American Health Information Community

Personalized Healthcare Workgroup #10

Monday, November 26, 2007

Disclaimer

The views expressed in written conference materials or publications and by speakers and moderators at HHS-sponsored conferences do not necessarily reflect the official policies of HHS; nor does mention of trade names, commercial practices, or organizations imply endorsement by the U.S. Government.

>> :

We’ll be doing a roll call in just a second.

>> Chris Weaver:

I believe we’re good to go.

>> Judy Sparrow:

Thank you, Chris. And welcome, everybody, to the 10th meeting of the Personalized Health Care Workgroup. Just a reminder, this is a FACA committee. The meeting is being Webcast over the Internet, and minutes will be made available following the conclusion of the meeting in a couple of days. Just a reminder for the Workgroup members to please speak clearly and distinctly and to identify yourselves before you speak so the transcriber can note who the comments of everybody. And also, please mute your telephones when they’re not in use. With that, I think, Chris, if you could just introduce those members on the telephone and then we’ll go around the room here.

>> Chris Weaver:

Sure, Judy. Let me give this a shot. We have people still logging in. But we have on the line, we have Kristin Brinner, Michelle Carrillo from PharmGKB, Mark Carroll from IHS Quest wait, no, I’m sorry; that’s IHS Beryl Crossley from Quest Diagnostics, William Farrow from National Human Genome Research Institute.... We have Emory Fry from DoD, John Glaser from Partners HealthCare.... We have Alan Guttmacher from National Human Genome Research Institute. We have Edward Helton from CDISC, Inc. We have Joyce Hernandez from Merck and Betsy Humphreys from NIH, Katherine Kolor from CDC.... We have Rebecca Kush also from CDISC, David Lobach from Duke University, and Rochelle Long from NIH, Joyce Mitchell from the University of Utah, Dina Paltoo from NIH, Ron I’m going to apologize for this pronunciation Przygodski from he’s in for Kupersmith from the VA.... Lisa Rovin from the FDA, Steven Teutsch from Merck, Mollie Ullman from Harvard Partners, Janet Warrington from Affymetrix, Dennis Williams from HRSA, Marc Williams from Intermountain Health, and Grant Wood from Intermountain Healthcare. I believe that’s it.

>> Judy Sparrow:

Thank you. Did we miss anybody on the line?

>> Ron Farkas:

I joined a minute late. It’s Ron Farkas from FDA.

>> Judy Sparrow:

Great. Thank you. And here in the room, we have...

>> Michele Puryear:

Michele Puryear, HRSA.

>> Paul Cusenza:

Paul Cusenza.

>> Greg Downing:

Greg Downing, Office of the Secretary.

>> Judy Sparrow:

And I think with that, we can turn it over to John Glaser, the Co-chair.

>> John Glaser:

All right. Thank you, Judy. Now, first of all, welcome to all of you. I hope you all had a wonderful Thanksgiving. Welcome back to the regular workweek. Doug is unable to join us today, so I will be chairing the call. And the first thing I’d like to do is, you should have the minutes sent to you as part of the email that went out. And I want to see if there are any corrections, suggestions, or clarifications that folks would like to have of those minutes. (Pause) Hearing none, they stand approved. And without further ado, Kristin, do you want to lead us on our prioritization of the Workgroup meeting for the 2009 use cases?

>> Kristin Brinner:

Yes. Thank you very much.

This document that we distributed with the Workgroup materials is quite similar to the prioritization for use case development document that we discussed at the previous Workgroup meeting. We simply did a slight reorganization based on people’s comments. And those were largely that newborn screening would be ranked first for use case development, followed by pharmacogenomics and then clinical decision support, with the understanding that there were several different kind of interconnecting needs, especially the need for clinical decision support in both the aspects of newborn screening and pharmacogenomics.

So I simply took this document, and after Lauren and I had some discussions with the use case team in terms of what would be most useful to them for our priorities, I broke them down a little bit differently, but somewhat similarly. You’ll notice that there is a PHC 5.0. That’s newborn screening. We kind of pulled that out to call it out specifically. Then we have Parts A through D below that, which really are not changed from the last ones. So we have the overall idea of newborn screening, and then we’ve called out some specific needs within that. And then I also went through the other workgroups’ priorities that we had available and these are undergoing a similar process in all of the other workgroups that we’re doing right now excuse me and tried to pull out priorities that could potentially have overlapping information needs. Specifically, for the newborn screening, there were several aspects of the Biosurveillance, which is now called the Population Health Workgroup, in terms of registries for diseases and exchanging health alerts, that could potentially have overlapping needs for the newborn screening. And similarly, I did that for pharmacogenomics in terms of potentially product labeling that would have pharmacogenetic information on it in terms of the lab results that the Electronic Health Records Workgroup put out. And then finally, the issue of clinical decision support had been highlighted as a priority by several other workgroups, so I simply took those priorities that specifically address clinical decision support and put them underneath ours, just so the use case team would note that there are many different overlapping priorities between the different workgroups, hopefully to kind of strengthen our position that these are high priorities for use case development.

And that’s pretty much it. I’d welcome anyone’s comments on this. I think, Lauren, it’s due in the next couple of days, so any comments now in the meeting or follow up with us by phone or email in terms of finalizing this document. But I hope it reflects the conversations and the comments we received largely from the last Workgroup meeting.

>> John Glaser:

So, any comments, questions for Kristin? (Pause) So Kristin, what happens next after these are set?

>> Kristin Brinner:

The Office of the National Coordinator will go through a process where they will pull all of the priorities into kind of an overarching document, and then they’ll go through and look at them to see which ones are overlapping, which ones are complementary, and kind of develop a final list of priorities that will then be presented to the AHIC itself for approval I believe either in January or February of this coming year. Lauren, is that correct?

>> Lauren Kim:

That’s our understanding as well.

>> Kristin Brinner:

Yeah, and so then the AHIC essentially will approve these priorities for use case development, and then the next round of use cases will start to be developed following that.

>> John Glaser:

Okay, so they’ll resolve the ones that you’ve pointed out may be overlapping. They’ll determine whether it’s the case or not the case.

>> Kristin Brinner:

Yes. Yes.

>> John Glaser:

Okay. And we’ll wait and see, presumably in January or February, what the final slate is that emerges from their deliberations.

>> Kristin Brinner:

Mm-hmm.

>> John Glaser:

Okay. Any other questions for Kristin, or comments? (Pause) Terrific. Well, Kristin, why don’t you then take us into the JAMIA submission document?

>> Kristin Brinner:

Sure. From several different conversations we’ve been having with people from both in the government and outside the government, a lot of people are very interested in what this Workgroup is doing and the products of the Workgroup. And so we thought it would be useful to communicate the work of this Group to the larger public those that have not necessarily been following the AHIC process very carefully. And we had spoken with some of the editors of JAMIA previously, and they thought it would be really useful for us to introduce the AHIC process, and more specifically the Personalized Health Care Workgroup, to their community that reads that journal, which is, I believe, very broadly read. And so we tried to generate a kind of summary document that would show the work that we’ve already done, introduce the general concepts and the vision for the Workgroup, talk about a little bit about how the work of the AHIC and the Workgroup then feeds into the standards development and certification process, the priorities that we may have, and then our future activities.

And so, basically, we generated this document largely from the priorities and the vision documents that you’ve already seen, as well as from the recommendations and some other documents that we have been generating internally. And this is our first pass. We have a contractor that has also been helping us with some of the writing on this. But we would like to have this document ready by the end of the year, I think, for submission. So that’s why at this point, and kind of earlier on in the editing process, we wanted to get your comments back on how useful you felt this document would be to the general not the general public, but the public that’s reading JAMIA, how well it represents the work of the Workgroup, and if you had any additional comments. And I think some of the comments that we’ve been getting back preliminarily have been that, potentially, we should focus a little less on the background and the process and focus more on the actual products that the Workgroup has already produced, such as the family history; core dataset; the confidentiality, privacy, and security document; etc. and the recommendations in the use case.

And, you know, we have that as a concept that we actually believe that we would like to have several different publications where one would be specific potentially to the family health history, another to the document the Confidentiality, Privacy, and Security Subgroup presented. So we absolutely believe and agree that these are documents that are important to publish. But and we also agree that maybe we should highlight those projects a little more in this document and then point to future publications from those as well.

So again, we welcome your comments on this document, and we’d be happy to engage in a conversation about this now as well as offline afterwards.

>> John Glaser:

Any feedback for Kristin?

>> Marc Williams:

This is Marc Williams. I thought it was excellent. I would probably take a contrary position. A lot of people that I talked to that are not engaged in this process are really they don’t understand the context of what we’re doing. So I think the for an initial publication, the emphasis on the background is quite important. I don’t disagree with the idea of at least, you know, giving a brief look in the future about what our first products are going to be, and I think the strategy of developing additional publications around these is a good one. Otherwise, I really didn’t have any substantive comments but did send some minor editorial issues off to you.

>> John Glaser:

Okay.

>> Dennis Williams:

This is Dennis Williams at HRSA. I also thought it was very well-done. And as an initial description of what we’re doing, I think it is very valuable.

>> John Glaser:

Terrific. Other thoughts for Kristin?

>> Betsy Humphreys:

This is Betsy Humphreys. I’m one of the people who pointed out that it seemed like it was at a very high process level and didn’t have as much substance as I would have liked to have seen. And I think possibly an approach that is halfway between where it is now and what I was suggesting might be useful. I think the problem to me is that there are an awful lot of government-sponsored committees meeting in rooms in various places, and some of them are talking to each other in those rooms and are not necessarily doing a lot of the substantive discussion and analysis of the issues. And I would not want this Group to sound like it falls in that latter category, because I don’t believe it does.

>> John Glaser:

Fair enough.

>> Greg Downing:

I think the this is Greg Downing. I think that the thank you for your comments. Some of the things you referenced were materials that were worked on by some of the Subgroup members, particularly in genetic tests, that I think we’ve been looking for the venue in which to pull some of the content into these discussions that may not you have to go through a lot of Web clicks to be able to find this stuff. So I think that one area where there were some tables developed around pre-analytical, analytical, and post-analytical considerations were very helpful to us. I think within the constraints of the manuscript submission requirements, I think we can adopt those.

The other point I’d like to make is that in discussions with some of the editorial staff at the journal really brought a light to us that we’d like to try to address through this, and maybe your final comments can point us to this direction. That is, how does this facilitate processes that are underway in many institutions to try to adapt information systems to meet these needs on the local level? So how does the work of this Group facilitate those processes? We think we’ve made some stabs at that. But one of the overarching aims of the paper, really, is to connect to communities that are struggling with this on their own level.

>> John Glaser:

Yeah. Other comments? (Pause) Yeah, I think this is a very nice piece of work, and I appreciate all the comments that folks have provided to this. I wouldn’t mind, you know, Betsy, per your comment, seeing more of the conclusions in this, sort of mindful about quite how to do that and stay within a page count or a word count, and also without stealing the thunder of subsequent pieces that come out here. What I don’t know is, when we sort of try that, whether in effect one has more than one piece here, you know, a piece that does describe the process for those who are not aware of the process; you know, a piece that does more goes perhaps more substantively into the use cases, and obviously, there’s the family history dataset and the CPS stuff. You know, it’s possible that JAMIA would be interested in a bundle along those lines. The downside to that is that to the degree that the subgroups wanted to put themselves somewhere else, because that may have been there may be other publications that are target for these things, I don’t want to unduly constrain where they present their work.

I think, Kristin, you know, we can try getting more of the conclusions in here, mindful of word count, and also not eviscerating the full publications that come out in some of these things and see what happened. I think the comment about “What does this mean?” or “How would this guide, you know, a place like Partners or anybody else Intermountain Healthcare trying to do this stuff?”, that I could see as a companion piece, so that might go into more detail about what how one takes use case work and some of the other work, and so here’s what some of the implementation challenges or clarity that’s now been introduced here.

>> Betsy Humphreys:

That sounds great, and I think we can work pretty rapidly to look at those different options and see how we think.

>> John Glaser:

You might touch base with the folks who have overseen the two workgroups on the family history and the CTS and see whether a sort of bundle along JAMIA fits with their vision or whether they have some other plans.

>> Betsy Humphreys:

Sounds great.

>> Greg Downing:

Hey, John, this is Greg. Just on a technical matter: The manner in which we had discussed with the chairs for handling authorship on this was that to address we think it’s important for everyone who has been involved with the Workgroup as a member to be included in this. It’s a large part of the effort was around visioning session and the work that’s gone on. So we will be asking everyone who has participated to, you know, approve the final comment, equal contributors.

>> John Glaser:

I think that would be terrific. Any other comments? (Pause) Well, it might be done, and I do think that publishing these will be helpful to the broad cause. So I’m glad to see we’re doing this.

Michele, I believe you’re up. You ready?

>> Michele Lloyd-Puryear:

Sure.

>> John Glaser:

Okay.

>> Michele Lloyd-Puryear:

Looks like we’re early, aren’t we?

>> John Glaser:

That’s all right. We’ve got plenty of running room.

>> Michele Lloyd-Puryear:

So the this will be short, no slides. The Newborn Screening Subgroup met for a conference call for a second time. And as reported before, the Co-chairs of that group, as a reminder, are Steven Downs from Regenstrief and Peter Van Dyck, who’s the administrator of the Maternal and Child Health Bureau at HRSA. The Subgroup is very large and grew with the addition, also, of hearing screening to make sure that we were looking at structure the infrastructure that surrounds hearing screening. There were also several the group agreed to even subdivide to tackle more issues, one being the identification of the analyzed or those entities, because they’re not always analyzed as you go from the newborn screening system from the laboratory to the diagnosis of what entities need to be identified to capture what codes or IDC-9 codes, but those things and I don’t know what they’re always called; sorry that are necessary for the communication process to be intact.

One of the other subgroups or sub-subgroups is looking at issues surrounding hearing screening and, for the geneticists on the line, may be interested in the genetic evaluation that’s going to be looked at more carefully also. That would be a sub-subgroup and to see if we can incorporate the mutational analysis of the genetic evaluation.

So there has been general agreement with the overall Newborn Screening Subgroup to produce a use case, and they’ll be working hard at that over the next month. And that will be presented to you guys in January. Is that right? In January. You want to add anything, Greg?

>> Greg Downing:

Yeah, I think that this there’s been a fair amount of background work and engagement with a number of the entities. I think the one of the encouraging pieces is that there’s been a lot of source documents on standards from a number of different communities that’s been brought together, and that’s represented a great deal of background work thus far. There’s been good discussion amongst the many stakeholder organizations and the public health laboratories and looking at this as a hybrid of one of the sort of unique areas where population based or public health information is utilized in primary care applications, and I think there’s a fair amount of interest in looking at how this model evolves, for some other comparative purposes, for how population information is integrated into personal health information that serves to support primary care decisionmaking processes.

To the point of sort of broadening the dissemination of this and getting brought input on the inputs into the use case development process, this is fairly similar to what we engaged in the Genetic Test Subgroup and has been more ongoing in the family history core dataset development processes. This around the newborn screening areas is a bit of a hybrid between these approaches, and I think, looking longitudinally, the efforts of this Subgroup will be to focus initially on we developed sort of action areas of 13 potential areas of focus and development in this group, and we analyzed those in the first meeting and have prioritized the first four of them that would be inputs into use case framework. One of the primary considerations for that is the data elements reported would have the capability to have quantitative measures and not just qualitative ones, being positive, negative, or repeat, but having absolute values in some context that’s interchangeable with the lab test results.

And there has, I think, been a good focus by the discussion thus far to focus on the IT aspects, and not so much on differences across different states, but focusing on different developing a pathway forward that’s beyond the use case development. The other priority areas will continue to be discussed and for future consideration in bringing to the Workgroup.

So I think our assessment from the staff’s position is that the Chairs and the, you know, Subgroup participants are on track to deliver specific recommendations and presentation in some depth at this Workgroup meeting in late January. Leading up to that, there will be some opportunities for public vetting and one of the other Secretary’s advisory groups, whose name I can never recall all of the details about, but has to do with many of the communities involved with hereditary disease screening and a number of the HRSA regional programs. So we’ve taken advantage, staff level, of being able to disseminate information about this, potentially building a contingency that would facilitate the use case development process longer-term-wise through our processes.

So I think the leadership has been excellent. The participation has been very high level of quality, and a number of you on the call today have participated in those, and I appreciate your efforts in taking this one up.

>> John Glaser:

Any questions or comments for Michele or Greg? (Pause) Fair enough. Michele, nicely done. Thank you. Greg, do you care to lead us in the next overview the next conversation to start the overview and then we’ll have the rest of the folks follow?

>> Greg Downing:

Yeah, I just want to make sure we have everyone on the call if I can just check real quick. Becky, Joyce I didn’t hear from Joyce. I know Ed’s on the call here.

>> Joyce Hernandez:

I’m here. I’m Joyce.

>> Greg Downing:

Okay, and then Michelle, and I think Rochelle has joined us. Is that correct?

>> Rochelle Long:

Yes, I’m here on the line.

>> Michelle Carrillo:

Yes, I’m also here.

>> Greg Downing:

Thank you. Can we have the slides?

Over the course of the last year or so from the first Workgroup meeting and our visioning session, aspects of pharmacogenomics have been relevant to the electronic health record in a number of different ways. This session today would like to try to put into context relative to some of the other work of this Workgroup.

The in that past year, there’s been a lot of clinical discussions about the applications of understanding pharmacogenomics in many contexts. We’ve had some presentations on the use of these and guiding therapeutic decisionmaking, some on screening. And in the backdrop, other policy decisions and programs have been ongoing that continue to yield new science. And many of your organizations are adapting to integrating pharmacogenomic applications overall. This is, as many people pointed out, not necessarily new science and understanding that there are differences between drug metabolism and application, and there have been implications of this science have been known for some time. The organization of the health care delivery system to accommodate differences in metabolism and distribution and many of the pharmacological aspects of this have not been borne out in the mainstream applications in clinical medicine to some degree to a large degree.

So on the next slide, Slide 3, basically the current Workgroup activities around genetic tests that we’ve addressed through recommendations and Workgroup discussions have thus far been focused primarily on single gene-type abnormalities, either through inherited forms of mutations acquired in cancer processes or biological variations that describe drug metabolism, such as the cytochrome P450 metabolism. A variety of drugs focused on Warfarin in a use case clinical decision support meeting back in September. And I think, consummate with a lot of other discussions ongoing in clinical research about the associations of targeted therapies around specific proteins and genes have been also discussed as well.

The aspect of newborn screening tests is subject to our current Workgroup activity. These often do not necessarily, by themselves, mean specific genetic test results, meaning a polymorphism, but it can often be the metabolite or protein that’s measured as a consequence of these. So these are other types of sort of genetic tests that are being integrated into our recommendations to AHIC.

We had a presentation, back in August at our meeting here, that focused on several other aspects of genetic tests that are being in the more in the development process for medical applications. And some of these are technologies that are utilizing multiple genes in the context of making a diagnosis or, in some cases, prognostic predictions on outcomes. These are in the context of looking at either micro-ray or gene expression patterns. In our early discussions with the AHIC Workgroup and others, we had identified a lot of standards activities that were ongoing on the technology side leading up to the development of a product for integration into clinical practice through clinical trials and other kinds of metabolic medical product review processes.

We have not had thorough discussions around index-type assays, where the introduction of multiple genes into a common output has been brought forward, and other types of genomic analysis, such as DNA methylation patterns, which are often referred to as epigenomics. These are not yet coming into clinical applications in the medical arena, and subsequently, we have not taken those on as a priority area for electronic health records at this juncture. But I think, as many of those of you who are involved with a number of research applications see that this is not too far downstream in terms of the needs for beginning to think about those applications.

So on the next page, Slide 4, some of our goals for genomic laboratory test information, at a high level, really represent the ability to represent data to be contiguous from basic research to translational research to clinical research. That does not mean that research data is being applied clinically. It’s the context that we don’t have to go back and resolve differences in whether a particular laboratory and assume the name for particular gene or polymorphism that in the clinical laboratory experience, that’s assigned to either a test that’s associated with particular type of drug that’s being metabolized by that enzyme. We’ve seen, in many aspects of clinical laboratories today, remnants of that that these are names that are used to describe tests, have a lot of variability, and are inconsistent across laboratories and even in the clinical world.

So the aspects towards the goal of at least the standards for electronic health records is to address the semantic interoperability of disparate entities that come together to perform these services, so that the results themselves are clear to the practitioners and even the consumers as to what they read about a particular research finding, a particular test that that nomenclature and the data standards around those ultimately translated into clinical practice in meaningful ways. There shouldn’t have to be a way to a need to, in the future, translate between the meaning of one test and the result test result and the units in which it’s expressed from a clinical research domain to clinical practice domain.

So there’s a continuum of standards development that’s also evolving from clinical research needs to clinical trials practice. And to facilitate that integration of molecular diagnostics into clinical care, one of the factors that’s often been pointed out is nomenclature issues or data standards to facilitate clinical adoption.

And so the purpose overall of this effort and this, I think, has been misunderstood in some contexts towards the adoption of clinical standards and the interactions between data standards from clinical research to clinical practice isn’t necessarily to facilitate the direct use of research data, but to enable communities to interrelate and share information as it’s developed. And this, in some ways, may be trying to overcome some of the obstacles to the integration of research findings into clinical practice. So I just want to underscore this, that this is not to sidestep or circumvent any practices that are currently engaged to review and establish validity and the evidence processes for assigning where a technology or genetic test is applied in clinical practice, but it’s to support the better understanding of how that information is meaningful in clinical practice ways.

And as we won’t talk about this today, but there’s growing interest in utilizing clinical research records and information for evidence development processes and review and safety assessments. Without having continuity of the data standards evolving in the clinical practice world, they’re consistent with those that are ongoing in the translational research world that this desire will not be achieved if there are not consistent standards development processes under way.

So the good news is that there’s been a lot of work done on the developmental side around as the technology platforms evolve to establish standards for either product development or medical product review purposes or what have you, and that we’ve also understood that in some presentations previously, that one of the reviewers or users consumers, if you will of this information is the FDA, as the as evidenced from the voluntary genomic submissions database processes that they use in looking at how genetic information submitted along with medical products is not necessarily used in medical product review but has been essential in helping understand the biologic information in the manifestations of how drugs are metabolized and distributed and the implications that may have in clinical practice about particular pharmaceuticals and other medical products. And we’ve also had a substantial input, back in our September meeting, about the development process of clinical genomic standards and HL7 messaging standards, and a fair amount of work is going on in the industry sector along the development of these standard capabilities that, in our assessment, hasn’t necessarily linked up yet with the clinical practice world. And Amnon Shabo made a nice presentation to us in the September meeting addressing some of these issues that are ongoing within HL7.

>> Janet Warrington:

This is Janet Warrington. I just wanted to thank you for mentioning the ways that this could actually improve the efficiency and effectiveness for the development work. One of the things that is of tremendous value in these discussions and this activity, I think, is raised awareness among all stakeholders. So for instance, I’m representing the developer side, and too frequently in the past, we’ve had to sit and think about what we think folks would want to use or how they would like to see the information. And some of these efforts that were initiated were really were really came out of having to move a project forward and go out and get information from the community. And so this activity that we’re engaged in, I think, is going to bring some transparency to this. And hopefully, it will reduce development time, and it will make it a much more efficient process. So rather than having these standards happen sort of in a de facto manner, you’ll have a lot of smart people in the room talking about the best way forward and getting alignment in the community and then moving forward in an efficient manner. Now I may not have stated that in the most clear and concise way, but hopefully you understand what I’m trying to say here is that this really lends itself to a more efficient system.

>> Greg Downing:

Yeah, thank you, Janet. I think, in our tour over the last year or so to many different organizations we’ll hear about some today the aspects of being able to the standards effort, I think, is sometimes misunderstood to mean you have to do things one way. I think the aspect of the standards in this context around electronic health records is necessarily being able to communicate your findings and allowing others to understand them and incorporate and integrate those into the kinds of work that they do. As we’ve seen this whole field in the research area sort of begin to take hold in other kinds of communities, whether it’s in asthma research or cardiovascular areas or immunology, clinical immunology, the aspect of having to sort of the genes have names assigned to them but the context how the assays are done and the other sort of contributing information about the types of technologies that are sort of being applied here there’s, I think, a growing recognition amongst the researchers as well that they’re having some challenges in being able to harmonize the clinical data and the clinical laboratory test information that is ascribed or associated with these genetic alterations.

So we’re going to hear, I think, a little bit about actually, I find this to be some of the most exciting stuff we’re doing as a Group. But I’m probably one of those mutants, as I was accused of being labeled recently. But the practical applications for the

>> John Glaser:

Love you any way, Greg. (Laugh)

>> :

I think you’re in good company here. (Laugh)

>> :

Remember, we’re all mutants. That’s the lesson of genomics.

>> :

I just want to know if his particular change or his particular aspect of being a mutant was detected in his newborn screening. (Laugh)

>> Greg Downing:

My mother can help you out on that.

>> :

Unless they decided to look the other way.

>> Greg Downing:

Well, we were all sitting around our Thanksgiving tables, I presume, over this last week sharing lots of anecdotes with our relatives about certain aspects of our personalities. To wit, the family history tool was not necessary. (Laugh)

>> John Glaser:

Garrison Keillor said, “Sitting around the Thanksgiving table is what makes you thankful that not everything is genetic.” (Laugh)

>> Greg Downing:

All right, well...

>> :

Okay, well, these mechanisms for benchmarking are important, and, you know, I share your enthusiasm, and I think most of the people who are on the phone today are on the phone because we recognize this the value in this.

>> Greg Downing:

I think one of the aspects of this new sort of horizon and future not knowing how it’s all going to play out is the clarity to which we need this Group, I believe, needs to help other communities in understanding how this is an important aspect towards evidence development and improving the quality of information overall, and very powerful decisions may lay at the hands of those who use this information, and that communicating our intention is probably something that I came to new levels of awareness. I don’t think Becky’s with us today, but

>> Becky Kush:

Yeah, I’m here.

>> Greg Downing:

Okay. Sorry. You’re usually at our table here today, and you’re not today, so I can usually look across the room and see from you whether we’re reaching the level of comprehension about the importance of things by just your reaction to them.

So I think, from a consumer’s viewpoint, that the aspects of knowing that the laboratory information, as it’s presented to providers, is consistent and accurate and has meaning is in facilitating the decisions with the patient, is really one of the things we’re driving for.

So as we’re in our prior conversation, our priority-setting aspects, we looked at one of the areas for the AHIC to consider, and at least have on notice that we’re looking at, is in this whole aspect of genetics beyond the polymorphisms and the single-point mutation types of tests that we reported on before, and looking at the aspects of the new types that we’ll be hearing a little bit about today. So our potential action items for the Workgroup is to sort of just to recast and put in perspective our job is looking at addressing interoperability or needs for standards and perhaps recognizing some of the work that’s been ongoing in other communities that facilitates or supports the aspects of electronic health records and clinical decision capabilities and also the identification of incentives, collaborations, and additional issues.

I think this is actually one of the hidden attributes of what this Workgroup has done in the past is help spawn some other collaborations and other related activities that otherwise just getting to know people may have been more of an obstacle. Included in your packet are a number of reference articles that I would encourage, on your next plane fight, you take a look at. And one recent aspect is the pharmacogenomics report that Marc and others from the Secretary’s Advisory Committee on Genetics, Health, and Society have worked on and, I believe, will be finalized. And there’s lots of references to the needs of health IT components in this process. The second one is a paper that Amnon contributed with a number of other people from HL7 that I found still to be a really good reference point and is not yet outdated. And the PharmGKB discussions today and the others by participants in the CDISC activities are also listed.

So without any further comment, we’ll go into the presentations today. I want to thank both of the groups for their hard work in putting these very fine presentations together. We’ve had some discussions with a number of our Workgroup members over the past, particularly FDA, involving members of the CDISC organization, and Becky, who’s heads that organization, is with us today and a number of members who’ve been working on clinical genomics interoperability standards with their relationship with HL7. Becky, So I don’t know if you and Ed have comments at the outset or how you want to handle the presentation, but we’re ready to begin.

>> Becky Kush:

Yeah. Thank you very much for having us on this call and this meeting with you. I really appreciate the opportunity. And I’m going to go through can you hear me all right? I’m going to go through an overview, kind of a general set of slides that talk about clinical research standards and how they relate to health care, and then I’m going to turn it over to Joyce, who is going to give some more of the details she and Mollie and then I’ll close with a few slides on where we’re going and what the vision is. And I’ve been on the HITSP board for the past few years, and I want to thank this Group in particular for appreciating the value of medical research to contribute to health care and the HIT initiative.

So we can just go ahead and go to the next slide, I think. You’ve seen the slide that lists all the people contributed to this. And this slide is just showing you, and I don’t need to tell you all this, but to set the stage that we live in a world that’s currently hindered by a lot of disbursed information. And in health care delivery, we’re trying to get to electronic health records and through the HIT initiatives, but still a lot of it is in paper medical records. And then we’ve got a whole area of basic clinical research, where a lot of the information ends up in publications, which are hard to get to, and it’s not structured information. And then we have protocols-driven clinical trials, where most of the information goes to regulatory authority to get approval for a drug. And all these worlds are living apart right now and are not really sharing the information, and the information is many times gathered more than once. And in clinical trials, often the same information is entered at least two, if not up to seven, times the same information into different systems before it finally makes it to FDA.

The next slide shows what I call the plight of the site, which is an investigator who’s doing research as well as health care. And if he or she sees a serious adverse event, there’s oftentimes a second thought as to whether or not that’s something they want to go through the whole reporting procedure to report, because currently, it’s spontaneous reporting. A lot of it is not mandatory. It doesn’t fit into the normal clinical care workflow, and it takes a lot of time.

So there are some things that CDISC is doing that I’ll cover at the very end, where we’re trying to reduce the approximately 30 minutes that it takes to do this reporting down to a matter of seconds. And we’ll see how it goes. But for clinicians doing research today, approximately 60% of the trials are on paper. And again, I mentioned that they’re the data are reentered a number of different times. And a lot of the data are collected, if they’re collected electronically, by what I call point solutions that don’t share data, so they have to re-enter data out of the medical record or an EHR into those point solutions. And an average site that’s active in clinical research may have up to 1012 different laptops or PCs that collect different data for different trials in different ways, not with the standard solution. So this is why a number of clinicians really hesitate before they get involved in clinical research today, and many of them do one trial and say that’s enough. So that’s the issue that we’re trying to address by making it more simple to reuse the data that they collect already for clinical care.

The next slide just shows how medical research does inform patient care and clinical decisions. But there you see all the different potential computer systems that we’re dealing with along with the paper and the multiple places where data are collected through EKGs and labs and how it’s really just very difficult to pull that together and integrate it. And what we’d rather see is on the next slide, where we have these two circles in a more concentric way where we’re doing medical research in a context that’s fits the workflow of a investigative site and that information can better inform patient care in a more rapid and efficient manner, as I heard someone say earlier on this call.

The next slide lists what we would like to see in a world where medical or clinical research is best informing clinical care. What you want to see is these researcher institutes that are doing patient care as well as research in the same workflow manner that we can collect and integrate and analyze the information for multiple centers and preferably around the world in an integrated fashion, that we access that information and integrate it in a standard format so that we can then use review tools. The tools that places like FDA want to use require that the data be in a standard format for them to use the tools. Otherwise they have to modify all that data so it’ll fit with the tools. So a lot of the tools that they’re using are detecting safety signals and such really require data in a standard format. And so that’s what CDISC is working towards, and that’s what would benefit from standards.

On the next slide I want to make the statement that in considering that if we’re going to do health information technology, that we need to consider a medical research in the plan now so that the initiatives converge and that they’re not going in two different directions when we get down the road with HIT, that we’ve considered the needs of medical research at the same time.

Okay. On the following slide, what you see is that we’re working towards improved data quality and improved patient safety, which, as I said, requires faster access to better information and the adoption of technology to get there. This also means that we need to streamline processes and that we have standards so that we can integrate the data and have interoperability amongst the different systems that we’re using out there today. So those two purple arrows are where the standards are working to try to provide the infrastructure to get to the ultimate goal of improving patient safety.

The next slide shows you this overview of CDISC, which is a global, open, multidisciplinary nonprofit organization. It was initiated approximately 10 years ago as a volunteer group and was incorporated as a nonprofit in 2000. We actually now have over 200 members as of last month, and those are corporate members. Those are organizational members, such as academic centers, global biopharmaceutical companies. We have all the top global biopharmaceutical companies as members. We have technology and service providers, institutional review boards, and others that have joined CDISC to commit to developing medical research clinical research standards. We have very active coordinating committees in Europe and Japan.

And the next slide shows what we have achieved. Can you switch to the next slide, please? So through an open-based consensus approach, CDISC has standards that are being used globally to support the electronic acquisition exchange submission and archiving of clinical research data and meta-data. And one of the key standards that we’ve developed is called the Study Data Tabulation Model. It is how the FDA would like to see the data that our submitted to them for regulatory submissions for approval of new pharmaceuticals, and that’s so that they can populate across the database, and that would allow them to look across their therapeutic agents and across sponsors to compare the way that they are looking at the different agencies in certain therapeutic areas. And that’s something that hasn’t been feasible in the past, because the submissions have not been in a standard format.

So that is now in a final FDA guidance, and it is they are now proposing it to be a regulation. The CDISC standards are available on the CDISC Web site, and they are being harmonized with health care standards. We’ve had an associate charter agreement with Health Level 7 since 2001 and a commitment to harmonize our standards.

And the next slide shows you what BRIDG is. BRIDG is the Biomedical Research Integrated Domain Group Model. We named it BRIDG mostly because we really liked the name and then we fit the words to it, but it is bridging a number of things. It’s bridging organizations, because the HL7-regulated clinical research information management technical committee is using BRIDG as their domain model. CDISC, FDA, and NCI are also doing the same. So this is a model that’s governed by those four organizations, and it’s an open model. It’s bridging standards, and it is the way that we, through CDISC, found to bridge the clinical health care standards with the health care standards of HL7 so we can have semantic interoperability between research and health care. So that is found on an open Web site, and we’re harmonizing the standards within CDISC and with HL7, including all the adverse standards from preapproval through post-marketing.

So with that introduction, I’m going to turn it over to Joyce to talk about specifically what CDISC CDISC felt that the clinical genomics area was bigger than we should bite off ourselves, and we’ve been partnering with HL7, since it was initiated within HL7 through the Clinical Genomics Special Interest Group that Phil and Joyce are members of. So Joyce, you want to take the next slide?

>> Joyce Hernandez:

Thank you, Becky. On the next slide, presenting there is one of the major goals, when we worked with the clinical genomics team is to come up with models that are consistent in terms of representing the pharmacogenomics data. So we are one of the major goals is to be able to take the pharmacogenomics data and integrate it with clinical data to facilitate data analysis. And some of the areas that we have identified as critical areas is to be able to help in decisionmaking when prescribing medications, so you have an idea of who are the good and bad responders in the patient population, and also identifying patients with specific disease biomarkers, and also helping to augment the decisionmaking in terms of medication when you have certain genetic mutations that indicate that someone might have a tendency to have an adverse event. So those are the type of areas that we were focusing on in terms of the standards.

On the next slide, we also are showing here that there’s another level of data integration that’s happening via the HL7 standards, which is to bring together genomics data that really, by itself, is made up of other standards. And we see there, on the left side, there are standards for like MAGE, which is for how the data is represented for array data. And there’s also BSML, which is the Bio-informatics Sequence Markup Language, which is used to represent the genotypic data. There’s also vocabularies that are used in genomics, like GO, which is a gene ontology. And then on the clinical side, with the clinical data, there’s imaging data that uses DICOM as a standard to represent the images and then describe them. There’s also LOINC and SNOMED, which are used for describing the type of tests that are being done for diagnostics. And so we’re through the HL7 standards, we’re kind of bridging that across, because we have a main model.

And that becomes more apparent on the next slide. Here we’re showing that we are actually also, because we have a what’s known as the HL7 model, we’re able to use that as the basis for doing an integration between the clinical data and the genomics data.

And on the next slide, I actually show some of the examples for that. We’re using a genetic locus and how and we’re showing there how we have information which represents the region of the gene; the sequence; significant findings, such as if there’s an amino acid H gene mutation; the phenotypic effects, such as the clinical interpretations all that information is represented in a set way that can be used both for research as well as medical practice. And then that information can be assessed through or accessed through different views, and so we have a clinical practice view, which can look at the information via physician and patient, and clinical indication, which means the diagnosis for that patient. On the research side, we can look and view similar information genotypic information through what are known as clinical research plans, such as protocol and investigator, and visits, which are planned points in time where the individuals come in for diagnostic testing or treatment.

On the next slide, I’m going to at this point, if Mollie is on the call....

>> Mollie Ullman:

Yes. Hi, Joyce.

>> Joyce Hernandez:

Hi. I’m going to let Mollie kind of explain some of the methods that were used for the developing of the HL7 clinical genomic standard as well as she’s going to share some of the actual uses of that standard in various organizations.

>> Mollie Ullman:

Thank you, Joyce. Hi, this is Mollie Ullman from Harvard Partners Center for Genetics and Genomics. And I think what’s coming across in the presentation today is that the HL7 Clinical Genomics Special Interest Group is really comprised of a cross-functional team with representatives both from the clinical research environment and the clinical environment. I’m here briefly to describe some of the clinical activities. And this slide really is depicting that within the clinical realm, we’re not only concerned with the visual representation of the data within the health record, but we’re also concerned with the machine-readable representation. This machine-readable representation is encapsulated and then either exposed or bubbled up via clinical decision support within the appropriate clinical context and guidelines.

Next slide, please. So if we take this to an example let’s say a pharmacogenomics example with cytochrome P450 we would actually bubble up or expose this information via clinical decision support, for instance, within a drug dosage calculator.

Next slide, please. Next slide. Oh, one up, please. Yes.

This is just a brief list of some of the pilot activities that are going on. We have two pilots that have been going on for multiple years now in the clinical environment, both with family history and clinical genomics, using an HL7 version 3 model as a payload in an HL7 version 2 message. We have a pilot that’s well under way between Harvard Partners Center for Genetics and Genomics, Partners Healthcare, Intermountain Healthcare, LOINC, and the National Library of Medicine, which is representing this clinical genomics model in an HL7 version 2 message entirely. This will be important, because this is a version that’s most commonly implemented in health care systems across the U.S. today. Concurrently with these pilots is, in the clinical research realm, the CDISC laboratory message, which is encapsulating the clinical genomics information within the pharmacogenomics model.

Next slide, please. This is just another representation of some of that work that’s going on. For instance, lessons learned from the first pilot are being rolled into implementation guides, and this implementation guide will cover both the clinical and clinical research findings and suggestions. This will be ready and published for comment via HL7 in December, in just a short period of time. Thank you.

>> Becky Kush:

Okay, I think I just have a few more slides to close this out, if you can show the next one. And thank you very much, Mollie and Joyce. And also, Phil’s on the phone, and he had a lot to contribute to this work as well.

Now you are on the right slide: “Health Care Link.” I’d just like to talk about can you go up one, please? There we go. Okay. This slide just shows what CDISC calls their Health Care Link, and it’s a general initiative that we’ve had going since about 2001, when we formed the charter agreement with HL7. And we successfully did a demonstration project that was done at Duke Clinical Research Institute with Duke Hospital to show that we could use electronic health records for both clinical care and clinical research. And that was the proof of concept which then developed into an integration profile that’s being implemented. It was developed through integrating the health care enterprise or IHE and demonstrated at HIMSS in five different use cases last February. And we’re currently now implementing this what’s called retrieve form for data capture, the integration profile in electronic health record systems for three different implementations that are going to happen this year and next.

The next slide shows some of these different use cases. So what’s happening today is, clinicians are needing to enter data several different times for multiple purposes. They have their safety data reporting. They do clinical trials through an electronic data capture system or paper, which is then the data are entered into a database at a sponsor site or a CRO. We have trial registries. We’re working with WHO for standard data to make sure that trials are registered and registries around the world or biosurveillance.

So these were use cases that were demonstrated at HIMSS this year, showing how we could take these multiple different electronic data capture systems and different implementations and tools. And if you look at the next slide, you’ve turned this into using electronic health record for multiple purposes, because what happens is, the form is pulled into the EHR that collects the data for these different purposes without having to end an EHR session that the investigator is participating in. So they can easily reuse the data for different purposes, and it’s all done in a very secure way to protect the patient identity and such.

So in closing, this last slide just shows our vision and what we’ve been working towards with standards. And we’d like to collect the data once, this once, from all these different sources and use it for multiple downstream purposes, including public registries, institutional review boards, contract research organizations, partner companies, regulatory authorities, as well as for use in the clinical care system. So what we’re trying to do is break down this wall and have real-time integration through standards, and HL7 and CDISC are participating with these together, and we’re also working with ISO. Okay, I think that’s it, Greg.

>> Greg Downing:

All right. Thanks, Becky. I hope that your Group can stay on for the next specific presentation and take questions at the end, if that’s okay with you.

>> Ed Helton:

Greg, we’ll be here. This is Ed.

>> Greg Downing:

I think the one thing I did want to acknowledge before moving on to the next presentation is, Mollie and others have continued to keep our use case teams aware of the use cases that you’ve been developing and some of the background documents for HL7. So we’ve appreciated having that content come forward. So before the next presentation, I wanted to give the opportunity to Rochelle Long, who has been attending some of our meetings in the past, addressing a number of the aspects that the NIH has been leading overall on pharmacogenomics, and she’d like to make some general comments and then introduce Michelle.

>> Rochelle Long:

Okay, thank you, Greg. I’m the Program Director at NIH for the Pharmacogenetics Research Network. That is a series of independently funded research groups, each of which has a different focus, different disease area, or different approach to pharmacogenetic studies. It includes studies in the cancer area, in the asthma area, in vascular diseases, etc. But they are united by one particular award, and that is the Pharmacogenetics and Pharmacogenomics Knowledge Base, or PharmGKB. And you’re going to hear a presentation on PharmGKB by Michelle Whirl Carrillo, who is the scientific curator there. Teri Klein may or may not have joined the line by then by now. She’s the Director of PharmGKB as a site, and Russell Altman is the Principal Investigator of the award at Stanford.

Now, PharmGKB was conceived as a research tool, so it you will see presented ways that different data types are united strictly for research purposes, at this point, so that people can see where the knowledge is linked, where there are gaps, and try to figure out, if you try and predict pharmacogenetic response, how people will respond to drugs, what are the polymorphisms you should look at, and how might they contribute. Or thinking about it another way, the research purpose was to link phenotypes and genotypes and try to go about developing some early standards in this field.

So you’re going to hear Michelle present PharmGKB, and I’m sure she’ll welcome your questions. I think the discussion thus far at this meeting is out of the level that we’re accustomed to having routinely in the Pharmacogenetics Research Network, because it’s not yet heading towards routine implementation. And yet many of the groups are actually interested in biobanking and develop the standardized electronic health records, of course, and how you move between those kind of information for research purposes and the research studies they’ve had funded to date. So this may represent the future of where our whole initiative is going.

But with that minimal introduction, I’ll go ahead and say Michelle should shoot away and present PharmGKB.

>> Michelle Carrillo:

Great. Thanks, Rochelle. I appreciate the introduction and thank you for the invitation to present about PharmGKB today.

As Rochelle said, we’re a basic research knowledge base, and so it is a little bit different. It will be a little different presentation from the previous one, which was really more at the clinical level. So can I have the next slide, please?

So I was going to start off with a little introduction of the pharmacogenomics, but basically that’s already been covered that’s why I’m here today how genetic variation impacts variation in drug response, and of course, this has long-reaching implications for personalized medicine. But right now, a lot of the basic research groups are interested in decreasing or eliminating adverse drug reactions by looking at genetic and genomic information. So Rochelle already gave you a little brief overview of the Pharmacogenetics Research Network, which was created at NIH, and so PharmGKB is one of these research groups, as Rochelle said.

Next slide, please. So PharmGKB’s mission in particular is to integrate, aggregate, and annotate pharmacogenomics data and knowledge, and PharmGKB separates data and knowledge along these lines. So, for example, the data we contain in PharmGKB consists of genotype data and phenotype data on subjects, on individual subjects, versus the knowledge that we collect at PharmGKB are acquired through by the scientific curators to create things such as pathways pharmacokinetic and pharmacodynamic drug pathways. We write summaries on the pharmacogenomic importance of particular genes or particular variants in these genes. And we also annotate literature publications that are found in PubMed or other sources. And actually, those literature annotations are what feed into the gene summaries and the drug pathways. So these are this is information that’s collected from the literature, and the curators synthesize this information from the literature into these summaries and pathways.

Next slide, please. I just wanted to show you I’m not sure how well anybody can see this, but this is the home page for PharmGKB. We’re a Web-based resource. Anybody can access PharmGKB. It’s meant for basic research. If you want to see you’re not required to log in to get to view PharmGKB the information in PharmGKB, unless you would like to see individual subject data, such as individual genotypes or individual phenotypes on a particular subject. In this case, we do ask you register with PharmGKB so that we make sure you’re a legitimate you have a legitimate interest in research. And also, to protect patient privacy and to follow HIPAA guidelines, we request that people have an ID before they view such data.

But the home page kind of gives you an overview of different types of knowledge and data. The little buttons across the top variant genes, literature drugs that’s all links to both the data and the knowledge in PharmGKB. And at the bottom of the screen is kind of a flow of pharmacokinetics and pharmacodynamics and how the genes interact with reactions to drugs and the variants that are part of those genes that affect the reactions to those drugs.

Next slide please. I’d like to apologize; I am coming down with a cold. Next slide, please. Okay.

PharmGKB accepts data submissions from the pharmacogenetics research network, from certain groups at NHLBI that have joined together and submit regularly to PharmGKB, but PharmGKB is also open to all scientific researchers in pharmacogenetics and pharmacogenomics. Right now, we have about 2030 different groups that submit to us are outside of the formal NHI groups that submit to us. Most of them are located in the United States, but we are trying to push for global submissions as well. We receive data submissions from geneticists all the way up to research clinicians. Next slide, please.

>> Greg Downing:

Michelle, this is Greg. If you I think that this is coming up in your subsequent slides, but if you could talk a little bit about if there’s an agreed upon format for the data and how it’s represented and any standards, activities that you’ve been working on if you could spend some detail on that when you come to those slides.

>> Michelle Carrillo:

Sure. Okay, great. Thanks. So we’re coming to that on this slide, just a little bit. So for genotype data in particular, PharmGKB has created a XML format for representing all of the snip-and-insertion/deletion/repeat data that people are submitting on for their genotype data on particular subjects. So I didn’t go into the XML file format in this particular presentation. We do have a paper that’s coming out in human mutation in the next few months that describes in detail the XML format. It was set up about 45 years ago. There weren’t very many well, actually there were none to our knowledge of genotype standards like that when we developed our XML format. So it was developed before DB Snip had an XML format or any of the other genotyping databases that are in existence right now.

I’m sorry, was that a question? Okay.

The XML file format is a difficult format for most researchers in small labs to deal with. You do need kind of a bio-informatics core or some kind of technical expertise to create the XML files. So we have an alternate way to submitting the data to PharmGKB, and these are using Web forms and Microsoft Excel templates that researchers can just fill out, and they fill out the Excel files with the appropriate fields and upload it to the Web site in a step-by-step it’s kind of a wizard that we set up for them to upload the files that way. We are also starting to get a lot of high throughput data from snip arrays, and that’s an entirely different type of data solution process. And we have set up mechanisms to be able to take the raw output from Affymetrix and Illumina and upload the information to PharmGKB.

I did want to say all the data that we get all genomic data is deidentified. Subjects are assigned unique PharmKGB numbers, but the numbers are given to the submitter, and the submitter is the only one that has the mappings between the PharmGKB IDs and whatever IDs are stored locally. PharmGKB has no access to local IDs. Next slide, please.

PharmGKB also focuses on phenotype data, and I know that’s not your focus right now in this particular Workgroup, but I just wanted to give a brief overview of the phenotype data that PharmGKB also collects. We have a wide range of phenotype data, so to PharmGKB, anything that’s not genomic or genetic data is considered phenotype data. And again, these are submitted to us by Excel templates that are filled out by the submitter, and the scientific curators at PharmGKB review the file. We have two different types of phenotype data that we take. One is a highly curated files, and the other gets minimal curation. For the highly curated phenotype files, the scientific staff spend a large amount of time reviewing all the fields, making sure everything is clearly defined, versus the minimal curator datasets we just reviewed to make sure that there’s nothing that affects patients’ privacy. Next slide, please.

The phenotype information is all deidentified. Again, any kind of information that would identify a person, such as their name, birth date, Social Security number, address, we have gotten files in the past with all this information in it. Those are obviously rejected right away and deleted off of any of our hard drives. We will not accept that kind of information. Next slide, please.

Some of the information is difficult to determine whether or not it infringes on patient privacy, such as where the data was collected, the type of illness the patients have, with a list of all the drugs that they’re also on. People could be identified from that type of information. So that gets a little bit tricky. But again, only registered visitors to PharmGKB are allowed to access the individual level data. Okay. Next slide, please.

Okay. So as Rochelle said, our basic science research tool we do not interact directly with any kind of electronic health records as it stands right now. All of our phenotype data that we would get from clinical research is submitted to us from the clinical researchers, but in whatever format they choose. We can take any kind of PDF file or whatever, but we do not take electronic health records directly, and we are not accustomed to dealing with those. So I just want to state that up front.

PharmGKB does acknowledge the need for genotype information to be used for critical decisions regarding drug prescriptions and so forth, and we agreed that standards are required. But I’d like to say that for genotyping data right now, we follow standards such as (inaudible) genes’ names gene names from HGNC, which is the Human Genome Nomenclature Committee. We do not have formal standards in place for representing the variants. This is something that we would like to work on, but it doesn’t exist right now in the basic research community. So just going through research publications, you’ll see that polymorphism names change from paper to paper, and it depends on who’s talking about the particular polymorphism. And so there are no standardized names for that to date. So I couldn’t make any kind of recommendations about how to deal with the actual genotype data itself to standardize these polymorphisms at this time, although it is definitely needed. Next slide, please.

Unfortunately, PharmGKB also does not currently use HL7. We have looked at the HL7 nomenclature, the terminology, but we have depended on the clinical researchers themselves to standardize their information as it’s submitted to us. And we have found, in the PGRN in particular, this has not been this is not a standard practice. And so we we’re not in a position right now to standardize any kind of clinical research, and I think that your initiative would be very valuable to the basic researchers as well as you guys looking at basic research in order to create your standards, because there’s not a lot of standards right now in pharmacogenomics in general. Next slide, please.

I’ll be happy to answer any questions that you have. I’m sorry that I can’t discuss in more detail HL7 or those kind of standards, but we do have other standards if you’d like to hear more about the XML or any of the standards that we use, like GO. We use GO terminologies and different terminologies to annotate genes at the gene level. I’d be happy to discuss it with you.

>> Rochelle Long:

Michelle, this is Rochelle. Do you feel capable of giving them a few sentences’ description off the top of your head about the Warfarin Data Consortium data sharing and also some disparate datasets into compliance?

>> Michelle Carrillo:

Sure, so absolutely. So that kind of so the Warfarin Consortium is a group of researchers all interested in the effects of pharmacogenomics in Warfarin. And PharmGKB has volunteered to collect data, both phenotype and genotype, from all of the different groups and try to standardize all the disparate datasets. I’d have to say that the most effort right now in the standardization has to do with the phenotype data, making sure that the different tests that are run the different types of information that groups collect regarding diet and so forth that those all use the same terminologies so that datasets from the different groups can be compared.

The genomic data is pretty straightforward for the Warfarin Committee. They’re looking for a few different genes, and the nomenclature for those genes are well-accepted in the research community. So there’s not quite as much work to integrate the genomic data.

>> Rochelle Long:

I would comment also, in that sense, PharmGKB was the responsive unit when a few selected investigators, some of whom are part of the pharmacogenetics network, some who are not, suggested the data-sharing exercise. About 20 different groups internationally came to participate and donate datasets for which there was one particular, singular, well-known, standardized phenotype. And that was INR.

>> Michelle Carrillo:

Right.

>> Rochelle Long:

So PharmGKB worked with them to standardize the exact formulation of the Warfarin they used, the dose, their data, to bring it into compliance, how they expressed weight or height, that sort of thing, concomitant medications and actually, it took the investigators saying, “We want the column headings to be this or that and the PharmGKB curators to get everything together to merge it into one dataset for analysis.” And indeed, they have research data on the five different subjects, and they’re going back and quality control-checking some of it now as they prepare to analyze it. And I think that is the case where it took a research effort to decide, “We want to standardize. Here’s our area.” You know, it’s a big universe, and it gave them a place to sort of start chomping down and working on it.

I think the whole discussion of standards has been much more complex than I ever fathomed when this started nearly a decade ago, from “Standardize first and then everybody submits to let everyone submit and see where the standards lie” through “Let everybody express things individually,” and that’s more of the dbGaP philosophy right now, a very different way to approach archiving phenotype and genotype data. So I’m seeking to learn from this whole discussion as well, and I hope those comments are helpful.

>> Greg Downing:

Well, thank you, Rochelle and Michelle. That this has been very, very helpful. And I think that everyone’s sort of beginning to ride this crest of this wave trying to figure out what direction to go in on standardized approaches.

I’d like to bring everyone back that was involved in this discussion and open up it up for questions or discussions overall. (Pause) If not, I think I had a few, and that you’ve pointed out at the end here, Michelle, that you saw this process unfolding in order to address this issue around a few polymorphisms and Warfarin. Do you have an idea, as these kinds of studies continue to unfold the association of drug metabolism, for example is that you is there a template that will evolve in terms of a standard practice, or is there, for each particular study of unfolding, a standards process we’ll have to go through for each kind of large database exercise?

>> :

Well, I’ll make one comment, and I’m not entirely certain this is answering your question as much as skirting it a bit. When I originally envisioned this, I thought there’d be more mergers of datasets from different sources for meta-analysis. I guess what I currently see as the capacity to collect large datasets in a single study, even on a single sample, have increased, there’s much more, within a single study, data collection and then analysis in that dataset. And remember, I’m speaking from the research side of things, and I really see a different standard being set as dbGaP setting up a study-by-study access, so much less towards standardizing to merged data, much more filed through a single dataset to extract as much value out of it as possible. And I feel like even the discussions are around patient privacy and ethics and access more than standard setting.

And I personally try to pick the brains of people at NCBI a little bit to figure out how that’s going to unfold. I don’t think I know the answers, nor do I think they know either. But all this research shifting and weighing almost has to come before we can move on to clinical decisionmaking. If you don’t know what it means, you don’t know what to recommend. And that’s where I’m I guess I wish things were more standardized than were. I wish people would get together and say, “Let’s conduct our experiments and our assays the same way,” but they’re not inclined to.

>> Michelle Carrillo:

It is very difficult to get I’m sorry; this is Michelle Carrillo to get researchers to agree on protocols and ways to collect their data and then conduct their tests. And I think that Warfarin showed us that the Warfarin Consortium showed us that there’s going to be some degree of having to define standards with each kind of project that comes through. I think the genotype data you can the genomic data can be standardized much more quickly. I hesitate to say “quickly and easily,” but it’s easier than the phenotype data. And that’s where most of the I think most of the difficulty comes in.

>> :

This is a question for both groups. If you knew that the bulk of your research participating institutions had electronic medical records that could be used for supporting research information collection and that had standardized genetic information in there, would you see any barriers to sort of adopting that type rather than sort of the re-entry issues that Becky addressed initially? Are there still barriers to utilizing the patient electronic patient records for clinical research purposes of the data extraction, even if those capabilities exist?

>> Phil Pochon:

This is Phil Pochon from Covance, and so I represent a central lab. One barrier I see is actually working through what is the relevant data for a particular test and how much granularity and very detailed information do you need about a specimen collection, specimen handling, the extraction process, all of the very fine grained information that in certain situations may be useful. And that is an area that, again, we’re just beginning to really get a sense of and perhaps talk about some standards of what is information that should just routinely accompany an analysis versus what may be there need to be there only if there is a significant question about analysis that truly needs a deep dive. And I think that’s another area that, before you begin to think about putting data on the electronic health record, one of the questions you want to ask is how much of this data and how fine of a grain of the detail of the data does the electronic health record need to have versus maybe simply getting a pointer to who performed this test and who should we contact for that very fine-grained information should we need it, you know, for a very specific question. And I think that is another question that is looming out there as well.

>> Marc Williams:

This is Marc Williams. I don’t mean to presume that we have all the answers here, but as Michelle was mentioning, issues relating research it seems to me that there is, under the purview of the Secretary, an opportunity to perhaps incent researchers to get on board in that if we were able to develop some ideas about how we may want to reflect standards in the submissions that would go to applications for NIH- or CDC-funded grants that that might be the place where we could start to really engage the community to make the necessary changes. I certainly understand that we’re not at the point where we really have existing standards that we could promulgate, but I think, in terms of a long-term strategy, it might be a reasonable thing to think about to try and get everybody on the same page.

>> Joyce Mitchell:

Hi, this is Joyce Mitchell from the University of Utah. And I wanted to make a comment on the previous speaker before Marc. It’s clearly is a case that you cannot keep every single detail of the data, but on the other hand, the current situation where, for a lot of at least the sequencing data and many of the other more new tests that the data is not kept at all in a way which is even connected with the electronic medical records so that if you, as a clinician, order a test from a pharm laboratory other than your own, which is usually the case, you just get a report back. You never actually get the sequence. So you’ve ordered a sequence, and you never get the sequence. And if you want to go back and analyze it at some point in the future, what is actually happening now, as people are looking at two- and three-gene combinations, you can’t do it. You have to reorder a test. So I would say that there is some basic information which needs to be kept no matter what and then some access ready access to other information to allow you to do reanalysis as you learn more.

>> Janet Warrington:

Hi, this is Janet Warrington at Affymetrix. The I’m not sure who just made that comment. It was an excellent comment, and I would just add that along with that sequence information, though, you need some way to benchmark the quality of the data that you’re looking at. And so besides getting the raw data, you would want to make sure that there was control information included with that data so that you have a good understanding of how that data was how that sequence was read and interpreted. And I think that gets back to incentives for researchers for developing a consensus around control materials and incentives for sharing those materials. One of the biggest concerns and complaints that I’ve heard from the (inaudible) diagnostic laboratories doing oncology tests right now is the lack of materials for proficiency testing. And it’s a pretty basic need that we’re going to have to find a way to address.

>> Ron Farkas:

This is Ron Farkas from the FDA. With the talk about, you know, sequence data allowing people to go back and see what the test actually showed, what comes to my mind, and what I think is what we’ll see is going to be important pharmacogenomic information, is PCR tests that kind of purport to tell you what the sequence is and but actually to narrow down what the sequence could be. And that, I’m fairly sure, will be relevant information used or, you know, referred to in drug labeling. And the interpretation of that data is complicated. There’s different tests for the same gene that gives somewhat different information that and it has certain inherent biases, even depending on the ethnicity of the patient tested.

And so I guess my overall point is that the clinically used tests are going to give information that is not that granular. I mean, again, ultimately, I suppose that it could be. But trying to figure all that out is very difficult.

>> Greg Downing:

Any comments on either of these presentations or related issues?

>> Marc Williams:

Yeah. This is Marc Williams again. I just to respond to the last speaker, you know, I think that the point is well-made, but I think it also strongly argues for, you know, being able to capture any of the relevant data, you know, whether it’s ethnicity, demographics, genotyping information, etc., in a structured fashion, so that as we begin to unravel this Gordian knot to be used predictive for prediction that we don’t have missing data that is not represented that we can’t pull in.

So I think that, you know, what we’re really talking about here is a foundational concept representing information so that we can actually then build the knowledge use the clinical decision support that was referred to earlier off of all relevant data.

>> Becky Kush:

This is Becky Kush, and this Data Tabulation Model that I talked about earlier is already covers a number of the domains for safety data that are submitted to FDA. And it is being proposed as a regulation for that submission. So we’ve been building off of that, adding domains as appropriate, which is what Joyce is doing with the genomics information. And we’ve also been defining, based on that, the data that should be collected, so and how that should be collected in a formatted way. So it covers things like demographics and all of the domains that would be relevant for safety assessment. And genomics gets a little more complicated, which is why it’s not done, although a lot of work has already been completed that Joyce and Phil and Mollie presented. But that base foundation is there. And through the Critical Path Initiative, we’re working on data collection standards, which should be out for public review in Q1 ‘08. So that’s what we’re using when we’re doing the Health Care Link and we pull the form into an electronic health record. And the more data that are available in a standard format, the better that can be prepopulated and save time on the part of the investigator.

>> :

This is maybe a little bit off the main scheme of our discussion, but those organizations that are part of CDISC and representing pharmaceutical development let me, if I may, ask what kind of a strategic opportunity exist for institutions in participating in research programs or protocols that have electronic health records. Does this make a major factor in how, in both public- and private-sector research enterprise activities that having a well-developed information structure for information gathering and the ability to adapt those information sources into clinical research protocol is this a major factor in how clinical programs are developed?

>> Becky Kush:

It’s a strategy for a number of them. Right now, they’re using these electronic data capture systems that are all point solutions. So strategically, they would like to see electronic health records be used for clinical research. And there is work that’s been done. It’s posted in a document on the CDISC Web site. It was done with FDA to make sure that these can be done in a way that follows the current requirements for regulations, the 21 CFR 11, or making sure that your data are appropriately stored and moved and that they follow the security and confidentiality policies and all of that.

So that work has made a lot of headway. The European Medicines Agency just referenced that now in a document they’ve now posted. So those kinds of infrastructure have been addressed, and now a lot of companies Pharma has a task force and the e-Clinical Forum are providing an EHR profile for clinical research that would allow an EHR to be certified for doing clinical research. And that’s out now. It will be out for review and valid, I think, in the second cycle of next year. So they have a document that’s being reviewed now that will then be valid through HL7. So it’s a functional profile that would allow for using EHRs for clinical research. It’s a strategy for a number of these companies, because they see the value and they see that they should make life easier for investigators who want to be practicing physicians and investigators.

>> Joyce Hernandez:

This is Joyce Hernandez. I just wanted to share a little bit. We’ve done a lot of collaboration with health care providers in terms of our research. And our approach was to use the Study Data Tabulation Model, because there’s not a real common EHR record format that’s published out there to be used. The problems we incurred were mainly due to health care organizations and to have very small IT staff, so to have them try and learn a new standard and reformat data is kind of a bit labor intensive for them. So that was more of a time aspect. But once we had the data formatted, it was easy to slow it up from the health care provider into our data center. But it would have helped to have had an HL7 EHR record format message that would be utilized by the health care IT, because it would basically, a lot of those systems support some sort of HL7 messaging, but they don’t support electronic health records at the moment.

>> :

I think that I don’t want to take up the rest of the time here for questions and comments, but a featured issue on electronic information exchange and commerce was focused on on “60 Minutes” last night. I’m not sure if any of you saw that, but the context of having efficient systems and with up-to-date standards will provide just that their message was that those organizations and companies that have participated in the standards-based practices have better assurances and quality of the privacy protections than those who don’t. And so the aspects of that you engender for your own sort of own purposes, the use of systems that are standardized there seems to be a lot of merit, at least in that particular nonhealth application, so using standards-based processes for exchanging health information this way.

>> Joyce Hernandez:

The other thing I want to question everybody is, don’t forget the vocabulary, because that’s very important. Everybody uses different standards, and there has to be a good mechanism to do the transformation from one vocabulary to another. So that would definitely ease in exchanging the information.

>> Greg Downing:

Any other comments or questions on these presentations?

>> Rochelle Long:

Greg, this is Rochelle. And (inaudible), ‘cause it’s not my area, but thoughts that come to mind include there are a number of academic medical centers right now really seriously getting into biobanking and using their population they see for research purposes, with very different models of how they’re retaining the information from the medical record and who has access to it. They’re clearly not standardized whether they scrub it and keep a version of it or keep it all intact, but they very much limit access. And again, from the academic medical center direction, it’s to enable and permit research as far as they can push the boundaries. I think, in particular, the Vanderbilt model is interesting, because they had a declaration that if they scrub the record enough, it’s not a human subject anymore and then link it to discarded samples that people have to opt out of the program. That is a possible way if you’re going through exploration as part of this Workgroup to look at the different models.

Alternatively, I’ve also been to HMO research network meetings where their primary purpose is “Treat the people first; do the research later.” And they have different ways that they’re keeping the records, I’m sure, more standardized for information flow, at least within the system less amenable to research, because less rich detail is being captured. But I don’t even know that there is complete agreement on what should be standardized, yet alone how to do it, when you throw in the aspect of for-research purposes. Of course, I come from a research background. But I guess I can envision some sort of hybrid model where you want as much information as possible for research, but when the conclusions are drawn, you want to standardize to a smaller number of information bits to get in the clinical decisionmaking cycle. And that is separate from the research information-gathering and analysis cycle.

>> Greg Downing:

Thank you. This has been a very interesting discussion, and I think it’s brought the two worlds a little closer together in terms of understanding of application more. Thank you all for putting your teams together and the presentations. But what we’ll owe you back, I think, are to keep you connected with how this Workgroup perceives the needs and potential ways in which the AHIC contributions and processes can facilitate the work that you’re doing, and we’d be happy to carry on the dialogue along with you afterwards.

So again, I want to say thank you. And then, John, I’ll pass it back to you for the next

>> John Glaser:

Yeah, thanks, Greg. First of all, very nice, thoughtful presentations along with the Q&A, and Greg, very ably led. Greg, what I think we might want to do next time you, Chris, and Doug and I get on the phone is see if we can put our arms around some work we might do. There’s some pretty broad terrain that was covered here. I don’t know that we could solve it all, but I suspect we could tackle portions of it. But I wouldn’t mind, you know, if we spent a little time scoping it and maybe outlining some possible work here and then getting back in front of the folks next time we have a call and see what folks think.

>> Greg Downing:

Yeah, I think that’s a good idea, John. A number of people on the Workgroup have volunteered while their arms have been twisted into various contortions

>> John Glaser:

Okay. (Laugh) In that holiday spirit.

>> Greg Downing:

with the (laugh) potential ways to formulate specific actionable, items, so...

>> John Glaser:

Okay. That’d be good. The next topic is regarding decision support. And let me give you a brief overview while you’re all looking at Document 7. This effort has gone from 0 to 60 in a remarkably short period of time. To just give you an overview of this, there will be a meeting of the Ad Hoc Clinical Decision Support Group in mid-January. I think the date has been set, although I don’t recall it right offhand. There will be a number of folks invited to that, including members from the various AHIC workgroups. I still need to have a little bit of a conversation with Doug to see who all from our particular Workgroup ought to be present in addition to myself, who will be chairing that meeting.

The intention of that meeting in January is to come up with a set of recommendations, or at least an initial potential set, for presentation to AHIC. Whatever those recommendations are, we will have a chance to look at them as a Workgroup to make sure they make sense to us and reflect the type of things that we want to do. But there’s a desire to get some recommendations into AHIC within Quarter 1 of 2008. There will be a call in December to organize that event and to get some sets of data put together and background pieces put together such that when the Group meets in January, it is got a, you know, coherent set of material in front of it.

What will be desired from each of the workgroups, including ourselves and the Personal Health Care Workgroup has been one of the lead advocates and movers behind this is a series of vignettes that’s what you see in the document you have in front of us that describe clinical decision support scenarios. And the Clinical Decision Support Work Ad Hoc Workgroup hopes to get a scenario such as these from all the Workgroup and to see if they can assemble and piece apart common threads that exist all across all of them in addition to unique elements, which might be quite specific to personal health care or specific to population health.

So one of the things we have been tasked or asked to give them is a set of vignettes; the second, to the degree that we have them if there are some recommendations that go on top of the vignettes that we are being asked to forward those to the ONC folks who are providing the staff support for this. There are clearly some recommendations we had in our letter to the Secretary back in July. There may be additional recommendations that are we have at this point in time.

So I just wanted to give you a brief overview. Again, in sort of mid-January, we get a preliminary set of recommendations, which we will have a chance to look at before they go off to AHIC. The presentation in front of AHIC will probably be done by some set of Workgroup members rather than the Decision Support Workgroup itself. So a sort of initial question for you all is, to the degree you’ve had a chance to look at these is comments and reactions to the scenarios that you have in front of you. There are a number of them. And there may be, in particular, at this point, areas that you all feel are notable omissions of vignettes that ought to be there or where a particular vignette is not an important one or sort of is off the mark by a good 1530 degrees here.

Let me just stop and see if there’s any question about the process itself and/or any feedback you all have at this point on the vignettes and/or specific comments or points you’d like to make. (Pause) Fair enough. If you have a chance in the couple weeks ahead and so between now and maybe second, third week in December to look at these vignettes and make comments, we’d love to see them. If there are forms of recommendation that you would like to make sure that the Clinical Decision Support Ad Hoc Workgroup takes into account in January, again, we look forward to seeing those. But again, you’ll have a chance to see the work product coming out of that before it moves its way into AHIC.

>> Marc Williams:

John?

>> John Glaser:

Yes?

>> Marc Williams:

This is Marc Williams. I noticed at the end of the document that you had some suggested topics. I don’t know if you’re looking for volunteers or not, but I’d be happy to do the colorectal screening vignette.

>> John Glaser:

I appreciate that, Marc, and we would be delighted to have you help put that together.

Still anybody who would be interested in the other tab there for newborn screening?

>> Michele Lloyd-Puryear:

Uh oh. I’ll help with newborn screening. This is Michele.

>> John Glaser:

Thank you, Michele and Marc.

>> Steve Teutsch:

John, this is Steve Teutsch. I was wondering if in terms of the vignette, are they intended to be directive as to what we think people should be doing? Or because for many of these things, there aren’t clear guidelines yet. There are

>> John Glaser:

Yeah, I think it’s their it’s intended to be illustrative of what is likely to go on or should be going on, etc. So they’re not purely today’s practice, obviously. And the particular vision they paint may or may not happen in a particular time period, but in order to make sure that we cover the waterfront of as best we can do today of potential clinical decisions for use to support personalized health care, these would be the scenarios. So I’m not sure I quite answered the question, other than we’re not here to make recommendations of what should occur. We’re here to say, “This is what decision support could very well look like in this area.”

>> Steve Teutsch:

Yeah, I think as long as it has sort of that caveat, some of these, I think, are pretty clear as to what probably should happen today, and others are pretty uncertain as to whether they should or shouldn’t happen. And as long as that’s clear that these are examples that one would want to work through, I think, then, they make sense.

>> John Glaser:

That’s fair, Steve, and we’ll make sure that the preamble points that out. We’re not here to recommend practice inherently in this exercise.

>> Marc Williams:

Yeah, and Steve, this is Marc again. I worked on a number of these vignettes, and essentially, I was just trying to represent important concepts that needed to be thought about from the perspective of what HIT would be asked to do, using things that were at least emerging. So I would basically support John’s approach.

>> Steve Teutsch:

No, I think that’s right, as long as we’re clear.

>> John Glaser:

Okay.

>> Greg Downing:

John, this is Greg. It might be worthwhile just mentioning, you know, the five sort of thematic areas that cross-cut the Workgroup’s interests and these vignettes. I could go through them if that would

>> John Glaser:

Why don’t you do that, Greg? ‘Cause I don’t have them in front of me.

>> Greg Downing:

So they this has come out from the sort of the cross-workgroup discussions of what are the common themes. And the vignettes, I think, help support where some of these issues come up. So and there’s no priority among these. So I’ll just mention them as we’re going forward on the through the list. The one and this has come out of in some context of a road maybe a roadmap is sort of a common warehouse, database, what have you, of common rules that are used in electronic health record-based clinical decision support tools. This is a repository of sorts, kind of just a concept around this avoiding duplication. Another concept is workflow clinical workflow in the how do the tools themselves integrate into the physician or provider and consumer’s practice base. Another one is around the medical evidence review or development or who certifies or credentials, what contacts the basis of the recommendations are made from. I think this is something that Steve’s been quite interested in. Another element was the patient-consumer interface, or consumer empowerment preferences for integration into them. So that’s for missing one, I think (inaudible) I think the way this meeting that John’s leading is structured is to address these common themes across the various vignettes and workflows.

>> John Glaser:

Yeah, there’s also been discussion of helping the smaller provider small physician office, small hospital manage the implementation and monitoring of these. So what will be elaborated on and we’ll flush out a little bit more in the next few month is some themes that appear to cut across, obviously informed by the vignettes, and their recommendations are likely to parallel the themes.

Any questions or comments?

>> Joyce Mitchell:

Hey, John, it’s Mitchell Joyce Mitchell from Utah. I’d be happy to work with Marc Williams on the colorectal screening recommendation.

>> John Glaser:

Terrific. And what we may very well do is get some preliminary versions of the recommendations stuff that might go into the January meeting for review prior to that meeting by all to see whether there or at least initial thoughts zeroing in on the right terrain.

>> :

And just FYI, the U.S. (inaudible) Service Task Force is just finishing its recommendations on colorectal cancer screening.

>> John Glaser:

Okay. Were there comments?

>> Mollie Ullman:

This is Mollie Ullman from Harvard Partners. Just looking briefly through the vignettes, particularly where the use cases are going to focus on pharmacogenomic guidance, it might be nice to have one that actually detects an allergy.

>> John Glaser:

Hmm. Right, Mollie. Okay.

Other comments? (Pause) Well, again, we look forward to your if you have comments or thoughts offline, feel free to send them our way. We look forward to them. We’re now into the Next Steps section. I think Greg, Kristin, I don’t know if you all’ve been keeping up, but I have a couple of thoughts. Why don’t I ask you to start here with things we owe these folks?

>> Kristin Brinner:

Sure. This is Kristin. Greg and I have been developing a list of things that are ongoing that we want to update you on and then some topics that we were also considering for Workgroup discussion in the upcoming year as we planned for the next several meetings in the spring. The first is an update on the paper that we are generating from the CPS documents that was presented to the Workgroup at the last meeting. And we’re working to kind of translate that paper into something that more resembles something we could submit for publication along the lines of the overview document that we discussed today for the Workgroup as a whole. And so we should be following up, I think, with the CPS Subgroup largely, but we’ll also be sharing it with the whole Workgroup as that process continues. And again, we hope to have that largely wrapped up, I believe, by the end of the year. The second update is that the detailed use case for personalized health care should be published for public comments by the end of December or the beginning of January, so the timeline has been pushed back somewhat from what we initially thought. I think there was just more work involved. But it should be coming within the next month or so, and so we will again, like when the prototype use case was published, we’ll try and arrange a time where the use case team can present it to the Workgroup for discussion, and that will also work on gathering people’s comments to submit them during the public comment period.

So those are updates on things that are ongoing. We also had two items that we have been following on the periphery a little bit that we thought may be of interest for future Workgroup consideration. So I just want to bring those two up. The first was that we saw, recently, there’s been some activities in the area of standards for population genomics, so things like the genome-wide association studies. There was an article published that NHGRI has given an almost $7 million grant to the Research Triangle Institute to develop standards for large genome-wide association and epidemiological studies. So that may be the concept of population genomics could be something that the Workgroup could take up in the future, and I just wanted to highlight that.

And the other topic that we have found that may be of interest has come out of two separate conversations, and that’s the area of genomics to better understand the biological underpinnings of transplantation. And we have been talking with the CDISC Group as they planned for their presentation today, and one of the things that they are working on is a family model based on genetics and genomics for tissue typing to better understand if a donor could be a match to a recipient. And then I believe we also have talked to Janet Warrington of Affymetrix about some studies ongoing at Scripps and USCF that are profiling blood by genotyping of both RNA, DNA resequencing, copy number to better understand donor recipient matches and also to follow the recipient over time to potentially have a better understanding of the biology behind rejection.

And so that’s two topics of population genomics and also transplantation genomics. There could be areas where we could look at those in terms of the standards that are being used to develop that information. We just wanted to highlight those and see if you had any comments, if you had any specific interests in either of those areas. And if so, we can work to bring those forward, you know, as the Workgroup continues its work through the spring of next year.

>> John Glaser:

Any immediate reaction? (Pause)Well, I think, Kristin, what we can do is at an upcoming meeting is put a little more trappings around those things and get some folks’ thoughts about that.

>> Kristin Brinner:

Sure.

>> John Glaser:

And then, coming out of today, we have the JAMIA’s revamping, along with the discussion of the sort of broader set of stuff here, and then a conversation to cycle back on some of the pharmacogenomic test work and see if we can get our hands around a potential scope here. Did I miss anything?

>> Greg Downing:

No, I think you’ve got it. And you know, I think just thinking back to where we started about a year and a half ago, the four areas of development for recommendations that we were charged with from AHIC were in genetic tests, the family history information, clinical decision support, and privacy and CPS kinds of considerations that the Workgroup has made considerable progress to all of those areas. So we’re very pleased with the products and the progress thus far.

>> John Glaser:

Excellent. I think, Greg, Kristin, thanks to your able leadership, frankly. So if not I guess we Judy, we should open up the public commentary section and see if we have anybody.

>> Judy Sparrow:

Yes, I think it’s time. So Chris, can you open up the lines for any questions?

>> Chris Weaver:

For those of you that have been following along on the Web, if you want to ask a question or make a comment, please dial the number that’s on the screen and then press star-1 to get in the queue to make a comment or ask a question. And for anybody who has been listening on the phone, all you need to do is just press star-1, and it’ll put you in the queue. If you guys want to make up any or make any wrap-up comments, we’ll wait to see if anybody queues up.

>> Marc Williams:

Hey, John, it’s Marc

>> John Glaser:

I don’t have any wrap-up comments other than, again, “Thank you all for being here and the presentations and the terrific conversation.” Anybody else have some final thoughts?

>> Marc Williams:

John, Marc Williams here. Greg had a nice reference list, but I wanted to let the members of the Group know about another report from Secretary’s Advisory Committee that’s just been released for public comment. It’s The Oversight of Genetic Testing. And there are a lot of issues relating to information transfer, clinical physician support, and that are also represented in that report, and we’d really appreciate people’s review and comment on that. We’re anticipating that we’re going to be getting a draft to the Secretary by the end of February, so public comment is open until December the 21.

>> John Glaser:

Thank you, Marc, and we’ll find a way to make sure people know how to find that. Any other comments?

>> :

The only other one, John, is that Secretary Leavitt will be in your backyard on Thursday this weekend giving a number of talks on personalized health care, and I think you’ll probably find common reference to the importance of inform health IT and information management to augment this. So just wanted to bring you he’s clearly referencing the work of the Workgroup and their contributions overall to personalized health care.

>> John Glaser:

We’re delighted to host him. I hope the weather holds for him.

>> Judy Sparrow:

Chris, any comments?

>> Chris Weaver:

We have none.

>> John Glaser:

Excellent. Well, again, I think we’re done. Thank you all. Welcome back to work. I know that I’ve heard, anyway, that the average person gains eight pounds between the beginning of Thanksgiving and the end of the year. I hope you did mightily in your efforts to meet that goal. (Laugh) And so without further ado, I think we are adjourned.

(General expressions of thanks)