIH did in the policy, which departed from the basic regulatory requirements, was to attach an expectation that the informed consent of individuals and those agreements that may have been made in terms of how data could be used in the future, how data could be shared, would remain attached to how data were distributed through the resource. All of the data, since de-identified, don't technically represent human subject, data, once it comes in, we maintain informed consent is an ethical -- as it went out, used by others.
In terms of implementation for the policy, the local institution consistent with general practices where the IRB is the authority for any study that happens, submit application to the NIH that the data set, all data within it are appropriate to come into the repository and be distributed into secondary investigators. They are specifically asked to an IRB review, elements of informed consent and state the consent is consistent with use coming through dbGAP. Also, again, assertions the PI will remove all the HIPAA identifiers so it can meet that standard of deidentification set forth in the policy. Any limitations on future data use are requested through the recertification, this can speak to issues around informed consents so if data were collected under an agreement where the data would only ever be used for cancer research, the researcher is aware of that, can only release or distribute to secondary investigators us doing cancer research, [indiscernible] going forward so we can respect the decisions of the local institution coming into the NIH.
In order to try and find, provide information to local institutions on these new responsibilities for the data that would be coming in to the resource, we did craft a points to consider document that discussed all the basic elements of the policy as well as some of the overview, the audience -- considered document wasn't in10 tended to be the IRBs, the back ground or science seeing, being asked to provide the certification. The points to consider walks through many of the elements in an formed consent the NIH felt were important, but not intended to serve as a check list, still leads to discretion of institution, what is appropriate and not appropriate, based on their own deliberations relative to the particular population in a given data set, relative to the institutional policies at their research institution.
Data access, two-tiered, public access pages available for anyone to look at to get basic high-level information on the studies within the data set. Again, to understand whether or not the data set might be interesting and relevant to the questions they would like to ask, but in order to get to the individual level coded data it had to come through a controlled access process to submit specific research use, proposed research use, reviewed by data access committee and decision theny made. The point of specific use was in order to have, for the data access committee to make determination about any limitations on data use provided by the local institution at the time of submission.
To try and gain some accountability for investigator practices, once they had the data, every request for data must come in cosigned by institutional official, taking responsibility and vouching for the credibility of the investigator coming in, acknowledging they know the vast investigator his the data, intending to use the data and in compliance with local policies they have put in place for how data use of this kind of genomic data, whole genome date a is used at their institutions, since different institutions have different policies about how they review the conduct of research which holds genome research and coded specimen information at their local sites.
I think I have gone through some of this already. The data access committees in terms of who they are, because all of the data reside within a government database they represent government record and only federal employees can make decisions about access to the data. So DACs are only federal staff, but are able to consult with anyone in the process of reviewing a document. They can bring in an expert in a particular population if they have concerning about potential [indiscernible] or bring in a scientific expert if they are not sure the particular use fits within the use provided by the organization. The other function the DAC has is to track the data use by those users they have approved within the database. Annual reports come in for all users where they talk about significant findings for the work they have had, any publications coming you out of it, any [indiscernible] that may have been noted and provides a way for the DACs to go back, make sure they are only working on the proposed use they submitted for approval at the time and not doing something else with the data.
The agreement between the secondary investigator, his or her institution and the NIH comes through the form of a data use certification. We have one common model template for all the committees to use for every data set that comes through the NIH, which was something we didn't have at the start. That is an improvement. Hopefully will make things easier for investigators and [indiscernible] to understand what they are agreeing to.
The conditions and terms are fairly obvious, federal and state law, only use data for things they said they will use it for; promise not to attempt to identify the study participants based on information from the NIH or by combining with any other data set they have access to public or otherwise. Importantly too, as I measure of transparency, everyone requesting access agrees to be identified on the dbGAP homepage, you can see every approved user, their institution and approved research use for that data, so the public can also see what's being done with the data and how it's being used.
The final two elements of the policiy speak to issues of scientific publication and intellectual property. For scientific publication the concept of this pre-publication broad access to data was something new for GWAS, moving beyond the genomics community, expecting to spend the time, energy, the PIs will spend many times over decades to develop cohorts now wanting to do GWAS, an embargo was put on, expected to be submitted as soon as quality control was complete, be made available for investigators to begin analyzing, but an agreement only the PI and direct collaborators would be able to publish -- for the first 12 months the data was available. This was implemented trying to highlight the embargo policy and dates attached to different versions of data sets, again on the homepage for dbGAP.
During those 12 months, however, anything else was appropriate to be done. So you could investigate it, thoroughly, write where are paper, just just couldn't submit the paper until after the 12 months expired.
In terms of intellectual property, we were a bit limited in terms of what we could do, wanted to stay within the bounds of existing NIH policies, respect the [indiscernible] principles. There was a broad consensus internally and through consultations with external experts in the area, the basic GWAS findings that would come out of the first round of studies were really pre-competitive, should remain in the public domain so everyone would have freedom to develop around and innovate around the basic findings. To try to substantiate that policy there are automated -- around the genotype analysis in the database so everyone can have them and they are out in the public domain to try to substantiate the fact that patents shouldn't be filed on those first-round findings.
This is further emphasized in the policy use statements and the certification, where investigators acknowledge this intent for the NIH, principle that the data remain in the public domain, as well as their institutions going forward.
Something else that was important and has proved to be vital to GWAS management and even within discussions in the community with how to go forward with genomic data sharing and repositories is a governance model. This model is both simple and complicated, depending on the level you are working, part of the design and has been helpful. At its core there's a senior oversight committee, reports directly on the NIH director, all the changes, as well as managing at the highest level how the policy is implemented across the NIH, chaired by Dr. Craig Greene at [indiscernible] other senior staff from the NIH director's office. For day-to-day issues, the -- two steering committees sit under the SOC and they are made up of senior staff and focus on two specific realms of issues, the technical standards steering committee focuses on programmatic, as well as technical issues around dbGAP and the participant protection and data management steering committee is constituted from the various state access committee chairs, as well as other experts at NIH and human subjects research and bioethics. That is where the core of the policy development and practices have developed as the DAC chairs have learned how tho do their jobs together, more of the framework of how NIH will do this across the board. They definitely inform and -- as issues arise so we have both leadership at the highest level making disiz decisions, as well as those staff on the grounds trying to implement and informing what the decisions are. I will stop there, point to our GWAS website that is under review, but hopefully will be a place we can have in information, and have transparency on the practices going forward for everyone that needs to interact, from the investigators, members of general public that hear we have the database of genomic data on thousands of individuals, and I will stop.
Thank you, Dr. Rodriguez. Any questions? Mark?
I may have missed this, but based on what you were talking about, sounds like if you are an investigate that wants to -- go through the data request, approval, all that, then you are able, sounds like, to download the data on to whatever your local resource is and use it under the terms of the agreement as opposed to the data residing within dbGAP, manipulated there, where it wouldn't move to a local type of server. Clearly, the advantage of having it centralized is you can develop audits, make sure people are staying where they are supposed to be, presumably there was a decision made as to why this model versus another is used. Could you comment on that?
There was a great deal of debate as to what model to use. The final decision was made because the statistical geneticists and people would be writing their own programs, it couldn't be done effectively within the space. Something security transmitted, put agreements in place in terms of security standards in place at the site for the data, and the IT officials for the institution are now -- one of the required signatures, though it doesn't actually get implemented that way, they are supposed to be aware of every request for access as well. So they are signing off again they have the capacity to protect the data the way the investigator is agreeing.
Question: Obviously you are requiring a report to come from the institution to say yes, we have been behaving, using the data the way we are supposed to, here are the results of that. Is there any opportunity for [indiscernible] to audit, if they suspect something is being used the way it isn't supposed to, to show us what you are doing? An IRS model. Yes, this is what you told me on your taxes, but is that in fact what your income was for the year.
We talked a good deal about setting up an audit program, we look looked at several models, the cost, the benefit of return on instituting such a policy was such that it was determined we would not go with an audit model to start with, unless we saw we had problems and in fact so much of what NIH does operates on this assurance model, the organize, and we trust you will do what you gre agree to do and if you don't there will be consequences.
Question: Yes, I am interested in the clinical phenotyping of the subjects. Are there standards for that? One of the problems we have are diseases that predominantly effect one organ system, but also something else. Where the clinical phenotyping may be partial because of the observer who originally created the data set, how are you addressing that issue?
So dbGAP set themselves up to accept any measure, however it was rmentd reported, to be open, because there was such variability. By putting protocols online for every study, you can see exactly how a blood pressure, for instance was measured in one study, know if that will be comparable to a blood pressure measure in another study. NIH has the Phoenix program, looking at building standards for phenotypic measures across the board but there are no requirements at this point been dbGAP for GWAS data.
Jim?
I was wondering how your deliberations, your model and all was affected by the Jacobs, nature of genetics paper about affecting -- inclusion -- the fact that it's possible to analyze -- the aggregate data and infer phenotype. Neils homer and -- the subsequent one.
Our policy hasn't been changed at all relative to Kevin's paper from this fall. We are having ongoing internal discussions about at what point we re-address the situation of is there any level of data that might be possible to be made public. We haven't yet come back to have the formal discussion with Kevin or his group, but we have definitely looked at the papers and the groups.
I have a question, lar a. You are in the process of changing, modifying the GWAS -- for genome -- you are in the process?
We are just starting to do that, some internal data collection to go to program staff, get information on what policies already existed for sequence data, how they would describe a sequencing project that would or would not be subject to search a policy. The sequence projects are a lot more -- pretty straightforward when you have it. We have that information now and are beginning to look at the various policy scenarios to put together around that, as well as some of the technical issues. It's harder to transmit all the sequence data and determine the appropriate point to release that. The sequence data come in, in a very different format and timeline than the GWAS data.
Do you have an idea of the timeline to roll that out?
Not officially. We hope to have a draft ready for leadership to consider by the spring, but I will never predict what the leadership will say about the draft, so --
Can I ask another question? Use of genetics and genomics for prediction necessitates having information over time. Have you planned for that in this database, that is, observational data that shows what happens?
That was again another reason why having codedded data was thought to be so useful, because there can and have been updates to different data sets. Framing ham, for example, had several updates and versions available, large cohort with another round of visits, data collection, you can go back in, sosh the that associate that with the data you already had. It can be a dynamic resource.
No more questions? Thank you, Dr. Rodriguez.
Now we will hear from Dr. Hoffman who will talk about a commercial model.
Dr. Hoffman: Just as there's no single academic, government model, there's no single commercial model for data sharing. My intent today is to provide a few examples of things we are doing in the commercial, electronic health record environment, to set the scene for the more effective exchange of genomic data, and then some examples from other domains outside of genetics that I think will serve as relevant examples of future transfer. How to facilitate data sharing.
We will begin with the comment that might seem out of center field at first and hopefully when I come back you will see why I am saying it. There are more virtual farmers in facebook farmville than there are real farmers in the United States. I will leave that out there, come back to it. It is a little bit provocative.
The topics I want to hit on today is that first, how can you generate high quality data during Parish the care to facilitate both data sharing and decision support? Then, secondly, we will talk about a few examples of what I call the data sharing ecosystem, there is no single model for how you go implement data sharing and there's actually strengths and weaknesses to a couple of the models out there.
I will share a couple of efforts in each of those. withinin Sirenner we are working on both sides of the puzzle, the patient care provider side, how can we enable genetic testing laboratories to capture data discreetly, but then we're also working to facilitate research using our deep knowledge and understanding of clinical processes and of clinical data architectures.
So to summarize at a very high level some of the key attributes of the tran health record and I should also point out there is no single electronic health record. There are multiple implementations. You can go to some of the organization that are prominently represented that have home grown EMRs, then the multiple commercial electronic health records, each designed around different principles, but most would agree capturing information, [indiscernible] is fundamental. Simplifying data retrieval, queries, analysis is key goal of moving to electronic health records, automating processes, reducing the opportunities for error, providing decision support capabilities, creating efficiencies and generating a body of data that can be analyzed whether for administrative, operational, clinical or scientific insights.
There's often some blurring between the electronic health record or electronic medical record, and the personal health record, which to me, the medical record is a legally binding system. If a physician is part of a malpractice suit, the assumption is there will be high-quality data in the system to be extracted, utilized in the discovery process; whereas the personal health record there's probably quite a bit more blurriness around the obligation there. There's often an expectation the two should be one. Likewise the expectation you can have one system an EHR, PHR and research system is something to scrutinize more carefully. It's not a perspective we try to promote. We believe there should be fire walling between the systems.
In informatics, sometimes we want people to think you couldn't do research in this model where information was stored on paper, but the fact of the matter is there is still a large amount of research done through manual chart abstraction, and the privacy issues there are very similar to those in the electronic world. In many ways more challenging because you have human beings pulling the paper charts out, see the name and re-enter that information into other systems. What we are trying to move towards is a fully automated digital system where clinical information, eventually genetic information is stored discreetly in a minable fashion.
A couple of the other presentations have referred to standardized vocabularies, ontologies, we have been very active in developing, deploying what we call the clinal informatics ontology. This is a vocabulary available, can be downloaded, creates standardizedded concepts that can be used to codify findings whether molecular diagnostics, genetics, other testing methodologies.
To give one example, probably the orphan -- cyto genetics, if there's anything that will put fear into the heart of an inform at sifts it would be a karyotype. We have put a bit of effort into putting the karyotype a minable resource. Because, let's say you are interested in a condition tied to band 21.2 in this example. If you were to do a purely text-based mining of that karyotype you would never find this patient's result. We drop out discreet concepts into the database from that, one of which is a conset of 21.2, because the beginning and end positions of the abnormality documented here and creates a minable resource, whether for research or decision support.
So these are just high-level examples to show that within diagnostic labs we are working towards systems that create that granular body of information so that as you get into data sharing your data is ready from the point of capture and you don't have to re-enter into another system.
The second theme I want to cover is representative data sharing models. One model, very familiar, what I would call the centralized data warehouse model. Increasingly, things are moving towards distributed models. From the electronic health record supplier perspective, we have a common architecture in use at thousands of healthcare delivery facilities and believe that architecture alone positions things to be used creatively for collaborative work So I will share an example of a project based data warehouse. We also have brought in technology to provide a consent-based -- web-based consent-driven system and I will show that briefly, then I will return to my social media comment.
The data architecture that -- there's a couple options embeddedded within the picture, within an organization the clinical care data is embeddedded into a database or the EMR, and that information can then be used by the physician or the CIO or CFO to make observations about how they are running the organization, so forth. It's also very feasible to migrate that information into a larger meta data warehouse, usually involves scrubbing the data of all HIPAA regulated identifiers. It's also normalized, so key part of any data merging activity, especially among non-affiliated organization, is mapping to a common vocabulary. That is a key part of what many aggregate data warehouses offer.
Moving to distributed models, if I just summarized the data warehouse model, the distributed model is instead of pulling data in, you push queries out to the sites. So in IT systems we think of operations jobs that run at midnight, so the impact is minimized. These will be routine processing but also can be queries that evaluate the data within that site. Then summaries of the findings, instead of the actual body of data can be sent to the organization managing the distributed project.
So we at sirenner are deploying what we call our research network where throughout our client base we can push packets of queries, so if you are interested in cystic fibrosis patients, you can work with us to sponsor a project, push these queries. The data remains at the local site, we don't really want the data from these type of initiatives in our hands under this model. Then I really don't like to make the comparison but helps it click. The analogy that resonates is Matchmaker.com, we match a trial sponsor to site that's have a candidate group of patients and that has value to the process so if you are looking for trial candidates you are not mining in territory you are never likely to find candidates.
In the public health domain we have taken the model a step further, working with the CDC, local health departments, reaching out to the entire client base, saying we would like to work with you. You will get a daily view updated every day showing how your organization compares to your state and national peers in terms of positive flu results, ininfluenza like indicators, so forth. In three months we rolled this out to 780 facilities, present in almost every state. We had 23 million records pass through the system for surveillance. The CDC gets updated information every day, the state and local stakeholders, it provides GIS level mapping and trending. It's very feasible to use this push model in a very rapid approach and I think a commercial company has the agility to do this type of thing very quickly.
As an adjunct derived benefit, a the lot of information comes out -- the [indiscernible] utilization. I don't know if -- there are people from Tennessee here, but if we compare Tennessee to the national norm of emergency department utilization, every day we can see that on even week days, 50% of healthcare is delivered in the emergency department in Tennessee. There's a lot of insight that's can be gained from this.
We also mentioned we recognize prospective research is an important model, I think you will be hearing about a consumer approach in the second talk. Our stance is that we want to let scientists do the science and provide the enabling technology to get to the science as quickly as possible. There was a company called first genetic trust, so we brought in the source code, patents to technology and it enabled patient controlledded disclosure of genetic information through this model, this web-based model.
The second to last slide is an example we participated in that pulls many of these topics together. Sirenner does the -- out-patient study, patient consented, enrolled, tracked longitudinally. They had the insight to track the HIV data, as well as prescriptions and lab data. One question looking at personalized medicine, how, when armed with genetic data, how well are physician utilizing that information. We did an analysis of the data, found using one scenario that using antiviral resistance, that if you mine the data as analogy of physician behavior, patients with the resisten the HIV genotype, 59 had contraindicated therapy initiated six months after that result was determined. I think that's evidence of the need for decision support.
I promised I would come back to the Farmville comment. Data sharing in the social media world is really completely, maybe doesn't -- I haven't heard it on the table yet, but if any of you are in facebook, used a single facebook application, I actually found one from NIH. When you sign up for a facebook application you are giving data from your profile and your friend's profile to anybody, to the organization launching that application. So if you are signing up for farmville and your friends are in the retinal blastoma support group, you are sharing their status with that, organizer. So I think that often technology quickly gets ahead of policy. I think one of the things GINA has going forward, defines how to protect patients from harm. That should be some consideration as we think through how rapidly evolving the various models are.
So, I think the aggregate data warehouse has both strengths and challenges in terms of there are -- if you haven't pulled the data you can't go back and add later. You have pull a lot of data or sack nice data quality. Distributed models are much more agile but involved a more limited amount of data. Social media, as yet untouched, but getting way ahead of things, but things are moving faster there than anywhere else. In my opinion the role of healthcare information technology is to serve as enabler, help any of the stakeholders in the process. With that, I will stop and address any questions.
Thank you. Any questions?
No one?
Okay. Thank you, Dr. Hoffman. Now we will have our last speaker who is Mr. Robert Shelton from Private Access, talking about the consumer controlled -- a consumer-controlled model.
I want to thank the committee for inviting me to make this presentation, and the staff for making it possible for me to be here.
When I received the topic I thought maybe the best thing to do was modify the first slide. I would like to submit instead of thinking what I will talk about is consumer-controlled, think about it as consumer-empowered. I would like to think about consumer-empowerment not as just the consumer, but consumer empowering the researcher. A lot of this is about ways the consumer, properly empowered can enable the researcher to go a lot farther than the researcher is able to do.
I decided as sitting in the audience I am only person without an MD or PhD behind my name. I thought I would talk about who am I, the parent of a prenatally diagnosedded genetic condition. This is him at 4 and a half. I selected that for a reason, I will get to that. 47 XXY, a proclivity toward Klein fieldser syndrome, a 1 in 600 instance in -- roughly 75% of the people who have the diagnosis are never diagnosed, life, from birth to death. One would assume, as in his case, it's a pretty mild condition. According to most recent statistics, 2007 study done in California, roughly 70% of the parents who receive a prenatal diagnosis of Klein felledder's syndrome will terminate the pregnancy in utero. I added this slide set as sitting in the audience because I really would like to bring the individual perception and perspective of individual patients to this committee and say that the kinds of subjects we are talking about in macro, millions, tens of millions of people really boil down to individuals and parents make bad decisions based on limited data set and fear and lots of things that you all know very well.
So, that led me to basically take off from work for -- with my day job, for about three years to become, first the director, then chairman of the board of the national disease organization that supports Kleinfelleder's syndrome and -- I selected this image because the organization had been in existence in 15 years. In 15 years we had never had a picture of a person with Kleinfelleder's syndrome on the website, a lot of the people on the condition are afraid of being recognized as having the condition.
So there's a tremendous privacy concern among that population. It's not one of the protected populations in many state laws, but it is very high-privacy concerned. In part because it's so mild and in part because there's significant stigmas. This picture actually now appears on the website. Someone called our organization and said you know, I have decided I will terminate the pregnancy, I assumed since the Down syndrome site, you go to -- you don't see pictures of people with the condition. I decided to go ahead and make the picture available so people would connect with a person who has this condition.
Information is power. The other thing that happened, you will see I am also an entrepreneur, and have founded a privacy technology company called private access. When we got started we focused the technology in private access on serving some of the needs we recognized through the disease advocacy area. Today, roughly $5 million invested, almost 500,000 lines of code, I will show you what was made possible. No one would recognize our company unless they have heard us at a conference or met us person to person. But we partner with organization that are already trusted organization. We think of trust as being an extension from human being to human being, not based on technology, but based on human relationships.
Partnerships are really vital to us. Our mission is we focus on creating an environment of trust. I talked about how we do that. So if I start in this slide on the upper right hand corner, I remember going to a conference at the Health 2.0 two years ago, Esther -- was speaking, referred to privacy as the giant hair ball clogging the drain of -- and we need to blow that, part. Can be viewed as a speed bump that is keeping data apart and hurting liquidity. Achievable goal to enable health information sharing. I think what we are really talking about is how to create an environment of trust. Inside an environment of trust you get speed and there's books written on this topic and borrowing the fast company quote, the new E connee begins with trust. We have to build with trust.
The way we focus on this is looking at creating what we call the perfect balance between privacy and access or accessibility to information. You see one patient on the side that has got the ability to leverage their words to a consortium of people that are really out to help them to achieve their health goals.
The need for speed is something that is a disease advocate and coming from that perspective, something that is just really critical. When you look at the Internet, I presented at the Electronic Patient Record conference, TPR, mid-last year actually. I used an example, did it live, of searching for a person based on attributes, a public person through Google. In a minute and 27 seconds, not knowing the person's name, we you found them, how to get in contact interest and booked them for a speaking engagement. In a minute and a half in Google we can locate people. In Match.com, or Monster.com we can do the same thing. I went through an example, in underred three minutes I was able to locate a person that matched the demographics, location, characteristics I was interested in finding for a date or for a job.
In healthcare, however, we have challenges. We take six months to a year to recruit for clinical trials, have terrible accrual rates in trials, an average of 15 years to develop drugs for diseases. Time is actually more critical in the healthcare area than in those other two domains, yet we have the worst ability to move things quickly in healthcare. I would like to submit part the of the reason is because of the trust factor we need to replace.
There was a study done by Case Western Reserve focusing on drieded blood spots from newborn screening. The question posedded was how willing are you to have your child's blood spot sample used for newborn screening for future research studies, done with permission, without. Willing, somewhat or unwilling. Over sniff 75% of the people granted permission, information used for research, but when you change the equation to denying them permission, what happens, all of those positives around granting permission changed dramatically and the opposition to using information for research increases dramatically. To me that's what happens if consumers are not asked about sharing their data. A lot of the technology we have proposed is set up to focus on that.
That study is not discrepant with the secondary literature. In fact, it is very consistent with the secondary literature and I am just going to slide through these because you all probably know this data, but the one that is the most compelling to me is from the Institute of Medicine study at the end, 57% of people would permit their personal health information to be used for research only if various privacy conditions are met and 38% of those, the total, the largest share of the 57, want to get information and notice on a case-by-case consent basis.
So how does that happen? If we think of the world as data seekers and data holders, a data seeker can get in contact with a data holder, a data holder can put out a query for who has my data, data I would be interested in, and a hold are could say hey, I have information. If the conditions are right, terms are right, I am willing to share it with you. The challenge is for that data holder to act quickly, that data holder needs to know, do I have the right to share this data with that, seeker, that particular seeker. That entails a determination of what's per per miss able under federal law, state law, there -- what would my patient think, there are legal and reputational risks entailed in that. Those answers, particularly in order to answer a search query like a Google or Match.com, particularly for that level of speed those answers have to be fast, reliable, containing the information about what that record-holder will be compensated for the information. Even if the compensation is in mill cents or pursuant to contractual relationship between them.
So what we are doing in Private Access is replacing those questions with an automated transaction-based system that is programmedded with an ontology of privacy that looks at each of those issues, the institutional law, the federal law, the state laws, and the personal privacy preferences expressed by the individuals to give that record holder, data holder, back that information in under a second or two. soso that would allow them to know red light, green light, yellow light, do I have the right to move that data to that seeker.
In order to power that, remember the title, consumer empowered, we look at tying the patient in through the ability to dynamically consent or decline access to the proposed sharing if their voice is permitted under the prevailing law of their state or federal law.
We also allow them to view the audit trail of associated activity. The data can be pushed, electronically, or in the form of FedEx pouch, U.S. mail. Doesn't have to be electronic, but what is conveyed back to the system is an audit trail of the actions taken and any dollars, amounts of money charged between the data holder and the data seeker.
So little fundamental architecture is what we spend the money and time developing over the last three years. To date we focused solutionsy directed to registries and biobanks, to select all our part of person's confidential information to be selected based on the needs. The first focus is to set up a consumer-centric site, in most cases we co-branded with the you trusted intermediary, with the disease organization as being a co-branded indication. So each case we are working with a trusted intermediary. We use a system of trusted guides to help the individuals set their privacy prerc preferences. We get to the level of granularity of speed and accessibility it's incredibly granular. Someone has to do a lot of reading and clicking of on and off buttons. The patients I am familiar with do not have the patience or proclivity to spend that amount of time. People's perspectives, trusted guides, we select three in each case, those guides reflect a perspective of the spectrum from I am in favor of a lot of sharing of data to I am in favor of very little sharing of data, I am very privacy concern, very accessibility oriented and we ask each of those guides to pretend they are talking to a person across the table from them saying vie high prief of as concerns, what would you tell me I should consider doing. Or I have low concerns, what would you tell me? We get a broad spectrum of perspectives on what should the settings be and we boil those down to in effect, permitting access, in a limitedded basis, a prior consent, permitting access with a dynamic consent. We set a series of stops along the way for each of those perspectives on each of the factors involved. Then we, as I said before, have a comprehensive audit log for each access to the data and when the IAG standards are adopted in HITSBE and in the standards required for, which chip certification, hopefully those will permit the audit trail to touch any EHR, any PHR that is standards compliant. The patient can go to one place and see the access to their data.
The last piece on this particular element, identity verification is vital, we have identity verification up front in the system, have written the privacy directed language, robust ontology, the audit you tracking and integrated the commerce features.
The initial applications we have built, focused on clinical research. So we are using these to help people locate -- help researchers locate patients for clinical trials who wish to be found. We call this application we built recruit source, based upon a researcher centric site, a researcher can go in, enter a natural language inquiry based on text match or UMLS language for their particular query. If they are looking for Tylenol they would find A seat min fen hit in the database. That results in, depending on privacy, a fully anonymized or fully personal identified record, based on the demographics, location of patients, 10-miles of a radius of a spot, a research cohort to put together, and if the patient said I want to be in a de-identified form, the HIPAA de-identification says 32 needs to be less than 50,000 people in the radius and we have less than 50,000 in the data we can't show the data. We use that switch to turn that off in accordance with the federal laws.
Finally, we have the dynamic consent tools built in from privacy layer, if a researcher says I saw you in a de-identified form, anonymous form, I am interested in you, you indicated you don't want me to know who you are, address, until I tell you about my research project, then the researcher can push that information through the switch they can send information back to the researcher, or say no, I read about it, talked to my doctor, we decided this is not something I want to do, or the yellow button, snooze, say I will wait for this answer for a while.
The first project we applied to was on the organization I chair, Kleinfelleder -- research, action, people didn't want Klein felledder's syndrome on return envelopes, we looked at 1200 patients, 9% IND qaited the system is easy to use, liked the experience, 75% said they would recommend to family and friends, and the experience would be the partnering with the trusted source was overwhelmingly what drove the patients to have an interest. We have a number of research projects under way, one is in the packet of materials you have been given, a project with the University of Michigan, focused on newborn screening blood spots. It was a challenge grant award for 200 or so of the challenge grants awarded. It is presently ongoing, looking at facilitating a state sponsored birth cohort to use the information for genetics testing, and looking at the use of consent for that purpose. This is the steps of the process. We are presently at the stage of creating systems and environments, and early in the summer we will begin with the pilots, recruitment for that study. We are very early in the study, but excited because all of our prior work came with working with disease organization and this is a general population as opposed to a specific disease organization.
We are pleased that we have strong support from some industry stakeholders. At the end of last year we announced a collaboration with Pfizer, Greg Simon, the leader of worldwide policy announced patients are the most important stakeholders in research, by merging respect for privacy with access to relevant actionable medical information we are giving patients more control over their destinies. This was echoed in the December by the CEO of Pfizer who in front of 650,000 people said, answering what he was excited about in terms of accelerating treatments, he said focusing on patient privacy and a technology that would ac El accelerate the ability to get in touch with patients who want us to get in touch with them. We are finding support for this. We focused on collaborations, most proud of genetic alliance, the public/private partnership, Pfizer, University of Michigan, other projects not announced yet with significant disease organization, working with a couple of government agencies on using this for their informatics grid type computing, and several HIE, health information exchange, regional health information organization for their applications and thern we are pleased to be on two sharp proposals for the recently announced IT initiative from the office of national coordinator where we -- one with Harvard/MIT talking about deidentification of information, and C-disk, the exchange of information -- hopefully we can, maybe in a year, come back give you a lot more data on how this works.
Thank you very much, Robert, and thanks for sharing your personal story. It helps remind us why we are here. Any questions?
Paul?
So, I understand your roll-out is limited, maybe you don't have data yet, but do you have a sense of when you put your processes in place do you have more uptake -- I mean, I don't know how you do the comparison, but a sense you will foster more research participation? absolutely, there's no question. The Harris poll has been done for a decade, the Institute of Medicine, 2009. Alan Weston is an advisory board member of our organization, we have seen his polling results, the raw data. His results mirror pretty closely what we are finding in the disease advocacy world.
He's a character in privacy, I believe.
He's certainly the dean of it. I think he wrote the definitive textbook in 1967 on the subject, privacy in a free society.
Thanks, you all.
Thank you, we are going to open up for discussion now. General discussion about all the topics. I am going to put up a couple of slides.
We heard a lot of information today, a lot of interesting information. I want to go back to two questions. What have we learned, the lessons learned, I will change that. And are there issues that warrant further policy considerations, and -- that SACGHS might consider, talk about what we learned, what stood out we are trying to aim towards best practices -- are there gaps, issues, -- where are we on this trajectory and where are the gaps where we might weigh in, or do we say these groups are doing a great job, we should move on? If we can talk about what we heard and, as Charmaine said, what is already there and working well. What are those needs going forward, and if there are those needs, is there a role for us to weigh in?
You need a mic -- not sure it's on.
Can we have all the speakers come up front?
I will finish my question. I am sort of struck by the fact that people like me, who are trying to develop large resources, also very focused, part of what makes them valuable is that they are epidemiologically sophisticated, or known populations. There are a lot of issues about, possibly, even though obviously everybody involved is a volunteer, and in some ways the private access, more the private access model, and yet the situation I con front is one where where I -- it's very expensive for me to update information about people in my cohort think about different research projects we might do, or different new emerging issues regarding privacy and confidentiality that come up. What I am trying to get to here, I guess, is it would be very helpful if within the context of the epidemiologyally defined populations we had the technology to allow the kind of communication that in some ways your system envisions, would be possible between researchers and participants. Right now that's prohibitive for us.
When I mentioned a couple of government agencies we are working with, it's based upon exactly that concern. When you were speaking earlier you talked about your definition of deidentification you used to establish the policy. Well, one thing I know about politics is politics is 50% plus 1 vote equals a change of policy. What happens if the change of policy on deidentification is different after you set up the database, does everybody have to be re-consented? There's tremendous challenges that cause the need for deaggregating the switch from the store of data. I think the metaphor I use is in a baseball game you have an umpire. The umpire does not care about the score of the game, does not care about what the batting average of the batter is, the pitcher's no-hitters, calling balls and strikes, all their job is. There's a role for that in this system, I believe. I don't think it ends up being one company. It ends up being multiple entities that play the role and the ones that play the best will end up prevailing. Our am bish yon is to be one of the parties playing the role and in addition to privacy patient -- a tremendous thicket exists by virtue of policies that is mind-enummingly complex. In order to address that, there's two things necessary. One is somehow or another developing a -- processing it, we have gone a long way towards that, but most of the processing -- I had a discussion last night with a top private as advocate, most of the resolution, A adjudication would yield Amber lights, not green, not red, somewhere in between. And we would say it's ambiguous. The law is ambiguous. The law we are putting in place pushes the ambiguity back to the policy makers, legislature, the institution, with the question, we're the umpire, not trying to set the policy. We are trying to say there's a challenge in processing this policy as it presently exists. Under a robust ontology. You as the policy maker need to decide do you believe the law should be permissive, [indiscernible] we can model, but once you make the decision, adopt it, if it changes next legislature, thern the rules change, and how the system processes.
I think there's a role to play, whether private access, [indiscernible], I think there's a role to play for that type of technology inside the architecture that reflects not just patients, but each of the persons who are stakeholders in the moving of data.
One of the things I heard from many speakers were a number of elegant solutions to some of the vexing problems. The other thing I was struck by was how under some of the efforts things are going to be put out into the public domain, [indiscernible] others, particularly struck by the idea that you could quantify the level of de-identification in a robust and repeatable way. I mean, I can think about, those of us that deal with IRBs, issues of privacy, de-identification. There's something about a repeatable quantifiable way, that is highly attractive. I can't imagine people wouldn't be interested in that. The point of that is that clearly there are innovative people coming up with solutions, is there a role for DHHS S, Secretary to aggregate, try them out, contribute data. In some ways it resembles what's being done on the GWAS side, we think there's important data, people are interested in it, we keep learning new things. That to me was the most important thing I took out of the discussion.
People aren't born 55 years old; and I know you've thought a lot about the developmental precursors of many of the diseases you are concerned about. How are you going to integrate the developmental aspects of ideologic cascades into your analysis over time, given you are starting with older, consenting patients. How do you see the developmental processes entering into the database as it grows and matures? Like adding children, reproductive outcomes, maternal histories, things that could be attached to the progression, emergence of the disease as people grow and develop.
For Katherine Dr. Schaefer?
We are beginning to build a preeing Nancy cohort now. I have been very fortunate working with an established cohort, the child health and development study cohort of some close to 20,000 live births that occurred in 1959 to 1967, part of an NIH funded study then when they had the foresight to store maternal serum, still available for analysis now. We used those samples in follow-up studies of schizophrenia, to investigate the role of maternal exposure to influenza, other viruses, show maternal exposure to influenza in the first half of pregnancy is associated with a significantly increased risk of schizophrenia. I am definitely a believer in the importance of better understanding of developmental contributions to adult diseases and have benefited, have seen very directly the importance of this. The prenatal period may be extremely important and unusual, unique in terms of the interaction of genes and environment and playing a role in adult health. If we neglect that we really won't have a full picture of the origins of adult health.
I don't know that I have a complete answer to your question, we would like to be able to to add children. In part it was a resource issue. We had a very limited amount of money to initially survey our population. In part it was the complications of asking parents to consent for children about being part of a very long-term study. We are now initiating a pregnancy cohort where we will get samples during pregnancy and intend to follow the children born of those pregnancies, and enroll them, get samples as well. That's about as well as I can do right now.
It caught my attention about the two speakers, Kaiser Permanente and Vanderbilt, the ways to recruit patients. Your system -- eight-page document, [indiscernible] tough to send DNA back, the Vanderbilt system is -- system. I am wondering if, NIH level or the federal government development, starting to develop best practices of what actually -- informed consent for these type of studies, where all the -- authentication, reidentification -- do they fully understand what they are consenting to.
I think that was for me. So we are not to the point in the evolution -- let alone genomics, concepts are changing so much around sharing of our data, what we are doing. We are srnlt trying to think about it for genomics, we have a resource on the website that has suggested language, which actually we have example language, not even suggested, people have IRB approved consent forms for different types of genetic studies and we are trying to make them available for other people to see how it works. We are hoping that will be a dynamic resource we can get comments on, can build from to have ongoing conversations with the community. At that point I am not sure we have a best practices. I know it's something the community wants a lot, they want to know the right way to do it. I don't know there is one right way to do it, it will always vary for the patient and subject population you are talking to. It's hard to quantify the risks right now, hard to know what to tell them.
Question: I think that's concerning. We are recruiting patients, we don't really know.
I will like to put Lara on the spot, as you look from a variety of federal agencies, in the beginning, different things we are concerned about, informed consent, storage, access and secondary uses, privacy, confidentiality, deidentification, handling sensitive data, the whole ball of wax in EHR. The question is, to what extent do you think -- we are already close to having good practices, if not best practices, where do you see the real gaps and where, do you think a committee like this could help shape the -- what are the key issues in moving the field forward. Or if there aren't any, that would be fine, too.
That's really not [indiscernible] part of the first question, you asked if we were close to good practices and my response is on which one of the many issues you mentioned.
Tell us if you think we are close on, which we have a lot of work to do.
Authorize work to do on There's work to do on all of them. We are still trying to understand what participants think. We don't have research data to know the risk tolerance. Do they care if we share their data this way. If they are trulial truistic, we may be worrying a lot about how we manage, how we share, not a concern. There are studies going on to collect that information. I think informed consent needs work, needs hard data to understand what people understand. What's the best way to communicate risk to them. That's at a much more granular level than this committee can do. It's hard to say. I am not -- things are moving in such an amorphous direction, I don't know if I could come together, get the one thing this committee some do. There's value in having different approaches go forward, and learning from them. So I am not sure I have an answer to your question.
Is it a role,y then to look at the different approaches, as Charmaine said earlier, identify what the different elements -- the different options, which look most promising?
There's some of that, one useful thing is to try and articulate principles useful to the group, trying any of the different solutions, if there are common principles that should be present in any model that went forward. That would be something that I think is at a high-enough level, providing leadership to the field to continue to try to experiment, modulate what is going forward now.
I think one issue, and Robert actually started to bring it up, but didn't bring up the entire context. Sharing patient data is very different depending on what you have been diagnosed with. So, I am a long-term cancer survivor -- doesn't have a stigma. Breast cancer had a stigma, all cancers have a stigma. Your son's disease still has a stigma attached to it. You can't say this is the way patients feel about sharing their data. People know because I am a cancer advocate, cancer survivor, but the disease is de-stigmatized. It's difficult to make blanket statements about privacy, sharing of anything when you are talking about a very large universe of diseases.
I couldn't gray more, and I would add mother vector to that, the stage of the disease. It's not just the disease, it's the disease state -- you are trying to say something --
Sorry, absolutely true. I work more in the cancer field, but very true. People are often less concerned about themselves than they are about family members, the impact to family members. We already -- we assume, if you are a survivor of serious disease, everybody knows you have it, you care about what happens to people who are close to you. It's stage, it's very, very nuanced, not something that you can just make a blanket generalization about.
I was going to go to the same place you just went. To build on it I would say before I came here, I always talk about my son, I think of my [indiscernible] dad. He's now 11, so before I came here I asked him whether he minded if I mentioned his having the condition. When he was five it was my decision. He's 11, not the age of consent legally, but I would like to know if he has a sensitivity about mentioning. It's starting to become his decision, not mine. There's a lot of nuances to all of this. The irony I see, from a patient advocate perspective, irony in our organization is I talked about the high termination rate. There's not a single person who ever terminated the pregnancy they were carrying. We have an organization, hotline for patient, parents that want to meet other parents, or children of other parents. One of these people decided to go forward with a pregnancy, scheduled for a termination on Tuesday morning, met them Sunday, changed their mind, has become a long-time friend, become close to us, we are invited to birthday parties, events, always 15 minutes early, using that time to remind us that no one in their family knows thirty their child has this condition. Their primary care physician doesn't know, they don't want him to, none of their educators know. Compare this to the person that comes to every national conference, outspoken, participating in clinical trials, tremendously active in a research context and context of -- for the condition -- same time parents, primary care, don't know, not who they believe the knowledge will help them.
>
And their case they do not want their son to be treated like they could not be doing something in baseball bid that is their particular concern with athletics. So every single person who would I have meant is a little bit different. One person has this and someone has something else. Policy wise as scientist and this committee, there are some fundamental policies that could be developed to empower that tremendous any clarity in society to take place. And to the gentleman's question that asked me as I was standing up at the podium when I believe would be a result of these policies I think it is a test of the hypothesis if the agency wanted to find out. What I believe would be the result is that more data would be shared in the paradox here, I believe is the Marco Foundation has done research to say the number one behavior is a failure to disclose and number two is destruction of fact. So maybe there is a way to encourage people so that they do not fail to disclose and do not lie about their circumstances but directed dated to go to the increases to help them or to help a family member or for altruistic purposes. The empowerment of that I believe will result in a proliferation of data, not a repression of data. That is a testable hypothesis and hopefully that is something that could be tested and demonstrated as, true.
I think what I would do is state the obvious. It seems the common theme that we're hearing is the notion of community engagement. If people feel -- if people are involved in decision-making the have some say and have a reason to participate and we know the research is out there but when it's a feeling of lack of control is with there is a lack of trust so I would say that is a very easy thing that I saw across all of the speakers today.
Anyone else?
I think going back because the question that we have about best practices, going back to what Laura said and sort of for Barbara is saying is the need for principles as opposed to Best practices per say. It was kind of hard to assimilate all of this information because there are some similarities in there are some differences but there are principles, a common thread of trust or engagement and privacy. I do not know whether we think at the committee that that is something that we may want to tackle in terms of coming up with principles. That probably could be applied across the board or should we wait for the Lewin Report? The to started the work with data sharing and theirs is a yearlong process and they will do interviews with various state clickers and then to a report. Do we want to wait? Would it be best to wait to that report to decide if SACGHS should do something or are there things that we should do now?
I think this is probably the worst time to come up with an answer to this question. If you want to ask the committee you may get an answer different today than tomorrow morning at 8:00 and people are fresh.
Yes. Everybody looks tired, I think. I think we will go ahead and close out the session or I guess, figure out -- maybe the steering group can come together to think about how we want to move forward.
We have all of this clinical data and clearly have research needs and producing the boundary between what constitutes clinical information systems will break down. And how one can build into the appropriate protections and still allow research to move forward. I still think those are some of the compelling issues that we're facing him and I am hearing some attempts to try to deal with in terms of the technologies and information about there. It seems to me there are still some policy issues and we're still getting focus groups and finding out where the boundaries are. There are still some very compelling issues and all of this bid I am having a hard time putting my finger exactly on what the next upswing before us and I also feel some lethargy here is keeping us from articulating this very well. At least a couple people have awoken to that comment so I think it is worthwhile that we ask you to sort of take it back and began to articulate what it might be based and what you have heard here or we think we need to go. The champion for this was Kevin before he got off the committee and maybe we should re-engage him. He is still on this task force to help us with some next steps. The real question is will we have enough to come out of the Lewin Report Cliff, I know that your reports are always brilliant. Do you think that we are well advised to wait until you complete your work or do you think there are some things --
Cliff, before you get started I want the group to know that you are still in the initial, you have done an initial scan of the literature but is it a little unfair but I'm glad that Steve called on you even know you are in the beginning stages but I just wanted to let the committee know that it's under way.
First of all, time internally cripple for Steve setting the bar ever so higher every time -- eternally grateful for Steve setting the bar ever so higher every time. What I can give is an approximate timeline P we have a draft literature reviewed to towards the end of April, I believe April 19. The final literature review about the end of May, May 31st. A draft final Report at the end of August in the revised final report the first week of October. That is are tying. I would defer to my task Officer Sandra Howard is here as well if she would like to speak, too pin what I would like to say to Dr. Teutsch and the panel. As you progress, and this particular issue becomes more clear to SACGHS and any particular areas that you would like to focus more per serving better, we can. We will proceed with our report on this schedule and three your staff and my boss can communicate about how we might adapt our reports and needs within the scope of our contract.
It sounds to me like the timing is actually good. Our next meeting is managing in by that time should have a pretty good picture of what you're finding and what they believe issues are. That could inform Charmaine and her group so that we could have a discussion about its next meeting pit and help us to begin to see where there are some information gaps, if there are in the.
If that is okay with Sandra Howard, that is okay with us. By then we would have a literature review and not the final report. The difference is that the final report will not only include the final points but also multiple interviews so that will be another chunk of input that will occur late spring, over the summer.
But you can potentially give us an update about where you are.
We would be glad to pay as I stressed, we would love to keep in contact with you.
I want to make sure that I heard from you correctly. It sounds like there may be some work to be done in the interim between now and June. So two questions be the first relates to whether or not there would be input now from the group that would direct the literature review that would be useful or whether that will not be useful at this point. The second thing is that clearly there is this interview process with key stakeholders and it sounds like a lot of them are sitting up front here it sounds like that would be a rule that the task force could presumably do is create a list of potential interviewees for that.
Yes, we are tasked with putting together a draft of such interviews first for a review by our task group officer and clearly input from you would be most welcome, yes.
How about the literature review question? Is the input at this point that would be useful or is this a process that is already going and we should not perturb it at this point?
It is already going and we do not mind. I know Sarah Carr may want to see the set of questions. I will be an example right now because any stiller piece of literature that we cannot do without on any of those questions, we would like to let us know.
So if those questions are able to be sure that would be something if I were running the task force I would be really interested in seeing them based on the work that I had been doing to say, wait a second, I am not sure whether this is actually well presented.
You are quite familiar with those questions of First.
Charmaine has seen -- I am not sure which staged the iterations appeared so we will keep our task force active. And listen closely to what is emerging from your work. We will hopefully be in a position to have a more concrete discussion, first of all, we were clean and second of all we're -- where we should be going in second of all where you should be going by June.
So we have heard a lot today about a different models. In some ways for collection of data also for sharing of data to some extent. I was wondering if it would be useful to hear more at some point -- it would be very helpful to me if I could report back, for example of corporate information about what has the experience and today? Is there any kind of collection summer of information about what has the experience been to date? What has their experience and like? Have there been problems was sort of results, are there other sort of the news for genomic data sharing and as we learn more about any of the experiences today that would help us kind of shape, you know, I learned a lot today about the structure but less about what is actually now happening since this policy has been in place and as you said it would be helpful to me to be able to reflect to people more realistically about what the experience has Ben and therefore better inform people about what the issues are when I am trying to reformulate a consent or something like that.
Could you be a little bit more specific about what you want? Are you talking about adverse events, a routine operations and how smoothly they go?
I think both of those things, you know, would be helpful. I am not aware of adverse events with respect to individual peaches and things like that. Gist of the work that has been published about the increasing recognition of the potential for reidentification so that is quite relevant for informing research subjects for example. But more generally how is it going? What has it been like from the standpoint of the different stakeholders' involved. The scientists and the people who state that is reposited.
Laura, could you work with the task force to put some of that information together?
Yes, we had a session similar to this the year before last we heard from institutions and investigators about their experience interacting with it and also putting together what the experience at NIH has been.
So some new world practical experience. And if you have some of that, I think that would be highly informative.
The reason that I lighted up to see something was Marc need some comments about virtual farmers and they spoke and real farmers and to think about the direction of technology and what is happening in the future because there is a wonderful video by Kevin Kelly talking about the fact that the Internet that we know about as of today came into span over ten years. And not this the past ten years but what the next ten years would entail. And what kinds of technologies could be if the www.continues on the trajectory that it's on. It seems to me that you could never predict the future but there is a tremendous amount of activity going on in the way that interfaces are connected and the way that the data can be required. So continuous glucose monitors or Nike shoes that collect biometrics that can feed the research so that the data is continuously updated and entered into the record without needing to have a patient provider and counter and that kind of information and dealing with the challenges in control of that information and how you build trust is something I hope you can take a look at.
Gwen.
I do not know if you were here in the morning but the theme of the date is Wayne Gretzky's hockey puck quote. And it is true, I sent my daughter is a reminder and she said, I did not look at my e-mail, I was on vacation. She is 13 and I think the hockey puck is a perfect metaphor.
That is a good way to end.
So many thanks to all of our guests and to Charmaine for leading us for. We will engage Cliff and look for it to have been more discussion in June.
We will keep piling on, I know that this is biologically challenging what we will do is go through several of our federal colleagues reports that were originally scheduled for tomorrow. I think that shows the flexibility of the federal workforce and their ability to help us in times of need. I appreciate that. The first Speaker -- I think you were focus is primarily on Healthy People 2020.
It is 5:00. It is biologically challenging for me to be here. Usually I am at the gym trying to get my muscles moving but his heart. And having a bunch of Feds at the end of the day speak to you is not a nice way to spend but if you have any chocolate next to you go ahead and eat it because you need that energy. I was given the task of talking about the healthy people objectives. I will tell you folks something that depressed me while I was putting that talk to get there but before we get to Healthy People 2020 I want to tell you why this is important. This is where the puck is trying to figure out how genomics fits in. I want to start very quickly with this translation gap and since some of the people are new, sort of where CDC is coming with this bid I do not have to be this point. You have heard this, the promise of technology is encouraging. We are about prediction, personalization and prevention. This illustrates to you the promise of the technology but the reality actually, Francis said back in 1999 that in 2010 we will have it all sold. He is not here to tell us why we do not have it sold in 2010. So there isn't evidence gap, translation cap and some people call it dilemmas or some people call it gap and it has to be filled with data and this is not discovery dated anymore. This is complicated data. You heard this morning from Marc about the transmission pathways and it gets more and more complicated the more and more I draw diagrams. You really need these other disciplines to come in and help us including clinical research, of behavioral research and health services research and what we need to do is figure out when to we have enough to make an evidence based recommendation. And the challenge is sort of what you heard this morning around the discussion for CER which appeared I will not dwell -- I will not dwell on this diagram and this committee is really on top of this because you are at the intersection of research, health, and society. A few years ago I took a look at the amount of research done in genetics and this translation space and I was depressed because out of 350,000 articles published in the literature of there was 2% or less in that space. This is a space that allows us to do evidence space recommendations or outcomes. There were two evidence based recommendations done by the task force one on BRC1 11 years after the gene was discovered and another one on hemochromatosis. And that was, we the minute we are not ready for population testing because we do not know what the natural history of this is. This is one reason why CDC started the process which I will talk about in just a minute. Since I spent so much time at NIH I want to figure out why we are here the way we are. So this is a paper that just appeared in the press a couple weeks ago on the investment in cancer genetic research portfolio and to cut a long story short there were about 1,000 extramural grants funded by NCI in 2007. A hundred 27 of which were purely discovery and 174 are really translation. There was only one funded research and out come at the population level. So we are not doing enough investment in this area and actually the numbers will even be more skewed out because 2007 was beginning the inflection point for GWAS so now we will be planting even more discovery research bids of this is a Safeway for why our office existed did not to bore you with too much detail it is all of these disciplines coming together to figure out an effective and responsible way of translation of these discoveries to improve population help. This is a good segue into Healthy People 2020. Our new director said in an interview in a journal on the Constitution in the beginning of a year, the single most important thing that public health can do to increase the degree to which decisions are made using good data. And those decisions did not have to be public health decisions. They could be clinical decisions or health services decisions. So this is a lot of the guiding principles behind CDC at close surveillance efforts and surveys. So that is what we have been trying to do for the last ten years is develop a portfolio of a number of projects. There is no time to go into them but we are trying to figure out if you actually there are more than 60 ongoing studies at the CDC to figure out what does genetic information mean for community health. What does it mean for this population or the population and a number of surveys in that space and that inflection point between resources and practice is sort of what we have worked on copper to meet with AHRQ and other groups through the initiative to need more to actions and to identify gaps for further research that would lead to further research to find the actions. And then to a number of collaborative initiatives and the workforce issues that we have funded a number of translation resources and programs to validate the genetic information into practice big refund the great state of Michigan and Janice can tell you more with what we're trying to do with the cancer genetic recommendations and there are a number of these programs that are being done. So this leads me to the 2020 objectives paid some of the tech line here is really important for us to think about is what gets measured gets done. Or, in other words, what gets measured gets funding because there is sort of have been your finger on the pulse of the nation's health. This is a activity that has been going on for years and they had the 2010 objectives and this is the beginning of a planning process for healthy people 2020 with four overarching goals and you can read them. And under that ecological model of the seas they have the determinants of Health and biology and genetics is one of them swept least we have achieved a certain stature in the lingo of how -- Healthy People. So that was encouraging. There is a federal interagency working group that has 55 members representing 24 HHS agencies and offices and includes non-HHS federal partners. The mission of this remark is a society in which all people live long, healthy lives. And there are these five missions statements which include improvements and priorities, increased public awareness, provides measurable objectives and goals. I would add to that in a minute which was quite depressing for me pinky to multiple sectors to take action to strengthen policies and improve practices and then identify critical research needs. You can see what I have been trying to do with in the CDC framework would this is a national effort that can actually helps genomics in a way. Anyone who wants to propose objectives means to fulfill these eight criteria. It has to be important and understandable to a broad audience. It has to be prevention oriented achievable through various interventions. It should provide action, useful and reflect issues of national importance. Measurable and address a range of issues. Build on past accretions of Healthy People, based on the best available scientific evidence and remove disparities. So I think each expectation should have a valid, reliable nationally represented Data source that could be state or national. You have to have base lending data and then you have to have an assurance of at least one additional data point through the decade in 2010. Each of tactical have to have its own target or methods. Each objective will be approved by the Federal Interagency Committee. In 2010 there were two over arching goals and one of them was health disparities and the other was on hold the line speed 28 focus areas and 467 specific objectives and no genomics focus area objectives were done under the newborn screening. I'm not putting this in that right now. There were some passing references in the genomics but nothing was measured or down for 2010. Now, we got depressed and we decided that we needed to have at least some proposal for 2020. So we propose to the Federal Council that it could be useful for the Healthy People 2020 initiative to develop a work group and objectives to help insure that rapidly advancing knowledge is translated into practice to maximize the benefits and minimize the horns. So they said, go ahead and do so Katie Kolar and Gurvaneet put together this and if you have any questions you can ask him since he is sitting at the table. So they promote evidence based practice in the first is to increase the knowledge base including more translation a research studies and evidence based recommendations. That was rejected by the federal panel because it was not measurable enough and did not fit the eight criteria that they put together but they accepted the second one which is the increasing implementation 14 no applications. I was quite depressed that there were only -- for genomic applications but I was quite depressed because they made the recommendation back in 2005 and I said to the group, there must be more that we can do by 2020. Right now we're thinking about what to do with this. At least in the space of these two conditions, it is very clear what you can measure. You can measure the increased the amount increased proportion of persons with newly diagnosed colorectal cancer and you can increase the proportion of women with a family history of breast or ovarian cancer who see genetic counseling. At least it's clear what needs to be done with these two conditions and that can drive both the data collection and implementation and national statewide. So this is sort of what happened in the interim period there were comments received and five objectives on the comments and I think SACGHS put together you're own comments. What happened since then is no changes in the proposed objectives but some will be incorporated in the near tip of the topic area. It is every intention that as new evidence based recommendations, space there will be added to this rather meager sort of genomics and population health. And this is how things are measured in terms of life saved and all of this technology, we're still in 2010 and I hope there will be more to discuss and use by 2020. The net steps are the final distribution of the public comments and it would surely be very useful for this committee to weigh in and tell this working group for they could add more genomics objectives and where some of these points of implementation can be sued thank you for the much.
Thank you appeared I believe in Tab 8 are the comments that we sent in on Healthy People. Any comments or questions? Anything that you want to say, Gurvaneet?
Marc?
I do not think that you should be that depressed. Over 2010 you have had an infinite improvment.
The promise is surely much greater than these two conditions.
Muin, when we submitted for 2010, they did not take any of them.
We got two.
It is not over till it's over. These will still get scrubbed a fair bit over the next few months. Released later this year sometimes.
In some sense this reflects something all of us when we really sit down to look at Genetics in the cold, hard light of day in comparison to a lot of the other things. I knew was looking at this in perspective to the coronary artery disease where we're putting work for a grant application and essentially the recommendations from Healthy People 2010 are moving unchanged because nothing happened in the interim. That is a frightening thought when you think about the overall process of preventive medicine in general in this country. So in some sense while I obviously have committed my career to this area and have heavily invested in think there is a lot of promise the reality is a lot of the way that we deliver health care in the system is problematic. Does not so much that we do not know what to do its that we don't know how to do is.
Speaking of heart disease I think one thing which we might hopefully integrate in these recommendations. The group in England has produced -- in the summer of 2008. Now HHS here has not considered any evidence based recommendations not seem to invite HHS and I think that is a very rigorous process and we might want to try to insert the recommendations or let me be the EGAPP fourth task force to look at that because we can save some lives in addition to the General presented strategies around coronary heart disease.
Thanks, Muin. If we need to find those things that are affected that can be done that will have a measurable impact. There were other objectives in there, not so much on genomics the other things that are pretty obscure. And the I suspect will fall by the wayside. I think part of our task is not just to talk about the hope of genomics but to gather the information that Muin was talking about so that we can have effective technologies that make a real difference and move into practice in -- and be part of the mainstream.
It seems to me that where we have granted -- this will not necessarily health us in this case but be that as it made there are a number of things where there is a relative underpinning of understanding in the family histories at least a contributor to its. So as a cross cutting kind of thing, the collection of that information particularly in the area of how that affects health behaviors. If something like that could be included there would be a high value to that.
Definitely what Marc is alluding it is at the recent conference for improving health. And if you want to talk about being depressed, at the conclusion of that report is that there was insufficient evidence that family history can improve helped. So what ties together, is that they are all [ Indiscernible ] conditions for which a family history is. Important and is part of the Cascade testing appellative spread but I think they were evaluating the role of family history in general as a tool for Health promotion and disease prevention and they called for more research of the type -- and the type that CDC had sponsored they're doing in randomized clinical trial to find out whether or not if you give people a recommendation based on their family history that they would do something. Believe it or not there are no clinical trials that put the family history in an evidential basis. So we will try to answer family history in any number of ways on the report but to have measurable things by 2020I am afraid we will stick with a single gene disorders unless there are some profiles that were sure over the next couple of years for which measurable things can be done.
Next comment, David?
I was going to comment in that sphre you could have a measurable outcomes for time to diagnosis for even more the common single genes. Even at this meeting with rare diseases yet that is a great frustration and there is so much anxiety created by a delay in diagnosis and that it's at the unevenness of health care in our country.
I think that is a good point because there are thousands of genetic conditions for which this may apply. I wonder if, Gurvaneet, your group has tackled the speed I wonder if there is a genetically to add something along the lines of earlier detection or earlier diagnosis for any genetic condition. They might come back and say, show was that this will actually improve outcomes. So I do not know if you have any comments on that but that is a great suggestion.
I think that is a process, Muin, that makes sense. The challenge is how do you define its? How much time would be ideal and how would a very across diseases and conditions? So I think that is a useful thing to explore and we get some standardization on. The challenge is putting this with Healthy People 2020 is there a way to extract the information from the current health care delivery system infrastructure and that would be the real challenge.
I think there is. That is, the simplest way is to have a survey of people who have the diagnosis made and how long it takes. There are population ways that you could approach eight, T1. Anyway, it is measurable.
Thanks for leading the discussion.
Let's turn to CMS and Jeff. They have been highly responsible for many recommendations and last week there was a meeting of four and the -- for anti-cancer therapes so they smack for being here.
Let me mention that there are none too relevant advisory committee meetings and I think thank my colleague Penny Teller just last month they meant to look to make sure that for genetic testing in particular not only the appropriate reference materials and challenge samples but also the survey infrastructure and data collection tools were available so that genetic testing the assays can take advantage of the same type of allegation that some the other laboratories get rid I just want to see if Penny is still here might wish to comment further on that.
I just kind of wanted to update -- we just initiated it appeared we had actually gotten the approval last year but because of the H1N1 epidemic all of the agency's work very busy. So we had the initial meeting in January and another will be scheduled for March. I am sure there will be a series of them. Except for the cytology proficiency which has been in the works for five years, we got some public comments back. That kind of opened the door to look at the proficiency program overall and, of course, genetic testing will be an issue but we kissed and she did it but I thought that we would share that.
Thank you.
Just about a week ago yesterday we had Dr. Goodman and Dr. Teutsch be part of a MEDCAC panel about from a -- about pharmacogenomic testing. We were very lucky to have none to distinguish people to health us understand some of the issues about whether the quality of evidence about pharmacogenomic testing actually improved outcomes because we are kind of interested in outcomes. We were very fortunate when Dr. Friedman not only proposed a great many lessons about some potential in this area but also gave us a vision of the future with some and could bring in the sequence to the pharmacy and get the appropriate drug. Also he was kind enough to point out that this area has received interest in the attention from some fairly high place elected officials in the past. We also were grateful to the group at Tufts about a specific tests that can be used. And in the interest of time, again, I am going to ask you to look potentially at the materials which I believe are or will be available on the table tomorrow and focus a little bit on the public comments that the committee heard. The first was echoed earlier today by the need by some especially the testing community to clarify what the responsibility is and really what CMS is interested in me a call for clinical utility studies. Second we are very much aware from public comments that there are significant barriers to clinical utility studies especially those that may turn out to be somewhat negative in terms of the potential role of these tests and outcomes. We also heard very clearly that some of the studies have been used for years and they are now considered standard of care. They are part of many clinical guidelines for the treatment of cancer. And that these are now being integrated by some of the large organizations like pharmacy benefit managers to make sure that patients for whom such drugs are prescribed have the private testing to make sure that the drugs will make sense. In addition we had a very interesting public comment from Dr. Novak in which he revealed to the committee that there are, indeed, a couple hundred laboratories for subscribe to CAP provisionally testing the studies that have signed up for HER2 testing studies. A somewhat smaller number perhaps because it is A new coremac program are signed the for KRAS testing. This reflects that laboratories are looking at this for an important area that they want to make sure about their accuracy and validity of testing. Finally he revealed that a majority of laboratories in the United States are interested and especially those better members of both CAP and AMP are interested in the first three tests. The MEDCAC panel as those of you know who have read some of our accounts of it essentially tells CMS what level of evidence we should have about the value of these tests in terms of determining clinical outcome benefit to patients. We use a fine point scale with a bill to the low degree of confidence -- number one is a relatively low degree of confidence. And a five is high confidence. The first question we asked, we asked the panel to tell us about their impressions of the level of confidence that pharmacogenomic testing affects health care outcomes. In these five particular situations that we know that there is testing out there and it does affect some cancer treatment. We set a barrier of 2.5 which is in the open less than some confidence to distinguish those tests with relatively larger confidence from those with less. This was about the question to the effect health outcomes. Clearly it is up by the panel to be supported by sufficient evidence to say with some confidence, with a high degree of confidence that there is an effect on patient outcome. The second is a follow up to the first. Does the panel believe that based on the evidence in this was presented by several groups as well as a packet of information that such tests improves health care outcomes and with these specific regions and let me mention that for BCR-ABL it was diagnoses and monitoring that, indeed, the HER2 and finally KRAS testing were enabling High testing where as BCR-ABL which was used to detect treatment failure mutations which would make a patient more liable to be unresponsive was not built at this time to be at the same level of confidence. The panel was also asked to suggest whether they had a level of confidence about the general -- of these findings and there was a Pyrrhic does fairly high degree of confidence there might as well as for general liability. Finally the panel was asked to talk about evidence gaps that they felt could improve the evidence that CMS would consider in looking at future covered stations and let me see we are not currently looking at any coverage decisions for any of these tests individually or as a group. And, in fact, the concerns about morbidities' especially things like polypharmacy and nutritional status to standardize the genome or if genotype or phentotype assignments. The importance of being able to maintain tissue and DNA sources and finally studies representing more tapirs patient groups were all mentioned. But I think as was mentioned earlier today the final question place to the panel by Dr. Goodwin was whether there were any particular high plains and almost unanimously the panel said the evidence providing additional information of clinical utility of these tests including quality of life outcomes would be welcome. Thank you very much to all of the panel members including the two presence here today to health CMS understand this issue better.
Thanks, Jeff.
Did you want to add anything? As he is the chair of the panel.
Cliff does many things.
One they had to do with the concern about many of the presenters about whether someone will always demand RCTs I think it was fortunate that we were looking at KRAS for example and I think all of you know and Marc may have addressed this. The evidence that impressed the panel by KRAS was based on retrospective some analyses of RCT data. So what we probe was, do you need more evidence or not and if you do what kind is eight? The answer that the panel gave is that yes, we to need more evidence but it may not be RCTs that was kind of a place that they arrived so far as the discussion that we had earlier about what comprises clinical utility in some of these instances.
I would say having captured that one of the way that they could do that with KRAS was because the evidence was zero for harm. So there could only be a benefit. Marc?
So two-point NIH would be I am curious and the list of evidence cast -- two points, and one would be, I am curious and the list of evidence gaps it is the idea that we focus on prevention of adverse events we're in the EGAPP report it showed that while, guess, if you have this polymorphous and you are more likely to have adverse events, however, your cancer responded a hell of a lot better to and if I was a patient I would be more willing to risk adverse events if my likelihood of cure was higher but we have not developed evidence to say where do we identify the balance between adverse events and potential efficacy at least in certain circumstances. So that was one comment/question. The other question that had was related to the information that you presented at the beginning. Could you give us a sense for those of us who worked hard on the overset report whether or not that was actually used as part of the decision to go more into the proficiency testing and on genetic testing?
When I came on board I read the oversight report. The whole thing. We considered it a good idea but it was not in the process of being presented. Like I mentioned before, it was a matter of timing in that the work that we have been putting into the cytology proficiency testing actually moved on and they actually approved the proposed rulemaking and allowed us to move on and propose an core need another work Group for the of the proficiency tests. And the fact that the committee's report stressed the importance considering genetic testing because it is very unique in the parameters and such. All of that will be discussed and probably be introduced in the work group. But his all preliminary right now. The work group will determine what the issues are and what the criteria are and such. We does kind of accord needed at this point.
Thanks a lot. Jeff, we are going to move on because we are beginning to boost folks. Gurvaneet, do you want to talk? You can talk it wherever you wish to give us an update on your activities?
I can make it a five minute presentation from here on [ laughter ]. So all of you have the slide set in front of you. I have two kind of the updates. Number one I will focus on EPC reports on the method project that we are referring to right now and then I will give you a little bit of an update on the BRC clinical support tool. So the EPC methods project that we are working on has them to different areas of focus one is on evaluation of remarks. What are the ways that we look at genetic tests? There have been several different from Merck's proposed and we wanted to try to synthesize what these are and what the limitations are. The other part of this project was just on the analytic validity which Andrea had raised earlier. So hopefully this report will health address some of those questions. If you were doing an evidence report how do you search for analytic validity and what is the quality rating criteria? How the you look at the evidence and how did you fill the gaps? So the initial conclusions, we have a draft report which should be finalized by next week but I can give you some highlights of what we have found so far. There are several different evaluation remarks, Fryback Thornbury and the more reason to frameworks were from the CDC, the ACCE project that was started in 2000 and the EGAPP working group. Each to a Merc has a different strengths and limitations and some of it is driven by who the ultimate audience of the assessment is. Is it the patients and providers? Is it the Pacers, the regulators, test developers. Our report is not focusing too much on the last two categories. One of the conclusions is that one single framework to meet everybody's needs and for all clinical scenarios is impossible. We are trying to come up with a small set of remarks that may be useful for most situations. We will see what the peer reviewers say. What came up was that the EGAPP comes closest but it needs some enhancements with some specific questions that they have not dealt with before. A different part of this project was on the analytic validity and if you have followed this there was many published data on analytic ability. It is often abound in the "gray" literature. So there are different credible sources of information. Some from federal websites or some from credential organizations. There was only one that was proposed by EGAPP that comes closest to genetic tests. The for diagnostic tests there are quite a few tools and you can see these here that have published their tolls paid one of the directions that removing the word is to come up with a new tool which is a checklist. It has right now 21 items. We will not have time to empirically test this tool but we're hoping that this will move the field for word in terms of something that people can agree on. Now I will switch gears and talk about the informations project. We have been working on this for about a year-and-a-half now appeared to cyclical decision support tool that will assist towards the task force recommendations the challenge for a primary care provider is to know which women is at high risk did it is very rare to have the information available to make that assessment and take it in a systematic format. So this tool is a web based tool that can be used in a primary-care practice that will have both the patient and the provider interface. Where once the patient bills in their family history, a risk assessment score will be generated with some guidance to the clinician what to do. There were different phases of this project. The first was to look at the difference evidence and talking with experts. The second phase which we are finishing up now is the usability testing of the tool and modifying that before we roll it out at a couple clinical sites and see how useful the tool is. And you have already heard about these reports. I will not talk about the family history or the EGAPP report. One of thing I want to tell the committee is that we will have a workshop at AHRQ later this month which will be assembling a small group of barely differs skill sets -- fairly diverse skill sets and hopefully move the dialogue for word and how can we overcome this. So we're looking towards a white paper and once we have a draft and input from this workshop we would be happy to share it with this committee and get their feedback. And one small of the is we're finishing up a randomized control trial that we had done on pharmacogenomics. It has taken about two years now and it compared two difference dosing calculators to see if there was any difference and out comes like tying -- like time and range. The report is still getting finalized but what I can say is there really is no conclusion on this project that will say we should start using this right now. And, of course, I will end on some of the funding announcements that we have come up. I would be happy to answer any questions that I can on the grand opportunities that we have on comparative attractiveness some of which are also asking for pharmacogenomics and other diagnostic tests. And I will end there.
Thanks, Gurvaneet.
Any questions for Gurvaneet.
Fortunately we have one more presenter and you have been incredibly patient. We talked a little bit about FDA this morning in conjunction with a force and we mentioned that they have been developing new mechanisms for getting reports of issues regarding lab developed tests and we're hoping that you will be able to say what that is.
Yes, I will go through the slides quickly. It should be fairly quick. If we can get them up. I will spend a couple minutes giving you a background to make sure that we are all on the same table here. Can I go to the next slide?
Okay. Thank you. So it just to remind you, as this committee knows, there are a few differences between what laboratory developed tests are getting looked at in terms of the dilatory oversight and what the FDA does in one of the things that we noticed is that is not only premarket look at the tests and to tell whether there is clinical quality or not but, in fact, there are a lot of post market or things that the FDA does that laboratory developed testing does because we're doing enforcement discretion. So one of the controls that we have is that we actually do surveillance. We do problem identification and correction for tests that are regulated by the FDA. What that allows us to do is we can really pick up problems and we use this as a tool that allows us to get manufacturers to correct issues that we see and prevent recurrence of adverse events and is the tool that we use quite frequently. So as it stands now manufacturers are responsible for reporting death, serious injury or malfunctions to the FDA and we look at the reports. We look for data trends. Sometimes we look for issues and there have been several cases in which the data has been quite useful in terms of the current surveillance of LDTs, the laboratories to not actually have to report malfunctions to us. It is not being enforced. That is not the only issue, we have had lack of a mechanism to really analyze and segregate the data because the way that we do that now for in vitro diagnostics that are cleared through the agency is that we give them a code based on that usually and then the data is analyzed based on that so we have experts that look at specific protocols and look for trends and issues in there. Since we do not have a protocol that says this would be a laboratory test, if somebody does report for some reason an issue with a laboratory developed test it would get lost among all of the other data that we have. So we're trying to create a mechanism where we have a protocol specific for laboratory developed tests and we will have an analyst assigned to that so they can look at any trends and issues that we see in laboratory developed tests. And lately how will this work? We are still not going to enforce our reporting from the laboratory developed tests but there is a mechanism for voluntary reports. So we are planning to advertise this and hope that people actually report any issues voluntarily so that we can go ahead and take a look at what kind of adverse events we are seeing among laboratories. The method for a voluntary report is called the MedWatch and can be used by anybody but you can go to this website and report adverse events or whatever you want and be will be able to actually then follow up on that. So I have here just -- If you want more information on MedWatch and where you would report on that is in the slide.
So for the reporting would there be a mechanism by which somebody say is knowledgeable about the fact that there is a laboratory developed test that has had problems would be able to report that to FDA or with all reporting have to come through the provider of that LDT.
By doing it through this voluntary MedWatch you can report directly to the FDA. We have a dual mechanism for reporting. Most reports come through the manufacturers because they're obligated to report to the FDA paid when we go and inspect them we make sure the are reporting but we get a fair number of reports from people who are interested or have been harmed or through the MedWatch and MedWatch is setup to do exactly just that.
When do you expect the LDT part of this to be up and running?
We expect this to be very quick. Within the next couple weeks or so. We already have started asking people to report. The biggest issue is, the weight that MedWatch and MDR works out there are actually contactors that take the reports and put them in the buckets. We need to train them and make sure they are putting them in the right buckets.
It is interesting what types of issues to you turnup and what are the consequences because they talk a lot about these harms.
And also the LDTs, I was wondering the process of all of these protocols. Will you seek input from the end users?
We can do that and we can actually collaborate with CMS if there is an issue CMS will health us look at it. The biggest issue that we have at this point is if we do not have a bucket -- we get a lot of MDRs. If we do not have a bucket that the contractor can put that into it really gets lost.
Because here with the LDT you are talking about culture and biology and microbiology and chemistry and biology. You are talking about genetic testing.
We get MDRs from the manufacturers themselves so we do know the range that we're talking about.
So maybe some input from the professional organizations and.
Are these data when these are reported and assuming that someone can do an analytic on them are the available publicly for 3Q?
Yes appeared some of it is public -- yes, some of it is public, yes.
I can give you an example. We have been falling glucose meters for a long time and we have begun to see a recurrence of deaths that were due to an interference of some drugs with some of the glucose meters and we actually have been able to analyze that we have been able to go back to the manufacturers and put a safety notice based on what we sought. So we do use these things in ways to prevent problems or to health solve problems.
Rate. I think this is a real step forward so thank you for that and I am interested in talking at more length.
So all of you have gist displayed incredible tolerance for a very long day so thank you for that.
A couple reminders, 7:30 a.m. tomorrow we will take up the patent and licensing report. And dinner for those were going tonight is at 7:30 p.m.. If you do not know where it is and want to meet in the lobby at about 7:25 p.m., people can meet there and walk over to get there. Other than that I think we will be adjourned and I looked toward to seeing you all very early tomorrow morning -- look forward to seeing you all very early tomorrow morning. Thanks so much.
[ event concluded ]