[This Transcript is Unedited]




November 28, 2007

Hilton Embassy Row Hotel
2015 Massachusetts Avenue, NW
Washington, DC

Proceedings by:
CASET Associates, Ltd.
10201 Lee Highway
Fairfax, Virginia 22030
(703) 352-0091

Table of Contents

  • Update from AHIC QWG on Quality Vision Roadmap


DR. CARR: I would like to call the meeting of the NCVHS Quality Workgroup together. Open it. Let’s go around the room with introductions. I’m Justine Carr, Beth Israel Deaconess Medical Center, member of the committee, and chair of the Quality Workgroup.

MS. FARQUHAR: Mary Beth Farquhar from AHRQ, liaison.

MS. MCCALL: Carol McCall, member of the Workgroup.

DR. W. SCANLON: Bill Scanlon, Health Policy R&D, member of the workgroup.

MS. MURRAY: Michelle Murray from the Office of the National Coordinator, HHS.

MS. ANDERSON: Kristine Martin Anderson, Booz Allen Hamilton, supporting the AHIC Quality Workgroup.

MS. GRANT: Erin Grant, Booz Allen Hamilton, supporting the AHIC Quality Workgroup.

DR. STEINDEL: Steve Steindel, Centers for Disease Control and Prevention, liaison to the Full Committee.

DR. FITZMAURICE: Michael Fitzmaurice, Agency for Healthcare Research and Quality, liaison to the Full Committee, staff to the Quality Workgroup and staff to the Standards and Security Subcommittee.

DR. LAND: Garland Land, with NAPHSIS, member of the Committee.

MS. JACKSON: Debbie Jackson, National Center for Health Statistics, CDC, committee staff.

DR. STEINWALD: Don Steinwald, Johns Hopkins University, member of the Quality Workgroup.

(Introductions off table.)

DR. CARR: Excellent. Thank you all for being here and for accommodating the changes in schedule. There are a couple of things that I want to mention, and then I’d like to move directly into AHIC work.

Number one is, as you know, we completed the letter on the June 19th hearing on quality reporting in the hybrid world of claims and electronic health record. The letter was approved by the Committee and the Executive Workgroup. We wanted to have a coordinated voice with AHIC Quality Workgroup and NCVHS Quality Workgroup. We have a little asynchrony in terms of moving things forward to the Secretary, so we’re still in a little bit in a limbo of when that’s going to officially go forward. I think we are going to follow up with Karen on that because we want to try to at least get that officially represented either through NCVHS or AHIC so that is doesn’t become old before it’s published. So, more to come on that.

The second is along those lines, we also have some other work that this group has done that did not get official publication status. I’m referring to the summary that we did in February of ’06. Does everybody have it — widening the vision for quality?

We don’t need to spend time on this, but it’s an interesting interim report. This came out after the 2004 – 23 candidate recommendations, moving quality reporting ahead. This report from February ’06 summarized hearings that we had had the year before about the current state. It’s very interesting because it predates the AHIC Quality Workgroup, and I think it tees up exactly what the issues were that we were hearing. So, this is a nice piece of work, but again, didn’t get official status.

Do you also have the update, yes, on the candidate recommendations? Again, we don’t need to spend our immediate time working on this, but this is the 2004 recommendations and the status of all of them, most of which have come to fruition. This report had two areas of focus: one was the claims attachments and what could be done with that claims data to make it more usable for quality; and the second was sort of anticipating electronic health records, and again, the kinds of issues that AHIC Quality Workgroup is dealing with today.

So, these are background documents, but after we’ve had the opportunity to go through the plan for the AHIC Quality Workgroup, I’d like to come back to these to say is there some kind of compendium or document that puts together here’s what has been done, here’s what we’re doing now, and here’s where we’re going in the future so that we can give these recognition and status and build upon them and have them publicly available. So, with that —

MS. JACKSON: To follow up on that and indicate that the staff is very supportive of trying to pull the items together. When you see these things happening various places – the Quality Workgroup has been the hub of so much cost-cutting information, so we are more than happy to provide that support.

DR. CARR: It is impressive. We’ve been going over this. I don’t think we even realized at the time the import of some of the things that were identified. But, it would be nice to do that, so that’s a goal. I’d like then to ask Christine to give us an update and open the discussion on how we can work together. Let me just be cognizant, just remind folks, our absolute latest time to close this meeting would be 9:15, but if we can finish it by 9:00 and give Don his full hour for the populations group, I’d like to do that.

Agenda Item: Update from AHIC QWG on Quality Vision Roadmap

MS. ANDERSON: Right before I have you look at the handout that we gave you, I just wanted to take a couple minutes and talk a little bit about what the AHIC Quality Workgroup was formed to do and what it’s done so far.

The Workgroup was approved to be formed by the AHIC in August of 2006, August first, I believe. At that time we had a specific charge and a broad charge, and we spent the first about year on a specific charge which was to make recommendations to the AHIC that specified how certified health information technology can capture, aggregate, and report data for a core set of ambulatory and in-patient measures. So, that was the AQA measure, the HQA measures, first addressing what could be done in the short-term to help to automate those and reduce the manual collection burden.

The first thing that the workgroup did — there’s a small snapshot summary of it on page two of your handout — was to write a vision that can be found on the AHIC website and articulated from the group of members of the Quality Workgroup, what they thought the future would look like in the mid-state and in the end-state. Then, the second major action was to write the first recommendation letter that was delivered to Secretary Levitt in March of 2007, which focused on how to get standards developed around the key data elements for quality measurement and reporting, to begin to think about workflow for how that data could be captured, to try to get some better understanding of what the hybrid data collection meant for that set of reporting. I was trying to think if there was anything else? I think there was one more but I’ll come back to that.

After we finished that work — of course, the work wasn’t finished, after the workgroup handed it off to some of the implementation bodies, NQF, AHRQ funded NQF to begin a group called “High Tech” which was an expert panel that was going to look at the data elements. We turned our attention to the broad charge. The broad charge is much broader. I’ll just read it to you quickly and that will help put the roadmap into context. It’s to make recommendations to the American Health Information Community so that health IT can provide the data needed for the development of quality measures that are useful to patients and others in the health care industry, automate the measurement and recording of a comprehensive current and future set of set of quality measures, accelerate the use of clinical decision support that can improve performance on those quality measures, and then also to make recommendations for how performance measures should align with the capabilities and limitations of health IT. So, it’s very broad.

In thinking in how to turn attention to that, what we did was narrow our focus to say, let’s not think about today’s quality measures, but let’s think about the future where we might have patient-centric longitudinal assessment of quality. Then, let’s do an evaluation of how health IT would need to evolve in parallel with the quality measurement community and try to understand what are some of the key building blocks that aren’t yet built.

So wrote a requirements analysis, which is also available on the website and then this, what you have on page three of your handout, is where we really want to focus our attention. This is really a summary of some of the information that came out of the requirements analysis that was somewhat a limited literature review and existing sources that we were aware of, as well as a set of interviews, et cetera.

This is now the Quality Workgroup’s work in progress for a roadmap specifically focused on developing the health IT capabilities to achieve the vision. So, it ties together the vision document and the requirements document.

I just want to tell you a little bit about why some of the things are on here and why they are grouped the way they are. The top three were identified by the Workgroup as potential accelerants, so saying we could achieve the vision faster if we had payment reform, or instance, et cetera, if the measure that was evolved to a place where it actually was able to support the vision and if there was a legal framework for data sharing.

We took then as an example, within each of these a stab at saying what would we like to see happen? Where are we today, and where would we like to be, what would be maybe a reasonable endpoint? In some cases, you could argue it is too aggressive or not aggressive enough. There are lots of opinions on that. But, this is where the Workgroup ended.

What’s in yellow as you look down, the data stewardship, the patient record maching, the provider entity record matching, the record de-identification locus, the data exchange and aggregation. Those were some more of the technical elements, building blocks, that we thought maybe there were pieces of today but that weren’t well enough established to support the future vision.

Then, we focus really in the green, really on the data itself. We are contemplating or discussing or trying to better understand whether or not having identified a quality data set would be potentially an accelerant even to the measure set evolution. There’s a series of conversations that we need to have about that. We’re not sure what the right terminology is, but the concept that you would pre-define your data elements that you would like to do the measurements from and be able to communicate much more effectively to the provider community about where you’re going and what data elements are important to have accurate and machine computable, et cetera.

Then, the data element standardization is really representing the process that’s already happening with a number of entities, the HITEP, NQS, HITSP, there’s an AMA NCQA collaborative working on some of these topics as well. We’re saying this needs to continue, but right now there’s not really a strong mechanism for how it continues or what the cycle looks like.

Then there was some focus on this idea that eventually we’re going to have to get back to thinking about the coding improvements especially as we think about moving to electronic health records. How do we really express the state of a disease, and will the future of quality assessment drive changes in coding that would be needed?

In blue at the bottom, and it’s only at the bottom because that’s where it seems to keep ending up on our priority list but it’s not because it’s not so important, is around clinical decision support for the patient and the provider.

We have in our broad charge a need to focus on clinical decision support. We’ve so far tabled it a little bit and are planning to come back to it after this next set of recommendations that we’re planning to give in early 2008, to focus more completely on clinical decision support. Which may have its own whole roadmap but we want to show it here to stress the fact that it’s tightly tied, because at the end of the day, we’re not really talking about measurement for measurement sake but improvement and how we get to real-time improvement and what would clinical decision support look like in the future.

So, I would say it’s not well-baked at this point but it’s very represent that it’s important and that we need to turn our attention to it.

So, Justine I want to ask, do you want me to walk through a definition a little bit more of what these are? I don’t want to read the page – what do you think would be best?

DR. CARR: I’d like to ask Steve to kind of —

DR. STEINDEL: Actually I have a question first before I’d like to say anything. I’m concerned about the way you characterize the yellow box. You refer to that as technical. It’s not technical. I would say virtually all of that box is policy decision.

MS. ANDERSON: Right, I agree with you. I made up the word as I was talking. It is very important. They are policy decisions that then have implications for how you’re going to —

DR. STEINDEL: Yes, like for example, matching patients to their records was something that NCVHS standards and security subcommittee looked at extensively – I think we closed the report a little over a year ago – and what we found out was the technology for doing that is very solid. There’s no problem at all. The big problem is where do you place the cutoff criteria and how do you resolve the ambiguities?

I point that out as a specific thing because you know, this dovetails with what Justine was talking about earlier with regard to this roadmap in that there are many areas of this roadmap that NCVHS has looked into, has material that it can provide – can point in directions that would be very helpful to the AHIC Quality Workgroup. It might be helpful for us, when Justine was talking about the compendium document, to look at that type of material also ahead, and then ask where can we help?

MS. ANDERSON: I think that’s a key point here. This page does not represent what the Workgroup is intends to do. One thing we’ve been gathering over the past four or five months is what are all the reports and who has already worked on these areas? We have another diagram – it feels like we are the diagram creator, that shows what entities are working on different elements. The list is long in terms of where work has been done but relatively short in where the decisions have been made. I think largely because of where it was. It is an evolving area.

What we would be interested in knowing is – two things – the AHIC Quality Workgroup is going to focus out this whole page our December 14th meeting, on trying to get some clarity on this quality dataset concept and data stewardship right now. Just thinking about how the AHIC Quality Workgroup can add value in those two areas.

So we will have some testimony and some discussions still on the 14th, in trying to tease out whether or not there’s anything we could say in the form of recommendations on those two areas come the beginning of January. What we’re also very interested in understanding is, where else and who else might take on some of the other topics and how we best coordinate so that we can report out given that these all have to work together to reach the end state.

DR. FITZMAURICE: I just want to ask you some things about stuff I really like in here. Single set of patients; what is patient-centric versus provider-centric with regard to a quality measure? I don’t understand that.

MS. ANDERSON: Well, the thinking is that if you were to — the longitudinal measures, some of the components that the patient cares most about are the outcomes for the patient over time. If we move more toward the longitudinal measure and have a greater focus on the results for the patient, then that set of quality measures would be more patient-centric than some of the current quality measures that are processed.

DR. FITZMAURICE: So my doctor would be graded on whether he/she has my blood pressure under control or my cholesterol under control over a five year period, so maybe a rolling five year quality measure. So, it still measures or evaluates the provider of care, but it uses a measure about me as opposed to grading me on gee, did you have a lot of fatty foods yesterday? You get a bad quality measure for yesterday.

MS. ANDERSON: Well, the first — I won’t comment on which measures because the first report on this that the National Quality Forum just put out on AMI and back pain is out for public comment until December 10th. That’s the current direction. But, we’re not judging what these measures would be but the two key phrases here are the single set. There is a great emphasis on how we, not that others wouldn’t have things they’re measuring and providers wouldn’t measure themselves, but that we have some agreement around when we’re doing public reporting.

DR. FITZMAURICE: So you are going to publicly report about my stuff. Stuff about me – Michael Fitzmaurice’s blood pressure or whether I got a beta blocker after my heart attack and whether that continued over the next three or four years. You aren’t going to identify me, right?


DR. FITZMAURICE: You are going to identify my provider.

MS. ANDERSON: This population based provider, yes. So patient centric is only in the design of the measure and not in the reporting of the measure.

DR. STEINDEL: Can I ask you a question regarding a word Mike used to characterize this and that was grading. Especially grading with respect to patient-oriented. As my two colleagues on this workgroup who I accidentally joined at breakfast can tell you, I ate a very healthy breakfast this morning — two eggs, sausage, et cetera. So, if you start quote/unquote “grading my doctor” on well my lipids are doing, that doesn’t do the doctor any good because I’m a bad boy.

DR. FITZMAURICE: Well, he’s not a very good educator, or she’s not a very good educator.

DR. STEINDEL: Oh, he’s a very good educator. I’m a bad boy.

MS. ANDERSON: I don’t think we’re going to resolve that.

DR. STEINDEL: I think that’s a very key issue on this. It comes up all the time with diabetes because we know how many people follow diabetes controlled diets.

DR. FITZMAURICE: Yes, but if I go see Larry and Larry convinces me to change my habits, he’s a better doctor, and I get a better result.

DR. GREEN: I’d never try to persuade you to change anything.



DR. CARR: Hold that on how you get your good grade as a provider. Larry do you have a comment?

DR. GREEN: I do. My brain’s going in about six ways at once trying to focus on yesterday, plus what I am hearing this morning. On page two here, the quality of workgroup vision summary, right in the middle of that blue box, says, “This evolution will require new efforts to collect, aggregate and analyze longitudinal data ..” I want to ask you about longitudinal data in another way. Then on the next page with the chart – you’ve got in the green area there, expanded data element standardization, coding improvements. We heard a report from an AHRQ supported conference yesterday about trying to establish the data standards for in particular, primary care, such that you can start with patient’s reasons for visits and actually aggregate over time into longitudinal data sets, episodes of care that would clinically meaningful. I’m using the word advisedly – I’m not talking about what a group can do to create statistically meaningful episodes of care. I’m interested in something that actually means something to Steve and his doctor, Mike and his doctor, and what they’re trying to accomplish.

What is the thinking going on in your group about what’s that going to look like? How is that going to get decided? Where is the action plan for getting to expanded data elements, standardization, and coding improvements for primary care that give us the longitudinal data?

MS. ANDERSON: Yes, I would say that this is an area that we’ve identified as a need and not drilled down on very deeply. We know that through our interviews, reviews, and knowledge that this is a key area. How we’re going to actually produce the data that would be meaningful for improvement along with measurement, is an area that has not been well tackled. So, all we have so far on page 12 and 13, is just an acknowledgement that we need to continue to move forward to this state, but we have no interim step for action that is going to get us there yet. This may be an area that we needs a little more tackling.

DR. GREEN: For the sake of discussion of the quality group, I’d like to nominate that for consideration as something that one of the partners for this might be NCVHS.

DR. CARR: Right, I want to bring this back to the focus of where are the commonalities that we can work together. I think you did mention that this isn’t all going to be all AHIC work, so that’s one thing, where it will all maps up to. But, as we look at this, let’s do two things.

First, let’s identify the areas where there is a commonality that we have worked in the past or are poised in the future. Then, I’d like to ask you to say more about data stewardship and sort of develop that more. Obviously, data stewardship is a key theme in the report on uses of health data that we’ve been talking about the last two days so that is certainly an area of overlap.

Steve mentioned patient record matching — we have work that we are prepared to share with you straightaway and provider entity record matching, I don’t think we heard about that precisely.

MS. ANDERSON: We were asked to distinguish between how do you know it’s the same patient and how do you know it is the same provider? And how the NPI doesn’t quite get us there and what might help us get there.

DR. STEINDEL: I would imagine the techniques through a technical point of view relatively speaking.

MS. ANDERSON: Right, and the policy decisions might be different.

DR. CARR: Then down to the quality data set, as Larry was mentioning, we heard yesterday about the IC, what is the acronym?

DR. GREEN: What it really is is a episode in a classification scheme starting with what patients care about.

DR. CARR: So that was a presentation to the folks in the meeting but expertise in Larry and interest on the committee. Then, coding improvements we discussed at the hearings that we held in June, the Quality Workgroup, this past year, and in the summary that we recently put together. Are there any other areas?

MS. ANDERSON: Can I say a word about the de-identification locus because I think it might overlap with work being done?

DR. STEINDEL: I think that that and also data exchange and aggregation we have done some work.

MS. ANDERSON: Because we’re struggling with – if you think longitudinal and you’re thinking there is by definition more data aggregation across entities, and there are implications for where it gets de-identified relative to the feedback loop, relative to the ability to do some of the analytics. There needs to be better clarity on how that might occur which might tie back to stewardship, some of these tie together.

DR. CARR: Can I pose a question on that because as you know, the report on uses of health data we are discussing these two days, builds heavily on HIPAA, and within HIPAA treatment payment operations permits use of identifiable data for quality. We understand that the absence of these identifiers undermines the ability to do quality analysis. So when you’re talking about de-identification, can you say a bit more about that? Where is it being de-identified?

DR. FITZMAURICE: Particularly for quality purposes.

MS. ANDERSON: We wrote some of this prior to that report, so we’re pleased because there was a debate about is longitudinal quality improvement outside the walls of individual covered entity but across covered entities going to be also considered to be included? There was a discussion about that and concern that if it had to be de-identified before it was brought together that the techniques for evaluating the data would be compromised.

DR. CARR: So we had a specific discussion on that yesterday, and in our recommendation seven, we say specifically that quality remains within operations as initially defined with HIPAA. What you’re saying is as it goes across entities and presumably our chain of trust and accountability would dictate that if you’re a business associate you get identifiable data so I think that should cover it.

MS. ANDERSON: What I glean from this discussion would have to be a business associate of all those entities, and then they may have to co-manage that entity in some way. I don’t mean manage in a literal sense, but being a steward for all the entities. There has been some clarity for us through the process that you’ve been going through on that committee.

DR. CARR: Mike, did you have a comment about this?

DR. FITZMAURICE: You can combine groups in different locations and treating different populations of patients as to organized health care arrangement. Then you can exchange individually identifiable patient data, probably the minimum necessary to accomplish your quality purpose. Even across countries you can do that under HIPAA.

DR. CARR: Okay, good. I wonder if we need to be clearer in the report?

DR. GREEN: Another question since you’re here going on in your discussions but integration of these data so that questions can be asked of it. Webster defines integration as a pulling together of what often appear to be disparate parts into a coherent group that has meaning. It seems to me in a very simplistic way, that we have to come to terms with where that aggregation is going to occur. It can’t just happen everywhere – it is going to have to happen at some place – some point.

Is the thinking that that’s going to be Microsoft’s Health Vault? Is it going to be that we have 300 million independent purchasing units that these will be aggregated and integrated around?

DR. CARR: Right, and that’s one of topics in the —

DR. GREEN: What does that happen?

MS. ANDERSON: That is one of the discussion points in the data aggregation element where we feel that there needs to be some more guidance on that. One of the things that the worker has acknowledged – it will happen at multiple levels.

DR. STEINDEL: This is aquestion I wanted to ask of Larry to make a point, and I’m glad you just said that. It’s not where does the aggregation occur in my mind. It’s what does the aggregation involve. If we take a look at the situation you mentioned yesterday, where you’re involved with I think he said 100 providers in quality measure. I assume they’re all in the same region.

DR. GREEN: Most of the time.

DR. STEINDEL: Yes. Everybody recognizes the need for this local exchange of information because when a person is getting health care, they generally go within their region. However, big you want to define the region. In the West, it could be geographically very large. In Atlanta, it’s very small. But, it’s still we’re confined to a set of providers.

There is a tremendous need for aggregation to occur, for information to occur, within that subset. Now, if I pick up from Atlanta and move to San Francisco, that doesn’t necessarily mean my data needs to be aggregated someplace on a national level for quality to be done. That means the system has to provide a means for my electronic data to transfer to my new provider in San Francisco where then the aggregation can occur for that area.

So, I think this gets to your what. What or how aggregation should occur, at what levels, what parts? You know this addresses some of the problems a lot of people have when they think about the NHIN because they think everybody’s data must be available instantaneously nationally both in aggregated and unaggregated formats.

DR. GREEN: Again to link this back to the report we heard yesterday from Karen and CMS, the CMS plan for five year EHR demonstrations and whatever this medical home thing is. That’s another thing. I really want to advocate for getting into the discussion about what role if any, this concept could play in helping us get these data organized in a way that they are usable. If you have a national provider identifier of some sort and then you can make it patient centric, there’s a patient. You can consistently identify the patient, then what Steve was talking about a regional, most of the health care is going to occur in a region, how do you get that defined? It could be whatever that patient says is their quote medical home. That could be a year fine focus whether it’s a hospitalization or an ED visit or whether it occurs in Hong Kong or in Atlanta or at the airport or whatever, that could be a repository target where this sort of stuff gets pulled together. That would effectively, regionalize it. It would typically get a substantial majority of the care that has been rendered in that setting. I don’t hear in the conversations an exclusive discussion about assumptions about how that is going to work or not. It strikes me as fertile ground.

DR. CARR: I’ve identified that as, so clearly that’s an area that we have expertise, interest, and it needs further development and we need to hear more about it.

MS. MCCALL: Data aggregation?

DR. CARR: Yes, well the data aggregation and the quality data set, those are two things. I mean, you know, your saying, your are not doing all of this. I don’t want us raising our hands, we are doing all this also. (Laughter.) If we can only hold back here a little bit.

DR. GREEN: Justine will do it.

DR. CARR: No thanks. No I don’t mean that. But I mean I’ve just learned a lot about how much work this is. (Laughter.) I’d like to go back to data stewardship and ask you to say more about that and what is it in particular that you’re looking to accomplish.

MS. FARQUHAR: Quick question about the data set. I know you’re going to talk about it on December 14th, and I’d be really remiss if I didn’t ask you. Is patient safety and disparities included in each of these data sets that you’re going to be discussing because it is not specifically inked out.

MS. ANDERSON: We are actually going to be at AHRQ for two hours on Friday with CMS discussing this topic. Caroline has put on the table all of those topic areas. We’re not even sure this box is defined right because how many quality data sets can you have if you start parsing it across settings? So, it really is, we’re not quite sure what this means yet.

DR. STEINDEL: Christine, I’ve been involved in some of those discussions. The HITSP level, and I would say the quick answer to that is an infinite number.

DR. CARR: Actually I’m going to diverge from stewardship for a moment just to talk about the — we had a little conversation yesterday and again kind of informally last evening, about the human factors piece of this as you are in a patient encounter and you have to reconcile medications and you have to enter these structured fields and you have to be all these other things and you have to get your e-prescribing, did you fill the prescription?

There are a number of electronic activities that are filling up the twenty minute space of a doctor-patient interaction. I would like to put that as a marker to, that as we have all of this electronic structure, what’s left for the physician or provider-patient interaction? I think that really needs to be studied because we’ve taken work that happens off hours at other times of the day, and we’ve crammed it into those twenty minutes. We want to make sure there is a chance to speak to each other.

So, Michael, one brief comment and we will go onto stewardship.


DR. FITZMAURICE: Steve mentioned HITSP. I want to compliment AHIC, ONC, and HITSP on the quality area. When the use cases are developed by a quality working group or by AHIC, ONC for further specification, and then to HITSP in a couple of places the data elements have been identified on a core data set and HITSP has been able to designate not only the standard but the data elements themselves and definitions. That happened for Biosurveillance and it happened for the Quality Workgroup, so that’s a compliment and order that that happened.

So, I would suggest that on every use case, you make sure that there’s a core data set developed by AHIC or ONC before you send it to HITSP. HITSP of course can add and maybe subtract from it, but it’s essential that the data that performed those functions be specified. The attack hit by HITSP and by the HITEP is that if you specify the data elements, they can be reused in a lot of different quality measures but they have to be specific and very well-defined.

DR. CARR: Okay, brief comment from Steve.

DR. STEINDEL: The brief comment is I believe that if the core data set is open for discussion. CDC is not a big proponent of definition of core data sets.

DR. CARR: We are not going to have that conversation right now.

DR. STEINDEL: I know, but if we are going to put that as a topic of discussion, it is an open topic.

DR. CARR: Yes. Larry, one brief comment.

DR. GREEN: We also yesterday made a request that AHIC develop a use case for primary care settings. We had a request that AHIC create a use case for primary care. That there is no primary care use case and the point was made that this was the largest platform in health care delivery in the country and it ought to have a use case.

DR. CARR: Yes, and not to overlook you, but the comment about patient experience of care.

MS. FARQUHAR: Well, if you guys can come Friday with regards to those data sets, other things that you might want to include is patient’s experience of care and then a way for patients to input their data into it.

My father, for example, has to call in his heart rate every so often and he does his blood pressure. They should have that input as well to be a little more patient-centric.

PARTICIPANT: Can we talk about stewardship?

DR. CARR: Thank you. Stewardship.

MS. ANDERSON: Well, we just as a planning group talked more about how the AHIC quality workgroup would address stewardship yesterday. So, you may be disappointed at how much clarity we actually have about this yet, but we have decided that we need to be talking about the practices and conventions for stewardship and start getting to the dialogue on how it might happen on the practical level. But, that is as far as we have gone.

So, we literally are going to take some testimony on December 14th, start the dialogue within the Quality Workgroup. There are some of the AQA leader’s, like George Ishum, that are part of the workgroup. I think there are some other folks that have done some work in that area that are coming on the 14th to try to address this issue of, how do you make it practical? What is stewardship? What are the functions that create some of the pieces that come from some other reports? Then, how might we move forward in this area?

Again, not with the eye on the Quality Workgroup doing it all, but perhaps it will have its own little roadmap that says here’s all the functions that we believe exist and here’s where we know work is being done on that and who might resolve that. But, that is as far as we have gone.

DR. CARR: In terms of stewardship, does that include the data integrity and quality for data aggregation?

MS. ANDERSON: It has been listed in the list. So it includes how do you know that the measures at the time that they are created are accurate to how do we know that the data. I mean people have thrown all of these issues into stewardship. I don’t think there has been a dialogue to say what belongs in/out, are there multiple different layers of stewards.

There have been discussions about certifying stewards which you’ll see on here. All of that is really just at the highest levels of discussion right now and I would say not conclusive. But, it is a big basket right now of things that could go into it.

DR. CARR: Right. So when spoke earlier on — last month we talked about would NCVHS be helpful in a role of holding hearings and so on about stewardship?

MS. ANDERSON: I think it would be great for you to be there next Friday and then maybe after we get some more clarity on some of the steps consider parsing off a piece that would be good. I think there’s going to be more than enough need for hearings across all these topics. They’re very disparate topics. Even if someone agrees in the concept of a steward, there certainly probably not agreement on what a steward would do or who would do it.

DR. CARR: Right. So the concept of stewardship, the boundaries, and whether it’s a person, an entity, a set of principals.

MS. ANDERSON: Right. What is it?

MS. MCCALL: Then, just to add one more thing is that with respect to all aspects of stewardship or is it with a soul focus on with respect to quality?

MS. ANDERSON: This discussion we focused on quality.

DR. CARR: What we’ve been talking about in the report is uses of health data – it is a broader concept.

MS. MCCALL: There are some places in the report where I can see where we might be able to focus in – I think it is actually in 7.1. whatever. But that is a nice thread of continuity.

DR. CARR: Alright. Anything more on stewardship or should we leave this? It’s still in the early stages, so December 14th you’re saying, we’ll know more.

MS. ANDERSON: Right, we’ll know a little more clarity ourselves.

DR. CARR: Okay, and just let me ask, in terms of the RFI that AHRQ put out. Did that bring us along at all?

MS. ANDERSON: It definitely gives the community viewpoint on what would be included and what — but at that point, it’s just a database of knowledge now for someone to act on or a group of entities to act on.

DR. CARR: Right, so taking this back to the beginning, as we’ve had this discussion, we have a lot of overlap. We’ve talked about overlap in data stewardship, patient record matching, the issue of record de-identification locus, but we’re saying that we believe that with quality, the de-identification should not play a role. I will add though that the whole issue of de-identified data is an area where NCVHS will be holding hearings because of the issues of sale of de-identified data. But, I don’t think that relates here. We would see that the quality initiative across these various entities would include identifiable data.

Quality data set — we have a particular interest in the ambulatory, the medical home, we heard a number of different things and we feel that we have some experience in that. On then the coding improvements, we summarize some in our in our letter or whatever it was, and our hearings from June 19th.

So, I’m looking at process and next steps. I think we need to learn more, hear more, and sort of be in dialogue over the coming weeks – hearing more about the stewardship and so on. But, I don’t know, Carol and Steve, we spoke this morning separately. I’d still like to see the body of work that we’ve done somehow have an existence in a public forum. We reference it as we look forward to the direction that NCVHS Quality Workgroup will take on in the coming year or two.

MS. MCCALL: What I was hoping to do with this work here today is to understand the future direction so that we could understand where this group might be able to – whether as a workgroup or NCVHS – might be able to assist in either taking testimony or through other activities.

What I would like to see happen in that assist is actually, think of it as packaging, but to actually look back and to look back at some of the things we’ve done whether it’s the ‘04 report, whether we’ve put together in ’06, and say this is what we have done. This is what resulted. This is what we’ve seen, and this is the work underway. And actually create that as a backdrop to say, and therefore, this is the work and our role in it that moves forward so that it actually strings together.

There’s a tremendous amount of continuity from work that we’ve done before to where we are now, and I think it’s within our power to show that continuity that we can do that. That’s really a comment as much to this workgroup as it is to you folks.

DR. CARR: It actually goes back to before ‘04 because of the Full Committee NCVHS that is true. So, I think that there’s history there.

DR. STEINDEL: Do you see this, Carol, as an informative document going forward like to the AHIC Quality Workgroup instead of like a recommendation letter to the Secretary?

MS. MCCALL: Yes, I think it’s to AHIC quality workgroup and I also think it’s to ourselves. I think it really is. The piece that might come forward would be our intentions of what it is that we plan to focus on. Then, all of this other information becomes the why. Why? You go well, because we seen things happen, we have made recommendations, and it got us to a certain spot. We then took testimony, we heard certain themes, we see a lot of work there, therefore these are the next logical steps kind of thing.

DR. CARR: One of the questions too is that as AHIC takes on a new configuration, do we know what the role of the AHIC Quality Workgroup will be going forward?

MS. MURRAY: We don’t know that.

DR. CARR: So, in some ways, I think this could help in the transition to at least say, as you go through your checklist of work that is out there needing to be done we could be saying, this is what we have done, this is areas that align with work that we see going forward, and help sort of parse out that we have in addition, we have a home for all of these initiatives.

DR. GREEN: What do you want to see happen?

MS. MURRAY: Yes, I think that is exactly what I want to see happen. That is what this roadmap is about and we would love to be working together on that roadmap.

DR. FITZMAURICE: Justine, would it be fair to say to then that the AHIC Quality workgroup is one of our drivers and as they identify issues and problems, that we can be looked upon as a resource that can help flesh out what the scope of the problem and what some possible answers are and then what the public feels about it. DR. DR. CARR: Yes, and I think Steve raised a good point this morning that the NCVHS Quality Workgroup has worked in two modes. One is sort of visioning and the other is sort of immediate reactive. The kind of work that you’ve been doing is building a rapid tempo, and it’s not the kind of work that this group can do. But, as we have done with the use of data, that would be the reactive one to your request, and also some of the other things we’re visioning. I think we are very helpful in sort of scoping out what was the landscape. Other there comments?

I’m going to move it to next steps because having said this I’m looking for volunteers around the table to help pull some of this together. Carol has been an architect in some of this. I’m looking to you for some help.

MS. MCCALL: Sure, Justine.

MS. ANDERSON: Justine, I also wanted to, just before we get off this, in addition to our meeting on the 14th, the AQA Data Sharing and Aggregation Workgroup still as they’ve begun and still has this keen interest in data stewardship and I expect they will also be looking for what’s the piece they will be able to move forward. So hopefully we can get to the point of some parsing.

DR. CARR: Right. I would say we’re not looking to own that. We’re looking to support and help inform it.

MS. ANDERSON: I don’t know when their next meeting is I was just more mentioning it so that you can follow them too.

DR. CARR: George Eisen(?) will be at the December 14th meeting. Just take my Blackberry and everybody fill it up.


DR. CARR: Looking to Carol and Marybeth —

DR. STEINDEL: I’m not volunteering. I wanted to point out and volunteer some other people. There is some archival work that needs to be done with NCVHS reports in this area, in particular the reports in the NHII Workgroup in the period of time between the formation of ONC and when we published the 2001 report because we held a lot of discussions in particular in the area of local data. This is what Billy Asnoff was in his position and pushing LHII’s, Local Health Information Exchange Networks, and we heard a lot, I don’t know how much we published, in the area of local data aggregation, et cetera.

MS. MCCALL: One may need to be refreshed a little bit.

DR. STEINDEL: That’s why I say we have to do research and see what was said.

MS. MCCALL: I wanted to go back to the conflict of the quality data set and data stewardship. Those seem so tightly aligned I know why you would be focusing on those simultaneously. But one seems to be about content and the other about process, and so they are so tightly aligned are they in fact one and the same?

MS. ANDERSON: I think on the quality data set, the primary question we’re trying to answer in the short-term is should we stop going measures to data and start going data to measures? So, trying to at least we’re trying to get to that point. How much support is there for that idea?

MS. MCCALL: Please explain that.

MS. ANDERSON: Meaning that right now, the way we’re, when we talk about standardizing from the health IT perspective, standardizing data in health IT, we’re saying here are the measures, here are the common data elements, and what are the standards for those data elements, how might they be represented in an electronic health record, how would you get them out of an electronic health record, et cetera. As opposed to identifying the set of data that you would like to be using for performing quality measures and the measures can still be performed but going right to how do we then make those machine computable, how do we capture that data, and have them measure development wheel tied to that in the sense of saying, for example, NQF.

If NQF, we’re endorsing a measure if it used those data set elements, it would require that it used them in a standardized form. If the measure deviated from that, either by adding other data elements, that they could justify. It doesn’t mean they couldn’t, but if they use other data elements, justify that the benefit exceeds the burden and sort of dialogue about whether those new data elements need to be standardized, as well.

So, just thinking about the way those cycles might work, we’re trying to talk through it, understand what the strengths/weaknesses are, is it a good idea/bad idea, is there a hybrid answer – as there always seems to be.

DR. STEINDEL: A specific example, a temporal type of quality measure where you’re looking at something after something occurred. That is very difficult to computerize because the before events, you don’t know where it might be recorded.

DR. GREEN: This exchange is a further argument for why we should have serious discussions about what would it take to define meaningful episodes of care. Because if you had any episodes of care as a unit of analysis as moving from data to measure, you have then got the bookends for an important event. You can calculate the value equation, you can know what that episode costs, what was spent on it, you can know the number of sites, you know where the data came from, how many visits, whether it went to an ED or not, whether there was an A1C done anytime in that episode.

DR. CARR: But you are still stuck with the same problem it is just another level.

DR. FITZMAURICE: We are addressing it in a practical sense at HITSP. We get the quality measures, we formulate the data elements, and we maintain that as a core data set. If they can be used by additional quality measures as they’re added, and then we use the same data. But, if they can’t be, then we’ll note the different data elements as they come in.

One of the hardest problems for us has been codifying the exclusions of what do you take out of the numerator and what do you take out of the denominator. The second hardest problem has been more precisely, what do you mean by, for example, time for the operation is it when it was sewn up, when the patient was wheeled out of the operating room, just more specificity on those lines help flesh this out.

MS. MCCALL: I guess I don’t see us actually playing a lot in that. That is a very technical realm. I wouldn’t necessarily put a star next to our work on that.

DR. FITZMAURICE: Except as Larry mentioned, you may see from testimony that there are some gaps for particular code sets could fill. We might advocate to the Secretary experimentation and use of particular code sets. DR. STEINDEL: Or we can point out that there are problems in the way some of these quality measures are defined from a measurement point of view.

MS. MCCALL: Right, I guess what I would stay on that one rather than sit through what every party that is held in order to get the few tidbits – that there be a mechanism to where those are gathered up and if there needed to be hearings or testimony gathered then that could be a role that we could play.

MS. FARQUHAR: Well, right now, NQF has now a 12 page submission form for all of there quality measures. It includes very specific specifications, and that is available to NQF who would be playing that role, I would think.

MS. MCCALL: So it may not be us. I’m just trying to — before we walk away from this — I want to make sure we understand what our role is. I don’t see a big one there. And on the data stewardship, I think there’s a natural synergy about what we’re talking about right now that could be more focused. So that one makes sense.

DR. CARR: I guess, right. I think again talking about the macro sort of big picture and the reactionary. I think with the ambulatory quality data set is sort of reactionary because we heard about that yesterday. I’m just trying to write next steps here so I’ll remember when I leave the room.

So, what the immediate plan is, and of course, we need to review this with Simon and the executive committee and so on. But, I think that what we’ve heard today really resonates very much with the direction we want to go in. We want to put together a summary of work done with a particular focus on stewardship and on matching patients with their records.

Steve mentioned some of the early work in NHIN. It’s an opportunity for us to summarize the work that’s been done and then identify areas that we would like to continue to support going forward. Again, it would be in particular, well, what did we say? It would be the stewardship. I mean what of these things would you recommend that we would play a role in going forward? I’m torn now between being too macro and too micro.

MS. MCCALL: You know what I think would be a good candidate after data stewardship actually is the data aggregation. I think that it’s going to follow naturally from some of the enhanced protections. I think it will follow naturally from some of the data stewardship. I’m just trying to see around corners a little bit. I don’t think it will be immediate, but when we start talking about de-identified and what we have already teed up for ourselves as sort of a second set of topics in respect to enhanced protection, a lot of that is about aggregation and we’re going to get into the same discussions about when it is for quality and quality improvement and research and other uses.

DR. CARR: Maybe there are two columns here. There’s what have we done historically and what do we see ourselves doing going forward. So, the things we’ve mentioned that we’ve done historically are the stewardship, the matching to the records, the NHIN work, the coding improvements. Then, what do we see going forward would be stewardship plus —

MS. MCCALL: As a follow-on, I’d think I’d like it to naturally link with what we’re doing in the other ad hoc group or as a full committee. But, I see that as a potential natural follow-on.

DR. CARR: Comments?

MS. ANDERSON: For one, I think there’s so much work to be done that it’s nice to hear anyone think about focusing on any one of it. There is a lot going on right now with aggregation that we will need evaluations. Aggregating the multiple — there’s the one project that’s doing all of the claims data from different foreign places – Brookings — and then there is aggregation going on at the value exchange level that’s kind of combining data sets.

There are a lot of pilots around that AHRQ is looking at combining clinical and electronic data. I think there’s a lot of work to be done in the aggregation area. I also think there’s going to be a tremendous amount of forward work on the coding improvement area that as we evolve how do we move these two along together. I frankly don’t see us, that’s not really IT oriented —

DR. CARR: I think that fits well with us especially — we have hearings next month on ICD-10 in January, and we have the work that’s been done. So, I would say that would be an area of future that builds on our past work, as well.

So, I’m going to summarize then that we will put together a document that summarizes our historical work on stewardship, matching patient records, NHIN and coding improvements.

We will forecast a role that we will raise our hand to play going forward on stewardship and coding improvement, plus or minus aggregation evaluation, not the doing but the kind of here’s all these different things, what might you do. But again —

MS. FARQUHAR: I’m a little cautionary is a possibility whereas I think data stewardship, you’re saying —

DR. CARR: Marybeth is saying it would be the kind of thing of not the doing it but the hearing of the different groups that are doing it and the pulling together.

MS. MCCALL: Exactly, and because the activities are so broad. It’s not to take testimony on any kind of thing that has the use of aggregation in it.

DR. CARR: We have distributed some of the earlier work and we recommend that to your reading. If you have additional thoughts, forward them along. We will circulate this. Carol is going to take the lead on pulling some of it together, maybe Marybeth can look through some of the earlier documents, and then we’ll follow-up after the December 14th meeting.

I want to thank you so much for coming and accommodating our changing schedule and thank everybody for their ideas and intentions. I would like to make sure that we have a Quality Workgroup membership list accurate in the e-mails, on the web, in the printed copy, so I’m looking to Cynthia to give us some help with that because I find the different lists in different places on the web for example.

(Whereupon, at 9:10 a.m., the meeting was adjourned.)