[This Transcript is Unedited]
DEPARTMENT OF HEALTH AND HUMAN SERVICES
NATIONAL COMMITTEE ON VITAL AND HEALTH STATISTICS
SUBCOMMITTEE ON QUALITY
June 11, 2009
National Center for Health Statistics
3311 Toledo Road
Hyattsville, MD 20782
CASET Associates, Ltd.
Fairfax, Virginia 22030
- Dr. Justine Carr
- Dr. William Scanlon
- Dr. Blackford Middleton
- Dr. Larry Green
- Dr. Michael Fitzmaurice
- Dr. Mark Hornbrook
- Ms. Carol McCall
- Mr. Matt Quinn
- Ms. Marjorie Greenberg
- Ms. Allison Viola
- Ms. Debbie Jackson
P R O C E E D I N G S (3:40 P.M.)
DR. CARR: Good morning. Thank you for getting here so early on this Thursday morning. I am Justin Carr, co-Chair of Quality Subcommittee and my other co-Chair, Paul Tang, unfortunately can not be here because he is attending his daughter’s graduation. Let’s go around the room and say who we are. So I am Justine Carr, Caritas Christi, member of the Committee, co-Chair of the Subcommittee and no conflicts.
DR. W. SCANLON: William Scanlon, Health Policy R&D, member of the Committee, member of the Subcommittee, and no conflicts.
DR. MIDDLETON: Blackford Middleton. Good morning, member of the Full Committee and the Subcommittee from Partners Healthcare, no conflicts.
MS. GREENBERG: I am Marjorie Greenberg, National Center for Health Statistics, CDC and Executive Secretary to the Committee.
MS. JACKSON: Debbie Jackson, National Committee on Health Statistics and CDC Committee staff.
MS. MCCALL: Carol McCall, Humana. I am a member of the Full Committee, member of the Subcommittee, and no conflicts.
MS. VIOLA: Allison Viola, American Health Information Management Association.
MR. QUINN: Matt Quinn from Agency for Healthcare Research and Quality.
DR. CARR: Good morning and thanks for being here. I think our goal today is to take stock of where we have been, the changes that have come about in the last few months, and an assessment of where we ought to be. I just thought I would start off by reminding us of what our charge is as the Subcommittee on Quality.
We are to focus on the first function outlined in the NCVHS charter, which is monitor the Nation’s health data needs and current approaches to meeting those needs, identify emerging health data issues, including methodologies and technologies of information systems, databases, networking that could improve the ability to meet those needs. A second focus is to identify strategies and opportunities for evolution from single-purpose, narrowly-focused, categorical health data collection strategies to more multi-purpose, integrated, shared data collection strategies.
As we had stated in our charge, a near-term priority was to concentrate on the data needs of person-centered health, the set of personal, professional, and community activities and services focused on health and well-being of individuals. So, specifically, the Subcommittee will identify emerging health data needs for a health system where the individual engages as a full participant. The second bullet is the Subcommittee will develop an understanding of the requirements of a health surveillance data set to support personal health tracking and improvement, which with privacy protection and effective and efficient data collection strategies can be used to assess the health status and needs of a population. That was in December.
I would like to just share a couple of introductory thoughts that Paul and I came to as we were meeting the night before last and he wanted to share since he couldn’t be here. So as we heard at our discussions throughout the day yesterday, clearly, the Meaningful Use hearings and all the activities that are evolving from there are very important. One of the things that Paul and I talked about and that really came through repeatedly through the day yesterday is the concept of meaningful measurement, particularly with regard to quality. Thinking in terms of where we need to get, in particular, with a focus on outcomes, I think we heard one of the speakers yesterday talking about the fact—or maybe Lisa Iezzoni—was talking about a lot of process measures and that we really need to focus more on outcomes and these outcomes need to be more than just lived or died.
Second, we want to think about where are the important areas where there are no quality measures? As Paul and I were thinking about it—you think about what are the major disease states, we have a great deal on cardiac disease but we don’t have anything on cancer or cancer pain or just pain in particular, so these were a couple of areas that we were thinking about.
I think I will just leave that out there as an introductory comment and just open—what I would like to do is open it up to areas that we believe we should focus on going forward. I have spoken with Marjorie about the possibility of us having a hearing on Monday, September 21st, the day before the Full Committee meets so I would like today to be a very focused activity to come up with where we think we need to go next and begin to plan toward that possibility of a hearing on that day. My final words will be just as we prioritize what we ought to be working on, I have been collecting the things that we have been saying over the last couple of meetings about what is important. One is to be consistent with our charter, to be data related, to be relevant to what is going on currently, to be non-duplicative but rather complementary to ongoing efforts, to anticipate the dynamics of health care reform, which is almost a changing landscape every day, to do something that will be timely in the value, to align with the skills of our subcommittee and our committee, and to be clear about who is the audience for what we will do and who is the customer and they may not be one and the same. I will stop with that and open up the discussion to the areas of potential focus for the Subcommittee.
DR. W. SCANLON: Actually, I did a little looking back, too, and I was looking back at the agenda for the meeting we were supposed to have on June the third. This Subcommittee anticipated the need to have a discussion of meaningful use and then, obviously, it was superseded by the Coordinators Office and the two days of hearings we have had. I feel like that, in some ways, was an essential step and the next step, in terms of our thinking, in part, is what the Coordinator does next. If you go back to the charge, which you described to us, how the goals that are described within the charge are served by the Meaningful Use requirement to come out seems to me one of the areas that we should be focused on.
DR. CARR: Can you say that again? How what?
DR. W. SCANLON: How the objectives that we describe in the charge—patient-centered health, population measurement—how that are served by Meaningful Use requirement is, I think, a logical step for us to be considering. We had the discussion yesterday in the Population Subcommittee—the landscape has changed dramatically and it has right now sort of changed in only a potential way. What really is going to determine how it shifts is the definition of Meaningful Use, the implementation of this. We have a lot of government regulations that are not sort of adhered to and so they don’t necessarily live up to their promise. The reality is implementation is a critical part of this so it seems to me that we need to look at this next step that is going to come out of the Coordinator’s Office and then think about where do we go from here?
The meaningful measurement part of this—there’s a question of our scope versus the other bodies because one of the difficulties is that other bodies haven’t yet produced these measures. In terms of gathering data, there is an issue that we, in some respects, deal with the logistics of getting the data but the measures, themselves, are not our province and the measures, themselves, are woefully inadequate. There is, in the post-acute world things that are better than mortality but they are not nearly comprehensive enough. We talked about, yesterday, this whole issue of improved functioning is one of the things that is measured but that is not possible for a lot of people. If you go beyond functioning, the only two measures I can think of are re-hospitalizations, which is getting a lot of attention now but there’s a lot of noise in that and the second thing is reduction in pain, which is a very subjective measure. Not that it is not important but it is still a very subject measure. We have to think about whose job is it and who is going to do this because I don’t think we can take on, basically, what organizations like NQF are doing, which is the development of measures.
Yesterday, Chuck was obviously not able to talk about the schedule that they are going to have but I think that schedule is important in terms of what we do next to be able to react to what that is.
MS. MCCALL: Just to build on that, I think you are right. To actually try to design the actual measures would certainly be beyond the scope of anything that we could do.
DR. CARR: Could I just clarify? That was not the intent. Paul, who is on ITEP with NQF said having a voice to say we have gaps and we haven’t addressed them—
MS. MCCALL: Right, and that is what I hear you saying or what I assumed you meant when you shared your comments with Paul. I think, much like we talk about Meaningful Use, I think the topic and headline of Meaningful Measurement is worth being first created—you know, the topic and people become aware of it. So if we were to have a hearing on that topic and have observations in gaps, not about specific measures but the fact that there are gaps and we need to think about them more broadly, that in and of itself is a goodness, number one.
Number two, when I think about—not specific measures but back to what Chuck was saying yesterday, there is an enterprise that is needed that we do not have. I think that there is this faith, there is this naïve belief that if we lay the pipes under the streets that we will automatically have really nice showerheads in our bathrooms and everything is going to work out. There is a lot of work to get to that point. I think it is worth igniting those dialogues.
DR. MIDDLETON: Following up on that excellent commentary, I see a broader array of issues. I guess I will just kind of just go through the whole kit and caboodle and then maybe we can dive into different parts, if that is okay, Justine?
To follow up on Carol’s comments, really we are at a juncture, though, at an inflection point or really a point in time we kind of need to mark and note. I’m reminded of the Flexner Report in 1910 or 1911—I’m forgetting. It is almost time for Flexner II, in fact, it is time for Flexner II. You think about kind of the state of the art of our health care delivery system infrastructure, our education and training programs with respect to their use of IT and knowledge-based tools, and then the connection to the patient and community and the population, in ways that we really haven’t before but in ways in which we are really fully enabled by this new ecosystem of an information technology infrastructure or the information ecology, if you will. So I would suggest one of our charges needs to be to look at the mission and to take a really broad view of what are quality data from the individual provider perspective, the personal perspective, the population perspective, and then the geographic or environmental perspective.
As we were reminded yesterday, determinants of premature mortality come from a wide array of sources and that is what our quality data need to reflect. They can’t just reflect the health care delivery system and its impact. It needs to, in fact, reflect the total impact on a person’s quality and a population’s quality. I think the point about then, secondarily, with that broader charge revision of where quality data need to come from and then we need to connect the dots between what is going on in terms of standardization in HITSP and the HIT Standards Committee under the direction of the Policy Committee. What are the relevant national standards for ontology, information model, terminology, concepts, attributes because every time we implement new information systems that has a different information model or different understanding of codes and terminology we create problems. We create patient safety problems. We create impediments to information sharing and exchange and we create impediments to the broad application of knowledge to support both individual and patient and provider decision making.
Next topic, I think, then is around the measurement science in quality and, in particular, a new field of measurement science and IT. We really don’t know how to measure well, reliably, and broadly, the impact of IT in health care delivery systems. We have done a wide variety of assessments. There is a growing evidence base but in terms of reliable, reproducible methods—the same kinds of methods we hold ourselves to when we introduce a new drug: dose-response curves, understandings of the impact and side effects, and post-marking surveillance for that matter.
We need to develop this measurement science for information technology to better understand impact and value and then to look back on health care reform and the initiatives that are coming to say this thrust in information technology makes sense but this one doesn’t or doesn’t have value, what have you. Part of that implementation science, I think, is this issue of getting at dose-response curves or IT interventions. We typically look at an IT intervention as success or failure, when, in fact, it is much more subtle. We really need to understand, if you will, how much of a response occurs when IT interventions are implemented and why we get partial response, why we get partial rejections of alerts and reminder, and why patients tune in to some forms of information and tune out some other forms.
So there is a whole discipline, I think, around this implementation science, measurement, knowledge, attitude, and practice but I suggest also, akin to the Flexner II, we need a Framingham II. We need somewhere to do a longitudinal study of IT. I propose Massachusetts. There is no bias involved, whatsoever. It is a convenient sample for me, personally, but, of course, at Framingham or Kaiser, actually, would be equally good in some ways and different in others. We need a place to say here is what happens over 20, 30, 40, 50, 60 years with implementation of broad-scale IT? Where do we actually impact modifiable risk factors? To borrow the Framingham and stretch it further, where can we actually materially impact on population outcomes broadly, for entire cohorts, et cetera? I would be very interested in us identifying this need and then helping the right agencies pull forward some funding to do this longitudinal study of IT.
I guess the last thing, then, is just to return to the Flexner II idea. I had a brief conversation with Brent James about this when he came to teach at Harvard in a course we do there. Who do we talk to to get Flexner II started? I think if we were to identify the need and sort of characterize the vision, we might approach the IOM and say, you know, let’s put it together, get some funding, and then say what is the bricks and mortar and technology bits and bytes infrastructure to accompany the Healthy People 2020 vision. In a way, I see Health People 2020 talking about all the right measurement and attributes and population issues but we haven’t talked about bricks and mortar, bits and bytes because those are changing. Patients are not coming to the bricks and mortar anymore, bits and bytes are going places that patient’s have access to that providers have nothing to do with. Health care will be determined, I think, finally, by a lot more than what we do along, as it has already been shown in the health care delivery system but we need to make sure that what patient’s experience outside the health care delivery system is beneficial and useful.
Thank you for indulging me with a long tirade.
MR. QUINN: I really appreciate what Dr. Middleton just described. I spent yesterday—why I wasn’t here was one of the projects that I am the Project Officer for, we started off as Population Health and Health IT. I inherited this project and when I got it I said we need to tie some closer boundaries because that is just all things to all people. Where we narrowed it down to was looking at individual practices and saying the population of concern here is the population of 1,500 or 2,000 or 3,000 patients that that individual clinician or that that practice is managing and looking at them in a different way. Jesse Singer, from New York City yesterday, described it I think the best way that providers aren’t used to looking at managing a panel but they are used to managing whoever is in front of them at the time. This is a huge mindset.
The explicit goal of the project—we’re going to have a report coming out soon and before the report we’re going to have some thoughts and findings. I am sure some people have already twittered them. The real focus here was coming up with a set of functionality, minimum functionality that an individual practice would need to support management of that panel. Not just knowing who they have but also identifying cohorts, sending messages to them, tracking them over time, et cetera. Notice I didn’t say electronic medical record or electronic health record or HIE. We took a technology agnostic approach to this.
Some of the folks in this meeting were some of the same folks who testified at our hearing on Meaningful Use, folks like Terri Magini(?), Terri Cohen(?), Mickey Tripathi. We had folks like John Ciderov(?) and Sarah Corley. It was really a good representation and some of the key things that came out of it were, number one, to understand that the current state of practices is that they are neither technologically, nor culturally, nor incented to manage that population that they have. Getting them to do that—technology is a help but many that already have the technology aren’t doing that work for a variety of reasons that have nothing to do with bits and bytes.
One of the things that we got from the group is that trying to hold them to that in a short amount of time is going to be really tough. Mickey Tripathi talked about only two of—I forget what the n was but it was huge—practices were able to produce measures in a timely basis from their EHR’s. So what some of these folks talked about was an incremental, phased, logical progression of measurement, starting with measures that are tied to behavior. Both Jesse Singer in New York City and Mickey Tripathi in Massachusetts look at such simple measures—not necessarily simple to collect—as whether practices that have EHR’s are collecting information in a structured way for key things tied to quality. For example, immunization, are they putting them in the notes or are they putting them in a structured format? I forget the numbers from yesterday but there was a lot of discussion about data models and how add-on, unstructured, inconsistent data models lead not only to changes in measurement but also challenges in quality and safety. There was a study that found that allergies, for example, were stored in many, many different places in an electronic health record.
What also came out of this is that there are deep concerns about the certification process focused on products. We took this approach because—and actually did a bunch of interviews with practices doing this sort of thing to really, truly try to understand what folks are doing in innovative ways and how they are putting together different pieces of technology to accomplish goals. We had hoped to find folks who were not using technology, who were using paper and processes in really innovative ways to accomplish these things but we found that that just was not the case, which I thought was really profound.
Finally, the mini-factors outside of technology, structure and incentives we’re hearing, people a little bit chaffed at where I set the scope boundaries for the project. We didn’t include all the issues with health information exchange. We didn’t include public reporting and CDC sort of stuff. We were really just looking at population health, defining that practice and what they need to do to help multi co-morbid patients in that population or prevent disease. There was agreement at the end as we looked at this that because this isn’t just a technology issue, it’s tied to structure and incentives, that these sort of measures need to be part of Meaningful Use but that that really won’t be enough and that this really has to be understood and the transformation to this really has to be understood and tied to incentives and broader health reforms. Some of the folks involved were big advocates of looking at qualification criteria for medical homes and tying this sort of thing and these sort of measurements to qualification for medical home or similar health reform-type measures.
DR. CARR: Thanks, Matt, for that update. Let me just summarize, again, where we are for everybody and then let’s go forward. Again, we started with our charter from December, which says that our Quality Subcommittee will concentrate on the data needs of person-centered health, this set of personal, professional, and community activities and services focused on the health and wellbeing of individuals. Then the two things are emerging health data needs for a health system where the individual engages as a full participant and, secondly, understanding of health surveillance data set to support personal health tracking and improvements, which, with proper privacy protection and effective and efficient data collection strategies, can be used to assess the health status and needs of a population.
I introduced some discussion that Paul and I had because Paul couldn’t be here about the possibility of looking at meaningful measurement gaps and just holding a hearing to identify where those gaps are—so oncology, pain, functional status—what are those things? Blackford brought up about we are at a point in time where we are really ready for a Flexner II report that incorporates training and information technology. I think you heard that. Then Matt is talking about looking at individual practices and what is needed to get current practices technologically and culturally incentivized. And Bill—I didn’t identify you—oh, what does the coordinator do next and how the objectives of the charge are served by Meaningful Use requirements so the landscape has changed and definition implementation of Meaningful Use will identify how much change and we need to look at the next step.
I guess my concern is that we ground ourselves in the scope of work that is not being addressed by another group, number one. Number two, that we take on something that is not so technically or clinically detailed that it precludes participation by all the members of our committee. I would just like to put that back there. Also, we talked a lot about the role of the Committee and the Subcommittee, teeing up issues for others to work with. If, in fact, there is a group that is already working on this, it sounds great. We heard yesterday a lot of things that are coming forward in terms of these groups that will help groups get going. I am really looking for the things that align with Quality, align with our charter, that are timely, that are not a three year initiative because we certainly have seen how quickly things can change in six weeks let alone six months.
So I am inviting now some feedback on the things that we have heard. I think Bill was saying maybe we can’t choose anything right now because there is so much in flux. Paul’s sense was that NQF, they endorse measures and we need quality measures in EHR’s and we need to broaden the perspective and horizon on that, hear from groups that have done it well, and groups that have a need but nothing has come forward. Marjorie.
MS. GREENBERG: All this is very stimulating and a challenge. It was a challenge yesterday with Populations, too, to see which of all these many directions really are the best to go forward. I was looking back at my notes from yesterday, I guess it must have been when Jodie and John Glazer were on the phone, apparently the Standards Committee has set up three groups. One is on clinical standards, which I know include e-prescribing, which does seem to me an overlap with the actual legislative responsibility of this committee but that works out. The second was on quality and performance measures. That seems to be a direct overlap here. So I’m kind of thinking—in a sense, I guess this is being taped but we were here first. I don’t think—I know we tried with some success to work collaboratively with the AHIC Quality group. We always seem to be at, if not different areas, different world views or something. Matt’s agreeing.
This quality work group is probably going to work on the actual measures that should be part of the Meaningful Use, maybe, definition or something. The more I think about it I really don’t think we should be too tentative here. We had a quality workgroup for some time now we have a quality subcommittee. We have experts here in quality. I don’t know what experts in quality are on the Standards Committee. Paul Tang, of course, is on the Policy Committee, not the Standards Committee, though I assume the two of them will be working together. But we could spend a year trying to see what they are going to do and in the meantime we could be doing something, which, if nothing else, could be useful to the other group.
In the last 35 minutes I went from saying well, they are doing measurement to saying well this is a big area, forget it. I’m not being so much substantive as process here but I think if there is an area—I mean I can see with the Standards Subcommittee, they really didn’t feel they could go ahead with the hearings they were having without the support of ONC. I’m not saying we want to go against anybody but I think it really is for you to decide what do you think would be useful to this quality workgroup, if they are going to have some responsibilities—I haven’t read much about them or know much about them—and put it together. That is just where I am.
DR. CARR: In response to that, the fact that Paul is on these committees and seeing what they are focusing on I think gave me—I felt that he was able to speak clearly as to what the needs are. My understanding is that there is a lot about measurement but sort of taking the 15,000 foot view and saying, yes, we’re measuring but we are measuring process not outcomes. Secondly, we are measuring where the street light is shining in the registries of American College of Cardiology, STS for thoracic surgeons, and oncology, which is a major health issue. We don’t have any measures for that and we have a lot of people doing it and they are struggling a great deal—you know, pain, functional status, all of that.
So, again, as I am looking at the tempo of things, I think that we have to decide are we going to hustle to try to keep up with what is being created or do we keep the role of reflecting on don’t forget about this or here is a gap that needs to be thought about. As an example, I think when we did secondary use or whatever we called it—enhanced protection, we stumbled onto de-identified data and we began to try to be prescriptive about that and then we said, no, it is enough for us to say that de-identified data needs greater attention and then the Center for Democracy and Technology took it and pulled together a very elite group to really nail it down and are coming out with a great report. I think that was a great synergy. We didn’t have the expertise and we didn’t have the tempo that could immediately go back and deal with that. I am just conscious of us making a contribution that synchronizes and synergizes with the work that is being done by the more granular initiatives. I think it is important to stay out of the detail.
MS. GREENBERG: I think, actually, you helped me crystallize my thinking here more that they are going to have to—this standards quality group—is going to have to deal with what can be done now because there are actually going to be quality measures that will be part of Meaningful Use definition for allocation of funds, et cetera. So they are going to have to be pretty much centered in the here and now and I think that this subcommittee can definitely be more focused on, as you said, gaps and where do we need to go.
DR. CARR: Just, again, I’m giving Paul a voice at this meeting, the kinds of folks he thought would add value to talk to would be ACC, STS for the cardiac/cardiac surgery things but NCQA, ADA, the Diabetic Association, also CMS, PQRI—what have we learned from that—AMA. What are the groups that have already done it? Then thinking about what are the gaps, the areas that are important areas that we haven’t talked about. I’m obviously pushing that agenda but I’m opening it back to all of you to respond.
DR. MIDDLETON: Justine, I think that sounds great. I think the tendency for us always is kind of the near-term agenda versus what is the overarching long range agenda. I hope we can keep sight of both because those are two distinct things. I think the ONC committees and that work is going to be necessarily near-term and we can reflect on that but also I think shouldn’t shy from saying here’s what you are not thinking about and what needs to be considered and needs to be part of the process in a very direct and useful way, collaboratively, obviously.
I did want to return, though, just to four points that you didn’t recapitulate from my earlier comment so I am just concerned that they may not have registered. The key thing, I think, for obviously the purview of NCVHS is data, statistics, national vision, and perspective. This idea of measurement in IT, who actually is going to take that up? Is that something we can learn about through a variety of discussions and make some recommendations? It’s all about quality. It’s all about data. If you don’t know how the IT is doing that, the data are always suspect. We really need to know how to guide folks in generating those data in a way that is replicable and verifiable, et cetera.
The second thing is this notion of implementation science, related to this sort of measurement of IT. The implementation science part is really freaking me out. We are going to throw a lot of money at a lot of people to implement a lot of technology and I can tell you from personal experience that everyone implements EMR differently, the results are dramatically different, and therefore, the value that arises in the quality of the data that comes from those environments and settings is all over the map. We would not tolerate this in any other field or industry. How do we sort of get at an implementation science, which supports, of course, the goals and objectives of transforming care with IT and, frankly, helps us figure out how to do it?
MS. GREENBERG: When Paul said to Chuck where are these people going to come from, he answered a different question. I think when we—I mean he said something about the workforce thing but you can’t be building the workforce for something you are going to fund in the next six months or something. I think where it might break down with the extension service analogy is—even though I grew up in Fargo, North Dakota, I don’t know that much about agriculture but I think the science of agriculture was better known at the time that they set up the extension service and then they continued to learn everything than the science, as you said, of HIT implementation and health is known now. So it’s kind of scary.
DR. CARR: Maybe we are talking about different segments. What I hear Blackford saying—hopefully, I will get it right this time—is that there is a continuum and we spend a lot of time talking about the EHR and that’s a kind of tool but to get Meaningful Use out of the EHR, we need to, first of all, have specific clinical information that can come out of it that can then be acted upon to improve health. What Blackford is saying is, yes, it is true you have to have the right measures but if you don’t implement those measure—if where you put them, structured or unstructured field, is not decided in advance, you can’t get that information out. You can’t trend it and aggregate it.
So the question is there is one piece about the infrastructure—the building it. Again, I go back to what we heard at Meaningful Use where the leapfrog group took a whole array of certified EHR’s and found completely different implementation in terms of quality of what they should expect. So one thing is that you have it and how do you implement it. What I think I’m hearing though is that we have a whole initiative about that as part of this with these enterprise groups or whatever they’re calling the transition groups—
DR. MIDDLETON: The Enterprise Information and Integration Research Centers, I think, will get at some of this but I’m afraid they won’t focus in a way that will be necessary on the implementation issues. It’s kind of like everything to the left or everything behind turning on the EMR. How is it designed? What does the data look like? How is it implemented? What is the training? Going 50 miles in a Cadillac versus a VW is a different experience for some people. How do we actually know that?
DR. CARR: I guess what I am going to try to do here is say all of these different things and where they are on the continuum and then we will come back to it to say is someone else doing it, is there an audience, do we have the expertise to be able to do it. Larry, we haven’t heard from you and then I guess we’ll just go around the table.
DR. GREEN: Thank you. I resonate with what everyone is saying. I would like to make two comments, actually. One is I want to go back to the measurement gap idea. I actually think that is a very good, practical conceptualization and way to go. I would like to advocate for us building on what we have already learned and heard from the hearings and work done in the last couple of years.
I can hear Barbara Starfield at the hearing saying we do not know how to measure comprehensiveness and we do not know how to measure patient experience. I have not heard anyone yet, this morning, refute those claims. To be unable to measure comprehensiveness is to deny health care reform’s promise of—you have got to get the comprehensiveness in order to achieve the promise of health care reform. So how do you measure that is a very big deal and I am unaware of anyone taking that on. It could translate immediately and directly into the work that the National Quality Forum and NCQA. A second measure that she mentioned is patient experience. I am less confident of that but that is what we heard her say and she is pretty expert about this. A third one that has been published since our last meeting in a couple of places is that we lack measures of adaptive capacity for change. To get to the quality agenda, to do what Blackford is advocating for with some discipline and dedication here is underlying this is the capacity to adapt. Adaptive capacity of systems is what I’m really talking about. Systems are made up of people, of course—so, anyway, I would like to support the notion that this committee might focus its work on measurement gaps. Measurement gaps that I would nominate for consideration are a measure of comprehensiveness, a measure of patient experience, and a measure of adaptive capacity.
There are two other trains running that present great opportunities to measure and improve quality that I have not heard anyone bring up in this meeting. I heard it yesterday afternoon in our Population Subcommittee meeting and I want to bring it up here also. The first one is maintenance of certification for physicians. I want to be quick to say that it is not my intention to be neglectful of all the team members and all the players that are important to getting quality health care and quality outcomes but we are witnessing the implementation of maintenance of certification as a bet that it is possible for physicians all over the country in all fields, all disciplines, to actually actively improve their practice and their care continuously, year by year.
What this does is bring the leverage of being a board certified specialist to the table to insist that practice by practice, physician by physician, year by year they measure something, do something to improve teat measure and re-measure it. This is an understandable concept. It is called Part 4. ABMS has adopted it. All the boards, every specialty, all 24 of them and all the subspecialties are under way in implementing it. It is being done on line. The standard bearer boards for this are internal medicine, family medicine and pediatrics. It is a long way down the pike and how maintenance of certification can improve quality is not really being examined externally anyway. I repeat, it is sort of a bet.
MS. GREENBERG: What was the word before certification?
DR. GREEN: Maintenance of certification. What this means is if you want to say you are a board certified plastic surgeon or a board certified orthopedist now you don’t get certified for life any more. You have to re-earn that certification repeatedly and to do so one of the requirements, and the shorthand for this is called Part 4 Maintenance of Certification, is you actually have to measure something in your practice, then you have to implement something to improve that and then re-measure it.
We are talking about 650,000 physicians participating in this and they are getting under way right now and the big three boards who are the standard bearers, they have the doctors that will induct a slight majority of every transaction with a patient today. This is a very large opportunity to improve quality.
The second on I would call attention to I am going to call the Sondik statistical development program. I was stunned at our Population Committee meeting yesterday when finally — I was very slow on this — it just never occurred to me that the National center for Health Statistics did not have its own robust unit constantly figuring out and striving to improve health statistics and their measurement. Overnight this has been haunting me. This idea now won’t leave me alone. It has sort of possessed me since six o’clock last night.
MS. GREENBERG: We have the unit. We just don’t have any money. We have an Office of Methodology but we have no money.
DR. GREEN: What I mean to say is this is not working. This is not in place. This is not doing what it should do particularly right now when things are fluid and stuff. To ask and answer the questions about how will we do health statistics in the future now — we have only talked at our last three meetings about what are we going to do about the 21st century vision. This is all under way.
This committee, it seems to me, could be a galvanizing focus on how we improve the quality of health statistics by using the new sources that are driven off of all these issues of certifying what and these new types of data or this new workforce to use and that sort of stuff. But where is that nidus of leadership for distilling information and useful measures in health statistics for the population and sub — we decided yesterday not to call them sub-populations. It is all the populations. Everyone in the population, you know?
Could we not make a contribution there in figuring out where the new statistics will come from, how they will be derived, how their errors will be estimated? What comes next in the evolution of health statistics? You would think that with a total remake of the way health care is practiced that there would have to be a remake of the way health statistics are done. And that is our territory.
To the extent that measuring that is crucial to the quality agenda, I don’t know anyone better positioned than the NCVHS to champion that.
DR. CARR: I am going to respond in one way, which is I agree with you on the importance of that. But as we think about the work for this committee, what you are describing is very granular — like what is the best way, how to do it and all that — which might exceed the skill set of the committee. On the other hand bringing this forward for the full committee to say that there needs to be a look at this involving the key stakeholders is I think something that we ought to mention.
DR. GREEN: Again, like Blackford, I really want to support first and foremost the notion of focusing on measurement gaps. I think that is really something we can move on forcefully and usefully and in a timely way. Then also, like Blackford, there are these larger issues that are floating by and I would like to flag those, too — maintenance of certification and advocacy for the National Center for Health Statistics in some way, to enable its existing office to be activated in a much more robust way. There should be a fellowship program here that brings people from all over the world in. They should be from the IT world, they should be form the mathematics world. This is a hard problem.
DR. CARR: I’d like this very important idea to come forward when we do our summary today and I will ask you to put this forward because I think it is beyond the scope of this subcommittee, but it is an important thing for the full committee.
Let me just recap. If we think of the continuum there is care delivery and then we have heard that it is changing, so a second thing is the model is changing, and there is concern about the competence of the physician, in part as indicated by the measurement, by their certification. So there is a measurement about physician competence.
There is a measurement about care delivery. So as the care delivery changes information technology is a part of it and in some cases it might be claims and in some cases it might be EHR, but the goal is better health for persons — acute treatment and preventive treatment. We are talking about all of these things. So I am putting it out there because at the end we are going to have to pick out where in the continuum we think our efforts will have the greatest impact.
I know there were a number of other hands. Matt, go ahead.
MR. QUINN: I just wanted to follow up a little bit form the experience at yesterday’s project on this, this AHRQ population health project, which I really define as proactive management of populations of patients. Again, Jesse Singer, from New York City, to quote him, said there aren’t enough enterprise customers. And somebody said you mean like health systems? And eh said, no, like New York City.
When you look at eth issue of implementing health IT and the implementation science of implementing health IT in an individual clinician’s office, and the capacity for change and the capacity for measurement, and the burden on that and the burden financially, and the way we have aligned incentives, and if you say no, we would rather the customer be New York City or Wichita, Kansas or Dayton, Ohio, it is a very, very different thing. And it also directly links to what we can measure about that implementation about the health of that community et cetera.
It’s a whole different mind set and the way that incentives are laid out today and the way that we are thinking about meaningful use is highly dependent on the smallest cog in the wheel. I read in Health Beat that CMS is planning to spend almost a billion dollars on being ready for ARRA and I don’t know if it is just the high tech provisions, and I don’t have the details of that, but I would hope that there is a relay robust evaluation plan, not just on are we tracking enough and who has implemented what? Really getting down into the weeds of understanding transformation at the practice level, transformation at the community level, and understanding what works and what doesn’t work as part of this real national transformation of care. Understanding those metrics and who is going to do it and all of that I think is vitally important.
DR. CARR: So we have care delivery system by competent clinicians. That is what we think we have. Then we have health information technology, which is here to facilitate batter care. WE also have HIT that facilitates the assessment of the care delivered. Critical to these functions are a proper implementation for one, and secondly the measures of care delivery and outcomes and measures of physician competence and maintenance of certification. Is this starting to sound like the areas we are talking about?
DR. W. SCANLON: I think it is more about the capacity of the IT than it is about the state of measurement today. I would almost want to stipulate that we have measurement gaps that are huge. And since you used the word strenuously, strenuously is not our job to work to fill those gaps. We can remind people of them, but when we sit here with the resources we have and I think of NQF and the resources that have been appropriated to them — Justine, you said earlier their job is in part to approve measures — I think a big part of their job should be to survey the landscape and to figure out where the gaps are and to be aggressively and proactively working to get those gaps filled by defining measures.
You talked about PQRI and STS and ACC — in my other life I spent a lot of time with those and it is not a pretty picture. So I am thinking of this hearing and how depressing it can be when you are done and where are you going to go? I guess you have also made reference to the expertise on this committee and the question is are we the right body to take it the next step after you get this sort of dismal status report. So I think that is a part of this.
I could see the investment for ARRA and the meaningful use, starting out in terms of trying to affect clinical care, the interaction between the providers and the patient, and not really pay necessarily enough attention to this whole issue of measurement. I have not sat down and seen this for myself, but I have heard many times about the VA system and how beneficial it is for the VA to have en electronic health record and then when they wanted to do quality measures they got pencil and paper out and they started abstracting. That is what we don’t want. That is not the measurement in IT that Blackford has been talking about. What we want is a system, and Matt made reference to it, too, that can actually produce data and not just data with respect to current measures. It is the issue of building blocks for the future measures. Are we moving in the direction that it’s the building blocks that are going to come out of these systems so that new measures can also be created.
We had a hearing two years ago where one of the themes was how hard it was for providers to deal with the multiple forms of measures that exist today and sort of how labor intensive that was. We have got to think about a more efficient way of gathering quality information, and it is not just standardizing current measures. It is anticipating that there are going to be new measures that haven’t been thought of but they are built off of the same kinds of clinical information. This is kind of where my experience is totally missing.
We had a discussion at that hearing a couple of years ago about the value of having lab information. I think I remember it was a discussion that we couldn’t do ejection fraction in a record because there wasn’t an agreement on it. Do I have the right example in terms of — and there is this whole issue and the functional status that we discussed yesterday. There are issues about, okay, if we would like to have it in a medical record that is easily retrievable it has to be somehow decodable. Where we are on that, which is to me sort of a critical part of the building block idea, is a real issue as well as sort of what the measures are going to be.
I think it is also something where we may be the right people to do this. NCVHS — what is the history of coding sets, minimum data sets, et cetera — it is the kind of thing where we would love it to be there, but it is not necessarily there. So this is a kind of tool that I think we should, a too gap, that we could potentially focus on with our resources and backgrounds.
DR. HORNBROOK: One of the things I am not hearing here is the cultural revolutions that occur inside physicians’ practices when they are faced with adopting and implementing an EMR. To oversimplify the Kaiser Permanente experience, the first revolution was what me, keyboard? The first behavioral barrier was age-related, so there was a wave of retirement of Permanente physicians because they didn’t want to do keyboarding because the EMR required them to do direct data entry, for things that for a long time had been done by clerks.
The second revolution is probably learning how to code for E for service standards. Our physicians were used to doing minimally adequate coding, good enough to get by in a system where you relay didn’t care that much about the pennies and small dollars of rounding errors. But Medicare does care and eventually when you lose money you care about making your billing, maximizing your billings. So the next thing is teaching all of our physicians to become fee-for-service billers, which is really the art of coding. And since they are doing the coding at the point of service they have to be taught how to pick the right diagnoses, how to pick the right procedures. So built into the EMR is a set of coaching materials for coding accuracy and coding appropriateness.
The third revolution is even more important and the most important one, which I think has got to be here, and that is the notion of physicians realizing they have a whole new tool to look at the practice of medicine from an entirely different perspective. And that is managing the collectivity of their work. The desktop that takes care of all of their patients at once, a panel management tool that allows them to take care of patients even when there is no data. So it is not just having the data there, it is also the absence of data that can be extremely critical for continuity of care, for follow-up, and also of patients going to the wrong place. The people get the data in there but they are going the wrong place. They went to the ER instead of doing a medical office follow-up visit when they were supposed to.
So what they are beginning to deal with then is getting control over their terms of practice by saying I have only so many resources, capital, time and otherwise, to give to my patients. Where am I going to deliver the most priority, and that has to do with every day but also things that pop up when they age to the point of criticality and you can’t let them just go along because the patient isn’t paying attention to it.
So that then means the physician is then designing part of their practice and this is the physician generically, physicians in collective are designing protocols that reach out to remind patients in an escalated fashion. So some patients only need one phone call or one e-mail. Other patients need a very strong coaching approach, and other patients you give up on.
But it is all about managing a set of resources against a set of competing priorities. And I am beginning to realize that the EMR is a tool to help them gain control over that rather than simply burning out and giving up.
DR. CARR: I am trying to sort of put all this together. So there is a care delivery system by competent physicians. That is where we began. Then we got HIT as a facilitator of care delivery, HIT as a facilitator of measurement of care, but below we have implementation of HIT is critical, redesign of work flow and culture is critical, which is what you were just saying.
DR. HORNBROOK: Right and if you want to use the old word, HIT to manage access. So you are going to prune off the patients who are overusing and you want to go out and reach out to the patients who are under-using. So it is in reach and then rationing — don’t quote me on rationing.
DR. CARR: Okay, so as we think about the areas, clearly understanding this as how does it facilitate care delivery. Understanding how it facilitates measurement. Then that measurement is – what is the measurement? What are the measures that we are capturing? Are they outcomes – are they single outcomes or are they continuum?
DR. HORNBROOK: In the health services world we have been very creative at creating a continuum of outcome measures based on available data. So you have some downstream process measures that are actually intermediate outcomes like; return to hospital too soon after discharge from the ER of the hospital. People being admitted in late stages of disease from ambulatory care when they should have been taken care of at an earlier stage when they got to the hospital or not gone to the hospital at all.
Then eventually you want some overall societal weighted health status which means that somebody has got to build into the HIT system a questionnaire that patients fill out on a regular basis.
DR. CARR: So it gets us over here to our goal is to improve care for better health for persons – for individuals. It is not acute care, it is not about patients, we talked about we wanted to be person centric. So these things facilitate care delivery and then this is care delivery we can measure to continue to improve it.
The things we have talked about today, I think we are all in these different boxes. I was up here saying measures of care, delivery, and outcomes are important so that if we have measured the right thing we can make an action. We have heard about the culture, what you were saying, that the implementation of HIT is critical – which is what Blackford is saying, and Mark was saying that the culture and the workflow.
MR. QUINN: I was just going to highlight that one of the shifts that we have seen is that facilitator of care delivery – the paradigm is changing sort of from health IT as a tool to make the hamster wheel go faster. So before health IT you could see 20 a day, and now with health IT you can see 30.
DR. CARR: In general, we are safer, we have fewer transcription errors.
DR. GREEN: It slows you down so much.
DR. CARR: Let’s not get into the details.
MR. QUINN: It is moving the paradigm from who is in front of me, whoever they may be who shows up, to managing that population of patients for the population.
DR. CARR: Okay, yes, right, I think population –
DR. GREEN: Justine, I like your diagram a lot and I think it is helping us a good deal here. That box under implementation of HIT is critical.
I just want to raise again perhaps for the groups rejection, that is implementation of the new statistics. It is new denominators, it is new biases, it is new errors, and it is new math.
DR. CARR: Right, new data – numerators, denominators.
DR. GREEN: To get to your goal here, how are you going to know whether you improve care and the health of persons or not? How are we going to get an answer to that is got to morph and change along with all of this. That is where NCHS really belongs.
DR. CARR: Move set of questions.
DR. GREEN: And new opportunities to get the answers to the questions that never existed before. This is what Blackford is conveying to me. We have to figure out how to seize these opportunities.
DR. CARR: Marjorie.
MS. GREENBERG: I would say improve care for better health for persons and populations.
DR. CARR: Right, good point.
MS. GREENBERG: I think Ed Sondik had something he wanted to say.
DR. SONDIK: If I were doing this I would start on the right and think about quality with the respect to the state of the population in the patient. So for example, prevention doesn’t show up in this in any obvious way. And prevention – it is sort of hard to think about prevention when you talk about care because you think about an intervention. So where is the interventions related to prevention, and yet there certainly are things that I think, we could talk about in terms of quality of prevention that is going on – as simple as the number of patients that have their blood pressure under control for example, which just doesn’t pop out for me from this kind of diagram as to where that would be.
I guess I kind of resonate what Larry was saying and what Bill was saying in terms of thinking about the state of the patient and what is happening to the patient and the quality of the interventions and the interactions. Then the question is – I like the idea on gaps – what don’t we know and then how do we go about finding that?
HIT is one way, electronic medical records is one way, but there may be other ways as well, in terms of particularly if you think about evaluating health reform, there may be other ways that need to be built into the “system”, system in quote, to give us this information. I see this as very HIT or EMR. That is how I am reading the HIT – EMR centric. I would think that with the mandate of this committee, it ought to be broader than that. It ought to be more information. And HIT or EMR, is one set of tools that will be used for this.
The other thing is, Justine, when you were going down that list, it struck me that it would be a real value to somehow be able to – and that is what you are doing here – but to organize that list in such a way so that you can see what depends on what. Some things related to for example, to the quality of the EMR implementation, I resonate to the point that is that the role of this committee? But identifying a gap – but why is it a gap? Well, it is a gap in terms of a problem or a gap later on in terms of our measurements of things about a patient, about the public, about the population. Do you see what I am trying to say?
DR. CARR: Yes, so I am putting it in these sort of categories so that we can begin to say what domain do we want to be in. All of these I think are relevant but I think we need to have a discipline about on the universal things where can this committee, with this expertise, add the greatest value synchronizing with other initiatives.
Mike Fitzmaurice, I know you had your hand up.
DR. FITZMAURICE: Some of this might have been said before but I have not said it, if you look at what happened, ARRAs come along, and now you have the Office of the National Coordinator very focused and has a timeline. The focus is on meaningful use, it is on getting the standards to support meaningful use and the certification criteria to support meaningful use, getting electronic health records certified and then a focus on implementation or adoption in 2011, 2013 and 2015. There are two FACA committees to advise them on that.
Generally NCVHS has a broader perspective with regard to the health system and with the regard to the health of the population. We looked at patient outcomes. We looked at measurement. We looked at quality measures.
I pick up on Ed’s comment, I would look at that chart and say does it make sense for us to do it if you take out HIT? If you take out HIT you reduce some of the overlap. The agenda of NCVHS is going to be negotiated with Harry, Marjorie, and Jim Scanlon with David Blumenthal because all of us are going to have to face questions of what is the overlap? Are you infringing on somebody else’s territory so you need to have agreement at the top.
What are some of the things that we could be doing? Well, look at who our constituents are – they are certainly CDC, AHRQ, CMS, other agencies. We have some expertise in economic cost analysis methods. While other people are focusing on comparative effectiveness, and we can focus on comparative effectiveness too, we can tackle some of the methodology problems in cost analysis. Yet, there is a risk involved but who else is doing it?
We can look at health reform. What are the incentives for change and what do we know about the incentives for change? We can look at quality measurement as it relates to value measurement. What are we getting with the resources that we are using and how can quality measures guide us in the right direction?
We can look at international lessons learned. What can we learn from other countries, particularly with regard to their use of quality measures and the responses of their health providers with information about those quality measures? I don’t know an awfully lot about that but it would be important to learn it.
I like Bill’s comment – in fact I haven’t heard any comments that I don’t like – but survey the landscape. What are the gaps in quality measurement? Different organizations have their own focus, their own mission. Whose mission is it to look at the gaps out there? Except for a market analysis, what can we provide to fill the gaps? Nobody is looking at the overall gaps in quality measurement and saying, somebody ought to do this. That could be our job as well.
I think there is opportunity and that we should not be so focused on health information technology, but focused on does it make sense without information technology and then as you add health information technology into it, we have something additional that can help advise David Blumenthal.
Or is there something that their FACA committees could be doing and we could give them suggestions on what they might look at?
DR. CARR: Thank you, Mike. I got back to what we said before, you are right, measures of quality of care are – we don’t even have measures – leave out EHR or any other thing. There are many, many other areas where we don’t have proper outcome measures of quality of care.
Again, I go back to oncology, it happens to be something that I am working on with somebody else. It is a real issue.
Okay, our time has drawn short. We have not come to closure because we are exploring this. Again, I go back to our charter in terms of the things that we committed to. So while I think Mike, many of the things you mentioned are absolutely worthy, I think they are outside of the scope of what we committed to in December – a little bit outside.
I think what we are going to have to do is conclude for today. I do believe that over the next week a number of things will be coming out in terms of directions of ARRA and so on. My guess would be that the best for us would be to plan a call in two or three weeks and get a sense whether we have a focus that we believe is timely, relevant, non-redundant, that we would want to anticipate having a hearing about in September.
Were there any other comments? Thank you everybody, for a very thoughtful, robust, and insightful comments. To be continued. We are adjourned.
(Whereupon, the subcommittee adjourned at 9:55 a.m.)