[This Transcript is Unedited]
Marriott Washington Hotel
1221 22nd St., NW
Washington, DC
Agenda Item: Call to Order, Welcome, Review of Agenda
DR. CARR: Good morning everyone and welcome to NCVHS Full Committee meeting. I will start by saying I am Justine Carr, Chair of the committee from Caritas Christi, Steward Health Care, no conflicts.
MS. GREENBERG: Good morning. I am Marjorie Greenberg from the National Center for Health Statistics, CDC and Executive Secretary to the committee.
DR. FRANCIS: I am Leslie Francis in law and philosophy at the University of Utah. I am a member of the full committee and I have no conflicts.
DR. WARREN: Judy Warren, University of Kansas School of Nursing, member of the committee, no conflicts.
DR. SCANLON: Bill Scanlon, National Health Policy Forum, member of the committee, no conflicts.
DR. GREEN: Larry Green, University of Colorado, member of the committee, no conflicts.
DR. FRIEDMAN: Charles Friedman, liaison to the committee, from the Office of the National Coordinator for Health Information Technology.
DR. EDWARDS: Brenda Edwards representing NIH and sitting in for Dr. Bob Croyle. I have no conflicts.
MS. GREENBERG: Thank you. Welcome. Thank you for coming today, Brenda. Only members have to declare conflicts. Though we staff are always conflicted.
MR. PARK: I am Todd Park, the Chief Technology Officer of HHS and happy to be here this morning.
MS. PRITTS: I am Joy Pritts. I am the Chief Privacy Officer, ONC, HHS.
MR. LAND: Garland Land, National Association for Public Health Statistics and Information Systems, member of the committee, no conflicts.
DR. OVERHAGE: Marc Overhage, Regenstrief Institute and Indiana Health Information Exchange, member of the committee, no conflicts.
DR. STEINWACHS: Don Steinwachs, Johns Hopkins University, member of the committee, no conflicts.
MS. MILAM: Sallie Milam, West Virginia Health Care Authority, member of the committee, no conflicts.
DR. MIDDLETON: Good morning, Blackford Middleton, Partners Healthcare System, Harvard Medical School, member of the committee, no conflicts.
MS. VIOLA: Allison Viola, American Health Information Management Association.
MS. SQUIRE: Marietta Squire, NCHS, staff to the committee.
MR. BIZARRO: Tom Bizzaro, First DataBank.
MS. KANAAN: Susan Kanaan, writer for the committee.
MS. JACKSON: Debbie Jackson, NCHS staff.
MS. SAVEKAS(?): Mauris Savekas, AMA.
MS. JAMISON: Missy Jamison, NCHS staff.
DR. BICKFORD: Carol Bickford, American Nurses Association.
MS. JONES: Katherine Jones, NCHS and staff to the committee.
DR. SUAREZ: I am Walter Suarez, member of the committee and no conflicts.
DR. CARR: Before we get to our agenda today, I want to take a few moments to say a few words about Cynthia Sydney. The last time we were here Cynthia was sitting at the table with everyone else and doing all the last minute things that we invariably end up needing to do with a smile on her face and just a wonderful, collaborative, generous spirit and as most of you know Cynthia died very suddenly, a couple of weeks ago. I for one -- I think all of us feel her absence very much today. We worked very closely in the Quality Subcommittee and every time I got off the phone I marveled at her patience because the things that we would ask for at the last minute, the meetings to be set up that were impossible, the materials to pull together that were scattered about and she would -- I know she would probably be rolling her eyes on the other end of the phone, but nevertheless in her voice was her usual congenial, collaborative lovely way. I think we were all enriched by her role with us and her spirit and her generosity and we miss her very much. We have sent our thoughts and prayers to her family.
MS. GREENBERG: If I could just thank the committee for your outpouring of sympathy and condolences and heartfelt comments that we received after Cynthia died and we did share those with the family and I know that it meant a lot to them.
Another thing that I just wanted to let you know was I think her last job for the committee or at least visible one, was staffing the first day of the quality subcommittee and then she wasn't feeling well and Katherine Jones suggested that she go home and she was very reluctantly to do so, but she told Katherine she really loved working with this committee. I wanted you to know that.
Agenda Item: Updates from the Department
DR. CARR: This morning we have a number of guests. Joy is going to give us an update on privacy issues from ONC and then Todd Park has joined us this morning and thank you so much. We are very excited that you are here and everyone here did their homework I know and view the community data initiative video and also some did extra credit work by viewing the October awards ceremony. We are excited to hear about that, but I think we will begin with Joy.
Agenda Item: Privacy, Joy Pritts, ONC
MS. PRITTS: Good morning. My report this morning is going to be fairly short. It is an update of where we were from the last time. And I will primarily address the issues that I call the outward phase in privacy and security issues as opposed to what we are doing programmatically because I think that is what is of most interest to this committee.
We have received a number of recommendations from the HIT policy committee which we continue to process within ONC. Our goal is to have some of those recommendations. We are trying to address them as a package not an isolation so that they make some coherent sense and I think our target date for having some resolution on those is probably for the whole package and moving forward is probably early next year.
We continue to get the policy committee continues to work on privacy and security issues particularly at the Tiger Team level which then sends its issues of course up, its recommendations up for the full committee to forward to us.
They currently have addressed recently transparency which is basically how information is presented to consumers about how their information is shared and whether that is adequate under the existing systems.
We continue to have what I would call somewhat of a little bit of an issue of identifying that we have a lot of protections in place now and that we are trying to focus on people on where the gaps are and what needs to be improved, not reinventing the entire wheel. I think this group has been pretty good about being brought back to those basic concepts.
In addition they are addressing provider identification. They were asked to address this primarily from some of our programs and from what I would call an organizational point of view. By that I mean if you are having one computer talk to another computer, what do you need to have to verify that the systems who are talking to each other are indeed who they say that they are. They have not drilled down to the level of a provider authentication within a health care system although I believe that they will be asked to do that by some of our groups.
Next, December 9, which is coming up quite quickly now, is the Tiger Team is having a public hearing on what they call patient matching. And what that is is the accurate way to match information with the correct patient. This is sometimes called patient ID, but that term can be a little controversial. They used a longer term which describes what they are interested in doing which is making sure that the right patient -- the right information is attached with the right patient. That is pretty much what we are doing with the Tiger Team.
And we continue to work with the cross agency, a task force on HIT. In particular there is a cyber security work group there and one of the focuses of that group has been on trying to find ways of shifting the burden of security from the end user, the provider farther up the chain so that when a provider logs onto a system, they aren't the ones who have to do all of the security work. It is something that we expect that as people who use computers on a daily basis and who do not update them on a daily basis, we suspect that this is going to be an area that is going to be problematic for providers and it is one that is getting quite a bit of attention right now.
We also continue to work with the other programs in HHS in particular some of the health care reform efforts that will involve a collection and use of a large amount of data such as the health insurance exchanges and the enrollment for work group.
And that is where we are right now. Thank you.
DR. CARR: Joy, thank you. Very helpful. One question. Can you say a little bit about the interaction between the state's health information exchange initiatives and ONC? In other words, as they are developing things, are they checking their approaches with recommendations coming out of the policy group and the Tiger Team?
MS. PRITTS: The cooperative agreement is fairly broadly written with the state HIE programs and does require them in their plans to address privacy and security in accordance with the framework which was set by ONC a number of years ago in accordance with law of course and those plans are being evaluated within our office. We are having an all grantee meeting soon where we have a number of sessions set with all of our grantees to discuss privacy and security approaches.
I think it is quite clear to everybody that we are behind on this issue and have been for a number of years on privacy and security. It is an issue that we have talked a lot about over time, but have not reached conclusion about because we really didn't have to in the past. And with ARRA it has become clear that these are decisions that now have to be made.
DR. CARR: Is there the sense that they will ultimately be convergence that models that were developed best practices will be able to be shared kind of thing?
MS. PRITTS: They are in the process right now of forming community of practices for best practices in privacy and security and in a lot of other areas. We are doing that in all of our grant programs.
I have been involved in a number of community practice organizations and sometimes the challenge there is making sure you have enough people participate who have actually been through the effort so that they can share their wisdom with the others. It is a little difficult in this area because there is limited expertise and what we have is spread fairly thin among a lot of these different projects.
DR. CARR: Other questions for Joy? We appreciate you taking the time to come by. Thanks very much Joy and thanks for all you are doing.
Todd, welcome.
Agenda Item: Community Health Data Initiative - Todd Park
MR. PARK: Good morning. It is a great pleasure to be here today. And actually hopefully the less I talk and the more you talk the better. We love for it to be a very interactive conversation. How much time do we have this morning?
DR. CARR: We have a lot of time if you want to use a lot of time. Chuck is going to give us an update, but -- 10:15.
MR. PARK: Wow. Fantastic. Very wonderful. All right. What I thought I would do -- I am a little under the weather so I apologize for the raspiness in my voice. I would love to take you through an overview of the Community Health Data Initiative and its current status and then open it up for discussion. Would that make sense and would actually love to answer questions along the way as well. Don't wait until the end if you have a question and of the presentation.
Is everyone here familiar with our open government plan, the Federal Government's Open Government Initiative. For those of you who aren't it was one of the first things the President did after being inaugurated on literally his first day where he issued an order, directed the Federal Government to become more transparent, participatory, and collaborative. Transparent meaning published data that helps the public hold the government accountable and generates social benefit. Participatory and collaborative meaning basically have the government engage much more actively with the world outside government to help take on the challenges the government is helping to take on.
What actually happened with the Open Government directive is unlike previous pronouncements that were similar in the past around reinventing government or changing government, et cetera. It is something that the government is actually taking incredibly seriously at multiple levels with a whole bunch of action happening. It has largely been outside the public eye maybe because it is just too boring and too prosaic. But for those of us in government it is incredibly exciting because it really does represent in a lot of agencies a real transformation how government thinks about itself.
One artifact of this progress is Open Government plans that every cabinet agency actually published April 7 of this year and has subsequently revised. Ours is actually found at HHS.gov/open. I would highly encourage you to check it out. It is really wonderful and very concrete set of steps that HHS is committed to taking to help make HHS more transparent and participatory and collaborative. It has the full backing of the secretary and deputy secretary and it is wonderful to actually have that level of engagement and also the White House as well.
One of the things the secretary says is that if you think about the things that HHS really wants to help do like help transform the health care system, help improve public health, help improve food safety, help combat pandemics and so forth that really the optimal way to do that is through publishing data and collaborating very actively with the outside world. If you don' do that then you are really not going to do those core missions well. Really open government to us is not just another thing that we are doing. It is actually the mode which we really want to do and are doing all the things that we need to do.
A flagship initiative in the open government plan is something that we are calling the Community Health Data Initiative which is really the brain child of Bill Corr, the deputy secretary, and I have been working very closely with him on how to actually craft it and make it roll forward.
The whole idea is actually to try to leverage the power of HHS' data in every more effective ways and efficient ways to help improve health in the United States. This committee knows better than anybody on the planet the kinds of data that HHS had, just an unbelievable array of data of all manner of topics related to health.
A traditional way of leveraging government data would be for the government to basically build a giant website, publish say a Mac with statistics on communities and that is actually a really valuable thing to do. But actually we thought that an even broader and deeper and more powerful play would be to actually take a page from the weather and specifically the play book that NOAA has been running, the National Oceanic and Atmospheric Administration, with weather data.
I think as everybody here knows NOAA supplies virtually all weather data to America which actually I didn't know until recently. And what they do is they take this data and they publish online increasingly in XML format, but downloadable, machine readable, free of charge and without -- constraint in continuous supply. And then a dizzying array of innovators actually inside and outside the government, take that data and turn it into an incredible array of uses and products and services inside that benefit the American people, everybody from weather researchers to weather.com to the weather channel to iPhone weather app developers to nightly weather newscasts and the sheet in your hotel room that says it is going to rain tomorrow. This is NOAA data manifesting itself in lots of different places helping people make better decisions and helping to improve the welfare of the public.
We said well what if actually instead of HHS thinking that we have all the data and let's just build all the apps of the data ourselves which HHS isn't necessarily good at and certainly can't do all of, why don't we actually make our data much more accessible, much more available like NOAA and actually help catalyze the emergence of an ecosystem a much broader, much more active ecosystem of supply and use of that data and sort of triggers a cascading wave of public benefit similar to what is happening with weather data.
What we did was we actually brought together on March 11 a group of about 45 public and -- innovators meeting convened by the Institute of Medicine. And they were a combination of two types of rock star. One type of rock star were people like Ed Sondik and Don Berwick who are just incredibly deep health care and health data people who have gotten more about health care in the last 10 minutes than most of us will ever learn for the rest of our lives. That was one.
The second class of rock star are people like Tim O'Reilly who has never been to a public health meeting in his life, but who is incredibly passionate about and incredibly gifted at how you take data and change actual lives of people with that data. For those of you who don't know Tim O'Reilly he is not the George Washington of the web, but probably maybe the Thomas Jefferson perhaps of the web. He is instrumental in the spread of the World Wide Web and helped coin web 2.0 as a term, open source software as a movement and is just very passionate about the power of data.
And interestingly Don Berwick and Tim O'Reilly who were both giants obviously in their respective fields and never meet. They didn't even know who each other were. And at the intersection of those two classes of rock star magic act -- happens when you take disparate disciplines and rock stars in each and basically innovative thinking. We said is this idea of HHS as the NOAA of health data an idiotic idea or is it one that actually makes sense. And actually the group said no, actually it makes a lot of sense. And the vision should be basically HHS and others making data freely accessible to the world obviously staying well clear of privacy trade secret confidentiality considerations, but basically making data broadly available.
And then encouraging a broad array of innovators, foundations, companies, researchers, entrepreneurs, government agencies, local folks, et cetera to take that data and turn it into applications used by consumers, patients, health care providers, employers, government agencies, mayors, civic leaders, journals, et cetera to help make better decisions, to help increase awareness of health, to help -- improve health, and to help inform those actions. That was the basic play. Release data -- innovators, turn it into really cool apps they can use -- used by the public and -- across the public to help improve health.
What we then actually did was have an impromptu call at the time, but we were so excited and we were so -- not be at a meeting where you basically talk a lot, people are excited and nothing happens because it just kind of dissolves. We said okay. We will check with the secretary, but we are pretty sure she will be cool with this. On June 2, beginning of June, we will hold a big public meeting at the Institute of Medicine co-hosted by the secretary and Harvey Fineberg and whichever one it is can actually build an act that you want to showcase that day will showcase it because actually through the course of the day this group of folks that Don Berwick and Tim O'Reilly has brainstormed about 20 different classes of application. It is like community health maps and dashboards and search engine integration, health data and various other things. And actually we said if you can build one of these apps that you have dreamed of, we would love to showcase you at this big public meeting at the IOM on June 2.
What happened next was an explosion of innovation. We actually took a bunch of the data that we have talking about, put it on a central website, the Community Health Data Initiative website, made it easy to find, easy to download, easy to work with and said here it is. We put it out actually a week after March 11 and we said to the innovators talk to your friends about it, talk to your friend's friends about it and see what you can do. And they proceeded to do a really incredible set of things.
Less than 90 days later on June 2 we had our big, huge meeting which a lot of you have seen on YouTube, Community Health Data forum that was co-hosted by Dr. Fineberg and Secretary Sebelius. It actually made the cover of Modern Healthcare. There is a really good article about it. And if you haven't seen it, you can go to YouTube, search for Community Health Data Initiative and watch the video of the plenary session which is about 2 hours, action packed I promise you, which actually demos about seven of the coolest apps that actually got build or developed in that 90-day span.
One interesting stat is that YouTube video has nearly 13,000 views at this point of a 2-hour video on HHS health data so is actually equal to I think a Jonas Brothers' Argentina concert video which is saying something. It really is something that is getting a lot of buzz and the video is a wonderful encapsulation of the day.
DR. FRIEDMAN: Question. You said you wanted to be interactive. How were the coolest apps determined?
MR. PARK: Very unscientific process basically. What we want to do -- there were actually over 20 applications that were either brand new or were new features on existing applications that got build. And we actually wanted to show a cross section of different kinds of uses. And actually I will show you that cross section right now. Not all of them, but just a selected subset.
One -- this is actually my favorite one. It may not have been the most technically flashy of what was actually shown, but actually felt like it was a real powerful objective space. This is actually something called the Network of Care for Healthy Communities and it is effectively a joint venture between a technology company called Trilogy, Berkeley's Healthy Communities Institute and the National Association of Counties. And the actual dashboard has been around for some time. They built a turbo-charged version of it for June 2 leveraging a bunch of data.
And what it effectively is is a software service offering that any county can utilize that deploys a community health dashboard website, public website for that county. It enables you to customize what indicators you want to actually show that informs your county in laypersons' terms. It actually explains what each indicator is. It tells you where you are versus benchmark set by your county public health official, what the trend line is.
Even more coolly than that, it actually connects you to an inventory of promising practices which have been curated by the Berkeley Healthy Communities' folks of interventions in other communities have executed or are engaging in to help move the needle on the indicator you care about and it classified in about three buckets. One is -- stuff that is rock solid evidence base stuff that has some emerging evidence and stuff that seems like a good idea, but is not yet in this space, but very, very interesting.
It also connects you to an inventory of pending federal and state legislation that impacts that you care about. It enables you to advocate for what you want with respect to the legislation.
Also if you are an individual just looking for help, it connects you to resources in your local community of how you can get help getting fit, achieving healthy weight, smoking, diabetes. It is a really wonderful community health information and action dashboard. It is actually delivered in a very efficient way and actually as we will talk about in a second is beginning to already spread across the country.
DR. TANG: What is the data source for that?
MR. PARK: It actually mines a whole bunch of different data sources. It actually mines a bunch of data that we have made available. It is also actually the California CHCF's data and other data. It also actually enables local public health officials to upload their own data as well.
DR. TANG: So the example in San Francisco's example they have a number of produce markets or fast food -- what is that kind of data come from?
MR. PARK: Actually it is a good question. I don't know if that -- is that part of the USDA food atlas?
DR. TANG: I don't know.
MR. PARK: The USDA food atlas actually turns out to have been one of the popular data sets we actually put out there. And the USDA actually put it out there some time ago.
Here is one interesting punch line. A lot of the data we put on our Community Health Data Initiative website, the initial one we put out a week after the meeting, a lot of it was actually data that had been public for a very long time. It just hadn't been in a form a developer could use. It hadn't been easy to find. And one interesting testament to that is we brought in a crème de la crème Silicon Valley to start thinking about this stuff. None of them knew -- none of them to a person knew that any of that data even existed. They were shocked by both the breadth and the depth and the quality of data that had been built over the years. USDA food atlas is one fantastic example which has an enormous amount of incredibly powerful data about access to healthy food and there is more data actually coming from USDA along those lines as well.
Another punch line actually is that the datasets the people are interested in are not just health care datasets, but maybe even more than that, data on things like access to food and other determinants that are very powerful that people --
DR. GREEN: Can you just keep talking about this a little more? Let's get a little more granular. About how many datasets did you put up from about how many federal agencies or departments or something like that? And given what you just said what was the difference maker? If these had been published for a long time, what was it that made the difference?
MR. PARK: It was actually in my opinion a very modest website. It was literally a web page on the CC website and we put up 10 to 15 datasets like hospital comparators -- how we put up the data that we supplied to the match kind of health rankings, David Kendrick's group, community health data indicators as a downloadable dataset which HRSA has worked with CC to compile, Healthy People, USDA food atlas, so on.
These were really initially community health oriented datasets. We have been expanding now actually as you will see in a second to broader sets of data that we are actually going to be publishing. But for this initial experimental phase those were really the datasets we put up there. And I think the difference is that actually it is incorrect to think of the Community Health Data Initiatives delivering those publishing data. The correct way to think about the deliverable isn't even actually a specific set of apps. The correct way to think about deliverables is the catalyzation of an ecosystem. It is an ecosystem of people and organizations that make data available that translate that data that turn that data into usable applications that are accessible to a much broader array of people than an Excel spreadsheet with a whole bunch of numbers in it which is not going to be something a consumer or an employer or a local mayor is possibly going to use.
And then actually also critically actively cultivating and sparking use and conversation about these applications which use the data among the stakeholders, consumers, employers, providers, mayors, public health officials. For example, at this meeting at the IOM in March we had not just data experts, but we had Nancy Brown from the American Heart Association, Valerie Brown, the head of National Association of Counties, Chip Johnson, the mayor of Hernando, Mississippi who is doing actually really interesting work on combating obesity, heads a provider system, et cetera.
To me actually publishing the data is about 10 percent of the work. Marketing the data, if you will, to the right groups of innovators and stakeholders so they begin to use it is actually 90 percent of the work. And if you asked me at the beginning of this, I would have just said it is 50/50. Not even close. It is 10/90.
One of the things that I think is an important lesson to learn is this is happening all across the government, efforts of published data. It is just publishing data on a website and saying okay, we are done. It is not the answer. That is not a transformative act. It is actually really being thoughtful about how you work as a member of the community to help catalyze an ecosystem. It is not even driven by you. Like this whole thing isn't driven or choreographed by us. It is increasingly driven and choreographed in a very self-propelled way by the community and we are just happy to be a data sugar daddy and an enthusiastic cheerleader for innovating of the data, but it is the ecosystem. It is all about ecosystem development. That is really the key.
DR. TANG: Let me talk about the last mile of the ecosystem. I totally believe what you said, but now I want to find out either what is out there. You think of the iPhone apps. I need to Google health for my -- what is my --
MR. PARK: So if you wouldn't mind actually -- let me actually finish -- these are phenomenal questions, but we may have some answers for a lot that may help spark more questions. Let's just go over this quickly. This is another example. One of Tim O'Reilly's fundamental principles in how data change people's lives is don't make people find the data. Make data find the people where people are already asking questions that data could help inform.
This is an example of integrated health data in the search. What Bing did is they basically took our hospital compare dataset and integrated it into search and there are certain set of flies have to be hit for it to materialize. If you type in New York Presbyterian Hospital actually pulls up the patient satisfaction rating with New York Presbyterian as reported by Medicare's hospital compare study versus state average. This is something again sort of illustrates the principle bringing the data to where people are. There was something like over 94 percent of Americans don't even know that Medicare has hospital compared data. The way we make it accessible is you have to basically go find a website which nobody knows about to use. It is a great website, but it is not necessarily the only way to look at the data. And so actually this is an example of how the data can actually reach out and touch many more people, help many more people who could benefit from it by bringing the data where people are.
Another example of that. Google built a hospital finder which actually -- their presentation was titled best places to have chest pain in America. They had Ronnie Zieger, a very gifted physician, to put this together just as a very quick prototype to match up Google maps with certain hospital compare metrics in terms of heart friendly and people friendly. It enabled to be much easier for people to interact with and understand.
iTriage is this really neat start up that helps patients with a variety of things that they need help with respect to making health care decisions and they took HRSA's newly available community health center, provider directory and uploaded into iTriage so you can actually find a local community health center using iTriage. It has been used according to them tens of thousands of times since they actually launched this on June 2 to help people find by care.
Here is an example of interest built by a start up in Boston. It is a game called Community Clash. And basically if you go to communityclash.com you can play it. It is kind of like community health black jack. What happens is you pick a city like Boston. You pick an opposing city like New York and you are dealt five cards face down. Each of them corresponds to an indicator like smoking, diabetes, et cetera. You are allowed to substitute two of the cards out for other indicators that you think Boston beats New York on and then you show your cards. It sounds silly, but I will tell you what. Of the 20 plus apps that were actually browseable by people on a floor show after the plenary session, the longest line by far was for this. I played it for 3 hours that night with my wife. It is strangely addictive. I have been in health care for probably the better part of 20 years. I learned things about community health I didn't know. For example, rates of mental illness in America are a lot higher than I thought they were. I kept trying to find a city where they weren't what I considered to be shockingly high and I couldn't. And that has actually suddenly but importantly adjusted my mind frame and will influence what I do in the future.
I submit to me -- have you all heard of Farmville? Farmville for those of you who don't know is a silly social networking game. It is not silly. It is fantastic. It is a social networking game on Facebook where you basically take care of a farm and you trade sheep and bales of hay and everything. It was started like less than 2 years ago. Do you know how many active monthly users Farmville has? 73 million. It went from 0 to 73 million active monthly users in basically 18 months.
I would submit to you that actually probably that Americans will become aware of and take action on community health issues is far more likely to be driven by a Farmville of health than by every American suddenly saying I have to go find my local public health website or public health community college lecture and get smart. This is just an incredibly interesting way to make community health information actionable entertaining and have people have fun but learn about community health at the same time.
PARTICIPANT: What is this one called?
MR. PARK: Community Clash. If you go to communityclash.com.
And then another one. Just one final example is Asthmapolis. This is actually an entrepreneur, one person in Madison, Wisconsin who attached a GPS device to an asthma inhaler and integrated with a web app. And what it basically does is that whenever you use the asthma inhaler it enables you to record the time and place of the attack which leads to some very interesting things. There is a study that is going to be coming out soon on Asthmapolis that show that in the target population increased asthma quality control levels significantly because you suddenly you realize that all your attacks are happening at work or they are happening every Tuesday when you walk past a cement plant in the morning.
There are also actually options to contribute your data to an anonymized crowd source map of asthma attacks that public health officials can use to actually pinpoint where stuff might be happening. It is -- example of crowd sourcing of data to help individuals and to potentially help public health officials as well.
DR. FRANCIS: Can I interrupt you just here for a second because this was the one and you used the word anonymized. This was the one that raised the most significant privacy question for me in particular because when I watched the actual video there were grafts of particular patients and a comment about one of them -- that is the patient that we really need to get on because that is the poorly controlled patient. And of course if it is fully anonymized you don't have any idea who that patient --
PARTICIPANT: Or voluntary.
DR. FRANCIS: It might have been somebody who you knew who the person was. But I also went on the website and I couldn't find a privacy policy for Asthmapolis and I did that in June or July. It may have been too early. I am just curious about just to put on the table because this is amazing stuff. This is not by means of a criticism, but a misstep could be really serious for trust. The question that I just want to ask you to reflect on at some point is who you are talking to about privacy, how you are thinking about privacy. Are there novel kinds of privacy questions that you see as emerging from this that we could help you with?
MR. PARK: Yes. The answer is absolutely yes. I don't need to reflect for even more than a second. We would love all the help we can get. In fact actually as I will talk about in a second as one of our next steps we are actually -- we in the IOM are making the June 2 meeting an annual affair of a health data meeting where we bring together people who can supply data, people who are turning data into apps and stakeholders who want to use the apps. And we are trying to actually figure out what the agenda should be for the next one other than that obviously perpetuating those conversations between those groups, but I think having a track on privacy considerations would be fantastic. And any recommendations that you might have on who might help us put that together that would be really terrific. That would be really wonderful.
Speaking of which. I know Joy is like one or two other things she is doing right now, but it would be fantastic if Joy could get involved and maybe mobilize community folks to come in and speak with folks. I think that would be wonderful.
DR. GREEN: Just let me chase that just for a second. Why don't you formulate particular questions that are timely and useful to you right now that in your imagination to say if NCVHS would explore that or get answers to those questions that would help us. Go for the specific questions that you think are pertinent. Could you do that?
MR. PARK: Yes. Absolutely. Here is one for sure. Where we actually have a relatively well-established framework is if we have HHS data and we are publishing it. There is a whole set of screens we go through to make sure that that doesn't violate privacy confidentiality -- stuff. The part which is a brand new frontier is this crowd sourcing of data. How should one think about a set of guidelines and practices to govern the development and use of applications the crowd source data from patients, data which could be incredibly valuable to those patients and incredibly valuable to other patients and to other stakeholders like public health officials, but do so in a way that doesn't inadvertently hit a land mine of some kind.
That is a specific question right now that I think I am mostly interested in precisely because it is the most unexplored and also precisely because this kind of application is really interesting and really powerful and there is more of these on the way. I can't speak for David Van Sickle, the entrepreneur, but I would think that he would be intensely interested in paying attention.
Here is actually another interesting data point. I think that there were actually folks in HHS that I think quite understandably had a bunch of questions about what liberating this data would actually do. It would actually there would be a bunch of uses of it that would actually proliferate that were grossly irresponsible.
One of the things that actually we have been discovering is that human self interest is a powerful force and that people like David Van Sickle or people like these innovators don't want to look like an idiot anymore than anybody else. They are actually really focused on, for example, getting smart about the data and what it means. They are really focused and will be focused on how to do this without tripping over something like privacy because they don't want to -- Asthmapolis doesn't want to get into a cross fire on privacy anymore than anybody does. I think they have a big appetite to learn.
I think the part of what we can do as a community is really meet them where they want to learn, convene these meetings like this next meeting and have the kind of opportunity for people who want to be helpful and help them to get help and get smart about privacy, get smart about -- and in develop essentially a whole new generation of health meteorologists of people who are more and more versed in health meteorology. That is a fantastic question and a fantastic outcome.
DR. FRANCIS: If you have other questions that you think would be burning like -- the crowd sourcing one is fantastic.
MR. PARK: It is a really good topic to dig into.
MS. PRITTS: Another one that I have had on my agenda to look at is the proliferation of public use -- and everybody does their own examination when they -- but to my knowledge but maybe you know differently who is doing the cross check as you are posting numerous --
MR. PARK: There is actually a group that White House has convened that is privacy, security something work group that specifically focused only on publication of data and their specific focused on the mosaic problem of sort of one plus one plus one equals. You should get involved in that actually and a bunch of your friends that you know already. That is another next step that we can --
DR. FRANCIS: That is a question I know state health departments are asking.
MR. PARK: Also on June 2 we launched something called the Health 2.0 Developer Challenge, more precisely an organization called Health 2.0 launched it and we are very supportive and they announced it at this meeting. If you go to health2challenge.org, it is basically a new website where any organization, government agency foundation, company, individual actually can issue a public challenge to build the best application that does X. And 12 challenges that have been issued to date by different organizations, two by HHS, private companies, foundations like Robert Wood Johnson, Markle Foundation, California HealthCare Foundation, West Wireless Health Institute, et cetera. And actually also on this site you can register to be a competitor. Over 400 developers have registered to be at least ongoing competitors, 95 teams from across the country competing. Actually and from places as far as India and Germany competing.
On October 7 at the Health 2.0 conference in San Francisco the first six challenge winners were announced by the sponsors of their respective challengers which were very exciting. And in addition Health 2.0 and others actually held a series of code-a-thons across Silicon Valley also called hack-a-thons to stimulate developer interest work in health data.
Have any of you actually been to a code-a-thon? The way that this works is essentially what you do is you say okay. We are going to actually convene a gathering of developers over a course of 24 to 72 hours to work on some problem. In this case the first hack-a-thon was held by a little outfit called the HealthTap in Silicon Valley and they said here is a whole bunch of health data from the government and from other people and you develop interest in learning more about this data and scrubbing in and working with it and building stuff with it, show up at our offices. They were expecting about 30 people and 200 people showed up.
What is actually happening -- Google actually held another code-a-thon the following weekend and another 200 people actually showed up. There is a growing buzz in the valley about health data and about how it is an incredibly exciting frontier of which to work and to build things that can help people. And code-a-thons and hack-a-thons help to do that.
Also October 7 -- have you all heard of Blue Button? Blue Button was actually also formerly launched by the Veterans Administration and HHS and really VA is the senior partner in that and we are the happy junior partner. But the whole gist of this is the following. I think as everybody knows the VA has the My HealtheVet portal and Medicare has them on medicare.gov portal where you can actually get an account, authenticate yourself and look at your own data.
All the Blue Button is it is kind of almost embarrassingly simple thing which is you have a button that is blue which you can save and you download a copy of your own data. Now you already have your own data. It was already there on your computer. All we are doing is actually bringing it to your form that is human readable and transportable electronically. But it actually turns out to be a big deal because for whatever reason this isn't necessarily common across the health care system for a variety of reasons and we are hopeful that our example can help through lead the way to encourage other people to do the same.
We have actually done basically zero marketing of this in any real sense. We announced it at the conference, et cetera, but to date over 120,000, actually now over 130,000 people have downloaded their own data which is amazing. There is still a long way to go because we enable Blue Button for a million vets who have accounts on My HealtheVet and for all 47 million Medicare beneficiaries. There is still a ways to go. But that being said it is actually interestingly -- an interesting step at MyMedicare.gov is that while 47 million Medicare beneficiaries actually have accounts that two orders of magnitude lower than that beneficiaries actually use MyMedicare.gov. We are hopeful that things like Blue Button may over time make MyMedicare.gov more useful so people can use it. It is a virtual cycle. It is another step toward liberating data in this case to individual people for their own individual use. That can be helpful.
DR. OVERHAGE: Before you leave that one, do you have any members yet on how many people have done it twice?
MR. PARK: No, but that is a fantastic question. I will get that number.
DR. OVERHAGE: It is just interesting.
MR. PARK: And of course there is a bunch of activity happening in the ecosystem. People actually building the ability to help people take -- it is a human -- file but it is a pretty ugly one of people building meters that make it into a much easier thing to understand. Also PHR folks actually building to upload the data, upload to other applications as well. It is actually really simple step that is kind of --
DR. FRANCIS: Is it actual health information or is it claims information?
MR. PARK: In the case of the VA it is actual health information. In the case of Medicare it is just your claims data.
DR. FRANCIS: It is just claims.
MR. PARK: Claims and demographic data -- yourself enter demographic data and your claims data. Medicare doesn't have your health data.
DR. TANG: That might be another area where privacy should look into because all of a sudden you have all this health data and you are exploring various apps which means -- companies -- now have access to -- those privacy -- probably more important.
MR. PARK: I agree absolutely. Markle Foundation has done a fantastic job with a working group promulgating a set of privacy policies and practices that they have actually got published. It would be great to actually I think key that up for the discussion as well.
DR. FRANCIS: Just to give you one illustration of that there is at least one state that requires computer professionals who have access to information about abuse to report that abuse. That statute I am sure was drafted with the thought of child porn and digitized porn, but there is nothing in it that says it doesn't apply to health information.
MR. PARK: Could you repeat that just one more time?
DR. FRANCIS: There is a state that requires computer professionals who have access to information about abuse to report that abuse, technology professionals in South Carolina. I don't think anybody ever thought of it, but I am sure it was aimed at the kiddy porn market. But it gives a list of professionals in the statute who in their professional capacity acquire information or require reporting abuse. And when I came across that one, I thought whoa.
MR. PARK: It would be wonderful if we do this globally is another question beyond the crowd sourcing question is that NCVHS could focus on Blue Button and maybe draft off of Markle work as a starting point to say does that set of guidelines practice make sense.
One of the things that is also beginning to happen is that private tech organizations are actually working on deploying Blue Buttons of their own on their own patient -- it would be a really good time to give all the different actors in a sense of what the best way to proceed is on this --
DR. FRANCIS: To give you another little privacy interface question, it is my understanding that -- there are a number of states now that have what are called all payer databases which are being used to generate episode of care, cost, data. I know my own State of Utah has one of those. And to date Medicare has not been willing to share their data citing privacy concerns. And I don't know what the base of that is. I just am a member of our health data committee and I know that they are continuing to try to get Medicare data. They have all the other payers in the state, but Medicare.
PARTICIPANT: I also know that it is for Colorado.
MR. PARK: We are very excited about what is going on and we just want to keep helping it proceed and part of that is actually filling it with evermore data that is made ever more accessible. Launching actually January 2011 is going to a new HHS Health Indicators Warehouse. This is being led by Linda Bilheimer at NCHS, but is being operated on behalf of all of HHS. It is going to be a single warehouse/website which contains over 2000 indicators of public health performance, health care utilization, quality, prevention, prevalence of disease in the American population, determinants of health, evidence-based interventions that can positively impact indicators all made available in a single warehouse at the aggregate data of national, state, healthcare market like HRR and county wherever that data is actually available. No microdata. No individual provider or patient level data. Just rates, metrics, geographic unit level.
All made available in a single warehouse with well-documented metadata, the ability to actually download the data in a structure form to use and also data made available via an API, an application program interface which is just a fancy word for a way for developers to basically plug into the data and be able to retrieve it in a really easy way to power other applications and websites. That will be launching January 2011.
We also actually of course have already had the compare data out for some time, but in September put APIs on it which makes it a lot easier to work with. The National Library of Medicine as you may know has also launched an API portal as of September 2010 that has a whole inventory of super awesome APIs to their data like clinicaltrials.gov and PubChem and RxNorm and MedlinePlus and all kinds of amazingly awesome datasets that are now much easier to work with thanks to these APIs and you can -- find them.
Healthcare.gov have you all heard about? It is a website that we have deployed that has an inventory of public and private health coverage programs. We are very proud about the website. It is a great public service, but I actually think that the data itself should not be found only on a government website. We are actually going to be making that data available as well to proliferate across the web.
And we also -- and this is going to sound like an incredibly geeky detail and a detail that is maybe even more profound than any of those. We are rewiring the processes within HHS to encourage the publication of data. This is actually all detailed in open government plan which has other datasets actually that we are also planning to publish or have already published.
But the idea -- it is going to be actually led by our HHS data council -- is for agencies to undergo a regular process where they survey their inventory of data based on their priorities and the secretary's priorities targeted additional datasets to actually make public. Also actually whenever there is new data investments being made to basically have as part of the evaluation criteria for the investment for making the data public unless there is good reason not to in which case that is of course okay, but to make it kind of a fundamental reflex and ongoing process to basically make data available in a machine readable public forum as opposed to having just be periodic bulging exercise that happens once in a blue moon.
And then finally to one of your points there is so much of this happening. It is really kind of hard to keep track. It requires like a PowerPoint presentation. We want to try to make it more institutionally viral and one way to do that is we are actually deploying a super site called healthdata.gov in January 2011. It is actually a community on the existing data.gov site that will be dedicated to health data. There will be a one-stop resource where you can actually find every federal datasets, not just from HHS, but also from USDA, EPA, Housing and Urban Development, et cetera, all downloadable of federal data and metadata about what that data actually is. Also a link to an apps expo that will be hosted actually by another organization because we can't host it, but that we want to actually have people be able to find which will have a cross section of super cool uses and applications have been built using the data.
An online community, a forum where people can actually talk about the data, ask questions about the data, raise issues with the data, request new data, talk about how they are using the data, which is totally not controlled by us. It is just going to be whatever the community actually decides to talk about.
And then in another section we are going to list additional data sites that we are not endorsing because we can't endorse them, but that are other sites. They are run by states or localities or KidsHealth.org which -- are just fantastic. Other data sites that provide data on the same basis we do. Free, without -- constraint, downloadable that we think that the ecosystem might find interesting.
DR. CARR: Larry, Walter, Mark, Garland, Paul.
DR. GREEN: Can you say more now about the federal agencies whose data will go to number five there, healthdata.gov specifically. Does it include the Department of Education, does it include the Department of Labor, and does it include the Department of Energy?
MR. PARK: Yes. Those specific departments I think aren't going to be as heavily represented initially and the datasets are put out there, but absolutely it will.
DR. GREEN: It will or it does.
MR. PARK: Healthdata.gov doesn't exist yet. Even when it does exist -- if you go to data.gov today, there are actually a ton of datasets from education, energy and labor. Healthdata.gov I think will have less representation because I think that there are additional datasets that they could make available that are more closely correlated with health that we would like to actually have them put out. The existing datasets they actually have available data.gov will be in healthdata.gov as well.
DR. GREEN: This is a good example conceptually though of something that seems to me needs attention. It is hard to name things that are more closely associated with health and education.
MR. PARK: That is absolutely 100 percent right. Agreed. Maybe the problem is that we are not just thinking extensively enough about drivers of health. Maybe there is actually a bunch of data that education already makes available that we should actually just include.
DR. CARR: We are going to -- did you want to say anything, Walter?
DR. SUAREZ: Great topic and we appreciate very much you coming. I have one comment and one question. The comment is there is federal data and healthdata.gov looks like it is going to focus on federal data for the most part.
MR. PARK: There will be another section that actually has links to other data sites.
DR. SUAREZ: Because there is an organization called the National Association of Health Data Organizations. I don't know if you are aware of it.
MR. PARK: The National Association of Health Data Organizations?
DR. SUAREZ: Yes, the National Association of Health Data Organizations -- 25 years in existence. We just celebrated --
MR. PARK: Are you part of it?
DR. SUAREZ: It is a national association so there are a number of organizations involved in it. It is called NAHDO.org. It has worked exactly on this precise topic which on most part on the local level or state level to bring in state data resources. In fact, it is interesting that, Leslie, you mentioned that All-Payer Claims Database because I was on a conference call that is why they had to come out with the All-Payer Claims Database effort. There is a whole host of resources and data that is also available.
But then the second question I have -- more than a comment is --
MR. PARK: What is the best way to reach out to them?
DR. SUAREZ: Denise Love is the executive director.
MR. PARK: Does anyone have contact info for her? And you are --
DR. SUAREZ: Walter Suarez.
MR. PARK: Thank you, Walter. That is wonderful.
DR. SUAREZ: It is actually on the website too. The other comment is -- with all this data marked and the super system of availability of data, is there any warning label that is on the dataset or would there be any disclaimer or some sort of a warning disclaimer on the healthdata.gov or any other places that advise people that are attempting to link this data for purpose of re-identifying individuals or others for that matter would be legal?
MR. PARK: Kind of long story short, we actually think that the disclaimer is actually too late. The data that we put on healthdata.gov which is going to be by definition without legal agreement is basically being available, downloadable on the public domain. It is on us to make sure that they would put on healthdata.gov doesn't actually get to the point where someone could use it to be there for anybody. For example, the Blue Button data is not being made available on healthdata.gov. The Blue Button data is made available only within your personal health record account. The health indicator warehouse data, for example, there are certain metrics of community health that you could not put in that warehouse because they could potentially be used to triangulate someone's identity and so forth. That is kind of the approach that we are taking is to make sure that the data we put on healthdata.gov is felt that -- can't be used to re-identify me.
DR. SUAREZ: The only concern of course is that data that is at some level of granularity available from healthdata.gov starting in January will pose a risk at some level of potentially allowing people to link that data with some other data from other place in some location or locality and re-identify individuals.
DR. CARR: Let me just interrupt. I think we have our timing off, Chuck. You needed to speak at 10 to be out of here at 10:15. Is that right? I think we are going to go until 10:15. I want to just jump in here for a second because yesterday executive subcommittee spent the day talking about emerging data sources. NCVHS is charged with advising the secretary on data, data uses, emerging data sources. Our homework was to review the June presentations. And actually we also heard from the IOM learning health systems who will be working with Chuck on thinking about when all this electronic data becomes available. What we will be talking about today is the convergence of things and I want to just give you a little headline so that this can be the beginning of a close relationship together because we are very much aligned.
The themes that came out of yesterday Jim Scanlon talked about the need for a dashboard on measures of how we are doing on health and health reform. And the discussion went then to communities. You could have it federally, but really what is happening in the community, what are people using, and how are they using it. In parallel with that we were hearing about the themes coming out of the learning health systems and the need for governance and oversight.
What we will be talking about today is just how we organize ourselves to now look forward to this new world, new epidemiology and how can we help in terms of holding hearings and pulling together experts on privacy, on data integrity, on uses of data. That is the direction that we are heading into. We are very excited about this and having you come here today and really bring it alive even more and point us to some of the other areas is very much -- excites us even more about this.
We have time probably for Garland. You have been waiting. Maybe we will let you conclude and I think we are going to maybe have to hold the questions.
MR. PARK: I will take like 30 seconds. I just wanted again to reiterate that we have this annual open health data meeting, the June meeting apprised with IOM next June would love to actually work with NCVHS.
DR. CARR: We would love to partner with you on that.
MR. PARK: -- topics. That would be wonderful.
PARTICIPANT: Do you have a date picked?
MR. PARK: June 9. We haven't publicly announced it yet, but that is kind of the tentative date. It is more than tentative. It is going to be the date. But there is actually going to be a whole bunch of work before that obviously. They will think about the agenda, think about who should be there and we would love to work with NCVHS. I am on that. What is the best way to do that?
DR. CARR: Why don't you contact me? Start with me. Justine Carr.
MR. PARK: Fantastic. Okay.
And then the second thing is actually -- there is also something else that we are doing as well. I just wanted -- heads up which is we are doing a beta of a health data journalism workshop. The University of Missouri has a center for computer-rated journalism and there is also a National Association of Investigative Reporters and Editors. And they are extremely interested in working with us on a workshop which will initially have about 30 or 40 journalists where they want to come in. We will learn about the health data, learn about the tools to use in data to develop their own tools, to generate stories, to generate officially their own kind of interactive utilities, to help get an edge to an increasingly tumultuous media environment. But there I think an important part of the ecosystem as well in terms of raising awareness, in terms of making stuff understandable, in terms of sparking action. That is actually very exciting to see that level of interest and for that community too.
DR. CARR: Excellent. I think we will conclude now. This has just been incredibly exciting to hear about, but just remarkable what you have accomplished in a short amount of time and transformative. And we will look forward to partnering with you in whatever ways that we can drawing upon our long experience in many of these areas of population health as well as privacy.
MR. PARK: Wonderful. Well, let me just say as well that this last -- what is it? It has only been about 45 minutes has been unbelievably productive. I have a full list of actually actionable concrete things to go do based on what you said, but I think speak to the fact that there is just an incredible amount of collective experience in this room that we should leverage for this.
I will not only take up the -- I will embrace it in a bear hug. I would love to work very closely with NCVHS on this. I will follow up with you -- that is fantastic. Thank you so much.
DR. CARR: Chuck, can you top this?
Agenda Item: IOM Preliminary Report - Dr. Charles Friedman, ONC
DR. FRIEDMAN: What I certainly can't do is a better job than Todd negotiating Windows 7. We are all XP users. This is a Minnesota Public Health Department thumb drive. This has contaminated the whole State of Minnesota --
Good morning everyone. This is my fourth or fifth update and each time I try to do something a little bit different, but this time I think to set up some of the discussion we are going to have later today. I am going to predominantly talk about the learning health system work and I apologize to those of you who were here yesterday. You are going to see some of the same slides that Claudia Grossman from the IOM showed, but you are going to get my narration which maybe a little bit different.
But before that let me just give you a few ONC highlights since this is the ONC update. I was trying to factor out of everything that is going on a few things that might be helpful for you to know and that there is a certain likelihood except for Paul Tang who runs all this stuff you would probably find useful to know.
First and foremost on this list is that even though 2013 seems like a long way away and 2015 seems even longer away for many of us, the advisory committee structure and predominantly the health IT policy committee has begun work refining its previous conceptualizations on meaningful use stages 2 and 3. There is a meaningful use workgroup under the health IT policy committee that has been created and it is having a series of public meetings and the next one of these in case you are interested is this Friday in Washington at the Humphrey building. These discussions are ongoing. They are preceding a pace.
The intent here is to formulate early enough a conception of what meaningful use stage 2 is going to be to give strong signals to the industry as to what it is they are going to have to be prepared to do more than they would get if they had to wait for the notice of proposed rulemaking to be their first queue to what the next stage of meaningful use is going to be.
The second highlight I wanted to share with you relates to the certification program. You may recall that certification is key to the whole high tech program. Meaningful use requires the use of a certified product and ONC has put in place a certification program which is open with regard to the number of certifying bodies that can participate in this process.
There are currently three so-called authorized testing and certification bodies or ATCBs in the local argot. I knew that, but it wasn't until I went to the website and I will give you the home page for all of our health IT information at the end of the talk how many actual certified products there were. There are approximately now 120 certified products that have been certified by one of the three of the ATCBs. And I leave you here the URL which will take you directly to the listing of the certified products. I don't know if this number was surprising to you, but I work at ONC and it was surprising to me because I had not been tracking this.
I am not going to go one by one and give you a specific highlight on the ONC grant programs, but I think it is safe to characterize and this is the language we use to characterize it that all of these grant programs are now in what might be called an early implementation phase. All of these programs now are up and running. These include of course the regional extension centers, programs, the workforce programs, the state grants for health information exchange which were referred to in an earlier conversation, the development of standards and interoperability framework and a certification program, and Joy's efforts that you heard about earlier, the development of a privacy and security framework.
We also have a set of transcendent programs in place because they address both the adoption of EHRs and the exchange of health information as pathways to meaningful use. This is the SHARP program for research centers on key research areas. The program of the 17 beacon communities to be demonstrations of what is possible through increased meaningful use in a community to address specific health care goals.
And then I added to this slide something that I really haven't put on it before and that is we are evolving a set of what we are calling key theme initiatives that are not reflected so much in major grant programs, but are reflected in increasingly organized efforts that we are pursuing around areas in need of pursuit. One example of this is clinical decision support which actually has been -- it is the first theme initiative that has been ongoing for some time.
Another example is usability. Another example is safety which is closely tied to usability and there was an announcement a few weeks ago about an IOM study that is going to be done for us specifically relating to usability. Rural health, disparities are other examples of these theme initiatives which are taking shape in different ways, but I think are very important complements to the other program activities you see on this slide.
Before I get into the warning health system aspect of this, let me just ask if you have any questions on this quick highlight film like the NFL's games in 3 minutes.
DR. FERRER: The IOM's study on usability. Can you elaborate a little bit on that?
DR. FRIEDMAN: It is going to be focused on actually what the government's role should be in providing safety oversight in relation to health information technology. It is really a policy exploration study, how best to do this. It is not a methodological, technical work about how to make EHRs safer. It is really more about how to put in place a structure that will enable EHRs to become safer.
DR. FERRER: And what timeframe?
DR. FRIEDMAN: I don't know. I don't know off hand. I don't want to give you a figure that is incorrect, but this is a formal IOM recommendation generated study and rarely are those completed in less than 14 to 18 months. But of course everything we do is done on the double quick. I would guess 12 to 18 months.
DR. FERRER: Of the 120 certifiable products, how many meaningful use criteria stage 2?
DR. FRIEDMAN: None of them would meet stage 2. These are certified products meaning that they have the functionality when put to appropriate use within a professional practice or hospital setting to support meaningful use. But meaningful use is not just about the technology. It is about what people do with it. These certified products are certified with reference to the meaningful use stage 1 criterion.
DR. GREEN: Chuck, as you know, we are very interested in the community's perspective about how all of this work benefits communities, local locations. Your Beacon Community Program -- that is what my question is about is what is your opinion about the appropriateness and the functional availability of the Beacon Community Programs to help us learn and understand what it is communities want from this.
DR. FRIEDMAN: I think the theory is perfect. We can't of course see into the murky future and guarantee perfect execution of the plan based on the theory, but the idea here is that a set of communities that was selected in part because they are more advanced toward meaningful use than others and is likely to get their first. Should those communities be able to demonstrate empirically supported improvements in health and health care quality as a result of their advanced state of meaningful use that will be an enormously valuable demonstration to the rest of the country of the value of health IT as a catalyst for better health care. The theory is good. Will these communities all meet their goals? That depends on a number of improbables in the murky future, but I am -- you asked for my opinion. My opinion is that this is an extraordinarily valuable program because there is nothing in IT having been in this business for 35 years that is more convincing than a demonstration of what is possible.
DR. TANG: A clarifying question of Jorge's. I think he asked about usability and you talked about the care of safety. There is not a separate IOM activity on it, right?
DR. FRIEDMAN: No. There is not an IOM activity on safety. Did you ask about usability or safety, Jorge?
DR. FERRER: Usability.
DR. FRIEDMAN: Okay. I am sorry. Thank you, Paul. I responded with regard to safety. Usability is a less formal program in the sense that we don't have an IOM study, but NIST is centrally involved in this activity and in fact last July on the same day that the meaningful use stage 1 final regulation was announced NIST held a very well attended except for all the ONC people who couldn't go because of the announcement workshop. It was attended by about 200 people on usability. I am quite sure there is a report out. In fact, I know as a fact you can find on the NIST website a report out on that. We are working very closely with NIST and with AHRQ and entities outside the government on this usability activity.
DR. FERRER: Chuck, will usability be a sort of issue criteria thing?
DR. FRIEDMAN: I can't predict what is going to happen. I am trying to think if there is anything specifically about -- that specifically targets usability there now. I don't recall. Does anybody know?
PARTICIPANT: Meaningful use?
DR. FRIEDMAN: No, in certification. To your point usability testing is something that is being investigated as part of this usability initiative and I will tell you that one of our SHARP sites, the site at the University of Texas, actually has set up a voluntary usability testing program as part of its activity. Building had been underway long before they were funded and 18 vendors have already signed up for voluntary testing in the UT usability lab. Whether it is required or not I think there is interest in more of a bottom up approach to usability where vendors might see it to their advantage to have their systems tested both as feedback to them, but also as a statement to the public that their system was tested and this is how it came out.
DR. FERRER: That is an artificial environment, right? You wouldn't be able to replicate true clinical processes there, right?
DR. CARR: Jorge, I am going to ask maybe that we take that off line so Chuck can get to the other part of his presentation.
DR. FRIEDMAN: But a quick answer. It is a laboratory environment. You always have this tradeoff between laboratory environments which give you kind of a precision and control and clinical environments in which this kind of work is hard to do in a way that enables exact comparisons between products.
Just very quickly turning to the Rapid Learning Health System. I think you all have seen a variant of this slide before which basically illustrates our vision of the ability in the future to interconnect data possessing entities of various ways, shapes and form for the purpose of quality study, public health activity, and research of various kinds which will require the activities of aggregation, analysis, and sharing all under an appropriate mechanism and governance that ensures confidentiality, privacy and security and enables this to function in a trusted way because talking about it is one thing and making it happen is quite another.
ONC turned to the Institute of Medicine to convene a series of workshops involving a lot of smart and savvy people to pursue this notion more deeply of a learning system. And this is how IOM defined it in words as they began to formulate the plans for the workshop, a system where science informatics and cultural line can generate new knowledge as an ongoing natural byproduct of the care experience and thus seamlessly refine it for whatever best practices for continuous improvement.
The workshop series as it unfolded over these three events that took place in July. I don't have the dates actually in mind. I know the first one was in July.
DR. CARR: I think July, September and October.
DR. FRIEDMAN: Thank you. It is going to result in a report, a short version of the report that will be released before the end of this year and the fact that IOM is able to turn it around this quickly reflects in part that this is not a formal recommendation generating report signed off by the National Academy of Sciences. It won't be specific and informative in various ways, but the findings will not take the form of formal NAS recommendations and the goals of the activity are seen on this slide.
The three workshops as they were conceptualized in a sequence the first one looked at opportunities, challenges, and the general vision and began to surface the issues that would arise in creating such a system.
The second rather ambitious we called the System After Next using the term coined several years ago by Tony Gorry really was future oriented and took a deeper dive into how we might actually build a system to do this. And then the third workshop really focused on strategies that will be necessary to get where we want to go.
And as the work unfolded, four themes that are probably not surprising to anyone who has thought about this naturally evolved: a technical aspect to this, an aspect of how you actually go from information to knowledge and put the knowledge to use to actually improve health and health care, how you engage patients and populations, and how you put a governance over this. It actually functions as a system.
One result of these deliberations I would really like to call out and from those of you who were here yesterday it was mentioned yesterday. We really were informed by a group of computer scientists who had really not engaged around these health issues, but were invited to this workshop. And they brought with them to the workshop this perspective of this learning health system as something that if it was going to be successful would have the attributes of an ultra-large system. This concept of an ultra-large system was formulated at a meeting in 2006 called by the Defense Department to really think about what are the challenges of maintaining systems of the magnitude of the ones the Defense Department is now maintaining.
There was at one point in the meeting and this was at the second meeting as the folks who were propounding this perspective were talking. There was like a moment of group epiphany. It is the only way I can describe it. As people realized that this framework was really something we really did have to pay attention to. It really captured a number of key features and key challenges of what we were facing going forward that had not really been articulated quite this way before, for example, that you have to get the standardization of a system like this just right. If you specify at the beginning too few standards, it doesn't cohere into a functioning system that will work in the manner that my first diagram suggested.
On the other hand, if you over standardize, you create a brittle structure that will crumble under the weight of its own complexity and that cannot easily change. It also was the system that dissuades instead of promotes innovation.
And I think the third revelation that really struck me as they were describing it is that a system as complex as this is going to fail in specific ways and you will have to design a system that can recover and learn from failure rather than a system that you expect to work perfectly all the time. It is a design for problems approach, but to be able to recover from them. Those are just some examples of the novel thinking that came out of this perspective and I think that will be very valuable going forward.
I and my colleague Adam Wong and David recently published our views on achieving a nationwide learning health system in Science Translational Medicine. Justine, if you would like, I could get a copy of this article to everyone because it is very highly correlated with what was discussed at the IOM but puts a little more with an ONC slant on this.
That is my overview. Of course much more detail on all of our programs and the matters I refer to as highlights can be found on healthit.hhs.gov. Any more time for questions or --
DR. CARR: I think we are out of time, Chuck, but thanks for this and also your input yesterday. Are you heading out now?
DR. FRIEDMAN: No.
DR. CARR: You are going to be here. Great. I think we will have a little bit more development when we come back at 10:45 and we can talk about the themes that emerged from yesterday. Take a break now for 10 minutes. Thank.
(Break)
DR. CARR: What I would like to do now is review what we talked about yesterday. You have copies at your desk and I will just -- is there a blue button? I am going to take you through yesterday. We were here in this room. Here are the folks who were in attendance outlined on your slide. We began -- well, we didn't begin. We need to begin as we think about what do we do, where do we go from here. We go back to what is our charter tell us as it says here: the objective and scope of activities of the committee. The committee shall assist and advise the secretary through the Department of Health and Human Services, data council on help data, statistics, privacy, national health policy, and strategy to best to address those issues. And I think Jim made the point yesterday that we are not a policy-making policy. We are an advisory group.
And secondly of course as we know we show a system advises in the department and the implementation of HIPAA.
As we said our current state is we have our mandated standards, our HIPAA work from Congress, and then we have an elective side. That is kind of how we thought about it yesterday. What are we required to do and then what is somewhat discretionary in area. And where we are today is that in our work in privacy that history dating back to pre-HIPAA and more recently we have been focusing on EHR and PHR. Population health. We have been talking about data linkage and of course we will be hearing more about the recent proposal from populations. And then quality we have been talking about meaningful measurement.
Grounding us back to our basics and I think really aligning very much with the conversations we have heard this morning remember that populations health is not just about disease, functional status and well being, but it is also affected by community attributes, context, place and time.
And similarly as we think about health and health care the health care provider is one dimension, personal health and other dimension, and population health, an important dimension.
Our goals then are to identify focus and themes for the next 18 to 24 months that make it timely, meaningful contribution to where the puck will be is a term we have used a lot over the years, but never more important today in this fast-moving environment. But also to align with national priorities and legislative mandates and to align with emerging issues in health and health care and also align with the resources and time frames of the committee members and meetings.
And so, our approach was to hear from JAMA about key priorities for HHS and then to examine emerging data sources and then potential impact of this new data.
Jim took us through the HHS strategic goals and objectives which included health care, public health promotion and protection, disease prevention, and emergency preparedness as well as human services and scientific research and development and the details are outlined on your slide for you.
And then emerging data sources. Obviously we had a rich discussion this morning. Digital health data and metadata generated from EHR adoption and meaningful use is one major theme that we have been discussing and the focus of the IOM meetings over the summer.
And the second as we have heard from Todd Park this morning, the liberation of HHS data and I guess I have to add crowd coursing there as well.
And so when we think about what is the potential impact of this new data in the rapid learning health system is one outcome and as Chuck read to us earlier one in which progress in science informatics and care culture align to generate new knowledge as an ongoing natural byproduct of the care experience and seamlessly refine and delivery best practices for continuous improvement in health and health care.
The IOM has been focused mostly on the emerging digital data out of electronic health records, but as we have heard today with the community data initiative is about novel juxtaposition of data elements from hhs.gov and other sources so not only health, but community resources and environmental influences. I put here one of the mash ups that they had correlating information for community near San Antonio where there is high child poverty rate. There is high teen birth rate. They looked at what are the available services of boys' and girls' club, acute care facilities. What are the federal assistant grants from HHS immunizations to state and local? And they did documented abstraction from committee minutes to identify key players.
Taking all these pieces was a way of identifying the current state and then opportunities for who to talk to and what to talk to them about, but obviously what we just heard from Todd was very dynamic, very exciting.
DR. FRANCIS: Justine, could I ask you a question? If you just go back a slide. This is really a question for Chuck because I missed a little bit of the discussion. The quote about a rapid learning health system one of which progress in science informatics and care culture align to generate new knowledge as an ongoing natural byproduct of the care experience and seamlessly refine and deliver best practices for continuous improvement in health and health care. My underlining question to you actually yesterday and what I just continue to have is to understand how much what the idea of a rapid learning health system is focuses on the health care experience and how much it focuses on overall improvements in health whether it is through bicycle paths in a community or the location of grocery stores. This actually goes to the new data.
DR. CARR: I think that if you go back one more slide what I was saying is there is one piece that the IOM has been focusing on coming out of EHRs all generated obviously by health care, but there is an emerging world coming out of the community data initiative that takes education information and so on.
DR. FRANCIS: I guess my question about the rapid learning health system is is it the narrower or is it everything.
DR. CARR: I think that is an excellent point. That is what we want to talk about because we have these two seemingly maybe parallel, but not really parallel emerging. Hold that thought and let me come back to it and let Chuck comment.
DR. FRIEDMAN: Just very quickly, Leslie. There is nothing in the infrastructure that we are contemplating that would automatically leave out any particular kind of data. While Justine points out very accurately, there has been a focus on taking advantage of this anticipated volume of information that is suddenly going to be in computable form in EHRs. That doesn't mean that it is the exclusive source of data for the learning system.
DR. CARR: And I think that and I will get to it, but good to talk about it here is that the meaningful use is just beginning. EHR uptake is in a very early stage. What these data are and when they are available is still a little bit down the road, but as we look at today, this morning, this minute we can see all of this other stuff that raises exactly the same issue, the same benefits by juxtaposing these data, integrating them and making them live. The same learning can come about. And similarly the concerns about privacy, data integrity, all of those kinds of stewardship or governance issues apply whether we are talking about what Todd is doing today or whether we are talking about what the EHR digital data health care data is.
I think we are on the same page. It is really a continuum and we don't have to wait for all that other data.
In the IOM report the priority action targets identified were making the case, getting stakeholder involvement, embedded quality measures, distributed clinical research, functionality standards, interoperability, identity resolution, technical acceleration, ultra-large system approach, test bed, and then governance and coordination and that is what we talked a lot about yesterday. That aligns a lot with what we have done. Chuck thinks stewardship seems like a holding back concept and I think we think of it more of an enabling concept. As we use stewardship or as was used in the IOM, the governance addresses all of these things of use of the data and proper use of the data on all fronts.
We talked in different arenas yesterday sometimes about the data then finally got our sequence right. It begins with data generating information, but then creating knowledge. And then as Tom said today and as we have talked about yesterday and Ed Sondik brought up, key thing is just because you have that knowledge do you have the transformation and what does it take for that.
As we talked about the continuum of learning getting to data, there are very real issues of standards and also metadata standards. On the information side there are issues around what is a relevant query and doing sound aggregation and intact datasets. On the knowledge side someone asked the question who declares it knowledge, when does that happen, and then with that knowledge how does education happen, communication and outreach themes that were reinforced very much by the former chairs of the committee at the 60th anniversary.
And then a term, I think -- was it Don Steinwachs or who said, connect the dots that we have all these pieces, but you connect the dots together of the information on of what is out there.
And then ultimately transformation. And I think Jim has talked about the fact that HHS is trying to build a dashboard to measure health care reform. Is it happening? And then is it impacting? A lot like what we said about the beacon communities. Are they doing things and is it making a difference?
Ed raised that question, Ed Sondik, of why -- how does transformation take place? What is the value proposition? What is the incentivization for change? How do you measure the impact? And does change come about from a top-down federal dashboard or does it really come about in the communities? Susan shared a little bit about Mendocino County and how that data is being used and making changes and Todd also shared in the Healthy People 2010.
Those are specific questions to those parts of the continuum, but then underlying all of it is governance, stewardship, privacy, security, data integrity, sound use of the data, and also a term I have later on of just the trust equation.
IOM report and NCVHS. How do these stay together? Go together? ONC plans to take a lead role in a coordinated effort to develop the infrastructure for a rapid learning system to build and extend meaningful use. And that effort, as Chuck pointed out, will need advice from many informed people, but a potential role for NCVHS given its traditional focus and mission on population health would be these themes of privacy and security, strong trust fabric, the standards work related to population health, governance, stewardship and education.
Then what are the emerging needs that NCVHS is looking at? Well, obviously we have our continued HIPAA and Affordable Care Act work. Jim asked could we contribute to the HHS dashboard development and the discussion that ensued was could we use local communities as a laboratory for data that is needed, useful, and available and learn how it can drive change.
Another area to contribute would be the governance aspect of the learning health systems. Again, as we look at communities and see how information and knowledge makes change at the same time we look at how they use that data and how did it -- what were the issues that came up with the data, lessons learned, and sort of evolving these same themes that are part of the IOM rapid learning health system.
Another thing that Jim actually mentioned was can NCVHS play a role in contributing to the privacy model definition for accountable care organizations that are pilots for which are due as early as January of 2012.
And then of course we have our assigned things. We have our elected things and then we have to always keep flexibility for the unexpected things.
Putting some of these things together I think one of the models that we were talking a lot about yesterday was the building on what populations has been working on: communities at learning health systems and the new epidemiology. If HHS wants to create the dashboard, can we look to the communities that have used data in their dashboards and have achieved improvements in their health or health care? What data did they use? How did they choose it? How was the community motivated to change? What were the obstacles and lessons learned? Again, getting that local knowledge and informing the national dashboard.
And then the second piece, can we drive governance and guiding principles in observing what has gone in these communities? How was privacy protected, trust established? How was data integrity established? What principles were applied to ensure cause and effect with the data? And how was the community educated on the findings?
DR. TANG: Paul Tang, Palo Alto Medical Foundation, member of the committee, no conflicts. This slide and the next slide have great needs and are you implying that these needs are addressable by NCVHS?
DR. CARR: Yes. We are trying to say these are things that align with the work that we have done already in terms of governance and so on.
DR. TANG: Let me just ask questions. Some of these and particularly the next slide have a lot to do with what I might call management versus health policy. Is that correct or not? And I am looking at Chuck to say well and what of these would be covered by the IOM study. I am trying to figure out what is already -- it seems like a lot of these things hopefully will be covered by the IOM study.
DR. FRIEDMAN: The IOM study is just getting the game started. The IOM study will --
DR. TANG: No, I am talking about the new one. The learning health systems.
DR. FRIEDMAN: That is the one that is completed.
DR. TANG: Okay. There is a new one.
DR. FRIEDMAN: The new one is safety.
DR. CARR: There is an ongoing theme of learning health system which has many subparts. The one that was related to the electronic health data held the meetings in July, September, and October.
DR. TANG: I will take that part offline. But some of these seem like something maybe an ONC would do, something maybe a private sector would. My question is some of these are more operational versus policy. Anybody else feel that is true?
DR. CARR: Just to be clear. We are not charged with policy, but building on what Jim was asking for yesterday was sort of input on what the dashboards at HHS is developing and from that initial conversation we had a lot of discussion about how does a dashboard help you. What can you expect from a dashboard? How do you know it is relevant when it shows improvement? How does that translate into actual improvement in the communities?
DR. TANG: And you think that is the work of this committee?
DR. CARR: Walter.
DR. SUAREZ: I think the role of the committee has been to advise the secretary on a number of issues, policy as well as operational. And one example of that with respect to the dashboard is advising on the content and the use of the dashboard itself and progressive addition of elements into the dashboard. It is not that we are going to run a dashboard. It is that we would provide advice on how to build or what to include in such a dashboard with respect to a metrics that allow the measurement of achievement of goals related to health reform, related to high tech, related to different things. That is how I see it as our role.
MS. MILAM: I think the questions are great. They are questions that a lot of us at the state level are grappling with today. They are essential questions that some folks have figured out pretty well, but there is no common base of best practice or recommendation to build from. And as we look at more data, make more data available, these questions are essential that they be answered and frameworks in place so that we do have trust. I think it is great.
DR. CARR: Either Don or Larry, do you want to speak to how this ties in with the proposal that you were developing from the population subcommittee?
DR. STEINWACHS: Larry keeps deferring to me. I guess I must be older. Is that the message I should take? That is it. This had all been initiated as you know by thinking about where is the potential for expanding population health information that people could use as you think about electronic health records, the American Community Survey that is ongoing, other data sources, and the fact that it is the community and there are a variety of definitions for communities sometimes, but it is the community that really is the place where you can work on eliminating disparities where you really understand the context in which health care, health and the population well being is.
It was a thing in terms of an effort to try to understand just what Sallie was talking about initially. There are communities around the country who are getting more engaged in wanting to look at and understand their own health data, why are they using that, how might health information exchanges and other mechanisms that exist in communities and states help actually provide information that is much more relevant to the community itself and that it talks about its popular experience. How do you address the denominator issues of some people have electronic health records, some don't. That might be matched through census data to try within a framework so you know what the missing pieces were even though you still have to learn more about the biases.
I think Larry and I and the subcommittee saw this as a beginning of a process to try to look at what is the potential when you look at where the puck is going for communities to actually have health data that would help them address issues that are important to them and would be the health data that could be rolled up if we had agreement on a dashboard and key elements. You could roll it up from the community to the state to a nation. That would be powerful if we could do that.
DR. MIDDLETON: I just wanted to add one other thought on this slide in particular which did come up yesterday and some of the folks yesterday have heard me opine on this already. Two related issues I think arise with the new epidemiology building upon comments that have been already. If we think about what Todd Park showed today, those mash ups, combining of data from many different disparate sources, the idea that there needs to be standardization of the metadata to describe the data becomes increasingly important. In a way we need to now have standards 2.0 that will allow us to maintain data integrity and do these combinations of data that are meaningful hopefully and centrally.
And that raises the second question. I think there is a new epidemiology also about how to defer or infer from these combinations of data that which makes sense from that which doesn't. I am just reminded of my epidemiology professor who said you can associate the number of telephone poles on Long Island with the incidence of cancer. Of course it is meaningless. We have to make sure we know the methods to arrive at reasonable inferences from mash ups because otherwise we will be chasing bunnies down the rat hole.
DR. CARR: I am reminded of Marc's observation of the growing incidents of no cough in Indiana.
DR. FRANCIS: I just wanted to comment that both yesterday and I was also talking earlier today with Larry that the plan would be and we will talk with the privacy committee this afternoon to actually meet jointly for part of our time with populations and talk about what are the privacy questions. Some of those emerged today in the discussion with Todd as well that as new combinations of data and new forms of data become available. An obvious one at the state level is the public use datasets that states make available. Many of those do it with different constraints. You have to promise you are not going to try to use it to re-identify, for example, then the federal Community Health Data Initiative. I am just using that as an illustration. All of those are tremendously important questions that appropriate practices need to be in place to have the relevant levels of trust so we can really use the data in its maximal capacity.
The thought would be to actually plan out for considerable -- this isn't something that you do in 24 hours to address these questions that are raised by community emphasis of useful data from the population side and from the privacy trust building side.
DR. CARR: We have Larry and then Chuck.
DR. GREEN: Could we go to your next slide? It may be now would be a timely propitious moment to address this just a little bit.
DR. CARR: Okay. Let me just see if Chuck has a comment to follow on that.
DR. FRIEDMAN: I appreciate that. Just to respond to the previous few comments, I think there is something enormously rich there. It struck me as Todd was talking about what a culture change this is. From a point of view of a scientist the notion that data might be treated as a kind of commodity that can be mashed up is something that gives one great pause because anyone who has done a study and collected data knows things about their own datasets. This leads to the sort of protectionist view that many scientist have about their data which at that extreme is probably not where you want to be either, but I think there are questions that Blackford was alluding to that go to this issue of data becoming potentially a commodity. And what would really take that too far in my opinion would be the absence of sufficiently rich metadata that would tell you not only what is his, but the quality of it and what it should and should not be used for.
I think the standards for those metadata are unless somebody knows something I don't undetermined at this point and that is a piece of work that is sitting there needing to be done and I think it goes to some of Paul's concerns about finding projects that are really important policy work as oppose to operational work. I think this hits the sweet spot at least in my opinion.
DR. CARR: Paul, is that -- good with that?
DR. GREEN: Let me grab both ends of this amazing spectrum and just make one wildly philosophical statement, but then I really want to get very granular about how we are going to function as a committee. The wildly philosophical statement is there is a strong case that has been made by decades by the epistemologists that there is no such thing as knowledge without context. And a lot of our conversation is about that fact or that thing we think we know. Knowledge as a lump is not really a thing. Knowledge only exists in context.
I will pick on Blackford. Chasing a rabbit down a rat hole is an example of knowledge in context and how they are critical to interact. Forget that.
Functionally both -- we need to unite our conversations today while we are in Washington meeting and our subcommittee meetings. We are having multi -- we need to unite our various conversations about our committee work. And one place that seems to me has matured and is ready for us to make up our mind about what we want to do is the realignment of our committee structure in some manner to help us be more effective, to help us get staff properly, to help us move expeditiously. And one opportunity is for the privacy committee and the populations committee to decide how to functionally work together for the next stretch of the road.
What we are framing up here today to do there is some urgent need for us to inextricably link the privacy work and the population subcommittee work. There are probably very good reasons to maintain the committee structure of whatever, but I would like to put a little bit of a stake in the ground that we should for the next at least few months if not 3 years begin to think increasingly that the populations and privacy committee overlap and moving in concert in some way.
There is important intersection here that I think we should capture and do a little reorganization to facilitate and enable and to get away from privacy working in one corner and populations working in another corner and then is trying to adjudicate some issues later on in the full committee or whatever. I would like to move that upstream if we can to where those two communities at least those two committees to begin to function more in concert.
DR. CARR: I will make a comment and then Walter. I think as we think about the work that we have done even just focusing on this theme obviously measurement is a key piece of it and that has been a lot of the work of the quality subcommittee and of course standards as well. This was the theme that came through many times yesterday of cross cutting topics that don't line up with one or another committee.
DR. SUAREZ: I just wanted to comment on the alignment. I think we are talking about aligning different things. The way I see it is we are trying to align domains with dimensions. Those are the two terms I guess I can think of to use because I can see how population health also needs to work or we in the data committee or the standards committee need to work more closely with population health as well.
There are some things I would call maybe dimensions that are the ones that we talked yesterday about: the population health and the personal health. And then there are domains that we apply to those dimensions to try to reflect on the priorities that we need to work on. And then there is going to be certainly teams that cut across those.
I think in going in your direction we came to that to this picture that is now being shown that describes the matrix. I think Leslie was the one that suggested because we originally started with three, the first three. They were data, privacy, and system performance. And she is just adding community health and I was having trouble trying to see how that is not so much a dimension around population health that cuts across the other one.
DR. CARR: Susan and I were struggling with this afterwards as well. Community health may be a lens that we look through or focus on. We had a couple of proposals of how we configure.
I think the real issue is that we enable cross functionality in our topics and that we have a theme that draws everyone in. I think one of the issues that we talked about is that with our kind of elective piece each subcommittee had freedom to choose an area of particular interest which may or may not align with what is going on in the other subcommittees and I think what we have found is the tempo and the burden. I don't mean that in a pejorative way, but the amount of work that you have to do in the amount of time you have to do it makes it very difficult to have a lot of projects going on at once. And I think we have had projects that a good hearing and then we stop and don't take it to the next level.
One of the underlying themes was to have a vision of space where NCVHS builds upon our history and expertise and applies it to contemporary emerging data issues.
MS. GREENBERG: First I want to thank you and Susan and others who contributed, Walter, everybody for this presentation because it is a very helpful and just to say that I thought yesterday was a really interesting important day to start thinking about where we want to go.
I feel that the current subcommittees sort of define the areas that the committee is working in and where we are looking for expertise. For example, Jim and I have been involved in recruiting new members and we hope that we will be successful there because we have a number of vacancies. It is helpful to know -- need to find people with population health knowledge. We would like to find people with privacy and security expertise. The standards area. We know we have -- that is an important area. And quality has increasingly seems to become an area.
That is helpful and I also think it is helpful to the outside world to say we have subcommittees in certain areas. It doesn't have to be these areas, but these seem to be the areas that we keep coming back to. But that doesn't mean we have to function with those subcommittees meeting separately or having siloed agendas at all. And I am open to whatever structure the committee feels is most useful for getting your work done.
But, for example, like this morning when the different privacy issues came up. I think there are areas that are being worked on by other committees, but there are clearly areas that this committee has expertise and experience in and is being looked to. It is useful to know that we have a focal point for privacy. We have a subcommittee on privacy and the department knows that that is one of our foci.
Population health, of course, that is identified with a whole history of the committee but so is data on basically health care delivery going back to classifications and uniform datasets and all of that. The standards area, which is a solid area that the committee has always been in for its 60 plus years. And I think quality is a little newer but it is obviously critical to all of this and to meaningful use, et cetera.
I see a certain value in having that as our public structure and then just -- at least that is where we are trying to fill holes in expertise and recruit members. And then you can decide however you want to organize this and however you want to address particular issues. That would sort of be my take on why I wouldn't throw out the baby with the bath water, but I would not feel constrained or like in a straight jacket from these subcommittees or other subcommittees that you might feel better define what your areas are. That is kind of from the exec sec's point of view and just take it or leave it.
DR. CARR: Thank you, Marjorie. Blackford, let me put your model up here.
DR. MIDDLETON: Thank you. It really is I think the group's model. I just happen to write it down. What I was thinking of those yesterday when listening to all the discussion was we certainly have some new areas of consideration, things that we might need to think about in the existing committee structure as defined today with the only slight tweak being -- I think quality is perhaps becoming data concept. We might want to think about performance because performance assessment around cost, value, quality, accountability, and maybe things we can noodle on and be useful around.
But then I won't read all the little thoughts there, but listening throughout the day many different things came up that might be the new considerations for the existing committee structure to think about.
Now next to the point, however, of organizational design to get things done I think we might think about a new structure wherein cross-cutting teams are formed sort of in a more agile manner or scrum like from software development world that draws together the relevant expertise for the particular topics or issues being addressed. The column on the left is the weakest part of this slide. Whatever the topic is would be what exists on the left, but then we could draw together membership for a swat team from all of the relevant committees. This is very attractive to me because I am not an expert in security or privacy. I know a little bit about standards, but there is a lot more that members of the committee actually know and is relevant to what we are thinking about even in quality.
Anyway, I think this is just a new model, fairly easy to do, fairly lightweight to implement, but it would be a different approach wherein the swat team should be the focus of the work effort.
DR. FRIEDMAN: Blackford, just a quick point of information. What do you mean by personalization?
DR. MIDDLETON: The idea that we are going to take a patient-centered view of everything: performance, quality, your normal test values for your serum sodium shouldn't be the population of referent normal value. It should be your normal values. All that kind of stuff. Not to mention genome, genotypes, et cetera.
DR. CARR: We will hear more about that this afternoon on the quality hearing.
I do like your idea about performance because it includes quality, but transcends it in a way and includes measurement and I think that that is very helpful.
DR. SUAREZ: If I may just make a point. I don't think these are alternative models when I looked at them. They are really saying the same thing in different ways. I think we need a structure. We need some defined subcommittees. We need the name of those. We need content of responsibilities of those and we need the ability to do cross-cutting activities as well. There are going to be responsibilities that a specific committee will be working on for a specific purpose. And then there are going to be activities that a committee can take the lead, but can form a swat team or -- teams, some sort of a group that will take expertise from the other committee. I don't really see it as --
DR. CARR: I wonder if we had leaders in each of these traditional subcommittees, but that the membership on a particular topic is fluid and that maybe you were on -- I was on standards with ICD-10 because I know a lot about ICD-9, but then when we get into metadata, I might not be the right person and that might not be the best use of my time. It is just hypothetical. I think my metadata is good.
If you sign up for -- and that has happened to me because I have been on just about every subcommittee except for one. And it is always that you get there and it is a theme that you are working on and then actually you don't have anything to contribute and oh by the way you can't get to your other subcommittee because they meet at the same time. Maybe we have leadership in each of those areas, but the membership is driven by the area of focus. Just another thought.
DR. FRANCIS: I think that we really need to combine those two. I think we need to have settled groups and we need to have fluidity of shifts. Unless you have one you are not going to be -- I actually think you have to have one to have the other as well.
DR. CARR: And the other thing that I think part of putting this all to paper in this way and maybe in part what Paul is reacting to is we really can't do all the things that we used to think we could do. It is just not possible. The depth of knowledge, the time commitment, and the organization to have this going on has really stressed the resources of the committee and at times interfered with the ability for everyone on the committee to be fully informed about the topic at hand. I think it bears noting that part of this is to stay thematic and build toward a theme and it may be different projects and require different expertise, but that we not have multiple directions because I don't think we can sustain it and I think that the times now call for a -- let's just leave it at that. We need to be able to do that.
DR. TANG: I sort of like Blackford's matrix here. I think some of our best work meaning it has been most deliverable and most impactful has been out of our ad hoc groups. If you say all the leaderships have to be there then you add -- you might as well work with the whole committee every time then. When you pick a group for a reason which includes diversity of opinion then you are probably going to get a much better result and also be passionate and be willing to spend the long hours that it took for each of these ad hoc groups.
Another reason why they were impactful was because they were relevant which means a relevant topic and a relevant timeframe.
What is attractive about this is there still is a roadmap for each of these kinds of things, but they aren't the governing -- those aren't the project product producing groups. The horizontals are more -- not that these are the only ones, but there are things that we could do that would be very relevant and very impactful and I think those are some of our best attributes of success.
DR. CARR: I agree and we talked about that yesterday that those initiatives and I think of ePrescribing and Secondary Uses as two examples. And they were self-selected groups. Not everybody either had the time or the skill set to participate in one or the other, but we found a way to do that. I agree. They have been engaging. They have been relevant and they have been enriched by the diversity of the number of participants.
MS. GREENBERG: I think ePrescribing actually was done by the standard subcommittee, but we had the functional requirements. We have the NHIN. We had the Secondary Uses. It certainly has been effective.
DR. CARR: Next steps. This was just a straw man. We recognize that we continue our legislative requirements. But my thought was that for the subcommittees and maybe we don't need to do this. Maybe we are doing it right now is to liberate on their roles in emerging themes. The new epidemiology is one. The governance of a learning health system not limited just to the digital health care information but really this community data initiative and beyond.
And then evaluate the current -- this is for the subcommittees to look at your current agenda and either reaffirm that direction, revise it, or retire it to ensure alignment with emerging themes. I am not quite sure what that means, but just to say if we are going to focus let's say on the new epidemiology and governance. Are there things that we are currently doing that align with that or do we need to redirect?
And then I think before we leave Washington we need to have a clear vision of what we are doing and how we are doing it.
DR. FRIEDMAN: I think this is great. Just from a very self-interested point of view I would like to broaden the notion of what the committee might do for the learning health system beyond governance. We might have coned down on governance because it was one fairly obvious thing. But I am hearing and one of my take aways from this discussion is that broadening the concept of the learning health system to include making best use of liberated data. I think as we were discussing before there is standards work to be done there. I think there is standards work relating to the basic functioning of a learning health system in a more limited sense of using AHR data that I don't think is going to get on the health IT standards committee agenda as soon as it needs to if it gets there at all. I think there may be special population health-related standards that could very productively come to this committee. It is a long way of saying let's not cone down on the governance and maybe let's broaden that wording health system concept to include new kinds of data.
DR. CARR: I was trying to say a little bit here and maybe I was saying governance guiding principles and thinking about how is data integrity established. What principles were applied to ensure cause and effect that was community educated. But you are right. Standards you would say --
DR. FRIEDMAN: Not taking that off the table, but putting something else up there.
DR. CARR: I was using a very broad definition of governance of sort of data integrity as well as uses of the data.
DR. TANG: Just like Chuck is commenting on the learning health system agenda, I might say that the community health liberation of data project or initiative or whatever would be an ad hoc group that was formed because it probably touches all of our committees for sure. That is another example of a project that I think. If you want to pick on learning health system as your project and Community Health Data Initiative as our project in terms of advising on information health policies and I thought that was in our charge.
DR. CARR: Advise on but not create, not develop policy.
DR. TANG: I understand, but I thought we were more --
DR. SUAREZ: I think another one would be the dashboard concept. I think that is one where there is an opportunity for cross cutting activities.
DR. CARR: Although IOM convened around it, I believe the community data initiative is learning health system here and now and a part of it can --
DR. TANG: I don't think it is the same. Learning health system is far more though it should be expanded provider centric. And I think the Community Health Data Initiative is really data 2.0 kind of initiative.
DR. CARR: But I think that -- just because ONC has -- this is a theme emerging out of that. Altogether it is the new way of learning. It is taking digitized information and aggregating it in different ways to get rapid learning on large communities.
DR. OVERHAGE: As you all know I am a lumper instead of a splitter so I tend to think of this stuff in chunks. I struggle sometimes with the way we categorize these things because at some level you are getting data and whether you are getting data about patients or at community levels of whatever, that is the set of issues.
There is the set of issues about how to manage that data in a usable format. There are meta issues to me like how do you figure out if somebody had a myocardial infarction or not. How do you want to define that? And it is not just whether they have an ICD-9 code or an ICD-10 code that says it probably because you get a very different cohort in different situations. That is a meta issue that bridges across whether it is for population health, clinical decision support or whatever.
There are the cross-cutting issues of privacy and security across these for the different use cases with probably different implications and issues because of who the end users are and what the access processes are.
When you talk about, for example, learning health care system, to me that is sort of a particular combination of the tinker toys to get somewhere.
DR. CARR: -- learning health system for one thing.
DR. OVERHAGE: And personalization is the same to me. HIPAA is -- I guess I struggle sometimes with the segmentations and wonder if focusing on these core building block pieces might help. It sort of gets to your point that nobody has the expertise. They are all cross cutting. How you get the data crosses population health and quality.
DR. MIDDLETON: I wouldn't suggest actually that this slide was never -- it is not intended to represent the actual topics of focus. These were sort of examples. Can I just finish the thought, Justine, if you don't mind going back for a second? The work flow could proceed, Marc, is obviously there can be columns of expertise where the roadmap for each column if you will. It says here the issues pertaining to whatever, the population roadmap standards, et cetera, and it is that subject matter expertise then which is distributed across these cross-cutting teams. And further a cross-cutting team can send back to the column something that needs to be focused on if you will sort of pure matrix style.
DR. OVERHAGE: And I guess what I was trying to sort out is what are those blocks that we want to think about.
DR. CARR: If we made this last specific here or broaden it because I think of the term as learning health system as all of the above, the data initiative as well as the ONC data. But if we talked about two things, the new epidemiology that Larry was talking about and then learning health system and sort of guidance tools of the road that are in this new environment that we need to think about.
PARTICIPANT: -- how are those different, for example -- epidemiology is what you need to be a learning health system which then -- it is part of the cycle I think.
DR. SUAREZ: I think part of what we are trying to do is identify work stream topical areas that include several aspects inside of them. For example, saying learning health care system or saying just learning health systems doesn't tell me anything. It just tells me that there is a whole host of activities within that can be tackled by the committee through work streams. Inside the learning health system there is governance and there is data standardization across the board. And there is system performance measurement.
Learning health system is one area that we can concentrate and form different swat teams and different topics within the learning health system. It is not that we were forming a swat team or whatever we call that into focusing on learning health systems, the same thing with the community health IT. And to me community health is very much tied to the new epidemiology concept as I think I remember hearing from Larry, his perspective on that. I think it is all part of that same concept. Looking at community health you have several things to look into it. It is not one single report on community health or on learning health.
DR. CARR: I guess it is thematic. I have both of these together that as we are doing one, we are also doing the other as we are learning about the new epidemiology, what are the data, how are they used, who is in charge and were they effective. At the same time we are deriving area gaps and opportunities for governance and oversight of these.
DR. SUAREZ: But I think we have to go down to identifying the topic that we would focus on a little bit. Again saying just new epidemiology doesn't help us in saying this is the work stream and this is the topics we would focus on. I think in my mind it is more -- we are going to become the committee where learning health system will be a major focus and there are several things inside there. When HHS wants to receive feedback from an advisory committee on learning health system, they know that this committee is now working on it. A lot of it is really being able to tell outside the world what we are doing and what we are going to be focusing on over the next 2 or 3 years so that they can perceive us as this is the place to come for advice. Learning health system is one. The dashboard is another one. Community health is another one. Personal health is perhaps another one. There are those kind of work streams that --
DR. CARR: I have Leslie then Marjorie then Paul.
DR. FRANCIS: One of the reasons I like both having the notion of some are traditional areas of expertise and having the sort of nimbler we are going to assemble groups that might address special issues is that I think we need to have the capacity to both be responsive and to surface new issues. I think without our traditionally identified groups we could lose the capacity to identify longer range issues.
I will give an example that actually has nothing to do with privacy. When I first came on the committee, Marjorie thought that I would also want to be very involved in populations because she knew I had written a lot of things about issues and justice. One of the questions that it seems to me that a populations committee might surface would be the question of because we are the advisory committee about vital and health statistics. It might to say look at all this wonderful new data that we are getting. There are a lot of groups that are left out in this data. We don't have it or we have it imperfectly. That is a question a populations committee might surface. Of course that might not be true, but if what we are doing is responding to with the 3-year timeframe, we might lose some of that.
DR. CARR: Marjorie then Paul.
MS. GREENBERG: I don't know whether it is because we are sitting next to each other because we went to the same college at the same time, but maybe it is a cohort effect. I want to support a few things. We seem to be on the same wavelength, Leslie and I on some of these things. I think this mixture is good and there seems to be a certain amount of consensus of these traditional areas where people know they can come to the committee and that we have expertise, but also going across areas.
In a way I see this -- well, first of all I feel it is within the tradition of the committee, which I like to hold, but to take something like this term learning health system and broaden it. I remember when we talked about this maybe in June or maybe it was September with you, Chuck, and we said does this really involve population health and everything. You said absolutely. But I also was really struck by that definition that was very health care oriented. That is fine.
These meetings have been what they are and as you said the IOM report will be what it is, but it is completely I think within the scope and traditionally something that the committee does to take something and then broaden it. And I think that is what we are talking about. Whatever comes out of the IOM report. I don't think that will -- it will just stimulate the committee further.
But I kind of see this almost as -- we talk about health 2.0 or the internet 2.0 or 3.0 almost as NHII 2.0. The committee now 10 years ago put out that NHII report. Now I think the NHII in a way from an operational point of view got sort of overtaken by the NHIN and then the electronic health records and everything. But you go back and you look and we agreed yesterday. Those three concentric circles and they are showing up in various places on these slides even if not in circles still simple but powerful that all of those areas need to be involved and I think that is what we are saying about this learning health system and the new epidemiology too. I think population health could have been community health. It can also be monitoring health at the national level. I think there are lots of components to it, same with personal health, and same with health care.
I see this all very -- it is not so much that it is linear. We don't want to be linear, but it is just building on it. And the same with this health statistics vision for the 21st century that there we were talking about influences on health and that is exactly what we are talking about now with this community health initiative. All the different kinds of data that need to be brought to bear and that is the community health initiative and it will feed in with the learning health environment too.
I am energized by this, but I think they all kind of come together.
DR. CARR: I think it is true when you keep going back to these diagrams and they are very prescient.
DR. TANG: -- this IOM committee because there is a follow on IOM committee and I just want to make sure. I looked it up and it is an ad hoc committee to the issues and urgent actions necessary to foster development of learning health care system design to generate and apply the best -- for collaborative health care choices for each patient and provider to drive the process of discovery is the natural -- it is the same sort of definition. In other words, there is an IOM committee now formed as a sequel to the roundtable exercise that Chuck described that is going to consider actually the topics. Other people consider this important too is what I am saying. I think there is another group that is going to tackle a lot of these questions.
DR. CARR: So two things. One is I agree with Marjorie that the information infrastructure transcends the health care stuff, but Chuck, you were specific in that you felt that this would be an appropriate follow on from that work to the NCVHS.
DR. FRIEDMAN: My thinking was and thanks Paul, I didn't know about that.
DR. TANG: I assumed that you had funded it.
DR. FRIEDMAN: No. There is so much going on and I might have known about it and forgot about it which is also possible. To the extent that as you put on one of your previous slides, Justine, I think there is going to emerge a natural leadership/coordination role for ONC in this going forward building this on top of and out of a number of things that are already government-initiated efforts. And to the extent that this is an advisory committee to the government, it makes sense organizationally as well as substantively in terms of what the topic is for the government agencies within HHS that are engaged in this to call on this committee for advise around specific issues that are sure to arise. I don't think the IOM can fill that role.
The IOM was perfect to do these studies for us and I am sure this other undertaking when I find out what it is will play an important role as well, but there is something that is a natural set up about an effort being coordinated by one government agency with lots of others involved and a federal advisory committee which is in the business to give advice to the government.
DR. CARR: Jorge. And also can you introduce yourself. I know you got here a little bit late.
DR. FERRER: I am Jorge Ferrer from the VA staff and the committee. Justine, could you go to slide number 20? Data. We have talked a lot about data this morning. Standards, metadata, inoperability. We do have to also I think put a place holder on capturing the data. That is that human computer interface that has driven a lot of the problems with usability to the point that Marc Overhage was talking about. We always talk about the data as sort of the -- but getting the data in the system is really painful, extremely painful -- for a lot of different people. It creates some real problems for people. I don't want to beat a dead horse because -- Chuck has been admitting that, which is great.
DR. CARR: We talked yesterday about the importance of having a customer. I think we heard a pretty resounding, enthusiastic response from Todd Park today in terms of our collaboration with him. Again, as we are trying to figure out what is IOM doing, I think Todd was pretty clear that we have expertise that would be very useful and there is a very clear roadmap ahead in the next couple of months of what is going to happen. I think that is important. I have Bill and then Sallie and then --
MR. SCANLON: Actually following up on what Jorge said, I would like to go back in part to where applications are under HIPAA and sort of in ACA. And it is not just -- the Accountable Care Act is not just a question of when NCVHS is mentioned in my mind. It is more a question of what is trying to be accomplished. I think of our role in some ways. What is HHS' set of responsibilities and do they have the data that they need to fulfill those responsibilities? Our role as an advisory body and data is to say here is what you need, here is maybe a vehicle for collecting it or gathering it and that that is critical to what we do.
The Affordable Care Act -- if you think about the premises that are in there in terms of creating ACOs, creating bundles for payment purposes, creating value-based purchases, there are huge data needs in order to accomplish those tasks, data needs that are not being met right now because we don't collect the necessary data. We don't have the methodologies to use the data even the data that we have optimally. There is this question of how do we move forward.
I think that there is a natural connection between what Blackford relabeled performance and standards because one of the vehicles for capture is what we specify through HIPAA as the data that are going to routinely flow and how we change what that routine flow is now because we have the potential for electronic health records I think is the key question for the immediate future because this is going to be the foundation that is going to allow the construction of all kinds of other things: dashboards, learning health systems, et cetera. Without the capture you don't go anywhere. You don't make progress unless you have the sufficient dataset to be able to build the kinds of applications that you want to have.
DR. CARR: Were you saying data -- data needed for assessment of health care reform or for --
MR. SCANLON: For performance. It is more than -- in the assessment of reform in kind of a macro sense. This is getting down to an individual level. This is down to the patient. Should we pay X dollars for this service? Do we know that that is the right thing to be doing and we don't have the ability to do that right now. And yet within the context of the health reform bill we are saying we should be able to have the context to do that. We should be able to say if these people are members of an ACO, this is what they are expected. This is what the experience should be. And did they have that experience. If not, we shouldn't pay for it as if they did. It is things like that.
An ACO is -- while they have gotten an incredible amount of attention, they are going to be the exception for the immediate future. We are not going to change the system overnight. This is going to remain largely a fee for service system and we need to think about -- within the bill there are provisions to change how we pay fee for service. The question is do we have the capacity to make those changes appropriately. And that is I think the real challenge that we have to face. We are not the policy committee in terms of having to say to HHS here is the new payment methodology that you should use. We are the data committee which is to say here are the data. Here is a way for you to get the data that are necessary to make the policy changes that are important for the future.
DR. CARR: Sallie and then Paul.
MS. MILAM: I have been thinking about Marc Overhage's comments in terms of lumping versus splitting. I tend to go down to the more elemental level. And I think just to look at some of the privacy mechanics might be useful as we parse these issues. When you think about the rapid learning health system and a lot of it being focused on treatment and clinical care, health and health care, we have HIPAA and we have a number of tiger teams focusing on helping to illuminate those rules around treatment. You have them around payment under the Affordable Care Act. You have ONC's privacy and security framework. And you have a lot of other work that NCVHS has done in the health care arena. And you have people already focused on doing that work.
You don't have another group focused on the privacy population health area which really impacts both the community data initiative as well as the rapid learning health system. You don't have a uniform de-identification standard. You have a variety of standards within the Federal Government at the state and community levels and people don't understand how to apply them. And this is an area where I think it would be fabulous to partner with population. Those folks have the expertise. What it comes down to is figuring out what are the best practices. What statistical disclosure limitation techniques are appropriate that epidemiologists help you with so that you do have privacy? And that is an intersection that some states have done really well, some communities have done very well that is really cross cutting and I don't know of another group who has really centralized and illuminated all of the best practices that are readily transportable to other environments.
Because as I have been looking into it and states and communities are grappling with this today, you have to see what other data is publicly available in your area. You need to understand your own population. You need to understand your own data. And it is difficult for people to pull all of that expertise together to understand people who figured it out, what worked for them, and how to just put out a framework so that people can liberate their own data.
Right now there is a lot of data collected, a lot of very rich data that has held up. People can't get it to researchers because they don't know how to even respond to research request. They want to apply minimum necessary so they don't give everything to the world, but what is appropriate? Then beyond that what can you give to a researcher? Do you want to better understand their request? Do you give them everything? What if you have already given them data that year? And then they have cumulative data. What if they want to publish a report? How much can be re-disclosed?
A lot of people have thought through this very well, but really I don't know of one go to source that has best practices or recommendations or standards that are really easily applied to different jurisdictions and I think this is an area that would really be very cross cutting through all of these different areas as opposed to continuing to stay more in the treatment area.
DR. TANG: I was going to ride on the data bandwagon from that side of the room because I think that feels more where one there is a gap. I don't know of anybody who is working on this. And two, it seems not consistent with -- congruent with NCVHS' charge.
It starts with part of Chuck's emotional impact about well it is one thing to describe mash ups which is a technical way of putting things together on a website, but it is not a technical way of understanding what the data means. To Jorge, well, it is really hard to get data in, yes, and also to have it understood not only by your own programs that are sucking the data up, but trying to get it out of the human but to try to pass it on.
There is incredible, as you know Justine, an incredible amount of FTs that are spent just trying to understand your own data from the applications that you either bought or created yourself. To think that that would transmit to someone else is really -- it is uncomforting in terms of what they are going to do with that data.
So then that goes to what Larry said which is the data and the context. Nobody is really looking at that or trying to appreciate it both the implications from a technical point of view, from an intellectual point of view and from an understanding what does it mean to the consumers and patients that are ultimately going to be affected by it either by viewing it on their own which is now being possible or through the intermediaries we call clinicians. There is a lot in between that and it seems like that is not something that IOM is working on. It is not something that the government is working on, et cetera. It is a gap and we are called the national committee on essentially on health data and it seems like we need to render at least raise the awareness of this issue as it really proliferates the use of the spread of this, et cetera.
That seems like something that there is a gap. We do have expertise because it relates to, again, all four of subgroups, et cetera, and it is very urgent that we deal with this.
DR. GREEN: I want to go back to functioning, Justine, just for a second. Paul, not that immediate comment but Paul Tang's comment back where we were asking questions of ourselves about overlap redundancy, that sort of stuff. I think that is an important principle that we attempt to follow anyway to my knowledge. I am not sure but given Justine using your word tempo. Given the tempo of developments, I am wondering if we don't need our standardized way of making inquiries of the relevant players when we start formulating explicit questions that we are going to explore to make sure that we are not stepping into someone else's territory and that sort of stuff is -- just from this morning.
I am -- just from the populations subcommittee's point of view, I feel like that we need to formulate explicit questions that are going to tackle and start getting the answers to. And I feel like we need help getting those questions right and we need help in knowing if those questions are already being asked and answered someplace else. Todd, Ed Sondik -- we could develop a list here. But it is almost as if we need a rapid means of determining if what we are headed toward is an alignment and not redundant. That needs to be systematized in my view to help us function better.
Before we plan a hearing, before we commission someone to do an environmental scan for us, it seems to me we should insert a step here where we -- literally within 2 or 3 days. There can be the right people polled and ask if this makes sense to Chuck. Has he already financed it? It is already being done somewhere else or not? And when all of those people say no, it is not being done and yes, it is important. No, it is not being done and yes, it is important and Marjorie looks at it and says that is in scope. That is a go.
Maybe I have been missing this for the last 3 and half years or so, but I am not seeing that done systematically as I think the tempo of the events requires.
DR. SUAREZ: We have people around this table. Paul Tang, Chuck -- Paul leads a lot of these initiatives -- there are people here that understand very well what the pressing needs are, short term, long term and the future. We really need to listen to the guidance because the paucity of effort. No. It is just redundant of no value. When Paul Tang says focus on something that nobody else is doing.
I think NCHS when it sat by itself, it was an easier place to live. And now you have all these other advice for committees within HHS are doing phenomenal work, but it is easy to tell people to collaborate, but what does that really mean. What is the product going to look like? I kind of always look at Paul because when he is chairing a lot of these committees, he knows exactly what the pressures are because he has to deliver them.
DR. WARREN: I guess I was just responding a little bit to Larry about the way around and asking permission or not asking permission, but is anybody else working on this. That sometimes the topics are coming so fast I don't think we have the time to do that. I think this committee needs to be informed enough to do it. There is just something about the committee asking permission to do its work that bothers me. I don't think we need to ask permission. One of the luxuries that we have that I think some of the other FACA committees don't have is we can set our own agenda as well as the agenda that the secretary and that the Congress sets for us. And I would like to keep that ability. And I think even if we do look at some of the same issues, I think the issues are so complex I don't think we will ever be redundant. It may even be a good thing to have different perspectives being focused on the same issue.
DR. GREEN: I did not intend for my comments to be construed to have anything to do with seeking permission. I think the independence as a committee really matters.
I would just rotate from what you said. We do not have the time to redo what someone else is already doing and we just didn't have the good sense to inquire whether that was being covered. Mine is just a highly functional, technical step that I think we should formalize to avoid wasting our time. It is not a requesting permission, but if someone else is studying. But we say but they are not going to bring the proper perspective. I think we go right ahead and do it, but we need -- I believe we have an acute need to know what is going on in a rapidly changing landscape. We talk about a rapid learning system. We need something like a rapid detection system so that we can actually align our work and coordinate our work effectively. That is all I am talking about.
MS. WARREN: -- is that we need to be an agile organization. If we have too many rules on how we go about doing things, we become not agile.
DR. CARR: I want to make us aware the fact that we have 15 minutes until we go to lunch. And in order to have the full hour of lunch we want to make sure that we are parsimonious and focus on our comments. With that, Blackford, your hand was up and then I really want to bring it back to what is the work that we need to do today so that by tomorrow we have a very clear roadmap, thematic direction, and configuration. I have three hands up. I am assuming that you are going to tell me what that answer to what we are going to do.
DR. MIDDLETON: I did want to say very quickly to build upon the prior two comments about not asking permission, but what are the requirements, if you will, from the world that we are aiming to try to address because that upstream perspective, if you will, tries to satisfy the customer and the need and relates to the next comment. How do we actually evaluate the NCVHS? Who do we ask are we doing a good job? Not that we need their permission, but that as a good committee of any kind or as a good worker of any kind, you should evaluate the product with the fruits of labor. I wonder who assesses our work. How are we evaluated? Who are the stakeholders specifically customers that actually would care so we can address the question. Are we relevant and really doing it in a formal and a discipline way? I think that would be a mark of a very high performance team that it would be evaluating itself. I think there is the upstream issue of getting requirements right, the evaluation process.
And then downstream there is this whole space that we don't address well. We think that our products are self validating and marketing in their own right and they are not. Any other team of this form, a high-performance team of this kind might have a marketing committee. Whatever you want to call it. Communication and dissemination. A lot of our major grant programs now have whole funding line designed for dissemination. I wonder what is our dissemination strategy as a component of this we look at NCVHS.
DR. CARR: Excellent. Leslie.
DR. FRANCIS: I wanted to say that I think in a way the most important part of our meeting is the final session where too often we have just sort of had report outs from subcommittees and people have then gone their separate ways. I would like to suggest that what the subcommittees are going to be doing is essentially in light of what we have talked about working out something that looks like this is our year's plan, maybe even a little less sketched out a 2-year plan, some of that sort. But what is utterly critical is that we have say an hour at the end where each subcommittee takes 10 minutes and then we talk about them together.
DR. CARR: Agree. Excellent. Someone else had their hand up. Marjorie. Whoever speaks let's just with a headline and we don't need to elaborate now. If we need to elaborate further, but headline, Marjorie.
MS. GREENBERG: I don't know Walter was next. I just wanted to say because we don't have time, but I think Blackford has raised some good points. We do have some ways that were required and we attempt to do to evaluate what happens to the committee's recommendations and all that. I know that Debbie and I and Jim who are involved in this would be delighted to engage further on his -- I am not answering them but to say that I think that is something that could be discussed in a future, executive subcommittee call or meeting.
DR. CARR: Great idea. We have heard some terrific ideas and we need to -- and I think we are close to the direction. I think this has been very valuable to sort of step away from where we were yesterday, build on that, and get focused and feedback.
MS. GREENBERG: What I am thinking is now the subcommittees really need to discuss this rather than trying to come to closure now. We can come back tomorrow.
DR. CARR: I think that was the original plan to take the discussion that we have heard today and come back with a direction that fits with the general themes that we have heard: a learning health system, our role in advising on data, our ability to assess the gaps and to ensure that the data is what we need to assess health and health care reform.
And then thinking also about what is going on now with the community data initiative and the privacy population privacy that Sallie brought up. I think these are wonderful ways of defining the areas that we are going to look at. I think take that back. We will meet this afternoon and we will spend time and I agree very strongly with you, Leslie, that that last hour or more is our most important time. We sometimes don't have as much time as we need. Let's think about that.
And again come out of the subcommittees with some bullets and some action steps so that we will have sufficient time to know exactly what we are going to do and when we are going to do it.
I think with that we will break for lunch a little bit early and we will come back here at 1:30 sharp.
(Whereupon, at 12:23 pm, recessed for lunch.)
DR. CARR: Welcome to the afternoon session of the National Committee on Vital and Health Statistics. Debbie, do we have a dinner check?
MS. GREENBERG: We need to know how many people are planning to come for dinner. It is at the West End Motel, one Washington Circle that we went to about a year ago.
DR. CARR: Located on New Hampshire Avenue between K and L Streets.
MS. GREENBERG: It is walking distance from here although the weather is -- how many are planning to come? About eight. Mark Hornbrook said he was coming. Nine.
DR. CARR: Next up is the quality letter. Blackford and Paul.
Agenda Item: Quality Letter - Quality Measures Roadmap - Dr. Blackford Middleton and Dr. Paul Tang
DR. TANG: We will tag team this. I will give an introduction and then Blackford will go over an initial read of our recommendations. We are not going to present it to for a vote because we want to go through the process of it going through the executive committee, et cetera, before coming here. This is a continuation of some of the things we have introduced in the past meetings and we will just go over some of these things and report on the hearing.
The first thing is what are the key questions. What is the issue that we are addressing? Why did we choose the issue? That is pretty easy. Justine said to. What did we hear from the testifiers? What are we recommending and why and some areas of controversy?
One of the reasons we embarked on this roadmap for quality measures is because I guess we are saying we are not completely happy with what we have. Why is it? Well, most of it is of course based on administration claims data, and a lot of it is very inward focused on the medical model of doing things. What is it that happens in a hospital? What is it that happens in an exam room? And very little is patient centered or understandable. Those are two things that we want to look into.
Also there is a lot of convenient sample bias which is well we know heart disease is a big thing. We can measure lots of things in there so a lot of our measures focus in on things like cardiovascular care or on preventing cancer. A lot of things by convenience not necessarily what is important to everybody.
We wanted to set up a roadmap looking at well what should we be measuring, for whom, and from whose perspective. That is sort of the perspective we took. We had a hearing and the goal was to focus in on the multiple of people who could be interested in quality measures, primarily concentrating on the folks who are really ultimately interested in the quality measures which are the consumer patients. It is nice to say. It is easy to say, but when you look at the existing quality measures with that in mind you go I am not sure that this really is helpful to people we are trying to address.
Knowing that those might be our primary target there are other people who depend on these measures and will depend on them even more so within a few years like the payors and the accreditors, et cetera. We want to take those folks' interests in mind, but probably from a secondary position rather than a primary. And as you know again if you think of HEDIS a lot of it has been driven primarily by let's say payors. Knowing that this is a different place, a different destination we are going to have to find a different roadmap to get to this different place.
The importance of identifying effective quality measurement is to give -- are several. One is the folks that are involved in the "care" delivery process like admissions. It is really important if we are trying to influence what they deliver that we measure something of interest to them. One of the top principles is let's come up with something that is very credible with this group.
The second piece is the health care reform which means we are going to be measured on more things than just individual processes and structure, but talking about population if you think of just ACOs by definition talking about populations like my colleagues --
And ultimately as I set up the whole discussion as we want to focus on the patient, what they are getting out of it, and what is their experience of this whole process. We want to take advantage of the stuff we heard this morning on the new sources of health information whether it is directly from the patients, from the providers, or other sources like CDC and other sources that generate health data or data that is relevant to health which can be more than the traditional form. Larry mentioned education, environmental issues, and social/cultural issues. All those things impact individual and population health. That is one of our primary motivations for even approaching this topic.
We took in a number -- we had a hearing and took in these various perspectives and tried to identify the gaps of where we want to go and a vision that I sort of painted from where we are. We realize that we needed to focus as I mentioned on more on the patient's perspective and take advantage of this new data.
I think what I am going to do is turn it over to Blackford who is going to go over some of the recommendations and some of the rationale or reasons for those recommendations. Again, this is just to expose you to some of these ideas, get your feedback, and then we will take it back to the subcommittee and work on a letter to present back to you.
DR. MIDDLETON: Thanks Paul. Going on to the next slide. There it is. Recommendation one. Let me just say thanks to Debbie who really marshaled the subcommittee all the way through this process in an accelerated fashion and was really super to work with. Much appreciated.
As Paul said, the backdrop here is really all about looking at quality measurement and the future of quality measurement in the light of emerging new data sources, new methods for quality assessment, and brought our construction of quality and the impact of both health care and other determinants of premature morbidity/mortality on outcomes to get a more comprehensive picture of health wellness and disease.
Recommendation one as Paul alluded to sets up this new focus on a patient orientation toward quality measurement. Well, you might think aren't all quality measures sort of patient oriented, but they are not really. It is about changing the hemoglobin A1C sufficiency metric from over 7.0 for the population to what is the hemoglobin A1C desired for this patient given their disease profile, their genetic makeup, their context of care and their social supports, community health, et cetera.
This patient orientation is really fundamentally a paradigm shift and that word came up in the middle of the hearings and our deliberations. We thought it was paradigm shift or C-Change and to get this right we have a lot of work to do to reorient things. Why current measures lack relevance to all stakeholders from this patient perspective particularly the patient of him or herself. Patient-oriented quality measures or existing quality measures aren't used much by patients. In fact, right now when I go buy a new car, I will look up the car on every website that has to do with autos and compare every feature of that car. And now we know that patients maybe surf health grades and stuff like that, but don't actually use all the data that they could in making their own individual health care decisions and what not.
At the same time we wanted to make sure that this patient focus didn't increase the burden on reporting for health systems and providers in particular. We want to make sure that is addressed as well.
DR. CARR: Do you want to cover all of them first?
DR. MIDDLETON: Either way.
DR. FRANCIS: I had a prior question that got alighted between you and Paul. I got on the website last night in preparation for this. I wanted to see if I could read the transcript and I couldn't find it. I wonder if you could tell us a little bit about who testified, something about -- the agenda was there, but -- and there were some slides from people and the Booz Hamilton report that you had, but I just would love a little bit more understanding of --
DR. MIDDLETON: I happen to have that handy, Paul.
DR. FRANCIS: That would be great.
DR. MIDDLETON: So the four panels were constructed based upon the big stakeholders that were identified and I think that was -- was that listed at the beginning? We viewed the key stakeholders as panel one focusing on what information and measures will consumers need to execute their functions as educated purchasers and stewards of their own health. It is a real consumer focus.
DR. FRANCIS: And who were the consumer groups who were there?
DR. MIDDLETON: Eva Powell, National Partnership for Women and Families, Judy Hibbard, Institute for Policy Research and Innovation at University of Oregon, and we had a couple of others. I have a draft agenda.
DR. GREENBERG: I can say that draft minutes have been distributed, but only to the subcommittee obviously and I just think putting in the back of our heads from a process point of view I think there is a transcript -- obviously the lead staff person who would have gotten the transcript posted probably as you know was Cynthia, but that is really not an excuse.
DR. MIDDLETON: I can give Leslie the sense today.
DR. FRANCIS: It would be really helpful.
DR. GREENBERG: In a way -- since we are not going to be voting on this letter as I understand it at this meeting, the minutes will be available, the transcript will be available, et cetera by the next meeting.
DR. MIDDLETON: I may have missed the set up if that wasn't said. This is trying to give the committee background information on why doing a quality roadmap letter. A draft letter has been constructed but we want to give you the back drop then give you testimony and minutes and then the letter for your review and deliberation of that. The next meeting we will fine tune.
Panel two was all about providers. What information and measures do providers need to improve quality and increase accountability? Fred Rachman, Chicago Community Health Services, Yael Harris from HRSA, Terri Collin from Indian Health Service.
Panel three. What do professional organizations, accreditation organizations and regulators need to assess clinical performance across the continuum? Kevin Weiss from the American Board of Medical Specialties, Margaret VanAmringe from the Joint Commission and Rebecca Lipner from ABIM.
And the last panel, what do payors and group purchasers need to measures of value and health care and use it for decisions about coverage? David Stumpf from UnitedHealth, Barry Bershow from Minnesota Community Measurement and we had one more. I am sorry I don't have it here.
PARTICIPANT: Arnie Milstein.
DR. MIDDLETON: And it was Arnie. Thanks.
MS. JACKSON: Thanks much to Matt Quinn who helped organized -- pulled it together -- amazing job out of AHRQ.
DR. MIDDLETON: The second of three recommendations has to do then with this idea of measurement of patient interaction with the health system and even performance aspects of the health system I think. The recommendation is improve measures of quality, cost, value, accountability, and experience that are patient-focused and assess the interactions of the patient with the health care team and health system.
There are reasons for this I think we have talked about and done already. Gaps in the existing measurement framework relating patients and providers and accountability in health care, providers identified gaps in standardized criteria for quality measurement, regulators need measures across sites and episodes, payors need methodologies to assess value and efficiency, and consumers want information and transparency about cost and cost needs to be assessed across the continuum of care with better consistency.
And the last recommendation is really thinking about all the ways in which these data might be used or these assessments might be used across the entire health care system in the oversight roles. For those stakeholders who are concerned about accreditation or regulation or licensure or certification hope that a new quality measurement system can help them with their assessments. The recommendation is regulators, certifying organizations, and payors should actively be involved in the coordination of information requirements to support creation of measures to assess quality, efficiency, and competency to ensure that they meet their goals in overseeing the health care system.
The reasons why: maximizing the utility of current and emerging data sources of rich clinical data in addition to administrative data, coordinated measurement from the provider's perspective to manage the burden of data gathering, and maintenance of certification needs being addressed and real time use of quality data in a learning health care system.
We had a whole bunch of recommendations actually which were boiled down to three with Justine's guidance and members of the committee. And we tried to keep them very punchy and sort of high level, but get the right words in there so that the full spectrum of issues that we see in a quality measure roadmap can actually be addressed. We can pick out and address.
Next I guess, Debbie, last thing is what is controversial about this or where is this really going to need NCVHS' attention. It does represent a paradigm shift and the full implications of that are not even yet appreciated. We will have to think all that through. Patient generated information or more generally new data sources can have variable quality and integrity to the data. We will have to figure out how to use new data sources appropriately.
The difference between internal and external measures of quality. Can they be harmonized so that what providers are measuring for their own performance assessment or quality assessment is the same or meaningful to what other stakeholders, patient, payor, accreditor, et cetera might need.
Making sure we don't increase the data collection burden so we don't have clinicians as data clerks, et cetera. Can we measure cost in a normative way across the system? I wouldn't say it quite like this, but how do we assess accountability so that we can not only determine how patients feel about their relationships to their providers, but assess also to the providers feel accountable and then make an assessment of accountability as a key construct of health care reform obviously ACOs, et cetera.
That is what we wanted to do today just present this background and open it up for discussion and then give you some homework between now and the next meeting.
DR. FRIEDMAN: Blackford, this is really interesting and really important and thought provoking. Just a comment about recommendation one which probably relates to something you are thinking but just didn't write down and I wonder if it is important to write it down and that is there is an enormous knowledge challenge posed by recommendation one which is the knowledge required to come up with the optimal hemoglobin A1C for Mrs. Smith as opposed to a uniform target for a whole aggregated group of folks. I think everyone around the table appreciates the magnitude of that challenge. And if that number isn't plus or minus some small amount, the right number for Mrs. Smith, it kind of deflates the purpose of the whole effort. I wonder if that challenge in actually making this work is large enough that it should be explicitly called out.
DR. TANG: One response to that, Chuck, is we have been thinking about this in different settings. Remember appendectomies at least before CT scans you needed to have something like 10 percent normal appendix when you operated because if you had zero then you were missing -- in the same way not everyone's -- your example is hemoglobin A1C. Even though the guidelines say less than seven, not everyone should be less than seven. But there is some statistical distribution of measures that should have some kind of mean. Whatever that is you can't have everyone less than seven because you probably are bringing some people making them --
The same kind of looking at it not just as a number and shooting for 100 percent that is an example of how you might turn this into a personalized goal -- I mean accommodate personalized goal --
DR. FRIEDMAN: I just want to be sure I understood the recommendation. I interpreted this recommendation to mean that every individual would based on some complicated algorithm have a goal for that individual which takes that individual's individuality into account. And the quality measure then for a provider, for example, would be how many of that providers' patients met their individual goal irrespective of what that goal happened to be. I am having trouble relating that interpretation of the construct to what you just said.
DR. TANG: At the individual's practitioner level they would be working with each patient with their goal, but would be measured from a population point of view or an ACO point of view or a society point of view based on a distribution around that goal. Does that make sense?
DR. FRIEDMAN: I get that. My whole comment was directed at the difficulty of establishing with validity the goal for each individual patient.
DR. MIDDLETON: I think the point is well taken because there is not yet a knowledge base to customize everyone's unique distribution for their own quality parameters. There is the statistical norm which we can, I think, start with, but this actually goes right to the new epidemiology. How do we combine all the new data sources as well as genetic profiles, personalized factors, as well as context, environment, behavior context, et cetera and make for you your right to hemoglobin A1C target or whatever the case may be. It is very exciting from a patient's perspective because it actually is what I need to know from a provider's perspective I think equally exciting because it is what is the reasonable target for this patient as opposed to some nonsense.
DR. CARR: Right now we have a list and it goes like this. Walter, Leslie, Garland, Mike, Bill and Jorge. I have a question too. Quick and to the point. A headline. Say it once. No need to repeat it. That is everybody.
DR. SUAREZ: I have two very short questions. The first one is about -- it seems that we have several initiatives going on around quality measurement. Certainly there is the whole effort around quality measurement for meaningful use stage 2 and 3 and a lot of discussion around what the measures and what the philosophy is. There is the national health care quality strategy and plan that HHS is developing and there is effort. How do you see this effort advancing both of the other two or complimenting the other two? I am thinking out of the first recommendation shifting the focus to measurement development that supports the needs of patient seems something good to do.
The other two I am struggling a little bit with how they really do the other efforts that are going on. That is my first question.
DR. MIDDLETON: I guess my quick reaction would be I think we of course aim to be informed by whatever the existing health care strategies for quality assessment are from whatever group, but NQF's role is to say well here is kind of the next set of issues that have to be worried about. NQF, for example, has done a great deal of work around the quality dataset and elaborating measure framework and all that abides by the sort of old now in the new paradigm the old approach to quality assessment, population base. This suggests that there is a new approach that measure developers and framework developers and others will have to consider. And that is the level that we are at. Pointing a new direction.
DR. SUAREZ: But this is recommending a new approach to be developed by someone or that we as a committee begin to then look at developing that approach.
DR. TANG: I think we are painting a roadmap for quality measures 2.0 and this is the expansion of the perspectives and the things you measure compared to where we are today. The proposals that we write are quality. We sort of develop a consensus driven roadmap for quality measures or for national policy at input. This is just advisory.
DR. MIDDLETON: It is a good question because this is establishing a new direction. All the different landmarks or weight stations on a roadmap have not yet been defined. One could elaborate a series of work products or request of agencies and suggestions. What I don't know -- this is the first time I have been involved in a letter like this. Do we actually go so far as to say ONC should do X and HRQ should do Y and NQF should do Z?
DR. FRANCIS: The question of how to measure quality in a more complex individualized way and the question of what information patients ought to have in order to make choices among providers seem to me to be very different questions. And as I looked at this I want to be sure they are kept separate because -- and I am not sure they are. Because when you say patients aren't using quality measures, they are going to health grades or something like that and they aren't doing it very often. Part of why they aren't doing it is that the choice set that patients have is not I want to try to pick the most efficient, very best care deliver. It is here I am. If I have health insurance with my employer provider insurance and a somewhat limited set of choices that that gives me.
In addition I am working in an environment in which in many states what might be the very most important information to me namely whether a particular physician has been sanctioned for misconduct is not available.
I think it is terrific to say we need to have far more personalized understandings of what quality measures are that aren't population based, but I wouldn't confuse this with the question of what is the information that patients need in their current context to make more rational choices about which doctor or nurse practitioner to go see.
MR. LAND: Actually my comments are following up on Leslie's comments. The information is a little dated. When I worked for the state health department in Missouri, we had legislation passed to develop consumer guides on health care providers. And when we did the literature review at that time, we learned that consumer guides actually are not used generally even to buy automobiles, dishwashers or anything, and certainly used even less frequently to make choices in terms of health care decisions.
Now it may have changed in the last 6 or 7 years. I don't know. With the availability of Google and so forth it might be different. I don't know what the literature says now, but that was certainly the case at that time. I doubt that it has changed much in terms of health care decisions as has already been mentioned.
I guess I am raising the same kind of question that Leslie was mentioning. What are the expectations if we are going to go in this direction for patient-specific performance measures? Is it for choice or is it for individual relationships with their physician? And if it is for choice, we just may be cautious that our expectations -- maybe they will be met in this new environment of where we are going with health care. I don't know. I just wonder if we are setting high expectations that there is no evidence to date that information is driving decisions by consumers.
DR. MIDDLETON: Just a quick thought maybe on these two comments. I think for sure there is a comingling of effects between how patients choose which providers to go to or which institutions to use or networks or even health care plans and that has something to do with convenience and access and my brother went there and he liked it, et cetera, but it is not completely distinct from, orthogonal to how good is the Brigham at labor and delivery. Or how good is this doctor more specifically with patients like me? And somehow we need to address both of these very important and distinct issues, but we need to find out how to do both in a way that is really patient driven and meets the patient's expectations both for choice as well as ongoing care assessment or quality management. I accept the point. I just don't think they are completely distinct.
DR. FRANCIS: I don't think they are completely distinct. I just want to make sure that our recommendation that is about an answer to one question isn't taken to be a recommendation that is an answer to another question.
DR. TANG: Can I respond as well? I think I -- listening to the comments that we may have confused you by saying that need to differentiate the individual care measure. The measure is used to deliver individual care versus the population measure. I think what we are saying is we need to be treating individuals with an individualized care plan. It is not the same thing as being able to relate all that to population measures. One way was this statistical distribution that I mentioned. Another way and this was stimulated by comments Eva Powell made is wanting to know not the risk adjusted score for somebody because I am much rather concerned about my risk population.
It is ironic. Maybe instead of reporting risk adjusted score which essentially makes everybody migrate toward the mean, maybe you need to publish stratified scores so that a patient, an individual can relate to the strata that they belong to. That is one way of making the same data and making it something I as an individual can relate to and that is more of where we are headed from a population point of view than saying how are we going to report your performance against personalized goal for each individual. Was that somewhat helpful? See what I am saying?
MS. GREENBERG: Well then also maybe you don't want to do as well as that group you happen to be in. We have to do as well as the other group.
DR. TANG: But you can decide.
MS. GREENBERG: But you can see.
DR. TANG: Right now you can't see any of that. That was part of the point that came out.
DR. CARR: My list goes Mike, Bill, Sallie, me, Jorge, Larry, Marjorie.
DR. FITZMAURICE: Recommendation one is shift the focus of measurement and measurement development and activities. The question I have is -- by the way, I think this has been a tough area and that you really -- a good job at laying it out. But I have a question with recommendation one. How do you do this? What secretary action is required? You are addressing the secretary.
The second one, develop and improve measures of quality cost value that are patient focused and assess the interactions of the patient. I wouldn't want us to lose a focus on how about the outcomes of the patient. Does the patient get better faster and can do things better?
Thirdly, regulators certifying organizations and payors should be actively involved. Find out how the secretary can do that. Invite him to the table and have them hold hands. But to ensure that they meet their goals. Suppose one goal is to maximize the number of patients that are seen and get better. Another goal is to maximize (?)31:51 from insuring them. Maybe they don't all meet their goals.
Tell us what the secretary should do and how it should be done. Tell us how should quality measures be changed and who should do it. Now maybe some of it is research.
DR. SCANLON: I had a similar reaction to Mike's in terms of what is really actionable here. And (?)32:23 was recommendation one. To be honest patient centric -- I had many ideas or definitions floating through my head. I got a new one today when you defined it and I think therefore it is very important that we actually be explicit about what we mean. I might have called it risk adjustment that I don't sort of use a measure that is not particular to an individual, but I sort of adjust the measure for the individual. That is one aspect of it.
I think that this issue of ideal data in terms of what people might want to use has to be juxtaposed with what is realistically feasible data. And the small sample problem is very real. If you want to start to extend the measures beyond certain diseases, your small sample problem becomes even more acute. And we are not talking about dealing with systems. For the foreseeable future we are going to be dealing with individual physicians for a very large share of our measures and so then what will that do to our small sample problem. If we send this letter to the secretary, we have to warn the secretary about these kinds of issues.
The second recommendation I feel like we are kicking the responsibility to the secretary. Figure these things out. There are a lot of issues that maybe came out of the hearing in terms of what is the pros and cons that need to be considered here because these are not easily reconcilable issues.
And the third one I have a very similar reaction to Mike is where is the secretary's authority to do anything about this third area which has been problematic. We have heard about this in multiple hearings over the years and not just in this committee, but in other settings where providers talk about the multiplicity of quality measures that are in some respects aiming at the same thing but with tweaks that make it very difficult to be compliant.
That is a problem that is systematic of the fact that there isn't any sort of one entity and control here to specify these measures and the secretary doesn't have that authority. We are talking about private insurers making requests. We are talking about ERISA plans making requests. Where is it that we might look to sort of try to correct this situation and I think that is one of the difficult things to think about?
DR. OVERHAGE: Can I have a response or a comment specifically?
DR. CARR: I think we have to give Sallie a chance. Hold that thought because I think we are all on the same thing and Sallie has been waiting.
MS. MILAM: When I look at the recommendations, one and two are very consumer focused. And recommendation three brings in other stakeholders. A consumer's view of the world would be very different than a payor. There is inherent tension and I guess it is somewhere between often times and sometimes they would view measurement very differently. And if this is the world of quality measurement, these recommendations encompass how we think quality should be. You have the payor in a very different position here. You have a payor making sure that information is available to support the data measurements identified by the consumer. I am wondering were payors involved in the hearing. Are they supportive? And how did these three isolated recommendations address the inherent tension and balance that needs to be in place so that everybody would accept these measures?
DR. MIDDLETON: I will take a crack at it and ask Paul to chime in as well. We did have payors on the fourth panel and their perspective was in a nutshell basically they desperately want more data for more sensitive assessments and quality and value particularly clinical data to combine with their readily available administrative data to make more fine tuned assessments of value of care rendered and services purchased.
On the tension front I think the tension is real and the goal is not to ameliorate or eliminate the tension, but rather to make sure that the needs of different stakeholders were all being -- as opposed to ignoring one or the other. It is acknowledging that there is a real tension, but can we find the data that meet the consumer needs as well as the other stakeholders as we have described.
MS. MILAM: But that is a much recommendation.
DR. CARR: Thank you, Sallie. Jorge.
DR. FERRER: I wanted to make a comment I guess from that experience of the VA and this goes to the point of Blackford and Paul brought up with (?)37:30. The VA within our medical home innovation construct when started looking at better ways of improving the care of the VA. We have gone so far starting with a simple diabetes registry. As that system gets sophisticated, it morphs into a care management tool and it has become a total population tool that can span almost any disease and incorporates the patient's input which is critical. You have to have the view of how is the patient expressing their outcome because as a surgeon you could say well he didn't have a surgical infection post 48 hours. Patients might not really care about that because he wants to know if he has gait deambulation and when can I -- so people look at these things very differently and if it is truly patient specific, you have to commit the system to capture the patient's view.
DR. MIDDLETON: Actually one of the most interesting presentations was on the patient activation model. The lady from Oregon -- it was terrific. It was clear that this is one way we might need to actually stratify assessments because patients who are nonactivated will have very different outcomes. We know that. But also very different assessments of their own quality. It is a subtlety that we did come forward.
DR. GREEN: I would like to observe mostly for Blackford and Paul in public here that listening to the comments makes me think that recommendation one and number three particularly are subject to reinterpretation and misinterpretation of what the intent was. Having attended the committee and listened to the conversation here has sort of gone off in another direction to some extent in my opinion. I would like to come back to what we learned in the hearing and make two suggestions about how we might lead into recommendation one and recommendation three with an intention of demonstrating fidelity to the results of the hearing.
Recommendation one really crystallized after the hearing discussion around this notion I believe that the work so far on the development of quality measures has largely been stimulated and driven by the needs and aspirations of the payors and the purchasers and the provider groups that want to use quality measures to advantage themselves in the market place.
Intent of recommendation one is to say the secretary you know what he is right. We need to enlarge the quality measurement development focus in a way that shifts it toward patients themselves so that the quality measures that are developed advantage patients as oppose to advantaging a payor. I thought that was one of the great insights from the work there. The discussion (?)40:49 about recommendation one sort of misses. That goes by that and we went somewhere else. We need to step this us.
DR. FRANCIS: Could I say? As the primary determinant of their own health and wellness that threw me off track there because if what it said was to support the quality concerns that matter to patients, it would have read very differently.
DR. TANG: That was there partly -- it is the new beneficiary. I was going to use the word target but that seemed to -- so really it is not only the patient as what happens to me and can I understand what happens to me, it is me as the primary determinant. We did actually -- those words were reflected in the comments and discussion is how can I be -- it is just like buying something. You are an active role. You intend to use this information, et cetera. We still can wordsmith it, but we want to keep that notion too.
DR. GREEN: If I can move to number three, there is a similar set up that might or might not help in the construction of the letter. But recommendation three I believe really came from us recognizing how much activity there is by groups that have not been full borne participants in quality measurement development to develop some sort of quality measure for their own use and this included certifying organizations and others.
I thought this recommendation built on our prior acceptance that we want to collect data once and use it often and that we don't want different groups creating different data elements that people have to collect in a different way for a different use, but that in fact as you make the roadmap -- remember we were trying to define a way forward for the secretary to follow. Once you do recommendation one and you shift in that direction, I thought recommendation three's primary purpose was to say you are going to have to create a space in which other people other than the payors and the providers and Medicare are defining what these measures are because there is a lot of people that are doing it and that needs to be harmonized because it would enhance the probability they are being used and used well and because they are so hard to develop. If we got a handful of really good ones, wouldn't that be fantastic and then we could use them in six different arenas. And I just don't think that idea is captured here enough for the committee to grasp.
DR. MIDDLETON: The point is well taken about harmonizing because that did come up.
DR. CARR: Marjorie then Marc then Bill.
MS. GREENBERG: Two sentences. First sentence. I actually thought to me --
PARTICIPANT: I am just going to interrupt for a second because we are not wordsmithing PowerPoint, right?
MS. GREENBERG: No. Two sentences from me. I had told Justine I had one sentence. I now have two sentences. The most interesting -- just maybe it needs more development. I think what Larry said just reminded me of it and triggered it. The most interesting sentence I saw and I don't know if it was in the PowerPoints or the actual draft letter, but what was that clinicians use different quality measures than the ones they have to report and that that represents a problem. Some of it you can understand, but that was very compelling to me.
My other sentence is that I am not quite sure whether it fits in as a fourth recommendation or a fits into the other three, but these recommendations just from the discussions we have had cry for a recommendation that research needs to be conducted in this area. It is not just a question obviously of creating or developing different measures. It seems that there is a lot we don't know about measures that would be more meaningful to patients that would take into account patient's issues, but also be scientifically accurate. Patients don't always want the best thing for themselves and I am among them.
I think that is what I have learned from our letter at the last meeting which I don't want to dwell on, but sometimes what we really need is more research and that is actionable whether it be through AHRQ or NIH or I don't know what, but that is actionable by the secretary I think if that we do need that and it sounds to me like we do.
DR. CARR: Thank you, Marjorie. Now I have Marc, Bill, Walter, and Leslie. We are planning to close this at 2:30. Marc.
DR. OVERHAGE: Just a brief comment on Bill's point. I think the secretary actually has pretty powerful levers to help influence the direction the quality measures go through what CMS does with Medicare and Medicaid as well as whatever influence they may have with DoD, VA and OPM.
The other comment -- I suspect Paul and Blackford have this in here and it is just buried in code, but I would like to know where to look for it is this notion that and I think it is probably in this first recommendation, but maybe as we go forward stands to be called out more is this ongoing challenge of we tend to measure things that we can measure. We tend to focus on things that rise to the top for various reasons that systematic sort of thought about. What are the five important things to do? And I say five explicitly. That may not be the right number but it is not 500 because trying to change the system in a system this way. You sort of micro-optimize it. If you have 500 measures for those and you don't get the system better and I don't quite know how to capture that and I suspect some of that don't look under the light post is in this first one. Look where it is important to patients and stakeholders but also wanted to capture that notion of I think there is a risk of too much breadth if we really want to make an impact.
DR. SCANLON: This relates both to what Marc just said and Larry. I think that I believe Medicare is a very important factor in terms of the development of quality measures. First of all, the Hospital Quality Initiative which there is real money attached to and then the PQRI. Medicare is not motivated by market share. It does not have to worry about market share. If given the right guidance in terms of what the right measures are, I think the program will respond and is the leverage that Marc indicates that the secretary has.
On top of that within the Affordable Care Act there are explicit provisions for you have to come up with pay for performance or value-based purchasings or the initiatives. Tying what we recommend to that I think makes it much more currently relevant, but it has to be done in the context of what is also currently feasible.
The concern is and the part here -- I am observing what has been happening with Medicare is that it is not payor driven as much as it has been provider driven. There have been people who are very supportive of physician pay for performance and we are incredibly disappointed after PQRI was enacted because they saw this scramble by every specialty to get some measure in there so that they could possibly qualify for their share of the pie. These are not things that are necessarily patient oriented. Is that antibiotic before surgery really that sort of critical? There has been a lot of evidence that says no.
It is this issue of the right measures and what the process is that is going to get us the right measures is not totally clear because if you are a Medicare program you have to involve the providers. They are the professionals. They have the knowledge. You can't viably in this country say this is what we are going to do. We know medicine better than everybody else.
DR. SUAREZ: Back in the early ‘90s when I was in Minnesota we started developing all these concepts of report cards for health plans back in those days and trying to understand it. There were two important concepts that I always remember. One was there is a very distinct and separate issue about quality measurement for improvement versus quality measurement for accountability. We always had that issue of what is it that gets reported outside to measure for accountability what we are doing, that is providers, versus how and what we measure internally for purposes of improvement, for purposes of patient safety, for purposes of a lot of reasons including not less payment and reimbursement of providers. Performance as in employee performance kind of a thing.
I think it is going to be and it has been highlighted, but it will be helpful to have that preamble discussion about the importance of distinguishing quality measurement or purposes of accountability versus improvement.
It always come to me back when we are talking about this because we have to do it back then and I think it is part of what the research could provide. We have heard across the discussion different aspects of measurements. As a patient, I want to know about the quality of providers and of payors. As a payor, I want to know about the performance of providers. As government, I want to know about everything.
When you are creating matrix of patient, provider, payor, accreditation, and government and cross it with the same kind of angle, you begin to identify who is looking for what. What are the needs for quality measurement and quality data from different perspectives? What is it that health plans want about providers and what is it that patients want about providers or health plans?
This has been done and documented probably, but there is new evidence about new needs, new expectations much more into the point that I think is in this particular which is I think a very important one is consumers decide tailored information relevant to their personal health care needs. Now are not sort of the overall measures of quality on a performance or a report card of a health plan that doesn't relate to me because that is just too general. I won't be able to relate it. It is much more personalized, tailored. I think that is where the research opportunity comes in trying to understand and identify the new needs, the emerging needs for quality measurement at these different points. Creating that kind of a matrix that helps understand so the framework around that is looking for quality information and what kind of quality information is being looked at will help.
DR. FRANCIS: The phrase the primary determinant of their own health and wellness is crucially ambiguous as I understand it. One understanding of it is that the patient has responsibilities for managing his or her health and the information that a patient would need to have in order to participate actively in that kind of management is one sort of information some of which we have, some of which we don't have now.
The second meaning of it is that what matters with respect to health or wellness is going to be different for different patients. That is the point Jorge made. It may matter to me how much mobility I have and I might rate my wellness differently depending on mobility.
It may matter to me whether I am treated abruptly or not because I might prefer a world in which I am treated nicely to a world in which I live longer. There are people who have. That is silly but it is true.
Be careful when you frame that about whether -- what you are really meaning to do is to say measures of health are different for different people or to say that what we are trying to do is reconstruct quality so that we allow for much more patient responsibility for health.
DR. CARR: Thank you. Thank you for pulling that summary together and the work that you have done in a short amount of time. I think it really afforded a great opportunity for discussion.
I wanted to just take a few minutes to talk about two things. One is how we do our work and then the other is just to restate the charge for the work groups this afternoon. We have had occasion to really look at what works and what doesn't in our various processes and we talked a bit about it yesterday and I will try to hit the high points. One is as you are a planning a hearing, the pre-work is important in terms of identifying who are the stakeholders, who are you having, all of that so that you can bring it to the full committee. You want to get some clarity. The subcommittee makes the plan, comes to the full committee. We will hear back tomorrow if there is some things in the works. That is an opportunity for someone else in the group to say don't forget to invite this person or that person or include this stakeholder. That is very important.
Second, at the hearings as we said we will have the transcripts and the PowerPoints and participation in the hearing. Ideally all the subcommittee members can be at the hearing, but if they can't a couple of things. Someone else can attend in their place or they can read the transcript, but we want to make sure that everybody is on the same page about the hearing.
After the hearing, the development. We have talked about trying to pull together these talking points so that we can have a meaningful discussion around it followed by the letter with the recommendations.
As the letter goes around, we talked yesterday about the challenges of version control and keeping track of comments. We want to make sure that we have a clear record of all of the comments and that they have been addressed when we are discussing it. Sometimes two people might say something that looks the same but it has very different meaning to them. Marjorie did a great job with the privacy letter of simply cataloguing every single comment and everyone was addressed and similarly with version control. We are going to look to staff to give us some help on version control so that we are all working on the same -- we are correcting the same version, not the one that came before.
Then when the work product is complete from the point of view of the subcommittee, it comes to the executive committee and that is an opportunity to identify any rough spots that may need more work, may need some assistance, and to determine whether it is ready to go to the full committee. When it gets to the full committee, they should have heard about it once at the time of the planning, the second time with the preview through the PowerPoint and we should be pretty much informed and have all the issues identified at that time.
The other thing that we talked about is having the opportunity to have some scheduled NCVHS time and staff is going to work to come up with probably a Friday afternoon a month that is NCVHS time. If you need to call your subcommittee together at that time, everybody will have blocked out more than one group needs it. It is probably possible as well. If we don't need it, we don't use it. But I think we have found that our schedules have been so challenging lately, but it delays the timeline.
Were there any other thoughts about our process or suggestions? Marjorie.
MS. GREENBERG: I may have phased out here, but did you mention that our processes are for letters to go to the exec after the subcommittee has signed off on them to go to the executive subcommittee.
DR. CARR: I did.
MS. GREENBERG: Also, we are going to try to do less and really focus on the recommendations and not wordsmith the other stuff.
DR. CARR: True. We don't want to do wordsmithing but as we talked about also the context around the recommendations is also important for everyone to see. I think it is a tension one way or the other. But I think if you follow this process and adhere to it and I am appreciative to Paul and Blackford because we talked about this process today and now this is what we are adhering to. We won't get to a letter tomorrow although we have time allocated tomorrow to talk more about this letter. It might be helpful to come back having heard what you heard today with here is the direction we are going in so that we are on the same page.
DR. TANG: That wouldn't go in the process though. I think there is value in the process you laid out for the subcommittee to work on some things --
DR. CARR: No, we wouldn't talk about a letter tomorrow, but I am offering you an opportunity to touch bases with the committee once more tomorrow. You have a subcommittee meeting this afternoon from 4 to 5. I will leave it to you. If you would like to use your time not for the letter but for further discussion, we can to do that. If you are all set, that is okay too, but the idea being that we would all be up to speed. It sounded like there is enough question about the direction that I would encourage you to use the time tomorrow to update the committee.
MS. GREENBERG: Just from the point of view of version control and all of this which is challenging to all of us in all of our realms I am sure, but we have established the SharePoint and some people find it useful. Some people don't. I think we can do a better job pushing it out to people, but also we might consider saying that when a subcommittee will say is commenting on a particular document. The version that is being -- you know well has been sent around by email. I know we get so many emails. In addition to maybe getting it by email the version being commented on we could post on SharePoint and even staff could do that. At least you know everyone is commenting on the same version because I think that often is -- we could sort of edge you into the SharePoint and see where it helps and where it doesn't, not as an alternative, but as a supplement maybe.
DR. CARR: Just the active copy available in SharePoint for anybody who wanted to check that.
DR. WARREN: Do we all get SharePoint accounts? We have them and then we -- I know I lost mine.
DR. CARR: You lost yours? Okay. We will look into that.
DR. TANG: Just an administrative question on SharePoint. Do you have to then detect something out and then it gets locked or how does that help with version control?
DR. CARR: No, I think what Marjorie was saying is that if we just keep posting the most recent one when you go to do your work, you could check and say am I working on the right version.
DR. TANG: The same problem with email.
MS. GREENBERG: You would have gotten it by email if you can go and check to make sure if you are working on the right version. Something to think about.
DR. CARR: Hold that thought about SharePoint. Maybe Blackford or frequent users can give us some guidance on that. I think that as you are planning your work, plan out when your calls are, when your drafts are going out, when they are due back and maybe just at least adhering to those dates.
I don't want to be overly prescriptive, but obviously we needed to come back after the last meeting with some clarity on processes and we will try this and see how it goes and thank you, everybody, for what you have done so far.
The rest of the afternoon is to develop what you heard this morning. I want to thank everybody for their very rich contributions to the discussion today. I think it was helpful for the work done yesterday to set a platform to set a direction, but really the development today was terrific and please try to recapture what your thoughts were at that meeting and come back with a plan that is exciting and actionable and meaningful.
MS. JACKSON: Where are people this afternoon for logistics? It works out I think pretty fortuitously that the populations and the quality subcommittees will be in the same room. We are going to start with populations. The original plan was for the populations to be here and for privacy to be in another room. But from the discussion earlier you were talking about having populations and privacy talk a bit before --
PARTICIPANT: We are going to start by meeting together.
MS. JACKSON: By meeting together. So we can use this same room to start now for the two subcommittees, populations and privacy, meet as long as you want for 20 to 30 minutes. And then privacy when they finish will go up to the third floor. Room 3015 is the break out room for the concurrent session here. And then at 4 o'clock then quality would come on into this room. We will use this as the base of operation room.
Standards tomorrow. What is the plan for Standards? Will they be in this room? Standards will be in this room tomorrow morning. For this evening we can gather about 6:15 or so for those of you going to the dinner we will be in the lobby. For those of you who are staying here. Otherwise we will just meet you over at the restaurant for the 6:30 seating.
DR. SUAREZ: For those on the phone they just send an email here. If they are meeting, do they dial a different number to go to a --
MS. JACKSON: No. The numbers were aligned. Populations and Quality -- that is a question about Privacy. Privacy was meeting with its own number. When you get back at 3 o'clock then privacy will plug in.
(Whereupon, at 2:40 pm, the Full Committee adjourned.)