American Health Information Community

Personalized Healthcare Workgroup Meeting

March 17, 2008

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.

>> Judy Sparrow:

Good afternoon and welcome everybody to the 13th meeting of the Personalized Healthcare Workgroup. Just a reminder, we're operating under the auspices of the Federal Advisory Committee Act, which means this is being broadcast over the Internet and the public will have an opportunity at the end of the meeting to offer public comment. Workgroup members on the phone please remember to mute your telephone lines when you're not speaking.

And on the phone today we have the two Cochairs, Dr.John Glaser and Dr. Doug Henley. We have Emory Fry from the Department of Defense. Kathy Hudson from Johns Hopkins. Betsy Humphreys from NLM. Lisa Rovin from the Food and Drug Administration. Steve Teutsch from Merck. Janet Warrington from Affymetrix. Marc Williams and Grant Wood from Intermountain Health. Allen Rudman also from the DOD. Katherine Kolor from CDC. Elizabeth Mansfield from the Food and Drug Administration. Dina Paltoo from NIH. Ronald Przygodski from the Veterans Affairs.

And in the room we have?

>> Mary Beth Bigley:

Mary Beth Bigley from the Office of the Surgeon General.

>> Greg Feero:

Greg Feero from the National Research Institute.

>> :

[[Inaudible]] from the Human Genome Research Institute.

>> Becky Fisher:

Becky Fisher.

>> Maren Scheuner:

Maren Scheuner from RAND.

>> Judy Sparrow:

Did I miss anybody on the telephone?

>> Clem McDonald:

Could I make a correction? I'm Betsy Humphreys.

>> Judy Sparrow:

You don't sound like Betsy.

>> Clem McDonald:

I have a hoarse throat. No, Betsy has asked me to substitute. I'm from NLM.

>> Judy Sparrow:

Is this Clem McDonald?

>> Clem McDonald:

Yes.

>> Judy Sparrow:

Anybody else I missed?

>> Dennis Williams:

Dennis Williams from HRSA.

>> Judy Sparrow:

All right, thank you.

>> Mollie Ullman:

Mollie Ullman, Harvard Partners.

>> Judy Sparrow:

Thank you.

>> Kristin Brinner:

Kristin Brinner is also here from HHS.

>> Judy Sparrow:

Sorry. And with that I'll turn it over to cochairs.

>> Doug Henley:

Thanks everybody. Again, John and I want to thank the workgroup members and the superb staff for their continued excellent work as members and staff of the workgroup. And you all do extremely good work and we're thankful to you for that.

The first item for our agenda today is approval of minutes from our February14th meeting that is before you for your discussion, for edits, corrections or deletions. Any comments from the workgroup?

(No response.)

Hearing none, then they will be approved. And the next item you'll see is the update on the presentations that John and I both made to the AHIC on Februarythe 26th regarding newborn screening.

Happy to say all those recommendations which many of you worked on in a very excellent fashion were well received by the AHIC and were approved and are now in the hands of the Secretary for his consideration for implementation.

Again, I think there was general consensus that they were excellent recommendations and important and necessary to move forward the charge that we have related to interoperability and standard setting in the area of newborn screening. John any comments that you have?

>> John Glaser:

No, I thought good questions were asked. It was uneventful approval and it was the quality of the work that Doug and I conveyed on your behalf.

>> Doug Henley:

Indeed. Any comments from the workgroup and/or staff about the AHIC presentation and those recommendations relative to newborn screening?

>> Kristin Brinner:

Doug, this is Kristin. I just wanted to give a quick update. We're going to have one final official meeting of the newborn screening subgroup that we had convened to develop the earlier draft of these recommendations to let them know as well how the presentation went. They'll be finalizing their kind of version of a use case that we can then, we'll then be presenting to the workgroup.

That hopefully will be used in the upcoming steps for prioritization of use cases by AHIC and then hopefully a newborn screening use case from the Office of National Coordinator can follow from that.

We're also working on an implementation document for the recommendations to make sure that we communicate to all the appropriate groups, and that should kind of follow along with the Secretary cost [inaudible] recommendations as well.

>> Doug Henley:

Thanks Kristin for the update and convey our thanks to the workgroup for a job well done.

Any other questions or comments? (No response.)

Okay. Then we get into our initial agenda, main agenda for today. And we're going to have a series of updates now and information from various folks regarding activities as it relates to family health history.

We're first going to hear from Mary Beth Bigley and Greg Feero there at the office in DC about an update they have on the data set and My Family History efforts. Then we'll be hearing from Vish Sankaran. And you have document No. 2 in your agenda about the federal health architecture efforts.

Then Dr.Terry Cullen will be speaking about what's going on in Indian Health Service relative to family health history. So let me turn it now to Mary Beth and Greg for their comments.

>> Mary Beth Bigley:

Hi, good afternoon everybody. This is Mary Beth Bigley. Greg and I are going to do this together. I'm going to start by giving a little bit of where we are today going back to the adoption of the recommendations back in July of last year.

And we put together a workgroup who, over a series of several months, came up with a minimum core data set document that was circulated and reviewed. And with the help of Indian health service, that minimum core data set became a requirements document, which I think has been shown at this workgroup before.

That document which has had really such great input from the workgroup as well as the members of this workgroup has moved forward in a lot of different arenas. One is it became part of the detailed use case, the personalized healthcare use case. It went out for public comment as part of that.

It did receive comments back. The majority of the comments were to clarify some of the information in the fields that were the data fields. But additionally there were several comments related to the functionality of how the fields and the data elements would relate to each other. So I think that the feedback that we've gotten so far will be helpful to then move this one step further in the process.

>> :

Could you clarify where that can be found now, the current version?

>> Mary Beth Bigley:

That’s a good question.

>> Kristin Brinner:

Mary Beth, I think it can be found on line with the use cases but we can send around a link.

>> :

Okay, appreciate it.

>> :

Is that the most recent version with the incorporated changes based on the public comment?

>> Kristin Brinner:

No it would just be the last version, and I believe they will be posting the final detailed use case as well as the final requirements document I think at the end of this week or next week. But it will be on the same website.

>> :

Excellent.

>> :

Which one….I have to confess it's hard to find things sometimes on these websites.

>> Kristin Brinner:

If you go to the AHIC website, along the righthand side there should be a thing that you can select that says Use Cases and it should say Personalized Healthcare. We can send a link out to the whole group.

>> :

Okay. Thank you.

>> Mary Beth Bigley:

If you want to look at it sooner, within the documents of the Personalized Healthcare Workgroup, it's in the archives of the meeting documents from probably two, three meetings ago.

So that as Kristin said will be hosted there. And then we're going to there's several activities right now to communicate that document both inside the federal system as well as outside the federal system. We have a workshop that was submitted and accepted at the HMR research conference which is next month. And that will be able to reach a population of people that have EHRs that are looking at how they want to incorporate family health history and these elements into their own EHR and to start working with this for the adoption of the specific data elements and functionality that's built into that document.

We've also submitted for workshop as part of the AMIA conference in the fall and a paper to AMIA also to get the word out that this document is being circulated and we'd like people to look at it, provide feedback so that as we move forward with the standardization it gets as much input from people using it as well as the electronic health systems vendors who might be looking to incorporate this into their electronic health systems.

There's been a series of accepted posters and different professional conferences that we've also had out there to again spread the idea of this family health history and the requirements document.

One thing I failed to mention is that the JAMIA workshop is going to be with the Indian Health Service in bringing in the Indian Health Service since they were a big part of putting together this requirements document and will be working with us to move forward as we implement this into electronic health systems. Greg?

>> Greg Feero:

Sure. So as Kristin mentioned, there are a large number of efforts right now ongoing to try to promote both the use of the core data requirements document as well as to potentially think creatively about how we might use existing tools to sort of move forward on the recommendations that followed in the use case after the recommendations suggested in the core data set be developed.

And there are currently a number of both federal and nonfederal partners who are interested in working with both the core data set as well as the tool. And on the NHGRIN, we have been working to develop a mechanism by which, with nonfederal partners, we can effectively share the tool in an open source way so that any modifications to the tool in the future get contributed back to essentially the public domain.

And I'm happy to report that in the relatively near future we'll be in a position to offer the tool out for nonfederal partners.

At this point in time we have also shared the tool, My Family Health Portrait with IHS in its entirety. And I think you'll hear more about this afternoon from Terry Cullen. And they are now working to link this to their EHR system directly.

The My Family Health Portrait tool has also been posted on the TRICARE website as well as the IHS website and is available for patient use. Although at this point in time it is not directly linked to either their EHR or PHR systems. It sort of acts as a standalone tool.

I think the most exciting upcoming event will be in the near future. April9th there will be a meeting of federal partners to discuss moving forward with pilot projects utilizing family history, both in terms of the core data set as well as My Family Health Portrait and perhaps the CDC's tool Family Health Ware as a starting point for pilot projects within the federal healthcare systems.

That meeting will include representatives from IHS, TRICARE, as well as the Veterans Health Administration.

So I think we've made very good progress. Much thanks goes to many of the folks around this room that have allowed this to move forward. And I'd particularly like to thank Greg Downing and his group for their efforts at coordinating the intersection of the various federal partners that are continuing to work in this area.

>> Douglas Henley:

Thanks to you both Mary Beth and Greg for a great update, a lot of activity going on.

Any comments or questions to Mary Beth or Greg from members of the workgroup?

>> Clem McDonald:

Just one question, there's an activity, there has been underway at HL7 about family history and a data model and a message, I believe, which is led by somebody from Harvard.

And Mollie, you might help me remember the name. But just wonder if that's been looked at or considered in all these thinking processes?

>> Mollie Ullman:

Yes, Clem, this is Mollie. Kevin Hughes, one of the cochairs for the HL7 clinical genomics SIG, led that work at HL7. He was actually pulled in and made a part of the family history subgroup work. So it has been harmonized.

>> Greg Feero:

They’re definitely in alignment. I think longterm one issue will be for this core data set to sort of be maintained, it will likely need to be adopted by one of the recognized standard development organizations for longterm maintenance, and that is something that we'll have to continue to look into as we move forward.

>> Doug Henley:

Good point. Other comments or questions? All right. Again, Mary Beth and Greg, thanks very much for the update. Let's move on now to Vish Sankaran from ONC who is going to give us an update on the federal health architecture as it relates to family history and so forth. Vish.

>> Vish Sankaran:

Thank you. Thank you very much. Thanks for having me here. Once again my name is Vish Sankaran and I'm the program director for the Federal Health Architecture Initiative an E initiative under the President's management agenda, which is managed within the Office of the National Coordinator.

I'm going to use a slightly modified slide deck which is very much focused on the federal agencies. It might not agree with what you have in those printouts. I hope most of you have access to these slides on line. So to start with, some of the federal healthcare challenges today all right.

>> Doug Henley:

The one now showing is our slide No. 2.

>> Vish Sankaran:

This is a revised slide deck. It’s more focused on the federal government. So let's talk about some of the challenges we have in the federal government today. Like if you look at Social Security, there is around 750,000 claims pending, and it takes over a year and a half to resolve those claims.

Look at the Wounded Warrior population within the Department of Defense, if you look at what's happened with Katrina, most of us know. But also the growing federal government share of the healthcare costs which is around 45 [[Inaudible]] 2.1 trillion dollar industry. These are some of the federal challenges.

Next slide, please. At the same time that's not the list. You have a lot more issues within the federal government that goes across multiple agencies that include all the operating divisions within HHS. Look at DOD, SSA, VA, Department of State, Bureau of Prisons, there are many, many activities related to health information but focusing on the need of interoperable health information related to many of the areas that are listed on this slide.

Next, please. What's driving the federal agencies? That's the first executive order. That's the creation of the Office of National Coordinator. There was another one in 2006 which was asking the agencies to implement the standard that comes through the AHIC, HITSP and recognition by the Secretary of HHS all the systems and technologies related to that as recognized by the Secretary.

Next, please. So what are the agencies doing today? Agencies have started the process of implementing these standards that come through the HITSP process. Also looking at certified products through CCHIT but also started the process of implementing what we call the trial implementation for 2008 related to the nationwide health information network activities.

How does this affect an agency today? So we talked about all these different technologies, what is happening at the national level and agencies are expected to implement these solutions. But how does that affect an organization as big as the federal government or as big as any of the agencies enterprise as such.

Then there is a requirement that comes to them. They start defining the needs. How does that matter to their organization, to their systems, architect-specific solutions. They also have to plan investment. As you know the federal government is civilian agencies are two years we have to start planning two years in advance and agencies like DOD five years in advance. That means how do we know something that becomes a use case, it's like what you're talking about in personalized healthcare, the use case that is being discussed today is required by the agencies to start implementing in 2010.

The budgeting cycle for 2010 is happening today. So agencies need to have that investment planning happening now. Once that is done, they go into implementing these solutions and also measuring progress. They need to report the progress of these implementations as per the Executive Order 13410. That's the Office of Management and Budget.

How are we doing that in the federal government? Federal health architecture, as I mentioned earlier, is an E government initiative. But it's brought together all the agencies in the federal government that have healthrelated activities. That moves across HHS, not working at HHS, CIO level but also reaching out to the CMS's, FDA and CDC and NIH and NCI, and any organization that has healthrelated activity within the federal government.

So how is making that happen for the agencies? As I mentioned earlier there are steps that agencies go through today to implement any solution. So health IT is one of those activities in the federal government today that the agencies are moving forward. Federal health architecture, the agencies have come together and they're defining specific tools and services to implement these solutions.

That includes educating the federal force. Providing investment guidance. So what does it mean for an agency at the capital planning level if they have to implement a standard related to personalized healthcare? And there needs to be some guidance given to these agencies and the investment guide is providing that kind of detail.

We have interoperability architecture. This is more of the techie talk where an agency, even if you take an agency like Department of Defense, what we're talking in health information technology are the changes are only part of their overall arching architecture, and we should provide, this group provide clear guidance to agencies on how these changes are going to affect the existing organizational architecture.

Standard adoption as you know, we bring the federal needs to the table to HITSP, to AHIC, to the different organizations toward the process to make sure the federal voice is coordinated to the extent possible and also advance some of the federal needs related to standards.

Everything, all the things that we talked about is going to lead us into what we are trying to achieve, which is secure exchange of interoperable health information, and there is an initiative within the federal sector called NHIN Connect that we are going to talk a little bit more in the coming slide.

So as many of you know, the Nationwide Health Information Network is a set of services that facilitate the exchange, secure exchange of information that follows the consumer supporting clinical decision and also supporting the public health needs.

As part of this process, in 2007, when the Office of the National Coordinator awarded nine contracts as part of the trial implementation of the Nationwide Health Information Network, the federal agencies came together and said we have to be part of this process and they founded what they call the Nationwide Health Information Network Connect Consortium, a group of federal agencies who agreed on a specific strategy to move things forward related to the exchange of information between federal agencies and also with the private sector.

Part of that activity or the charter stated that, like we called the six principles, of which like there are two key things which I want to emphasize here. One is reuse. That means we are not going to go build everything from scratch. We already have initiatives in the federal government like the DOD VA exchange, the National Cancer Institute is exchanging information. CDC is exchanging information through [inaudible], the Department of Homeland Security through the NBIS project is exchanging information. So the expertise is already there in the federal government. How do you reuse as many components as possible to achieve the broader goal?

So as part of the process, in 2008 federal agencies, there are a few agencies who have agreed to exchange information with the private sector and among themselves for the trial implementation in September and in December.

Okay. So I did go through some of the high level details of what FHA is doing but why is it important for the federal sector? As you know, the Wounded Warrior is a topic of discussion nowadays, and these are some of the three business areas that FHA will be addressing related to the federal agencies in 2008. Wounded Warrior, health information exchange needs, disability benefits, as I mentioned earlier takes a year and a half to process some of these claims today.

How do we speed up that process using HIE technologies and part of being part of Nationwide Health Information Network and also some of the emergency response needs related to the disaster victims. Obviously, IT is not a total solution here, but it's an important tool that can be used by the private and the public sector to advance some of the citizen needs. There's a key message here that the patient is waiting every day we spend trying to advance some of these needs. It's important to keep in mind that there are, these are some of the federal aspects but also as citizens these are things that we all need to advance together forward.

That's pretty much my slide presentation. But I'd also like to talk a little bit about what's happening related to the family history work that FHA will be helping Greg and his team advance forward. As you all know, there's a meeting scheduled for the 9th. We're having discussions with DOD and VA and IHS to move some of the specific needs of these agencies and how the portal or the existing portal that we have, how that can be integrated into these agencies EHR or PHR. Those discussions will have a clear direction as of the 9th. And the goal is, if possible, to have some demonstration by end of this year. By the end of this year.

And thank you very much.

>> Doug Henley:

Thank you, Vish, very much. Any comments or questions from the workgroup?

(No response.)

Great, very good. Well, let's go on then to Dr.Terry Cullen, who is going to give us an update on what the Indian Health Service is doing relative to family health history as well.

>> Theresa Cullen:

I want to also introduce Chris Lamer, who’s a pharmacist and medical infomaticist in the room with me. So next slide.

For those of you not familiar with Indian Health Service, we're an op div within the Department of Health and Human Services. Our budget comes from the Department of Interior. We are legislatively authorized to provide a comprehensive health service delivery system for approximately 1.9 million federally recognized American Indians and Alaska Natives. We provide services in 36 states.

Next slide. Per our model our mission and goal is really to raise and you'll see these attributes of health, the physical, mental, social and spiritual health. We attempt to do that in a communityoriented primary care model which is in fact why we were so interested in this family history work.

The vast majority of our patients receive birthtodeath care through the Indian Health Service with only approximately 20% at any time moving outside the system. They'll be referred outside the system for a larger number of their healthcare needs, but they actually don't move outside the system.

Next slide. This slide gives you a sense of where the American Indian Alaska Native population is in the states. You'll see we're predominantly in the southwest. Though a large percentage of our patients do also live in Alaska. Our programs are run both directly by the Indian Health Service, and then they are compacted or contracted to what we call Tribally Operated Health Programs. You can see the distribution of the programs here.

In addition, we operate 34 urban Indian health programs in, in fact, cities where there's a larger percentage of American Indians and Alaska Natives.

Next slide. Some of our remote areas. This is obviously in Alaska.

Next slide. And this is another one from the southwest.

Next slide, please. Now, our PMS electronic health record is over 40 years old. It's a sister or sibling to Vista from the VA. In fact we leverage highly off of the VA. Their gold applications, pharmacy, labs, we use extensively. We obviously extend what Vista provides into the areas of obviously children, women's health, obstetrics, care where the VA will obviously not ever be responsible for providing care.

Right now our PMS is, application is used at over 300 sites, both tribal and those operated directly by the federal government.

Next slide. Can you guys advance the slides?

>> Judy Sparrow:

We're working on it. Just keep talking. It will catch up.

>> Theresa Cullen:

I'm sorry for that. Let me talk to you. This next slide shows us where our database lives, which is in the patient care component. It includes some of the 60 applications.

The critical part of this is to recognize that we do have a meta data registry and a data model that's comparable to the VA. Though on the backside we store data differently. We have a V file, so we have a file structure from an enterprise architecture that's different from the VA. And it's really being geared to exclusively ensure that we're able to meet the needs that are birth to death needs. I'm going to keep going and assume you have the PowerPoint.

The next slide is a copy of our electronic health record. It's a screen shot of what our EHR is able to do from our web itself.

The next slide that we have in this group shows the family history. Traditionally, we have recorded family history for the last 20 to 30 years.

We've recorded it in a fairly nonstandard way. And I'm going to show you how that's been. It's been recorded in the problem list. So even though we have a V file, a visit file that's family history, from a provider perspective, all you ever saw was if it had been added to the problem list.

So maternal, family history of breast cancer.

By extending, by working with the family history subgroup, we're now in the process of extending how we're recording that data. In fact, we're electing to not record it in the family, in the problem list itself.

Next slide. What you'll see on the next slide is the documentation of the family history. This is a screen capture from our graphical user interface. We, just like the VA, record in a standardized fashion, either in ICD, CBT code or whatever is in there. You'll see that as ICD. That allows us to query. But in addition similar comparable to the VA you can also write your own narrative.

So when the system goes in and queries and presents the data, you can ask either their representative data to be presented to you either as the ICD or V code I'm going to show you on the next slide, or the narrative that's been written by the physician.

In addition, you can see whether personal history, family history and obviously the active and inactive problems that everybody else has.

The next slide shows how you can use V codes to document. For those of you familiar with V codes, there are V codes for some family history. They're not they obviously don't include every medical condition that you might want to record. They in fact do include the short list that is part of the Surgeon General's work in terms of family history and the diseases that have been focused there.

The next slide just shows you ICare, which is a screen shot from our integrated case management application. The reason we're showing you this is you're going to see later in about two to three minutes that it's a place where we'll also be displaying this data.

The next slide really talks about how we combined standardization and our PMS. As you know, and as Vish alluded to the Indian Health Services, is now functioning to a large extent under the auspices of the FHA and the Office of the National Coordinator. In terms of the master you have to follow which is the standardization master.

So what we're attempting to do is anything we move forward we move forward in the HITSP or HL7 or whatever can be there.

So if you go to the next slide, familyhistory.hhs.gov. You're familiar with this screen shot because it's been presented, I'm sure.

But what you'll see on the next slide, this was added to the IHS home website. Now the caveat to all this is we recently completed a review of patientbased health access patientbased access to the Internet. Only 22 percent of our patients throughout the country are reporting Internetbased access. Clearly we're seeing higher levels of penetration in more urban areas, but in rural areas we're in fact seeing rates in the single digits.

So we are embracing the ability to do webbased capture of family history as part of a continuum of appropriate technology for what's most appropriate for our patients.

If you go to the next slide, you just see this is a screen scrape of the family health history multistakeholder workgroup. This was alluded to. We've done a crosswalk and gap analysis looking at what are the fields that were recommended by this workgroup and what we need to then, how we need to change our meta data registry to incorporate the fields that we traditionally up till now had either not captured in the standard fashion or captured or not captured at all.

The next slide really talks about what I alluded to before that this new field for family history, what happens right now for our patients is that we have a patientbased health summary similar to the VA health summary. And up till now the family history data was hodge-podged into the problem list.

It's now pulled out as a family history subsection, just like we have a surgical history subsection. In addition to that, we believe that we're on the way of incorporating that family history into what we call a wellness summary which is a patient health summary that's given to the patient at the end of their visit that incorporates information that we believe is important for them in terms of improving their wellness and their commitment to adhering to certain standards.

The next slide shows you that this incorporation and really our approach to this is to try to incorporate what's in the requirements summary document that was alluded to as well as the Surgeon General's portal. And you can see here birth status relationship, diagnosis.

These are things we are in the process of either have right now in our PMS or will be included. The future ideas relative ethnicity, percentage of ethnicity, we in fact interestingly enough capture that right now for American Indian Alaska Native enrollment, because it federally recognized tribes have what they call certificate of Indian blood that gives you the specific percentage, 25 percent of one tribe, 25 percent of another tribe.

So we already capture that, but we capture that only for American Indian and Alaska Native identification.

Our plan is to extend that to get that to relate to other ethnic data, though we're struggling with exactly how to deal with that entity and lab results.

Obviously our system includes lab results. But we don't have markers right now other than the very common ones like [indiscernible]. We're in the process of trying to figure out how to include that.

Next slide then shows you a sense of what we want to do with this integration. So our goal is to integrate the electronic health record which we just focused on ICare which is a subset of data in the electronic health record that's used to track population health both from a quality perspective and from a community perspective and the personal health record and make sure we can feed these together.

And somebody talked earlier about that we are trying we've gotten the full code for My Personal Health History, and we're trying to figure out how we're going to use that within our PMS.

Our problem relates to the Internet usage. So what we I think what we've come to conclude is that for us probably the first step is that it's going to be at the time of the patient visit to give the patient access to fill in that part of the health record and then data dump into our PMS in a standard fashion.

In the meantime, because we don't have pedigree availability, which is a huge part of that application, you can see on the proposed display, this is just a display panel from the electronic health record. So until we can leverage the pedigree display that was done in that application from the Surgeon General's office, we're probably going to elect to do just a graphic display like this so that we don't misuse work that needs to happen.

And here you'll see relationship birth status. These are all in standard formats that you don't see at this end. But obviously in the back end, single has a code set and it all ends up in code set. The same for status, because there's a limited number of options you can choose as the provider.

If we have the ICD, you'll see that it's listed there and the age of death.

And that's about it for us. Our goal is really to increase awareness in our patient population to increase access to personal health tools and then to ensure there's integration into our continuum of appropriate technology tools for the electronic health record.

>> Doug Henley:

Terry, this is Doug Henley. Thank you very much for an excellent presentation. I was impressed with your initial solution for recording pertinent family history as part of the problem list.

Not necessarily the best way to do it but certainly a very functional way to do it in the absence of a separate section. But glad to see that you're moving to a separate section at the present time. That should work out quite well and certainly brings up the need for some standardized tools in terms of methods to collect this type of information, which obviously we're very interested in trying to help with.

Any other questions to Dr.Cullen or her colleagues or questions that the workgroup may have?

>> Marc Williams:

This is Marc Williams from Intermountain. I appreciate the comments regarding the use of problem lists and the idea it's probably best to get away from that, but I thought it would be important to share our experience relating to this, because we found that even though we had a family history web form that a lot of our providers were in fact intentionally entering family history information into the problem list.

When we inquired about this, it was because of the way the information actually displayed on the patient worksheet, that if they entered it into the web form then they would have to click out of the worksheet at the encounter to go to that information and navigate through the web form.

If they attempt to import the data from the web form into the worksheet, it was not presented in a very functional way. And the advantage for them at least having it in the problem list was that it would automatically populate their worksheet so they could incorporate that information into every visit.

So I think the solution obviously isn't too use the problem list but to be cognizant of how the end user wants to have that information presented. Related to that, none of our users have any interest at all in seeing pedigrees.

They are very used to seeing this information presented in the tabular format. While I think it will be important to build in functionality to display a pedigree, substantial group of primary care physicians at our place say we're just not going to look at it in that format, we'd much rather have it presented in tabular format.

>> Doug Henley:

I think those are great comments, Marc, thanks. Other comments or questions for the workgroup?

>> Becky Fisher:

This is Becky Fisher. I have a comment for Dr.Cullen. Pertaining to the limited Internet access of her patient base. Was I understanding right when you said you were thinking about giving the patient access at the point of care so that they could use the Internet to build their family history?

>> Theresa Cullen:

What we have available right now is we have a pilot demonstration project. It's probably beyond a pilot project it's been a couple of years where we have kiosks with Internet access available in some of our larger facilities.

And the splash screen that can default to that is IHS website which is where my personal health record now is linked to. So some of the initial internal discussion has been to let people do that.

I don't know how familiar you are with that but you can print out your own pedigree. The problem is you can't stash it. You can't store it. So for us at least we don't store anything on the kiosk because it's a public site. But we are obviously we're really struggling with Internet access for our patients and how to move along.

We in fact I, in fact, did not believe the digital divide was going to be as great.

>> Becky Fisher:

I'm sure you're finding it to be quite deep. I just wanted to pass along that I noticed an article in the most recent bulletin of the Medical Library Association where they used this kind of venue to capture a request from the patient and open up a feedback loop that would allow someone to contact that patient.

And I thought that was worth mentioning.

>> Theresa Cullen:

Thank you.

>> Doug Henley:

Other comments or questions from the workgroup to Dr.Cullen? Okay. Very good. Well, thank you, Terry, very much for an excellent report.

Let's go ahead then, I'll turn it next to my colleague John Glaser to take us through the next agenda item.

>> John Glaser:

Thank you, Doug. Which you should have received, is version 3.2 of the draft recommendations regarding clinical decision support.

As you may recall, we have seen or you all have seen an earlier version of this several weeks ago. And during the time between now and then that draft had been through a wide variety of the working groups. This may be one of the first times that a particular set of recommendations have traversed all of the AHIC work groups.

You'll see the results in front of you of the feedback we received from them and also some thoughts along the way from the ad hoc group that's been working on this.

Several of you on the phone are also members of the CDS ad hoc group.

I will touch briefly on the document. There are no slides to go with this, and point out probably the most significant change that's on here.

If you have the document in front of you, just to remind us collectively on the bottom of page 1, the goals for clinical decision support and the recognition that there are a variety of environmental factors that would influence how quickly and effectively CDS has adopted ranging from priorities of care for the country sets as a whole to incentive mechanisms that reward providers for delivering high quality and safe care.

If we move to the second page, to sort of point out we won't necessarily return to it that within the federal government a CDS collabratory is being organized, it will bring together the CDS efforts of a range of agencies within the federal government to facilitate sharing of information, identification of challenges and the development of approaches to dealing with a number of those challenges.

From that we go into sort of the three core recommendations that are part of this document.

The first, which is on the bottom of page 2, points out the fact that and it traverses into the top of page 3 where you can see recommendation 1.1 and 1.2, is to essentially to the national quality forum does some work to identify key health priorities for the country or conditions and also zeros in on a consensus set of metrics by which we can assess our progress in addressing those healthcare conditions, is that we're recommending the CDS collabratory. Again, the government agency or the group working with government activities should develop an approach to determining to what degree CDS can contribute to advancing those, our progress in those high priority areas, along with a way of providing a scorecard or a set of metrics by which we could evaluate to what degree CDS is having an effect here.

Recommendation 1.2 sort of extends that to a degree, and it says that, again, that the collabratory will work to coordinate a variety of efforts within the federal government to advance progress made in addressing these targets of national healthcare conditions.

So 1.1 and 1.2 are in the context of both a road map or game plan for lack of a better word for the country and is making sure that efforts within the federal government are aligned and supporting our collective progress in removing both the cost and improving the quality of care delivered in those particular priority conditions.

Number two is perhaps where most of the change is made, although the change is more organizational and structural. Before you'd seen a variety of recommendations in areas such as a national repository for the collection of decision support rules and logic and approaches or recommendations regarding vetting that logic to make sure that it had good defense behind it.

Recommendations regarding looking at particularly the ability to apply clinical decision support in ambulatory settings. Also recommendations regarding incorporation of the electronic health record. It was recognized that a lot of those recommendations still need some further discussion and elaboration and that is being proposed here is the creation is something called the CDS alliance, which would be a public private partnership that might be initially funded under contract to the ONC and would include people from the private sector but also people from the federal government that would work together to take the sets of recommendations that you saw in an earlier version of this document and to advance those.

So here the recommendation set is smaller. It's essentially 2.1, which we are recommending we create this CDS alliance. You can see at the bottom of page 3 a list of, for example, suggested representation from some of the activities that have come out of the Office of National Coordinator such as HITSP and CCHIT along with recommendations from a number of agencies represented by folks on the phone.

Then turning the page, onto page 4, recommendation 2.2, is in effect suggesting an agenda for that alliance. So that the alliance work to advance all of the bullets that are listed here and realizing the alliance per se does not have authority or power but can act as a vehicle by which a wide variety of activities in the private sector and in the public sector can be coordinated in advance and common challenges addressed and recommendations might come out of there that are directed back to the federal government and are directed also to the private sector.

So this is perhaps a big core thing which is to create this alliance as a private public partnership and to suggest an agenda for that alliance and its members to carry forward.

Section three, which is on the bottom of page 4 and continues into page 5, has four recommendations, which are targeted to specific federal agencies. You can see some recommendations to AHRQ regarding research further and leveraging CDS.

Recommendation 3.3 is a recommendation to HITSP regarding a personal attribute minimum data set and then 3.4 is to suggest to both CMS and AHRQ that such the demonstrations that are occurring that those demonstrations explicitly consider and incorporate CDS activities into their efforts to improve care through whatever demonstrations are being set up here.

You'll notice a lot of blank insert dates. The recommendations 3.1 through 3.4 are currently being vetted and discussed with the agencies that are listed here to make sure that they're comfortable with them and to the degree we're going to provide a date to make sure we have their thoughts when a realistic timeframe for those particular dates is.

So this is what has been done. And what follows in page 6 through the end of the document are the specific recommendations that you all have seen before, and again these are provided as background to the CDS alliance which we would hope would incorporate that into its agenda.

So we are almost at the tail end in the ad hoc working group of vetting this with the various workgroups. We have an ad hoc CDS workgroup meeting next week to take in your feedback today and the feedback of all the workgroups and incorporate that and any other changes they feel are appropriate into this particular document.

We're still targeting to the degree this is all coming together well a presentation of this at the April AHIC meeting. And assuming that it is looked upon favorably there, we will engage in the conversations with both the ad hoc workgroup and also the ONC folks about examining the possibility of putting the alliance of causing it to be or to set it up sometime this fall and that would proceed in parallel with the efforts going on within the Health and Human Services and the federal government broadly to establish the CDS collabratory.

So again this is the second time you've seen this. Although some changes have obviously gone on since we chatted last. Why don't I just see if there are any comments or questions or concerns regarding this.

>> Clem McDonald:

This is Clem McDonald. Can I comment on two parts?

>> John Glaser:

Of course.

>> Clem McDonald:

First, I'm worried the discussion of the decision support so focuses on the logic for making the decision, and doesn't whisper about the real problem, getting the data in so that a computer can do anything and in an accurate form.

I mean the simplest example is hemoglobin A1C where the rule is trivial and it comes from the lab, comes from the doctor's office, the office practice machines don't have interfaces and it goes on and on.

If we don't at least pay attention to that we'll build these beautiful machines that won't have anything to do because they won't get any data to feed them.

And then the second one related to the creating the new organization, I just see so much overlap between what it has to do, and although there's a much, maybe a stronger computer nerd side to the alliance, then what the quality, the HQA has to do, they have to pick out the rules, they have to figure out data that can make them work. And whether it couldn't be conceivable to make them to want to be a subgroup or somehow connected. So they're kind of bound in reality.

>> John Glaser:

Yeah, fair points, Clem. I know certainly your experience and the experience of the other partners is more than the rules. That is getting the data in doing it in a work flow that makes sense. It's avoiding fatigue on the part of the doctor, there’s a wide variety of implementation stuff which is nontrivial.

So I think it is recognized by the workgroup and some of the material in the back refers to that. You're right. If you look purely at the way the sequence that it's presented on page 4 it starts heavy on thisrepository and making sure that it's of good quality rules, et cetera.

And so it sort of leads in with a rulecentric approach here and I think it's a fair point that you're raising. And whether we re-sequence the recommendations or the prose emphasize you understand and so do we and so do a lot of other people about you can have the fanciest and best repository in the sky, but if the data is either they're not going to put it in or not put it in in a structured way or happens in the wrong way in the work flow we would have done all this for naught.

And this isn't the first time for the comment. I think we had to think through this when the ad hoc workgroup gets together, why have a separate thing. There's enough meetings and enough dissipation of energy by a whole lot of folks here, can you fuse this into something else?

Which might leverage the talents of people being provided here to defray some of the costs of doing this, et cetera. I think that's a fair point. I think one of the things we'll have to sort of sort through is there's a number of candidates for that. I'm not quite sure where the right ones are.

And one thing you could imagine, though, rather than force that issue early is you spend some period of time perhaps the duration of the calendar year thinking that through.

>> Clem McDonald:

The AQA really has the viewpoint after the fact sort of what do we have, what can you figure out? And the CDS has the viewpoint of, well, the provider's there and they have to see what do we have we can deal with in real time.

But the rules should kind of the good rules all are the same ones. I mean the good content that they have to work rules out of are the same thing. At least think about that.

>> John Glaser:

Fair point. Other comments, suggestions, critiques?

>> Grant Wood:

John, this is Grant Wood with Intermountain Healthcare. During the CDS ad hoc telephone call, I brought up the Morningside initiative. I wonder if you could comment on that with this group also?

>> John Glaser:

Yeah, you may some of you may be aware that there is an initiative I think being overseen and at point Patrick is running although AMIA, American Medical Infomatics Association, is quite close to that, of looking at the creation of a repository and this is sort of the first bullet under 2.2. That initiative is proceeding and there's some analogous work under the AHRQ grant that's going on here.

It's less clear to me to what degree this is going on in the private sector. So I think Grant what we would recognize here is that that is underway and should continue and what we would propose to do in this alliance or whatever form that it takes is to make sure that there is coordination and discussion of that initiative along with anything else that is comparable within the federal government and anything else going on in the private sector.

For example, it's not clear to me when I talked to my CIO colleagues that they have a clue that the Morningside initiative is trying to do or what it is all about. So I think one of the things we can do is to sort of ensure that that initiative is more broadly understood in the industry than it might currently be. And that it is the dialogue as appropriate between, within and between the federal government and the private sector is going on as it ought to.

I don't know if that answered your question or not.

>> Grant Wood:

Yes, it does.

>> Greg Feero:

This is Greg Feero. One quick question. Who besides Greg Downing, will there be anybody from the CDS group at the April9th meeting regarding family history? Because I think consistently people have been commenting that not only do they want their family history they want some interpretation of their family history and systems.

This may be an area where proof of principle could be demonstrated for some of the

>> John Glaser:

That’s a good idea. I don't know, Greg, if you're on the phone. I know he had to duck out for a bit of time during this call. If he's still ducked out. I'll follow up with him and make sure we link that up. That's a good idea.

Any comments, questions? Well, I appreciate your willingness to see this twice. And we will keep you posted as this moves forward. And again for those of you who are both a member of this workgroup but also the clinical decision support workgroup, I appreciate your efforts on both fronts.

So without further ado, I believe next we have some I know review of some work and analysis done by the RAND Corporation looking at the genetic and genomic information in the EHR.

Maren, are you ready to go?

>> Maren Scheuner:

Yes.

>> Alison Gary:

Just before you begin, I wanted to remind everybody that if your slides are running a bit slow, if you press F5 on your screen it will refresh and should speed up your slides again. Continue. Sorry.

>> John Glaser:

Maren, whenever you're ready.

>> Maren Scheuner:

Okay. I'm ready.

>> John Glaser:

Now what do I do? (Laughter)

>> Maren Scheuner:

I think I spoke to this group about a year ago when I was just initiating this project, interviewing a bunch of folks about genetics content and electronic health records. So it's nice to be back to review the followup on what we've learned.

Next slide, please. The next one. So our aim's to gain an in depth understanding of state of the art in EHR products regarding collection and organization of genetic genomic information, related clinical decision support and stakeholder needs and interests.

Next. So this is a descriptive study. We have semi structured interviews of four stakeholder groups. Primary care physicians. These were general internists and family practice physicians, medical geneticists, genetic counselors and electronic health records representatives.

The interviewing occurred between March and December of last year. And we found these folks through different listservs of professional organizations. And we excluded clinicians if they were not from one of the stakeholder groups. There was one of those.

Or if they didn't use an electronic health record. And the electronic health record people, representatives that we spoke with, we excluded those who did not have an ambulatory or inpatient EHR product. And there were two of them.

The interviews lasted about 60 minutes. They were recorded in full and then fully transcribed. And then codes were developed by two of the investigators on the project that represented key themes in response to each question asked.

And then those codes were subsequently applied to each transcript tied to additional individuals on the process.

Next slide, please. So the people we spoke to, we talked to 10 EHR providers. Actually, there were 13 individuals representing 10 different products. Eight were commercial vendors and two were from integrated health systems.

We talked to 18 primary care physicians. 16 medical geneticists and 12 genetic counselors.

Next. We asked questions in three main areas relating to family history content in the EHR, genetic test ordering and results with the EHR and then what I call the issue questions or broader overarching type questions. I'll go over each of these on the following slides.

The family history questions were the following: Does your electronic medical record include an electronic health record include family history information, how is family history collected, how is it entered? How is it organized? Is there clinical decision support? And is security and access permission different for family history.

Next. Okay. So interestingly, two of the medical geneticists and one of the counselors said that EHR did not include family history. So these folks, the remainder of, obviously we didn't continue with that line of questioning for these individuals.

One of the medical geneticists said there's no linkage to any system that collects any meaningful family history.

So I probably could guess that this person did have some type of family history capture in their EHR but it wasn't, it just didn't qualify in their opinion.

Genetic counselor said we are still taking a paper pedigree, and then that is scanned into the EMR. So that was her description of and basically the reason why she said that the family history was not included.

And then the last geneticist here says the EMR does not collect family history.

So all of the clinicians said family history could be collected in a non-structured format, as did eight of the 10 EHR representatives. Primary care physician says there's a box that you can type in if the patient has seen genetics. They have a very good text entry of their family history and then it says see pedigree. And it wasn't quite clear where that pedigree often was.

But unless they have seen a genetics it's pretty limited as far as information in that box.

One of the vendors said to capture everything in the structured way is too painful. And the counselor said supposedly the family practice doctor is supposed to populate that but it doesn't always happen.

And this came up kind of frequently in speaking with the genetic counselors. They kind of felt that that was the domain of the primary care physician and they weren't sure if they could actually go into that family history text box, or whatever it was, and enter family history information.

Next slide, please. 35 percent of the respondents said that family history was also collected in a structured format. So just under half there.

A medical geneticist said that practitioners have the option and continue to exercise it to put most of their family history in our EMR in the form of text within comprehensive clinical notes.

And so this was something that I heard pretty often was this idea that even though their electronic health record was able to capture both structured and unstructured text format, that most physicians seem to prefer the text format.

Another geneticist said family history is captured both in free text which is by far and away the majority of that information. But it is also beginning to be captured in a couple of areas with coded information.

When it came to who provided or who entered the information, almost everyone said it was the physician. A few, 11 percent of the respondents said patients provided the family history electronically. These were folks who also had information and asked if they had a personal health record that was available to their patients and this is a subset of those that said that patients could provide family history in that way.

However, one of the counselors who had this available said you couldn't really use the data there and just dump it into the EMR, because you wouldn't know where to put it. So there wasn't an easy way to represent that information in the EHR even if the personal health record did capture it.

Next slide, please. Organization of the family history. So about half of the respondents said that the EHR had a family history tab. So very specific place where you would go and look or enter the information. And then the other half said there was no special section whatsoever. And that it was basically documented within the clinical note on that day's encounter.

And in probing around these questions, the next the primary care physician said it's not easy to find it doesn't pop out as one of the major things. And this was pretty commonly stated among the respondents. When it came to pedigree drawing, everyone I talked to said it wasn't possible to do that within the EHR.

Next slide. With respect to clinical decision support for family history, all but one respondent who was an electronic health record representative said that they did not have clinical decision support that could assess familial risk or make recommendations based on family history.

The EHR representative who said they did described it as within our template system, if you say premature history of whatever condition, there are clinical reminders that will ask you to maybe check something else. So but this would be something that the user would have to kind of set up front and integrate into their system. It was a customized approach but it was definitely possible.

Five percent said family history populated the problem list. So that and often and more often physicians would indicate that when they would find a family history they would say family history of colon cancer in the problem list. But these folks said it automatically would go there and populate it once you entered it.

Next. Security and access permission for family history. There were some interesting answers here. So 11 percent said family history was treated differently in their electronic health record, different from other medical information. And the way they viewed this question differs, which you'll see.

So geneticists said that pedigree remains in the shadow charts. The counselors would often say we're getting information on Uncle Harry, and they're not convinced privacy will be preserved in the EMR. These are clearly concerns about privacy of the information.

A counselor said: We put a lock on it and that lock makes it so it is only seen by physicians in our confidentiality group. Our concern is because the pedigree may contain detailed information about people who are not our patients.

Another geneticist said: If you add a patient portal that's when I start wondering how secure it's going to be. I have no trouble with the patient wanting the medical records they're sitting in front of me, I just have them sign a consent and I print off whatever they want from it. But I wonder actually having access to it of who can enter and who can't. I don't know enough about it to know if it can be secure enough, and I think that's why we haven't added it to the patient portal yet.

So, again, just some concerns about privacy, especially with the first respondents about having information about someone else other than the patient of record in the record.

>> :

But all those concerns came from genetic specialists, none of those came from

>> Maren Scheuner:

No, there were two, I didn't give all the responses. There were one or two primary care physicians who also mentioned it as well.

>> :

But a minority then?

>> Maren Scheuner:

Correct. And, actually, as we'll see around this question of challenges and barriers, I'll just say now it's really the genetic counselors who see it as a greater concern than the medical geneticist. It's interesting because I was going to lump the medical geneticist and genetic counselors as one stakeholder group. But when I looked at their responses, it was clear that it didn't make sense.

So genetic test question: Does your EMR include genetic tests, the results, any way of special handling, is there clinical decision support and again the security and access question.

Next slide, please. Okay. So let's see here. Only two or four percent of the 58 respondents or whatever it was, said that all the genetic tests that they needed to order were in the test order menu of their electronic health record. So a very small minority here.

14 percent said some genetic tests were in the test order menu. 20 percent said, and these are not mutually exclusive responses, I need to say, that some or all genetic tests are ordered within the EHR as miscellaneous tests. That are typed in to a text field or written on paper and scanned. So clearly that would make things difficult for some kind of clinical decision support. And then 25 percent said genetic tests could not be ordered at all.

Next slide. With respect to clinical decision support, it was really only one geneticist and one counselor said they had it. And the following is what they said. The geneticist said yes, it is limited. I don't know if they're planning to get the genetics really integrated into the whole lab system. For one patient I did see GH. It's saying here microarray GCH sent to an academic outside lab and results received by the genetics clinic.

So I don't know most of you may not think of this as clinical decision support. It's really how the sample was tracked within the system and it even went outside the system.

The counselor, I think, is closer to clinical decision support. She said: That is where the OB/GYN alert thing comes in. I don't know how much they're doing it for all genetic test results but I know for some things alerts get put out. So there must have been some and I don't know, she couldn't recollect exactly what it was but she thought there was something there around there around an OB/GYN-related test.

Next question. I’m sorry, next slide. The security and access commission. Okay. So here, 14 percent of respondents said genetic test results were treated differently from other lab results in the EHR. One of the primary care physicians said: Genetic tests we do are sent out. I forget now even where they were going to be within the EHR, under miscellaneous reports I think and scanned in or something like that.

So I think this response was more about the organization of where the information would go and that's why it was different. It wasn't really and accessing it was different because of the organization of the information. It wasn't so much about privacy concerns.

A geneticist said yes because most of the reports are maintained at the genetic center and not entered electronically. So again, kind of a process sort of thing. A geneticist saying we still keep a separate chart for BRC mutation carriers. We still have patients who don't want their physicians to know their mutation status and they don't want their insurance company to know.

So here clearly, there are some concerns around privacy and the genetic shadow chart. Genetic counselor from Oregon said that Oregon Genetic Privacy Act says we're supposed to have a patient sign a specific release of information saying they understand that a genetic test is being performed and we have the ability to disclose that information to the referring provider or her insurance company, and they have to consent specifically for those.

So regarding just with whom to share and so forth, this counselor was aware of the specific state law that she needed to abide by.

Next slide. Few more answers here. A geneticist said we might click on a result for urine amino acids and we might in that case get a broken glass reminder: This is a genetic test. So then an audit trail is set in motion. Where with a cytogenetic test we don't get that broken glass. I haven’t dug deeper to find out what the source of confusion is. Clearly, it's a genetic test. So why are urine amino acids a genetic test in the EMR but not these other test results?

A genetic counselor says we don't allow genetic tests to go into EMR until we've informed the patient. And we've always locked it from the PHR. I have been hearing recently, though, that the people want it in their patient portal view, especially to share with other family members.

And a genetic counselor said: I was told that things like Huntington and BRC test results are different. At this point those are not ones they scan in. When I ask the reference lab for paper copies of those results that are not in the EMR and I ask how would they know not to release that result or not pass them on to other people, they were just: Oh, they just know that. That didn't sound encouraging. So there were policies and protocols in place within this health system around these, specific genetic tests. And this counselor didn't seem too confident that those policies were being followed.

Next slide. So the issues questions that we discussed, the first one is what are the barriers or challenges if any to include family history and genetic test information in electronic health record? We asked if the clinicians have vendors or product managers responded to your genetic genomic medicine needs and what data elements would be most useful.

We asked of the EHR representatives: How has the market reacted to the increasing discussion of genetic genomic medicine and how has this translated to changes in data input or other requirements. What's the impact of EHRs and health IT in genomic medicine and how do you see this changing. And vice versa, what's the impact of genomic medicine on EHRs.

Next slide. So in pooling all of the respondents together, because I'll try and point out to you where I think there were some groups that kind of sit out. So with respect to challenges, over half of the people indicated time and resources to enter family history. And it's interesting, most of the challenges related to family history and not genetic tests there. So time and resources to enter, number one, challenge. Getting complete and accurate family history. Number two, at 39 percent. Clinicians don't understand how to use family history, 30 percent.

Privacy concerns, medical discrimination, 29 percent. And again, I believe the counselors, this was I believe their top concern. So it made number four when I pulled everybody together.

Lack of demand for genetics content. 25 percent. Lack of pedigree drawing. 25 percent. And lack of standards, 23 percent. These were standards around terminology and interoperability and rules for clinical decision support; any kind of standards.

Next. And of the clinicians where I asked if the vendors had met their genetics needs, only two respondents said yes. And both were medical geneticists. One said, yeah, but I think it would be good to have the pedigree drawing in the system. Getting tests into the system has been very easy. We asked can you please give us a separate entry button for the Newman syndrome test instead of putting it under miscellaneous molecular diagnosis. They say sure and they do it.

I have to say this was very different from pretty much everyone else I spoke to, but it was encouraging to hear. Another geneticist said yes they're working on it and the vendor they're working with is definitely interested. So only two of all the respondents said yes. But, anyway, next slide.

So again of the clinicians what elements or requirements will be most useful? So and remembering Marc's comment that at Intermountain pedigree drawing was not really high on the list, I do have this broken out by the type of respondent. But you'll see here the number one thing was pedigree drawing. And I know it was the first choice of the medical geneticists and the genetic counselors in terms of added capability, but it ended up as the secondrated amongst my primary care physician respondents. Their number one was clinical decision support, which is a number two of both the counselor and the geneticists.

So and then that patient portal, was really driven by the geneticists and primary care docs and not so much the counselors, again having standards in place. Family history, better organized and more accessible, genetic tests better organized and more accessible and ability to update family history.

I think there were 20 different I've given you the top seven. There was a range here. But I think that covers the majority.

Next slide, please. The vendors and folks from the health systems that are dealing with EHRs. So have there been demands to change things within the EHR for genetics. And basically demands for genetic content the in the EMR have been minimal to none so seven of the 10 respondents said that. Five of the 10 did say that the vendors are trying to improve capture of family history and genetic tests so it’s kind of they realize this is something that is probably important but at the same time they just want to provide basics to their clients today, just making sure that getting all the basic requirements in there and getting it uptodate, and that the fourth most common response was genetics, is or will be a future direction. So only three.

What is the impact of genetics on electronic health records and HIT? So I'm showing you here the four different stakeholder groups. And you'll see that it looks like the primary care in the medical geneticists thought that there currently is an impact of genetics on the EHR and health IT. And pretty much a good balance said there will be.

And then the responses that the next three basically covered most people saying as genetic knowledge increases it will be a driver of EMR content and capability. It will change the EMR.

So seven of the 10 EHR reps and 20 to 30 percent of the other clinicians. Genetics will improve the [[Inaudible]] management prevention and our ability to personalize healthcare. Six of the 10 EHR reps and about 40 to 50 percent of the clinicians said that.

Then genetics will encourage creative development for privacy security and access permission and again you see the counselors here, 25 percent said that. And there are a few of the others.

Next. And then we kind of flip the question around asking about the impact of the electronic health record and health information technology on genetics. And here we see a few felt that there currently is or there will be genomics can't be done without electronic health records was said by four of the ten electronic health record vendors and about 15 to, with the counselors, up to 33 percent of them.

So it is seemingly important to those. Will improve ability to manage genetic information. So a good percentage of respondents, about 40 to 50 percent. Will enable decision support. Will facilitate genetics research. Will facilitate uptake of genetics in practice. And then will raise ethical concerns about privacy and confidentiality. And here it was the counselors were the only ones. It was about half of them who felt this way.

Next. So what we learned. So from in speaking with these I think it was 58 people we've learned that EHRs allow collection of family history in a non-structured way, which seemed to be preferred by many of the clinicians. About half allow collection of structured family history data.

About half describe the specific area within the EHR for family history. Yet time and resources are the number one challenge to entering family history. And along with this I think a lot of people spoke about the patient portal as a way to potentially remedy this. None of the EHRs had pedigree drawing capabilities.

Genetic tests could be ordered using the EHR by about half. But often this was done as a miscellaneous test with just writing in a text box or scanning in a piece of a paper with the request. Clinical decision support is really not available around family history interpretation or genetic tests results or orders.

For most security and access permission was the same as other medical information. But we saw that 11 to 14 percent, 11 for family history, 14 percent for genetic tests where there were some differences in the way the information was handled.

Privacy concerns do exist, especially among genetic counselors. A few clinicians felt the EMR met their genetic medicine needs. Geneticists perceived this as a low priority, and EHR reps said demand for genetics genomics content was low currently.

I have one more. Clinicians want pedigree drawing and clinical decision support. Genetics will be a driver of EHR content will improve our ability to provide personalized healthcare and EHRs will enable management of genetic data research and clinical decision support. And I think that's my last slide.

Yes, I want to acknowledge my coinvestigators, Honda Reese, Robin Mylie, Benjamin Kim, Sarah Olmstead and Stephanie [inaudible]. Thank you. Any questions?

>> John Glaser:

Maren, that was terrific. Looks like we're in a fairly primitive and idiosyncratic state at this point. Any questions or comments for Maren?

>> Marc Williams:

Yeah, this is Marc. I just wanted to respond to the pedigree issue and say that I don't think our findings are really inconsistent. Our ascertainment of primary care physicians for interviews were those that were actually showing evidence of use and inclusion of family history in electronic health records. So using our available ways to enter family history and our questions were basically: How are you using it? And what things would be helpful to lower the barriers.

And your group, of course, was ascertained from a different method. So I don't think it's surprising that there were some differences around that specific issue.

I will say, though, that our group definitely agreed with your group that clinical decision support, in other words tell me what to do with this information was their number one thing.

>> Maren Scheuner:

Absolutely. Among the primary care that was it. It was almost like the pedigree drawing would be nice. But what primary care needed was what do I do tell me what I do with this family history. Whereas, the genetics folks, I mean that's all we do is draw pedigrees. So it would be nice if something else did it for us. Good, Marc, thanks.

>> Clem McDonald:

Could I contribute or ask a question further along this line? Clem McDowell. The question might be restated: What good is it. That's not meant to be pejorative. I'm here to take care of this patient. I know if I lower cholesterol I save their life. I know if I give them a flu shot I'm going to help them a whole lot.

What do we know in evidence about the use of the family history across I mean in some areas I think we do know breast cancer, et cetera. But if we sorted that out we might be able to make better judgments. So where do we know that as the field knows that it's definitely useful for the patient on that visit? You know, we know it will be as we sort it all out and get better data. But we're still kind of on the way to Nirvana.

>> Maren Scheuner:

Thanks Clem. I could have provided some more detailed responses around that and maybe I will for the paper for sure where it was these were the clinicians who recognized that there are these high risk families that they probably are missing. I mean, I talked to a couple primary care docs who said I've never referred a patient for a genetic test. I've been in practice 20 years.

So while I was speaking with them, I mean they clearly felt that it's very likely that they just aren't identifying these patients in their practice. And so it's probably mostly those high risk things that are more commonly known about today. You're right. Like BRCA 1 and 2 and HMPTC and those types of high risk families.

>> Clem McDowell:

It would be useful somewhere along the lines if somebody would highlight those things that are really clearly we ought to get on the ball and do. And other things that we have to wait until the research sorts out.

>> Marc Williams:

Actually, this is Marc, one of the reasons that the CDC tool picked the disorders that it picked was based on the presence of at least some evidence that we might be able to impact or make changes. And Maren was actually involved in that. I'll let her speak to it in more detail if she wishes to. But I also just wanted to mention that in our interviews of the primary care physicians, there was another way to kind of get at this question.

And that is, clearly these physicians were using the information. And so it's interesting to me that no one has ever published anything about asking primary care physicians how they actually use family history information in patient care.

And so we did that as a qualitative research study and found out some very interesting things that we're in the process of putting together and will be publishing.

So there is, you know, the evidencebased utility that I think you're referencing, but there's also support of the current standard of practice for which there may or may not be high level evidence that may still be of some value to the physician.

So we're trying to look at it from both ways. But I think your point is well taken. We have not been as intentional about collecting the utility evidence as we should. And I think that that's an issue that definitely needs to be addressed.

>> Clem McDowell:

Look forward to your paper.

>> :

Couple of quick questions. The EHR people that you talked to, do they really represent sort of the in the weeds side of the EHR or were they the business case folks? Because it seems like there's a disconnect between the future plans and the activities versus their importance they ascribe to it.

>> Maren Scheuner:

So I spoke of the 13 people representing the 10, I have it here, I shut it off, it was five CMOs. One CMO/CEO. Three or four product managers. And then at the two integrated health systems I talked with kind of IT specialist who was all about kind of developing and receiving requests and so forth.

But I think amongst the commercial vendors, primarily I talked to the chief medical officers who were out there trying to sell their product and dealing with clients regularly.

So I mean in some of them I talked to were these very large I broke it off at the sides, so I think I had under 10,000 users to between 10 and 20,000 users of their product was another couple. And then greater than 20,000 users was two or three.

And the larger vendors had research like whole strategic genetics research people within their organization. But it wasn't what they were selling today. It wasn't what they were getting, people clamoring for genetics content in the electronic health records.

And most of them I asked who their users were and primarily primary care as you would imagine. But then the big firms also had, you know, all the specialty derivations or by anyone.

>> :

I would think as they look at this kind of product, those folks would probably in their minds weighting very heavily the primary care perspective just in terms of the percentage of their likely sales, unless they're catering to the medical genetics side of the world.

>> Maren Scheuner:

Right. I do think though that amongst primary care it's just this low level of awareness of genetics, and it really coming to mind. But if you sit down and talk to somebody for 60 minutes, yeah, then you might get them thinking.

So I think that they just don't think about it. It is a low priority. I mean that repeatedly came up over and over, even by the genetics people, because most of them were in academic centers working within departments where it was the internal medicine and surgery and OB that were driving the development of their EHR at their institution and genetics is last to be implemented. So typical.

>> John Glaser:

Other questions or comments?

>> Douglas Henley:

How many of the interviews were in small and medium sized practices where EHRs have been implemented?

>> Maren Scheuner:

I have that. Let me open… So the size

>> Douglas Henley:

Less than 10 docs.

>> Maren Scheuner:

I interviewed ten solo practitioners who were primary care docs and quite honestly I don't know how many, how large some of the other practice sizes are. I can't tell you off the top of my head but I could probably go back and figure that out.

>> Douglas Henley:

Well, certainly speaking for the American Academy of Family Physicians, and I know the numbers from our colleagues at ACP as well that roughly 80 percent of our members and their members are in practices of less than 10 docs and most are in the range of three to five docs. Only about 18 percent of ours are in solo practice now. Still a good percentage, though.

And while 37 percent of our members have already implemented electronic health records and another 17 percent are doing so as we speak, and that's a high number and well ahead of the general physician curve, we also know that it's just difficult to implement and adopt the darn technology, number one, because often as hard as the vendors may try, it doesn't fit with work flow.

Then once they get it and get it adopted and implemented and life is better and chaos has gone away, they still aren't turning all the switches on. So even if you had the best genetic information embedded in the most functional and work flow focused manner, they still may not be turning the switches on yet because they've just gone through the chaos of implementation. But we're certainly trying to move them in that direction.

>> Maren Scheuner:

I think maybe it’s just a real simple thing, this idea of where this information is. And I found the IHS presentation really interesting where they've now put family history in its own place, is how I understood. And I think that might be one first step for getting this a little further along. But right now I heard over and over that most people really weren't even sure where to go to look for the most recent family history that had been entered by anyone within their EHR. It was frustrating.

>> John Glaser:

I think from our own physicians here at Partners, there's a long way from today to the future that we're talking about, beyond the IT EHR PHRs are designed well to incorporate this data and to do decision support in the right way and capture family history in a way that doesn't take a lot of time. I don't think our docs know what tests to order. I think the reimbursement is a little uneven in terms of whether they'll get paid for them. I'm not sure they always know what to do with the results of all this stuff. I think we've got some ground to cover nationally on this front.

I'm struck by, I suspect, the series of product design decisions that will be made in the years ahead in this room which might be some good ones and some bad ones. If you try to grapple with how best to store, represent, for example if you had to have a browser that listed 50 genetic test results for a patient how in the world do you present that to a doc in a way that's comprehensible.

There's some ground to cover here. It's striking me how early this is.

Other comments and questions?

Maren, this is terrific. I want to thank you for the time for sharing, all the people for taking the time to share this with us. It’s very interesting stuff.

The next agenda item that we have is the next steps.

I thought there was one I'd put out on the table for all of you. I think as we may have mentioned at a prior meeting, we are being asked as we transition from AHIC 1.0 to 2.0 that we take our recommendation activity and get it to a point where we more or less bring them to closure by sometime this summer.

So I think what we may want to do with you all next time we get on the phone is talk about the work that's in flight and what might be necessary to bring it to a reasonable closure or at least temporary closure point sometime this summer, such that as the transition occurs, we can let that transition play through before we pick up potentially the workgroup activities under AHIC 2.0. So just to let you know that will be coming up.

The other question, I don't think we need to answer it today, we'll seed the question with you all before the discussion next time we get together, is assuming we bring a number of initiatives to a good resting point this summer, is whether there would be value and utility that you all would see to continue to have calls throughout the fall and early winter.

And the focus of discussion could be less on developing recommendations since we need to give the AHIC transition some time to play through, and more on examining implementation status. We've talked about, for example, earlier in the call today, some family history work. So that will be, I think, just plant that idea with you all and we'll return to that in a subsequent call.

This way we give you all a break for several months or whether there are in fact areas that ought to continue to be part of a conversation that we collectively have as we collectively pursue our interests in advancing personalized healthcare through the use of EHRs and PHRs all reasonably connected to each other.

I think those are the major next steps we have. Both a project plan, so to speak, but also some agenda discussions for the duration of the second half of this calendar year.

Any other comments or thoughts on next steps?

>> Kristin Brinner:

John, this is Kristin. I just wanted to jump in and talk about some upcoming meetings. The first which has been mentioned several times already on this call, which is the April9th Federal Health Architecture Stakeholder meeting. And we should be able to update you on that meeting actually at the May workgroup meeting. We hopefully will have presentations from some of the stakeholders as a result of that meeting.

In terms of upcoming PHC workgroup meetings itself, I think John you covered it really well. At the next meeting I think we should devote a significant amount of time to planning for the future of this workgroup and potentially developing not necessarily a official road map but something to pass on to AHIC 2.0 to let them know these are the things we feel we've accomplished well. These are areas you might want to look to in the future.

We'll also be having an update from Mike Watson of the American College of Medical Genetics. That's kind of a follow on of Maren's presentation. The AMCG recently had a workshop on genetic test and electronic health records.

I think he's going to have an update on that for us as well. We're also planning to have some testimony in the area of pharmacogenomics, and this is part of our pharmacogenomics subgroup activity that we're currently doing. We plan to have recommendations ready to present to you at the May workgroup meeting so at the April workgroup meeting we're going to have presentations from a couple of different groups. Someone at Marshfield is going to be discussing how they use electronic health records and clinical research and we'll also be hearing about a collaboration between Mayo Clinic and Medco, again using electronic health records and clinical research specifically around Warfarin and anticoagulation clinics. Those are things I have on my radar coming up for the workgroup in the next couple of months.

>> John Glaser:

Terrific, Kristin. Other comments or thoughts?

Fair enough. Judy, do you want to see if we have public comment.

>> Judy Sparrow:

Allison, can you check with the public, please?

>> Alison Gary:

Sure, not a problem. For those on line, you'll see a slide on how to call in to comment or ask a question. If you're already on the phone, just press star 1 on your phone right now to make a comment. Any wrapup comments while we're waiting for the public?

>> John Glaser:

Not that I have. Doug, how about yourself?

>> Douglas Henley:

Only go Tar Heels.

>> John Glaser:

The knife's already in my chest and you're twisting it.

>> Douglas Henley:

Sorry, John, just had to.

>> John Glaser:

I appreciate that. I appreciate that. We'll win out west and we’ll see you in the Final Four.

>> Marc Williams:

I think there's just one other thing that's coming up in the future that might be worth mentioning. And Greg Feero can comment more on this. But this relates to Clem's comments about evidence. A number of us met a couple of weeks ago for a planning meeting for an NIH State of the Science conference on family history.

And that is going to take place next year I believe in August the meeting itself at which point there will be sort of a report that will come out that will talk about evidence gaps and I think that will be highly useful for the work that we're doing. So while it's a long ways off, it is an activity that I think is going to be very important. And there may be some individuals relating to the topics that we're considering for that may be contacted relating to possibly being a content speaker.

And so if Greg wants to add anything, great. Otherwise, I am going for Wisconsin, thanks.

>> John Glaser:

Okay. We can have our own NCAA pool on this call.

>> :

For those of you who believe in long range planning I think the dates of the 24th and 26th of August2009 for the state of the science conference will be held here in Bethesda. It will bring together what we hope to be a comprehensive report on sort of the use of family history as a screening tool in the primary care setting. Conducted by AHRQ with a panel of approximately 21 speakers that will discuss various aspects of the use of family history in care, ranging from its ability to provide accurate risk estimates to its use for clinical decision support to potential ethical, legal and social issues that might be raised by the routine use of family history as a screening tool.

A number of individuals I think in this room are likely to be tapped as potential experts at least in the room and on the phone to speak at this conference and it will open to the public for all to attend.

>> John Glaser:

Terrific, thank you.

>> Alison Gary:

And we don’t have any comments from the public.

>> John Glaser:

Anything else, folks? As always, I want to thank our speakers who we appreciate very much. We all know you've taken the time to be with us. And also as we always do express our appreciation to Greg and Kristin, other members of the staff for the extraordinary work in setting this thing up.

So without further ado, I wish you all the best. Have a great week and I look forward to our next conversation.

>> :

Talk to you in April.

>> :

Thanks a lot.