[This Transcript is Unedited]
Department of Health and Human Services
National Committee on Vital and Health Statistics
June 14, 2016
Capital Hilton Hotel
Federal Room A
1001 16 St., SW
- Call to Order, Review Agenda – Walter Suarez, MD, Chair
- Updates/Reports from Committee Members – Walter Suarez, Chair
- Updates from the Department
- Jim Scanlon, ASPE
- Charlie Rothwell, MS, NCHS
- Shana Olshan, CMS
- Elise Anthony, ONC
- Standards Subcommittee – Ob Soonthornsima and Alix Goss, Co-Chairs
- Public Health Data Landscape
- Susan Queen, PhD
- Marcie Cynamon, MA
- Clarice Brown
- Hyon Shin, Census Bureau
- Nicholas Jones, Census Bureau
- Laura Conn, MPH, CDC
- Work Group on Data Access and Use
- Vickie Mays, PhD, MSPH
- Helga Rippen, MD, PhD
- Joshua Rosenthal, PhD
- Public Comment
P R O C E E D I N G S
DR. SUAREZ: Good morning everyone. We are going to get started with this meeting of the National Committee on Vital and Health Statistics. We are connected already. We have people on the line as well. We are going to move ahead.
We are going to go through our regular introduction process in just a second. We will verify that we have indeed a quorum to conduct business and take action. But before we get started, I wanted to take a minute and pause for a second and reflect and really take a minute of silence in honor and in remembrance of the victims of the recent event in Orlando. We are going to take a minute to pause and really want to honor them. Thank you very much for this moment.
We are going to move forward and begin our meeting. We are going to start with introductions. We are going to go with introductions first of the members of the committee. If you could state your name, organization, and your membership in different subcommittees or activities and then if you have any conflict of interest in any of the issues we are going to address today.
I will start with myself. My name is Walter Suarez. I work for Kaiser Permanente. I am a member of the Full Committee, chair of the National Committee and member of all of the subcommittees and the work group. I do not have any conflicts.
MR. SCANLON: Good morning everyone. Jim Scanlon. I am the deputy assistant secretary for Planning and Evaluation at HHS. I am the executive staff director for the committee.
DR. MAYS: Good morning everyone. Vickie Mays, University of California Los Angeles. I am a member of the Full Committee, a member of Privacy and the Review. I have no conflicts.
MS. GOSS: Good morning. I am Alix Goss. I am an independent consultant. I am a co-chair of the Review Committee and the Standard Subcommittee and a member of the Full Committee and I have no conflicts. I plan a couple of other groups, but not consistently.
MS. KLOSS: Good morning. I am Linda Kloss, a health information management consultant. I am a member of the Full Committee, chair of the Privacy, Confidentiality, and Security Subcommittee, hope soon to be co-chair, member of the Standard Subcommittee and the Review Committee and no conflicts.
MR. LANDEN: Good morning. I am Rich Landen. I am with Harris QuadraMed and I am a member of the Full Committee, member of the Review Committee and the Standard Committee. No conflicts.
DR. O’GRADY: Good morning. I am Michael O’Grady. I am a senior fellow at North University of Chicago and principal of O’Grady Health Policy LLC. I am a member of the Full Committee and the Subcommittee on Population Health. I have no conflicts.
DR. EVANS: Good morning. I am Barbara Evans. I am a law professor at the University of Houston and I am a member of the Full Committee and the Privacy Subcommittee and I do not have any conflicts.
DR. CORNELIUS: Good morning. I am Llewellyn Cornelius. I am a faculty member at the University of Georgia Athens. I am a member of the Full Committee and Population Health Subcommittee. I have no conflicts.
DR. PHILLIPS: Bob Phillips. I am the vice president for Research and Policy at the American Board of Family Medicine. I am a member of the committee, the Subcommittees on Public Health and Privacy and no conflicts.
MR. COUSSOULE: I am Nick Coussoule. I am the senior vice president and chief information officer with Blue Cross Blue Shield of Tennessee. I am a member of the Full Committee, the Standard Subcommittee and the Privacy Subcommittee. I have no conflicts.
DR. RIPPEN: Good morning. My name is Helga Rippen. I am CMO of Allergy and I am on the Full Committee and the Subcommittee on Data, Population Health and Privacy. I have no conflicts.
DR. CHANDERRAJ: Good morning. I am Raj Chanderraj, a practicing cardiologist with Nevada Heart and Vascular Center, member of the Full Committee, no conflicts.
MS. LOVE: Denise Love, executive director of the National Association of Health Data Organizations and co-chair of APCD Council, which is a non-legal entity joint collaboration with the University of New Hampshire. I am a member of the Full Committee and I try to be a good member of the Subcommittee on Population and Standards and no conflicts.
DR. COHEN: My name is Bruce Cohen. I am from Massachusetts. I am a member of the Full Committee, co-chair of Population Health and on the Data Work Group. No conflicts.
MS. HINES: Rebecca Hines, Executive Secretary of the Committee. I am with the National Center for Health Statistics.
DR. SUAREZ: Do we have members on the phone?
DR. SOONTHORNSIMA: Good morning. This is Ob Soonthornsima, vice president of IT Specialty Systems at CVS Health. No conflicts.
DR. SUAREZ: Anyone else on the phone?
DR. LINCOLN: I am Mike Lincoln from Department of Veterans Affairs. I have no conflicts. Privacy Subcommittee.
DR. SUAREZ: Any other members on the phone? Any other members of the National Committee? We can confirm that we have a quorum, our quorum given our full complement of 18 members is to have at least ten members and we have close to 15. We have a quorum. We can proceed. Thank you very much.
We are going to go quickly through introductions of members of the public and staff as well.
MS. SQUIRE: I am Marietta Squire and staff to the Committee.
MS. LOHSE: Gwen Lohse, CAQH CORE.
DR. LAZARUS: Steve Lazarus, Boundary Information Group representing CAQH CORE. I will let you know that the public phone number has no audio.
MS. WEBER: Erin Weber, CAQH CORE.
MS. SMITH: Gwen Smith, Veterans Health Administration.
MS. WILLIAMSON: Michelle Williamson, CDC’s National Center for Health Statistics.
MS. KNAPP: Katie Knapp, Veterans Health Administration.
MS. TUREK: Kelley Turek, America’s Health Insurance Plans.
DR. BRETT: Kate Brett, NCHS and lead staff for population health.
MS. NARCISI: Jean Narcisi, American Dental Association and WEDI chair.
MR. STELLAR: Charles Stellar, WEDI.
MS. JONES: Katherine Jones, NCHS and staff to the Committee.
MS. QUEEN: Susan Queen, NCHS.
MS. KOCHER: Gail Kocher, Blue Cross Blue Shield Association.
MR. DECARLO: Mike DeCarlo, Blue Cross Blue Shield Association.
MS. JACKSON: Debbie Jackson, National Center for Health Statistics, CDC Committee Staff.
MS. DEUTSCH: Terri Deutsch, CMS.
DR. BURKE-BEBEE: Suzie Bebee, ASPE and staff to the Committee.
MR. RODE: Dan Rode, independent consultant and educator.
MS. OLSHAN: Shana Olshan, CMS.
DR. SUAREZ: I think we have everyone. We are going to check to make sure that audio for the public is working. The public’s audio comes through the WebEx. What kind of audio do they get? Some report that there is no audio on the WebEx. It is now working. Thank you.
We are going to start with our agenda. This is a very full agenda as well as a full week of meetings. It is one of the more intense weeks we probably ever had. We very much appreciate all the work that everybody has been doing in preparing for this. But we are now interested in including in our agenda a very first short opportunity for committee members to report any updates or activities that involve them with a representation of the National Committee. I know Bill Stead, who is not today with us, but will be later, has presented on behalf of the National Committee to the HHS data leads and it was a very good conversation, very important one about the activities that the National Committee has been doing on population health and data access. Vickie was also present and participated in that discussion. I just wanted to briefly report that it is a good connection and liaison and reach out to others.
I know there are other committee members. I am going to give an opportunity to any committee member that wants to express a few thoughts about any activities they are involved that relates to the National Committee.
MS. GOSS: Good morning again. This past spring I had the pleasure of presenting our review committee letter that we sent to the secretary on February 29 with several organizations. I presented it at the HIPAA Summit in March and then at the Pittsburgh Business Group on House Legislative Forum. There is a definite interest with the letter and opportunity to not only promote the good work of the Full Committee, but also get some education out there about who the committee is, what role we play, our history. A lot of people are not really aware of us. There has been really great feedback from the opportunities to present those.
DR. SUAREZ: Excellent. Thank you so much, Alix. Anyone else?
MR. LANDEN: Rich Landen. I have an upcoming speaking engagement with the National Plan Automation Group, which is an annual meeting of a group of EDI professional middle managers of the Blue Cross Blue Shield Association Systems of Plans that is scheduled for September. If anyone has questions that would be appropriate for me to seek feedback from that group. Again, these are health insurance plan EDI middle management that does a lot in the day-to-day EDI exchange with the provider community. If there is any particular feedback that would be appreciated from that group, let me know and I will incorporate that into my agenda.
DR. SUAREZ: Thank you very much, Rich. Just a quick reminder for everyone. This is procedural and part of our responsibilities under the FACA law. We all certainly present and participate and are engaged in speaking activities and lots of meetings. If you are invited to present on behalf of the National Committee or will be presenting about the National Committee, it is important that you all communicate to us particularly Rebecca and me so that we have the appropriate process in place to do that.
DR. HINES: And especially because if you are representing the Committee, you cannot accept travel or anything like that. Then we would have to cover it. We would have to assess whether that makes sense. If you are going to be speaking on the committee’s behalf, we need to talk. Thanks.
DR. SUAREZ: Thank you. We are going to go to our updates from the department. I am going to turn to Jim.
MR. SCANLON: Thank you. Good morning again everyone. I want to bring everyone up to date on some personnel changes in HHS, a couple of policy initiatives that I wanted to bring the committee’s attention – data and metrics and then bring you up to date on some of the data policy activities within HHS.
On the personnel front, our ASPE – Richard Frank has left and returned to Harvard Medical School after about seven and a half years here at HHS. That is a long leave of absence. But at any rate, he has returned there. He is the professor of economics.
Katie Martin, who came to HHS as a counselor to the secretary, is our new ASPE chief. She is an analyst. She understands the kinds of things we are interested in. Her specialty is health policy and financing. She just started this week.
I think I reported previously that many of you know Mary Wakefield. She is the acting deputy secretary and she has been nominated to be deputy secretary. She was the head of HRSA. Many of you know her program.
Joel Cohen. Many of you know Joel. He is the director of the Center for Financing, Access and Cost Trends at AHRQ. He has taken over Steve Cohen’s position.
At SAMHSA, Daryl Kade – I think many of you know Daryl. She is the director of the Center for Behavioral Health Statistics and Quality at SAMHSA. She is the new director of that program as well.
I will say a brief word about budget, not that we know that much, kind of reading the tea leaves. But we are almost in the fourth quarter of the year 2016. July, August, and September. That is really our fourth quarter. The new fiscal year begins on October 1. Congress is working on that appropriation. We are getting a look at what the Senate is thinking about. We do not have much in the House. But, again, we are hoping for the statistical – of course, statistical activities we are hoping to at least maintain the core capability and potentially even a little bit more possible. But, again, it is really reading the tea leaves. No one really knows.
As you know, we already think about the fiscal year 2018 budget. Obviously, that will be a transition year. There will be a new president who actually presents that budget in probably February. OMB and this is typical simply asked that the agencies prepare more or less a current services budget, in other words, take fiscal 2016 and allow for salary increases and other things that are allowable. But otherwise, it would be more or less level. But that is not a policy budget. We call that a non-policy budget.
And then when the new president’s team arrives, whoever it is, and during the transition period, they will then add, subtract, create new initiatives, whatever the team feels the policy should be. That would be the budget that the president presents probably in February or so. This year is a little different obviously for obvious reasons. But, again, it does make it a little more difficult for our statistical programs and others to plan.
We had a roundtable at our data council meeting last week. All we really could do is – we assumed that current services level – what would the portfolio look like for a major statistical – of course, statistical programs and so on. But, again, it is planning, not yet a budget.
I will say that some of our earlier investments are coming to fruition. You will remember the collaboration between NCHS and SAMHSA for emergency room survey. We had two of them before. We tried to find a way to integrate those two. It looks like it is coming to fruition. We will be getting some data there. It seems to becoming operational before long.
In the area of policy, as I mentioned before and we should probably have our strategic plan summary in all of your briefing books for every meeting. We have the HHS and all federal agencies have a strategic plan from 2014 to 2018. It has four major goals and about 21 objectives. The objectives all have measures in terms of how we are doing. We can reach briefing for the committee. We have a one-page summary that shows that those goals and those objectives are, which we could make available to you. That is the umbrella. HHS strategy. It is all of the specific activities within HHS cascade from those very high-level goals and objectives. And obviously data, capability, data for decision making, privacy and confidentiality, metrics to measure performance and for decision making are a big part of all of those objectives.
Health reform. We will be coming up to the next open enrollment period in the fall. We have really gotten it looks like the uninsurance rate down to historic low levels. I think probably at the lowest level that we have seen since we started measuring in Health Interview Survey. Maybe 1957 – probably later. That was even before Medicare. Clearly, the uninsurance rate is going down. There are a lot of other aspects, as you know. Access to care and health insurance.
Let me say a word about monitoring the implementation and impact of health reform. The strategy there, as we talked previously, is we are using administrative data where we can from the health insurance exchanges and states, some of the regulatory data to monitor enrollment. That is more or less the numerator. And then we had been publishing during the open enrollment periods monthly figures on the enrollment numbers. I think CMS is taking that over now. Interestingly, we tried to disaggregate some of those enrollment numbers by race, ethnicity, and other factors. We are getting more and more progress there.
On the other side of the monitoring, we have been relying on our national surveys. As I mentioned previously, we have made a number of enhancements to all of our major – most of the major surveys. Before ACA was implemented and we have been modifying since then and we have reached the point where we are now getting quarterly data on enrollment, which lags about – it is a six-month lag, but it is a quarterly data on the National Health Interview Survey, which is really very – it is just immensely helpful.
We look, as everyone else does, at pulling data at the gallop surveys and these web-based surveys and everything else. They have very low response rates. You really do not know what you are getting. Sometimes they might be right by chance or they might be right because they are good, but it is hard to know. There was no gold standard until we started getting the HIS data. Now, at least, we have reached the point where every quarter we getting the updates from the Health Interview Survey on health insurance rates, insurance rates, access to care measures, a number of things that you really need to see how things are going. We are fortunate that we have reached that stage now.
I think Charlie can tell you more about this, but one of the initiatives at HHS across HHS now is an attempt to consolidate our research data census. You will remember we have talked previously that NCHS has a research data center. In fact, we have a location in the Humphrey Building of an NCHS Research Data Center. Charlie will tell you more about this. But after an uphill battle, getting everybody to consolidate, I think there is actually a move now, a fairly successful move in pulling together in NCHS on using that research data center to make it easier for everybody for a lot of reasons.
I wanted to mention on the public health side that there is a – the assistant secretary for health, Karen De Salvo – many of you know may know about this. She has tried to turn attention to the public health sector. They are working on a concept called Public Health 3.0. Many have heard about this. She is having listening sessions across the country. This is how it is conceptualized. Public Health 1.0 was more or less the – this is modern public health now. This is not Florence Nightingale and – modern public health. The first Public Health 1.0 was basically the basic improvements in sanitation, public health and the first scientific antibiotics and medicine and things like that.
The second stage, Public Health 2.0 – they have sort of characterized as the point when public health, building on those successes, begin to focus on the capabilities and the core functions of public health, more professionalization and led to accreditation boards and so on and a focus on preventing chronic conditions, not just infectious conditions.
But Public Health 3.0 is a concept that says – it really looks at social determinants of public health and it involves official public health agencies more to reach out and collaborate and leverage some of the social dimensions, I will not say economic dimensions of public health, so it involves links with housing, links with education, links with place-based amenities and nutrition and food and so on.
The idea would be that – the listening sessions are how can public health do this. It is largely a focus on local and community health. That is where much of this goes on. There is a role for states as well. It is not just governmental agencies. And a big part of that besides leadership – much of this takes place because a person, as you know, takes the lead, a champion and there is some sort of spark in the community that has – besides that, there is infrastructure. There is data and metrics, which actually are a big engine for moving this along. I can provide more information, but there will be – and I can send you the little journal article about what this is. It was in the APHA Journal. You will be hearing more about that.
I think the idea is to take this concept across the country to a number of listening sessions and then to see how it can translate into some sort of a report or set recommendations for everyone to move this forward. Fairly recent. Well, actually, Karen has been working on that for a while.
A couple of other things just really quickly. We have been – again, in the interest of not only collecting data and analyzing it, but making it available. We have a couple of activities where we are trying to pull these things together. I will just mention a couple.
One of them is our HHS IT Policy Steering Committee. It has issued a policy of data sharing first. And nothing new here. The idea is that when you are planning and trying to justify new investments and health IT and they often refer to our surveys as health IT investments and they are to some extent. That there should be a philosophy of thinking from the very beginning how you are going to analyze and make the data available appropriately. If it is confidential, you think of how you might – there are ways to do that too. But it should not be an afterthought where everything is done. Somebody asked for this data. That is at a policy level. It does not say how to do it. It says think about this when you are putting together investments.
On the more policy end of it, we have a couple of things going on. We were asked to create under the data council, an administrative data working group. We put the group together. We have 20 members throughout HHS. We have Census folks and others as well. The goal here is to try to lend some policy and procedural best practices to sharing and providing access to administrative data across HHS. We have some well-known administrative data sets. As you know, Medicare, with fairly elaborate procedures for access to care.
Then we have others where every time you try to get the data, it is something new. It is a one off created. There does not seem to be any continuity or policy or procedure. We brought everybody together. I think we are looking at what kind of policy would we want and best practices across the administrative data assets of HHS. It is not just an inventory. I think the inventory is the last thing we do. This is actually thinking about what can we learn from the better programs. You have to think about having data available at the beginning, not just at the end. But it takes people to document the data, to assure the quality, to curate and so on. Most of the agencies do not have the resources. It is a somewhat complicated problem, but at any rate, there are ways that some agencies can learn from others about how to do this and maybe even share some resources if possible. We just started the group. We will be looking at some recommendations there.
I did want to mention and I will close with this. A law was passed, you will remember, last fall. It creates a federal commission on evidence-based policy making. There will be 18 members. They were supposed to have their first meeting here in June. The executive branch gets to name three. The minority and minority leadership and the House and the Senate each get to name three. Some of them have been named already. Katharine Abraham, who is at the University of Maryland – she was at OMB and statistical policy for a while. I think she is a co-chair. Bob Groves, who was the former director of the Census Bureau. He has been named as a member. Sherry Glade, Mike and I’s old colleague. She has been named a member as well. The various folks who can name people are starting to name them now. But I think when they have a quorum, they can have a meeting.
But there are a number of things to look at. It is to look across all the federal agencies, both the administrative data and survey data and to make some recommendations about how it could be made more useful and more easily accessible.
I think even one of the provisions is that they should come forward with the idea of – they should look at the idea of the Census Bureau hosting some sort of a platform. A clearinghouse, they call it, for administrative data. I do not they have had their first meeting yet, but I think now that they have a quorum, they will be having that.
There are privacy folks. I think you know Latanya Sweeney. There are some other privacy folks around the group as well. This is not just let’s just push through data access. This is meant to be a balance between the laws and the policies governing data, but again it should not be so hard to get. It should not be climbing a mountain each time you want some of this administrative data. If you are not an aficionado, it is just very hard. We will see. Eighteen months is their lifespan. They are supposed to come forward. It should be trans-political administrations. We will see. It has members from both parties. We will see where they all come out.
Just very quickly. We have started a couple of projects on the vulnerable population side. The secretary asked us to look at is there any way we can improve the data on American Indians, Native American populations and particularly tribal data, which is very hard to do. We have also been asked to look at again on social determinants. It is the integration of the intersection between criminal justice involves populations and health. As I said, previously, we are looking at, as you are, how can we introduce social determinants of health from the metric side and from the collaboration side as well. Let me stop there.
DR. SUAREZ: Thank you so much. Always fascinating level of – extent of which the work at the department – thank you so much for that. Any comments? We have time for two or three questions.
DR. MAYS: As always, thanks for a great report. I wanted to ask a couple of questions about the evidence policy making and the length of their time. Are they intersecting with any of the stuff that we do?
MR. SCANLON: They are a commission. They will have a full-time staff and a full-time executive director. They have not really planned their agenda yet. I can send you the statute if everybody is interested. A good strategy for them would be to look at what do we already know and what committees and whatever mechanisms we have to give advice. We could share information with them.
What we will do I think is – we are having our evaluation and evidence-based office and ASPE be the contact with that group at OMB. OMB and Census – I think the statistics agencies are actually being asked to support it – the Census will probably get the big part of the bill. We will see what we can do there. I think that will be a good idea. I would hate to see them spin wheels worrying about things that have largely been settled. That is a good idea.
DR. O’ GRADY: Two questions. One, you said that you have new and better numbers on the uninsured and I did not know whether you were in a position to share what those actual estimates were.
MR. SCANLON: I can actually provide that to you. We have uninsurance rate down below – do you remember, Charlie? I should have brought that with me. 18 and above. I think we got the uninsured down to under 10 percent. It is really remarkable. Again, at the trends, it is a little hard to know how far to go back because Medicare started in ’65. I do not have that – under ten. I did not bring the report with me. It is on the HIS website.
Of course, there are always differences in subpopulation groups, but pretty much everybody is doing pretty well. We will get another report – I am thinking it is probably September, Charlie, on the – what we saw from the HIS a month ago was the full year 2015 calendar year and we do quarterly findings as well. We will see the first quarter for 2016, calendar quarter for 2016, in a couple of months. This is the early release program that ASPE and NCHS have created.
DR. O’GRADY: One other question on this commission. Is there a website so we can see – has the whole commission been appointed yet? They are still in flux. There is no real source to go to other than the media.
MR. SCANLON: It is mostly email so far. I do not think they have a website yet. I can share with you – OMB is letting us know who has been named. They will have a website. They will obviously have to meet. We will share that with everyone. But they are still trying to pull themselves together. Until they get a quorum, which I think they are nearing now —
DR. O’GRADY: It sounds like these are mostly the administration of appointees rather than congressional appointees at this point. Is that correct?
MR. SCANLON: Yes. I think the director of OMB – he might serve on there as well. And then there is some rule about who can serve. They seem to have somewhat of a – not a quota, but they seem to give more freedom to academic appointments than other probably what they think might be partisan.
DR. SUAREZ: This is a federal commission on evidence-based health policy?
MR. SCANLON: No. All policy. There is some thought that this arose out of disagreements about how to measure income and poverty. There clearly are members who – that I have seen so far that come from that world, but it looks pretty good so far. There are folks here who understand the need for – if this is belief and anecdote. This is not anecdote decision making. This is supposed to be serious evidence based whatever the nature of the evidence would be. They will look at what this sort of state of the art now. What are the dimensions of privacy and confidentiality and data sharing? Are there obstacles to moving forward? What would promote the use of those? Again, this should not be a naïve group. I think they understand there are laws and other statutes. Privacy is a big deal. But there are ways to do this as you all know.
DR. LOVE: They could recommend new data or is it just looking at what exists.
MR. SCANLON: Yes. Within that general area, they could recommend what they think would move the field along.
DR. LOVE: Unique identifier?
MR. SCANLON: I do not think we are going there.
DR. SUAREZ: Thank you so much, Jim, for the report. We are going to move on and invite Charlie to talk to us about the national meeting. Welcome Charlie to the national meeting.
MR. ROTHWELL: Thank you. I am Charlie Rothwell. I am proud to be the director of the National Center for Health Statistics, also, always surprised that I am. First of all, I would like to thank you all for giving up your time to come and help the federal government do its business. The federal government is more than just those who are in the government itself. It is all of us. We really appreciate your time and your guidance and you really do, looking at the agenda that you are facing – it is really amazing what you all are involved in.
Also, it is amazing when I heard that folks do not know about this committee. This is one of the oldest standing committees that the department has. That just shows you that just because you are around for a while does not mean that people understand who you are. Based on that, I would like to say a little bit about NCHS for those of you who do not who we are. Organizationally, the National Center is located in the Centers for Disease Control, but we are a unique organization in federal government. Unlike many other countries, we have a decentralized federal statistical system. There are 13 statistical agencies throughout government. Justice basically has a statistically agency. Transportation does. Obviously, the Bureau of Census is the big dog. Bureau of Labor Statistics and NCHS – the energy information agency is a federal statistical agency. One of the oldest federal statistical agencies is in the ag department. Agricultural statistics has long been a staple of federal data that has made a difference at the local level.
NCHS has its own data system. We do that through our surveys. The surveys are done like the Health Interview Survey that Jim just talked about, which is a personal interview survey that is done. We also have – that collects the health status of the United States. We also have something called NHANES where we have trailers as well as interviews that go out and do a variety of measurements that are useful in so many areas where it has to do with a measurement of obesity, measures of hypertension, controlled hypertension, uncontrolled hypertension, and a variety of other physical measurements as well as sometimes measurements in the household itself for an environmental perspective. This is a much smaller survey of about 5000 or so interviews a year.
We also have a survey that is done for monitoring family formation in the United States, which is the National Survey of Family Growth. We have a family of health care surveys. You will be hearing more about later in the day and the changes there.
We have really the beginning of NCHS and really a part of Public Health 1.0. That is the vital statistic system of the United States, the collection of information from birth and death and fetal death certificates. I would like just to stop with that point and not to be in disagreement at all with what the department is suggesting, but that we cannot forget about the importance of Public Health 1.0. I think even new things are needed. The new things that are taking place that we are concerned about whether it is opioid overdose or whether it is Zika that require issues of sanitation and capabilities and vital statistics reporting in order to be able to monitor what is going on as well as ameliorate those types of activities. The infrastructure in public health is still all important.
NCHS has a staff of about 500 folks. Our budget basically goes to support our staff and to support our data collection activities. When Jim mentioned the Health Interview Survey, which is a combined effort between us and the Bureau of Census. The Bureau of Census does the data collection for us under our requirements. And then we have a variety of other private entities, for example, Westat is our data collection contractor on our NHANES survey.
Most of our money goes to either data collection or to our staff. We do not really have a major grant program. We have a very small – it is tiny – that we just started again of only about $300,000. For the most part, that is where our money is and that is where our effort is in collecting data and disseminating it.
We have a budget of about in excess of $160 million as far as appropriations go and depending upon the year and how successful we are, we can get up to $30 million additional throughout the department to ask specific questions on our surveys and to do specific things to improve our surveys and responsiveness. We are located in Hyattsville, Maryland right outside of the University of Maryland as well as Research Triangle Park. That is who we are.
The budget situation. Right now, for FY17 upcoming, the only thing we know is the senate mark. Their NCHS was cut by about $4.4 million. Basically, last year NCHS received its first increase. I cannot remember when. That was to help support vital statistics and improving its timeliness and data quality. It would seem that at least from the Senate version that might be cut. It is hard to know what will happen with the House and what will come out at the end.
For the most part though, NCHS has been flat lined for a number of years. That is the reason why I talk to you about where our money goes. For the most part, salaries have been flat lined for a number of years. All of a sudden we got a small increase in salaries in the federal government. The cost of data collection with our contractors has continued to go up. We are feeling a pinch I must say. Also, there is a suggested increase for salaries of federal employees. In case you do not know it, when Congress says they are going to do that increase, they do not give you the money. That comes out of your appropriated dollars that you have already had. Even though I want my staff to be able to live appropriately, it does hurt when we have those increases.
I think I have said this to you before. If you ever hear a federal director say that they are just fine from the budget perspective – well, you will never hear that. But we are not in terrible shape. We are not in as bad shape as some agencies are. But I am concerned for our future.
Some of the things that we are doing right now that you will be hearing about later. One is the redesign of the Health Interview Survey. That is a tricky proposition. We have already done the redesign of the actual survey sampling scheme. That is already in place and has been successfully accomplished. The next thing is to look at how we can reduce responded burden from those who are participating in the Health Interview Survey by focusing it more. By that, we put a goal on to try to reduce the amount of time it takes for a survey by 30 percent. If we do not put something like this out there as a goal, it will not happen.
A survey can easily take an hour and a half to do. Sometimes if you have a family or you have an individual with a variety of chronic diseases, it can take much longer in order to get that information. That is a terrific burden on someone to provide that information. As it gets longer, you sometimes wonder about the quality of the information as you go through that survey. We are trying to focus it. You will be hearing from my staff, Marcy Cinnamon and others, about what the process has been. Nobody wants change, especially if they have been using certain types of items on the Health Interview Survey.
The other interesting thing that has happened or that we need to keep our eye on relative to the redesign is that there are many really important health indicators or health issues that you would think why are we may be considering taking them off the Health Interview Survey. If in fact the survey is not large enough, in order to really delve into what we are trying to measure then in fact is this an appropriate mechanism to collect this information, which is needed.
The last thing we need to do is try to collect information or say we are collecting information that really cannot be used for the purposes that we are touting it to be. We need to make sure that what we are collecting we can be able to drill down and look at demographic differentials and see where our problems are from a health status perspective.
This is not going to be easy sledding. We are hoping though to have the redesign put in place. As of January 1, 2018, we will have a new survey in the field. We are going to try to as best we can by the end of this year make the final decisions with all our contributors to decide what that survey is going to look like and then we will be working with the Census to modify the data systems that actually – CAPI systems that actually collect the information. We probably should have done this five to eight years ago. It is long overdue. We have had a survey creep and it is time to refocus.
Another area that is of change and you will hear about later is in our health care surveys. One of the things that has happened in this administration is the obvious expansion of electronic health records. Our health care surveys for the most part have been dependent upon abstracting information from medical records. It seems sort of stupid to abstract from electronic records. The other problem you have though is if you just go out and collect electronic records. What do you have when you put it all together and what are the standards and what are the problems between vendors and even within vendors? Our staff are working on that. I think we are on the cusp of something really great. We would have I think some of the – instead of having a very small focused survey, we could have some of the largest – probably the largest survey that we have in NCHS will be in our health care area. The problem is, again, how do we standardize this information so we are assured that we are really providing appropriate information out.
Vitals. I think the last time I talked to you, I talked about some of the improvements so I am not going to bore you with that. Suffice it to say that we are publishing vitals now on events five months after of the actual event itself whether it is a birth or a death. We have already published information for 2015 on vitals for births and deaths and major causes of death. We are reporting on a quarterly basis. There is probably within the next three to four years, I think there is a chance that we can report within three months, maybe within two months and be able to say something about where we are from vital statistics perspectives in the country. That is really important.
But really the important part of this will be that we can now go back to the providers of this information and ask them did you really mean this. Is this really correct information that you are providing? We can do that now at the time of data collection. This is going to make hopefully a great improvement in what we are collecting in the vital statistic system, something which has been needed for decades and decades.
I would like to say that much of this improvement has been because of states’ interests in improving their systems that provide us that type of information. Without the state support, this would not have happened.
Jim talked about response rates of some of the private surveys. Federal government is having real problems with response rates. Private industry would love to have our response rates. But from our perspective, we are getting in trouble. I can remember when we thought that we should – when we dipped below 80 percent response rate, we were in trouble. Now we wished we were there. NHANES has some of the lowest response rates that we have ever had. Part of it I think has to do with just with what people think of government right now. Part of it has to do with people just being too busy. Somehow we have to get over this because you can only do so much adjusting at some point in time and then are you really having representative surveys. I think this is a huge challenge for us, not just NCHS, but all the federal statistical agencies. How we do that through different modes of data collection, through administrative type of record collections, we can no longer just look at our surveys. We are going to have to expand into other areas.
The other thing that has happened is as Health Interview Survey has become more responsive, as vital statistics has become more response and I think when health care becomes more responsive, we are coming under more pressure by our folks throughout the department to share our information and share that information before it should be shared. Our independence is being tested. I think that is in a way a good thing because if it is not being tested that means that you do not mean it. What you are collecting is not useful. But, for example, with health insurance, with, for example, information that we are now putting out on suicide and opioid overdose and a variety of other things that are very current, folks want to know what we are coming out with before we can really say. This is something that I think is going to be of a continual struggle. It is a good struggle to have, but it makes my life a little difficult at times.
I think from a staffing perspective, what we are able to do – once we are able to get somebody in the door, we have outstanding staff. I look at the younger staff. I think they are as qualified or more qualified than any of my older staff. That being said, our HR system is broken. We have this attitude in federal government that you should be honored to apply to us. No. We should be going out and recruiting. We should be saying we need you. We want you. We need to court people. It is a new day. Somehow we are going to have to change our recruiting in order to do that. But we have a lot to sell. It is just that it takes a great deal of patience for someone to apply to federal government and wait and wait and wait and wait and then finally get word. If it is a negative word, that really is upsetting if you have waited for months and months and months and then it is a negative because somebody else got selected. This is not just affecting NCHS. It affects all of government.
Jim mentioned the research data center. I think this talks about the future and talks about response rates. It talks about areas that are health related, but not what you normally think. We need to be able to link our data files. For example, NCHS right now is linking our Health Interview Surveys, our vitals, and NHANES with HUD data. HUD data could be considered potentially as something that might improve health outcomes of families. I do not know. But unless we link it and we look at what the differentials are, it really might be something that could explain some of the differentials that we have.
In order to be able to do that data linkage, you have to have a safe house to do it. That is what a research data center is. It is basically taking our data and taking it out of our normal computer systems, putting it in a protected environment where there is no interaction with other systems and allowing you to be able to link data together and then be able to analyze it in such a way that you still will protect privacy. What we have done is SAMHSA and AHRQ and NCHS have come together and we will be having a single research data center approach for the department, run by NCHS, but sponsored by the rest of the department. Hopefully that will improve things like data linkages, use of surveys with administrative data systems that we could not have used before because of issues of privacy and confidentiality. I am very upbeat on that.
With that, I will just be quiet and see if you have any questions for me.
DR. SUAREZ: Thank you so much, Charlie. Any questions?
MS. LOVE: That is great, Charlie. Thank you. As I sit and listen to you reflect on the future, you said a couple of things that resounded with me. One of them is think of new ways to collect data so that you free up resources. I just sit and think of all the administrative data in the states that emergency departments, surgery centers that may be are underutilized. Is there a day in the future maybe the next 10 or 20 years that some of those data could populate federal data sets as well? I am thinking of a hybrid, vital statistics model that frees up then your resources to go after things that states are not getting or are not positioned to get. I am just trying to project our new way of thinking for that administrative data primarily.
MR. ROTHWELL: I think that this is an area where the states or communities could really come together and teach federal government something. If we could find the funds to support this and maybe with folks like RWJ and others, what are the data sets out there that could be used to monitor things at the local and state level and that we might be able to use from a federal perspective?
I would hate for the federal government to come in with the plan. I think that would be a bad mistake. We would do it. But I am not sure it would be what is really needed.
MS. LOVE: Well, I am sure there is a lot here that would have to be sorted out. I do think some of these data systems at the local and state levels that have taken decades to build are on shaky ground themselves. I am concerned just like you are for the federal data investments that that infrastructure is getting pretty old and creaky. I am just putting it out there that maybe there are new ways to collaborate or think about it.
DR. SUAREZ: We are going to go in this order. Bill, Raj, Linda, and Vickie. Then we will turn to our next presenters. Bill, if you could introduce yourself.
DR. STEAD: Bill Stead, Vanderbilt University, member of the Full Committee, co-chair of Population Health, no conflicts. Charlie, I second Denise. The clarity of the vision you are beginning to lay out. I wonder, as I hear it, the efforts like mining or linking data from HUD and the efforts like mining the electronic health record are a bit distinct from the efforts to rethink not just the items in the survey, but given response rates, what can you survey? I wonder if someone is working out how do we manage those two efforts so they help each other so that we could use the data mining part to get much less precise data much more broadly and use it in a way to in essence give us the context, the target, the surveys if there is a way to do that that would be stable to some degree over time. What are your thoughts on that?
MR. ROTHWELL: I think we would have to have the administrative data that would be out there that we could do the data mining on. I agree with you. It is just that right now we do not have that administrative data that will allow us to do the data mining whether it is data from CMS, whether it is data from IRS, whether whatever it might be. By the way, the commission that is being developed – my guess is legislation is going to come out of this. It almost has to if in fact we are going to change how we do business because, for example, NCHS cannot get access to some of its sister federal statistical agencies data and vice versa even though we have all the same protections and all that because of some of our legislation precludes that from happening for some good reasons, but there are ways around that. I am hoping that the commission would be looking at this and be able to come up with ways that we can share both our statistical data and our administrative data in a protected environment.
DR. CHANDERRAJ: That was a good presentation. I think there is a big part of participation in the surveys. It is very abysmally low. I was wondering if education can improve that like the smoking campaign that we initiated. A lot of funding and paying the salaries can go into the education part and then have more people participate in it.
MR. ROTHWELL: Well, there is an organization of directors of federal statistical agencies that meet once a month. One of the things we are looking at is how do we – maybe we should be doing some advertising on what these surveys do for the public. I think just NCHS by itself forgetting about the other statistical agencies we can make a great case whether it had to do with lead and gasoline, whether it has to do with obesity, whether it has to do with emerging diseases, you name it. Health insurance coverage. We can make a case for why it is that this is your chance to participate as a citizen in guiding where your resources are going for in health. We could do that for the federal statistical community as a whole. We are looking at how that might take place. If it does not take place soon then somehow NCHS is going to have to find some money to make our case I think by ourselves, but hopefully we can do a combined campaign.
By the way, our National Survey of Family Growth just to give you an idea, we collect some of the most personal information about people’s sexual practices. I think the response rate last month was around 71 percent. That is still very good. That being said, the slope is like this. The slope for NHANES is like this. We are going to quickly reach a point where we have to make a difference and see if people will participate.
I will just say that when you have local government or state government in some states not even supporting you because you are a federal activity, where are we? I will not say which state, but it is a big one.
DR. SUAREZ: We have several more people. We only have really time for three more questions. I am going to ask some of you. We need to move to the next item. We have Linda, Vickie and Bruce and if you can be brief, I appreciate that.
MS. KLOSS: I will be brief. I am very interested in your experience with accessing information from electronic health records. Two questions. What is the timetable to be able to share some of what you are learning both in terms of data findings, but maybe the process? What can be generally learned – what the experience has been and your approach?
MR. ROTHWELL: You will be hearing about that this afternoon from my staff. You will not be hearing when the – they will be giving an indication of when the results will be coming out, but they will certainly be able to tell you a little bit about what the problems are and how our approaches are going.
One of the interesting things that have taken place is that it is not easy to go into a physician’s office or a hospital and ask to abstract records. You take up space. You take up people’s time. You cause a problem. When we have gone to electronic health records and we have been involved with the meaningful use, that is, providers get credit for participating in our surveys, it is unbelievable the numbers who are willing to provide us not just with the data within our sample frame, but everything.
One of the interesting things that we will be facing is that we might in the future have data well beyond our initial sample frame. How do we use that? It will be a huge data set. It will not necessarily be representative of the United States, but it might be very representative of specific areas of health care. It is going to be a fun time. I will let my staff who know more than I do respond to your question.
DR. MAYS: Thank you for a great report as usual especially in terms of how quickly data is getting out and how much data we are getting.
I have a couple of questions. One is I do not know if you have seen or are familiar with the Data for Health report by the Robert Wood Johnson Foundation. There are several recommendations in there that they call for, which would be a very different type of operation for CDC. For example, they want – and they specifically call out CDC to have an investment in data exemplars for states. They really want to see major changes in terms of the privacy regulation. They want monies going into education. It is almost what Raj was asking education so that people will actually give more of their data, have more trust about that data. I am wondering what you think about that because it is one of the things that we are going to be talking about in our work group.
The other question I have is the data releases are great that are coming out so quickly. What I am trying to get a sense of is if there is a standard of what it is that is released. The suicide data came out and everybody was clamoring over the data, but it had no race. In the past, we have actually had race in terms of the press releases that went out. People were running around asking of academics to go in and use the data and give them even more information. I am just wondering when there are data releases if there are a set of variables by which all those releases are subjected to try and include. I know you cannot sometimes because of numbers, but in terms of race. In the past, that has actually been one of the things that you have been able to do. It made me think about data standards in terms of the data releases.
MR. ROTHWELL: I must admit. I have heard of the RWJ report, but I have not looked at it. I really cannot respond. That sounds like some excellent ideas, but I cannot respond.
On the issue of fast data releases and then how much information is there, that is – one of the problems we face is that if we are going to release information, how much can we say? When we are releasing vital statistics, we are releasing let’s say 12, 13, 14 major causes of death. We are going to be this coming year for 2016, we are going to be doing high risk first. How much can you really say by with the amount of records that we have in and with the adjustments we make? How can we differentiate by age or race and that type of thing? I think we are going to get better because the states are getting much better in providing us the information on a quicker basis. That means that we are getting more and more – when we are looking at a quarter of information, we really have a larger population group to look at that is representative of the country unless we can say something about race groups or age groups.
DR. QUEEN: What came out on the suicide, for example, was a data brief. We have internally a set of limitations on how many figures, how much narrative so for that particular report, for example. This makes me think. Maybe we need to have something that is up on the web that more clearly or easily defines here is what a data brief is and more will be coming because more is in the works for that specifically because there was not enough room and there was much more information that they wanted to put out.
MR. ROTHWELL: We have to put our data out though on public use files. There are ways that people can look at the data much more in depth than what we might do in a data brief.
The other thing that relates is when we looked at suicide and we also looked at opioid overdose deaths, we did something by use of data visualization tools where we put the data out that you could play with it in a variety of ways and that really spawned a lot of interest. I think we are going to have to have better data access capabilities behind the actual reports that we have.
There is an issue of the space that we have in several of our reports. In order to be able to get them out quickly, they can only say a few things. Also, we have a restriction as far as what we really can say in depth when we are looking at preliminary data whether it is from HIS or whether it is from vitals at this point.
DR. COHEN: Charlie, I really need to commend you for the incredible work you have done leading NCHS. Your vision and perseverance has really been phenomenal. It is really appreciated not only in this committee and my former life in the Massachusetts Health Department and connection of vitals. Thank you so much for everything that you have done.
I know that the budget limits create priorities. The work you are doing around data collection redesign, timeliness and linkage is fantastic. There is one area that I really would like to see NCHS go back to some of its earlier roots and perhaps the commission would provide the funding and opportunity to do that. That is really to develop tools and methods for folks to use at the state and local level to use the data and to create companion data sets that mirror what NCHS is able to do at a federal and regional level. NCHS has always been the heart of statistical expertise in the health sector in this country. It would be wonderful if we could figure out a way to spread that expertise so others can take advantage of the knowledge at NCHS. I hope there is an opportunity to leverage some new funds or increase the priority of developing tools and methods that can be used to deal with the amazing amount of data whether it is visualization, developing local surveys based on existing models.
We have talked about it before. The statistical note for health planners that really was incredibly well used way back when in the ’70s when comprehensive health planning was at its height. I would love to see NCHS recommit to providing some of those kinds of things.
MR. ROTHWELL: I would agree. One of the problems that we are facing is we have a growth of health indicators. When you take a look at Healthy People 2020, it has some 1200 indicators. That is ten times what it was when we first started. The same staff who used to do what you are talking about is trying to keep track of healthy people objectives.
What I am getting at is I think there will be ways other than just funding that could help us in that end. You talk about state systems and all. One of the things about improving our responsiveness in vitals is you do see some of the problems that states are having or organizations are having in providing the information. What we found recently that should not be a surprise is the medical examiners and coroners in our country are hurting. They have no representatives. NCHS has been trying to be a representative for them, but we need a stronger representative for them because they cannot get their tox reports out. If they cannot get their tox reports done and a variety of other activities done then we cannot really measure what opioid overdoses are, what suicide rates are or any of the other examiner or coroner events that take place.
Once we are speeding up our data systems, we are seeing where we are hitting walls and where those walls are is where we need to improve our infrastructure in areas that we did not realize before. I am just using that as an example. Some of the comments on our weak state systems are really very true. We need to take a look at those that are really critical to us.
DR. SUAREZ: Thank you so much, Charlie, again for joining us. It is a really fascinating discussion.
We are going to move to Shana Olshan from CMS.
MS. OLSHAN: Good morning, Dr. Suarez and all the members of the committee. As always, it is a pleasure to be with you here today this morning. You clearly have a very full agenda. I look forward to listening to the upcoming discussions at this meeting, in particular, the one on the standards of committee letters.
We have taken the Review Committee recommendations in the February letter that you all sent us very seriously. I want to report on our status about two of the broad recommendations that you included in the letter. The first is the need for additional education and the second is the need to strengthen our enforcement.
As I have mentioned in the past to you all, education and outreach about administrative simplification adopted standards and operating rules is critical to increasing compliance and ultimately reducing costs, increasing efficiencies and lessening burden. And the Review Committee letter confirmed that position. It was also our experience with the ICD-10 implementation.
I am happy to announce that in about a week, our new and improved administrative simplification segment of the CMS website will go live. It is not perfect and it is just the first phase of our improvements, but it is definitely a step up from the current site. We have updated the content. We have renamed sections and added new pages to the section of the CMS website. We tried to make it much more logical than it is today.
We believe that this new format will improve the accessibility of information about administrative simplification. Once the updated pages have gone live, I will make sure that we share this information with the Full Committee. We would appreciate your feedback on it as we will be continuing to make modifications in the future. As I said, it is just the first phase of improvements.
With regard to your enforcement recommendation, CMS has designed a new and enhanced web-based HIPAA compliant enforcement tool to replace the existing Administrative Simplification Enforcement Tool, otherwise known as ASET. Since we are not terribly original at naming things, the new tool is called the Administrative Simplification Enforcement and Testing Tool, also known as ASETT or as we like to call it ASET-T so we can differentiate between the two.
The new tool will be released very shortly, close in time with the updated website. ASET-T allows individuals and organizations to file HIPAA complaints for alleged violations of the HIPAA and the ACA transactions, code sets, unique identifiers, and operating rules.
This new system has added enhanced security features for complainants when submitting their complaints, improve usability and it maintains the existing tool that we have for testing transactions with or without filing a complaint. We encourage all covered entities to use this testing tool to assist them in determining their own level of compliance with transactions. It is clear to me that the whole industry is really ready for an attachment standard. You have made that clear over the years as well. And the upcoming letter from the committee will definitely assist us in making 2016 the year of attachments.
At the recent work group for electronic data interchange annual meeting, there are many presentations and listening opportunities for CMS to gather information and feedback and attachments. Many members of this committee and members here in the public today were also at that meeting and attended and participated. We are working steadily on a regulation and are optimistic that HHS will finally move forward with adopting an attachment standard. Stay tuned.
Since these are CMS remarks, I wanted to mention a couple of other initiatives that may be of interest to the committee. The Health Care Innovation Awards is a CMS innovation center initiative that tests new payment and service delivery models and aims to find better ways to deliver care and bring down cost for Medicare, Medicaid, and the Children’s Health Insurance enrollees.
Last month, CMS released the second annual independent evaluation reports of round one of these awards. Overall, these evaluation reports show a wide range of experience that have resulted in tangible benefits for patients and it helped inform CMS in the development of new payment and service delivery models. Where data are available, these reports describe preliminary impact estimates on key outcome measures such as hospitalizations and readmissions. A number of awardees showed favorable results on one or more measures of cost, hospitalizations, readmissions and emergency room visits.
One such awardee, I will just give you one example, was Innovative Oncology Business Solutions, Inc. In its community oncology medical home, it reached more than 2100 cancer patients through seven community oncology practices throughout the United States. Through its comprehensive and coordinated oncology care, the model established pathways that allowed providers to identify and manage symptoms in real time. It improved patient access to providers through same-day appointments and extended night and weekend office hours and it provided disease management guidance for providers to improve treatment decision making, symptom recognition, and assistance with the patient’s self-care, pain management and caregiver support.
The evaluation report showed that this award demonstrated a significant reduction in hospital readmissions and emergency room visits. In addition, the qualitative findings suggest that staff highly value the triage pathways for making their work flow more efficient and that patients greatly appreciate weekend hours and the increased capacity for urgent care visits during the day. In fact, elements of this model were incorporated into the design of the oncology care model.
Another one of the awardees is the Diabetes Prevention Program is the first example of a preventive service model from the CMS innovation center that has become eligible for expansion into the Medicare program as a whole. The CMS Office of the Actuary certified that expansion of the Diabetes Prevention Program would reduce net Medicare spending. The expansion was also determined to improve the quality of patient care without limiting coverage or benefits.
CMS has also worked for many years to improve how Medicare pays for durable medical equipment, prosthetics, orthotics and supplies or as we like to call it DMEPOS with a goal of insuring that people with Medicare get the equipment that they need while reducing cost for beneficiaries and for taxpayers.
Since 2011, we have operated the DMEPOS competitive bidding program to set payment amounts for covered DMEPOS items in certain areas of the country. In fact, between 2011 and 2013, the program saved more than $580 million in the nine markets that were included in the round one rebid and that number comes from the end of its three-year contract period.
Between July 2013 and June 2015, Medicare has saved approximately $3.6 billion after the first two years of round two and the national mail order program. Round two was most of the country.
Health monitoring data indicate that its implementation is going smoothly with few inquiries or complaints and no negative impact on beneficiary health outcomes. By all measures, the DMEPOS competitive bidding program has been a great success both for beneficiaries and for taxpayers. At the beginning of this year in compliance with the statute, we phased in new rates for non-competitive bidding areas based on the information from the DMEPOS competitive bidding program. We phased in these new rates with a blend of 50 percent of the unadjusted payment rates and 50 percent of the adjusted payment rates on January 1 of this year.
We are also using the same real-time monitoring system we used in the competitive bidding areas to ensure that beneficiaries are receiving the equipment that they need. This data monitoring tracks access to items and services and a number of clinical outcome measures such as mortality, hospitalizations, and emergency room visits.
Last week CMS posted monitoring data that shows our efforts succeeded in saving the Medicare program money while continuing to provide equipment to those who need it. The data reveals that suppliers in these non-competitive bidding areas have continued to accept the new adjusted DMEPOS payment rates as payment in full.
For the first four months in 2016, suppliers accepted these new rates as payment in full for 99.88 percent of the claims submitted compared to 99.87 percent for the first four months of 2015. This suggests that the adjusted fee schedule rates continue to be more than adequate to cover the cost of furnishing the DMEPOS items in all areas.
I know that you have a session tomorrow on the Medicare Incentive Program so I will not provide any updates on it today, but please save any questions you have on that topic for the subject matter expert Dr. Yong.
Thank you for inviting me to speak with you all today and giving me the opportunity to share these updates with you. This committee plays a critical role for the department and I am, as I said earlier, particularly interested in the discussion you are going to have just before lunch about the recommendation letters on attachments and Phase IV Operating Rules. They will do a lot to helping my group continue to do its work. At this time, I am happy to answer questions.
DR. SUAREZ: Thank you. We have questions.
MS. GOSS: Thank you. That was a particularly informative and I appreciate the updates. It is really great to hear that we are making improvements to the educational resources and to the tools that people can use for validating their compliance aspects. Those were two of eight recommendations in the February 29 letter to the secretary including aspects related to covered entity definition, expansion, and a few other transaction specific activities. Could you offer an update on those other recommendations?
MS. OLSHAN: Well, we continue to study them all and the covered entity one in particular. It is one that is also included in section 10109 of the Affordable Care Act, which is something we do need to continue to work with this committee on and hopefully as you go over your work plan for the remainder of this year, there may be a way to squeeze it in, maybe in one of your last meetings so we can have further discussions on that particular issue. It is an issue that comes up a lot in a lot of environments. It is a tricky one given the way the law is worded and we need to think creatively to see if there is anything we can do absent a change in statute on that topic. But we do continue to look at all of the many recommendations that are in the Review Committee letter and we are trying to make progress on them. But at this point, I cannot give any further updates. But I assure you that at future meetings, I will give whatever updates I can.
MS. GOSS: Thank you. We are also planning to have an exciting conversation to frame up our upcoming discussions around predictability, roadmap development, development of standards, efficiencies, and promulgation efficiencies. I am very much looking forward to the work plan discussion. There is a lot on our plate and not enough hours in the day. We will be prioritizing those. Hopefully, you will be able to also listen in on that conversation.
MS. OLSHAN: That will be great. Thank you.
DR. SUAREZ: I do have a couple of quick questions for you. The first one is about ICD-10. I know we made a major transition to ICD-10. It looks like everything is working well. Just a quick note about any issues that you might have noticed about the implementation of ICD-10. Particularly, as we approach October 1 of this year when a new set of codes – all the codes that have been frozen during the transition period are coming up. If you have any thoughts about that. That is the first question.
MS. OLSHAN: As you said, the ICD-10 implementation has continued to appear have gone surprisingly and wonderfully well. I am very happy about that. In fact, I think AHIMA recently published a study. While there was some loss in productivity, there was only a minimal change in the coding accuracy, which I think is fabulous considering the number of additional codes that were implemented on October 1. I think that bodes very well for the future and for what will happen come October 1 when that pent up demand is released with the updated code set.
The fact that the industry handled the transition as well as it did I think we are going to be fine personally with this change. I think it is people think it is daunting, but they have proven they can handle this. I think that is great. Medicaid claims continue to process swimmingly. As I mentioned in the past, on average, there are 4.6 million claims a day processed by Medicare Fee for Service. That does not include all the Medicare Advantage payments that are made. By and large, things have gone extraordinarily well. I am optimistic.
DR. SUAREZ: The other question is about two pending actions. One is on health plan compliance certification regulations and the other one is the Health Plan ID. If you could say a couple of words about those.
MS. OLSHAN: Two of my favorite topics. As I mentioned, last time I was here, we are working on a second NPRM for certification of compliance. I was hoping it would have already been released. Processes being what they are and things not ever moving as fast as you would like them to be, it is still not out, but we do continue to work on it.
In the original Notice of Proposed Rulemaking, we had to link the Health Plan ID into the process and that is why the two are moving hand in hand and how we are working through that. We have evaluated all the comments that came into the RFI. I think we have made our decision about how to approve, but we still are going through that process internally before we can announce anything on that.
DR. SUAREZ: Thank you very much. Any other questions from anyone on the committee?
DR. PHILLIPS: Thanks so much for being here. I learned that the BETOS codes have gone away. The way that claims data were categorized to help look at expenses in different categories. I only recently discovered that that went away in January. I wondered what the reasoning was or what is the plan for that community.
MS. OLSHAN: I actually cannot answer your question, but I can take that down and get an answer back to you. I personally have not worked with the BETOS codes in a very long time. I will have to check with our data area and get back to you on that.
DR. SUAREZ: — additional question. We heard last time the major transition that CMS will be going through replacing the Social Security Number with a new number. Could you give us a quick update on timing and when this is going to be happening?
MS. OLSHAN: There is not a whole lot to add to what I said at the last time I was here. The law says that all new cards need to be issued by April of 2019. The new cards will have on them a replacement for the current number we have, which is Social Security Number based. It is going to be replaced with a randomly generated Medicare beneficiary identifier. There have been listening sessions with various stakeholders around the whole industry. Beneficiaries, providers, plans, everyone you can think of. There have been many listening sessions to date. There will be more – one and done exercise. Stay tuned. Keep your ears to the ground about those.
We do have to update both at the federal level, CMS, SSA, and at the state levels about 75 systems in order to achieve this change. A lot of people think it is simple. The VA did it very easily. It is an extremely complex change for us to make. It is important that we do it right. When we do mail the over 60 million cards that it is done the right way, the right time. The cards are issued securely and they meet the needs and that the community at large knows how to work with that. To that end, they are planning a big educational push in 2018 through 2019 and beyond once the cards are issued so that everyone can really understand.
There will be some transition period. They are still working out the details of that transition period. You can probably bill using the old number and the new number at least for some period of time, but those details are forthcoming.
DR. SUAREZ: Thank you so much. Thank you so much again also for the continued partnership between CMS and the National Committee. It has been really wonderful.
MS. OLSHAN: Thank you. As I said, I look forward to the rest of your discussions.
DR. SUAREZ: We are going to move to Elise Anthony. I think she is on the phone. Elise, can you hear us?
MS. ANTHONY: I know I am running a little bit behind so I will try to catch us up a little bit. I will use my New York speed to help us do that.
One, I wanted to start by giving some updates on what ONC is working on, particularly from the policy perspective, which is the team that I lead. But before I do that, I wanted to just talk about our annual meeting. We recently about two weeks ago now had our ONC annual meeting. For those who were not able to participate, it was really a wonderful opportunity for us to bring together and hear from health IT stakeholders from across the landscape, discussing different components of the health IT world, including related to privacy and security, which I think Lucia Savage will be providing some updates on tomorrow as well.
And then also with regards to patient engagements, we had some sessions on opioid response and opportunities that might exist from a health IT perspective to aid in those efforts and precision medicine was also discussed. It was a wonderful opportunity, a three-day opportunity, including our consumer day to really engage with stakeholders and hear about what is happening on the ground. Thanks to all those who may be on the line or in attendance who were able to participate.
To start, I wanted to talk a little bit about our rulemaking activities. Rulemaking activities. A lot of the work that we are doing relates to updates to our rule structure and then also working with our partners. Some might recall that in October 2015, we released the final rule on the 2015 edition health IT certification criteria. That is our latest update, our latest edition. We have had previous ones, the 2011 edition, the 2014 edition and the 2014 released two edition. These are the rules that govern certification criteria for purposes of a health IT product that is looking to come to ONC for certification.
It includes standards and criteria related to everything from demographics, being able to take that in to how you can spend or receive information as well as information on CPOE standards related to computer provider order entry. We also included elements that are a little bit new in terms of social determinants. Some of the criteria are optional, but they provide opportunities for developers to have a standardized way of taking in information that look at beyond the clinical sphere, including criteria related to depression, for example. And then also we have updated some of the assignments related to demographics including sexual orientation and social type information as well as deeper race ethnicity standards. And all of those, we believe, are helpful to addressing health disparities and generally providing more health IT opportunities for both providers and developers. That rule is final.
That builds into our overall work across the program and working with our HHS partners. For example, we have been working with CMS on their MACRA and MIPS proposed law. I will not steal the thunder. I think you are getting a presentation on that tomorrow. But generally, across the program, the quality payment program, which is noted in the MACRA rule, there are certified EHR technology requirements that Congress included in the MACRA law itself. We have worked with CMS to bring those through in the rule. Whether you are an APM, an advanced payment model, or whether you are more in the Meritus entertainment system world, there are certification requirements. Those build on the foundation of the 2015 edition rules. We have been thinking about the 2015 edition rule at the foundation for MACRA and MIPS in the proposed rule and we have been working very closely with our CMS partners in that regard.
One other rule that we have been working on is the enhanced oversight and accountability proposed rule. That proposed rule was issued in May 2016. The comment period is over and we are now reviewing the comments and thinking about what the final rule would look like, which we are targeted to publish in fall 2016.
The enhanced oversight and accountability proposed rule is a rule that allows additional opportunities for ONC to engage in looking at products once they are out in the field, for example. Once a product is certified, ONC’s rule would be expanded to encompass the ability to directly review health IT certified under the program and where necessary to take corrective action including the suspension and termination of certified health IT.
Our goal here is to work with developers. If there is a non-conformity of some sort found with the product that has been certified, our goal would be to put together a corrective action plan to help address that so that the product that is out is working as it should be once it – testing lab. We want the products generally to be able to operate as they were certified and as the product operated in the testing lab. That is an overarching theme that you see across the program. Even in our 2015 edition final rule, we included updated surveillance requirements that look at that as well that provide an opportunity for us to look at the products once it enters the world of health IT. Once it goes out and implementation has started.
That oversight rule – also two additional provisions. One is it would allow means for ONC to have direct oversight and the National Voluntary Laboratory Accreditation Program, the testing labs that are related to the program that that would align them with existing policies that exists for other areas as well.
And then also it would require ONC-ACBs to make identifiable surveillance results publicly available on their websites on a quarterly basis. These are proposals that are included. As I said, we are in the deliberative process of putting together the rules. I cannot really expand too much more upon that as we are reviewing the comments. As I said, our hope is to have a final rule out in the fall.
Moving on to some additional activities at ONC, we also do a lot of work as most may know with Federal Advisory Committees related to health information technology. That is the Health Information Technology Policy Committee and the Health information Technology Standards Committee. There are some activities that are ongoing now that I did want to mention. One is looking at June. The Policy Committee has a couple of charges before them. One is that they are actually looking at the MACRA proposed rule that was issued by CMS in terms of the health information technology component. That includes the quality measures and the quality aspects of the programs. The advanced care information category of that program, which is similar to the meaningful use program for eligible providers that existed before QPP, but again I do not want to take up too much time. I know you guys are getting a full presentation on that. Needless to say, FACA is looking at the QPP or the quality payment program. I am going to try not to use too many acronyms today.
A couple of other charges that are before the Federal Advisory Committee are the ISA. ISA is the Interoperable Standard Advisory. The 2015 edition rule includes certification criteria. The ISA is the best available standard that is out at a time on particular areas. They are related to interoperability. That is a process that is not included in the rule, but is a resource for those across the health information technology landscape to be able to see what does the government think are good standards for any particular criteria. It could be related to care plans. It could be related to demographic. That process happens – the ISA is updated on a yearly basis. And the Federal Advisory Committee helps to contribute to that update process. That is also ongoing at Federal Advisory Committee.
A couple of other areas are the Interoperability Taskforce. And the purpose of that taskforce is to provide recommendations are the most impassable policy, technical and public/private approaches that could be implemented and improves interoperability experience. We think that is a big piece of interoperability. It is not just a technical component, but how are they working. What is the workflow? How do we make sure that when implementation happens that interoperability is also happening? That is part of the activity.
One other I wanted to note is consumer playbook. I will talk about that in a little bit. But the Federal Advisory Committee is also looking at that contributing to that process.
I also want to talk about Zika. The Zika Virus, as many know, the response is ongoing and CDC obviously has been very engaged with that including providing clinical guidance. ONC has been working with CDC to take that clinical guidance and think about what are the health information technology components that developers should be aware of and working on.
What we have done is we are having ongoing conversations and working with the health information technology developer community to translate the CDC clinical guidance into guidance for developers. We are also working with developing Zika-related vocabularies and have developed the series of webinars for health IT developers.
Some of the things in thinking about turning that clinical guidance into technical guidance, as many know, the Zika response affects women, but also can affect men as well. Thinking about the questions that are asked that is included in the health information technology system and how that translates over to the system operations is what we have been working on. A series of questions like did you travel to an area with active Zika transmission. Are you a resident of such an area or is there plan travel that might occur?
And then proceeding forward, questions about pregnancy. Are you known to be pregnant? Is it possible that you might be pregnant? And then how to also bring in state health departments as part of that. It also includes questions regarding testing. Thinking about how those questions can be translated into technical guidance. That is an ongoing activity that we are very excited about and we hope will help the Zika response.
Another component obviously is interoperability and that is a lot of what we think about here at ONC. The interoperability pledge was released earlier this year. Right now, companies that provide 90 percent of electronic health records used by hospitals as well as the top five largest health care systems in the country have agreed to implement three core commitments as we call interoperability pledge. One is related to consumer access to help consumers easily and securely access their electronic health data to a desired location and learn how the information can be used as well as to be assured that that information will be effectively and safely used to benefit their health and that of their community.
There is also a commitment on information blocking and ensuring transparency. And there the goal is to help provide how to share individual’s health information for care with other providers and their patients whenever submitted by law and to not block electronic health information.
And the last commitment is related to standards, implementing federally recognized, national interoperability standards, policies, guidance and practices for electronic health information and adopting best practices including those related to privacy and security. Those are the three pledges. Those are an umbrella to much of our work at ONC.
One area I want to talk about in particular is the transparency component of that commitment. One of the recent developments in that space is that we have recently released a website, healthIT.gov/transparency. The goal of that website is not new information, but to take information that is included on what we call our certified health IT products list or affectionately the CHPL. Take information from there and put it an easily used interface for providers and other stakeholders to look at. It is not new information. Just putting all the information in the same place related to transparency so that providers are able to more easily compare information about different products and to see what developers have said about their products related to transparency.
What are some of the requirements? In the 2015 edition, we included a requirement for mandatory disclosure statements. That means the developers must fully disclose all known material types of cost and limitations, including technical and contractual restrictions that a user may encounter when implementing or using the developer’s technology. That is a requirement for 2015 edition products as well as for 2014 edition products that have already been released. The transparency website I mentioned includes links and information regarding those disclosures.
We have also included in the 2015 edition of transparency attestation and the developer has to take attestation and they can decide whether they want to affirmatively respond to the attestation or they want to decline the attestation. But basically the attestation states whether the developer will take additional above and beyond the disclosure requirement adjustments and to take additional voluntary actions to promote transparency regarding their product.
We are also working on surveillance and that is that our ACBs or the authorized certification bodies are conducting ongoing surveillance of the new disclosure and transparency requirements are stated as well as surveillance of the program overall in terms of the certification criteria and requirements of the certification program. Transparency is a key component. And what we want to do is we want to enable increased transparency and conversations between the provider who is purchasing or using a product and the developer who is developing that product. We think by these transparency requirements it is one step in the right direction to increase that conversation so that the products are ultimately beneficial to the provider and to the patient.
We also are doing our part in terms of transparency. A part of that is making this information available as I said through the transparency website, but also updates we have done that we have made to the CHPL, which is that product system, all the products that are certified under our program. We look forward to – important resources for the health information technology stakeholder community. Those will be updated on an ongoing basis.
With that, I hope I did not talk too fast. Those are some updates generally about some of the activities that are ongoing at ONC. If there are any questions, I am happy to take them.
DR. SUAREZ: Thanks so much for your remarks. We do have a question coming from Bruce.
DR. COHEN: Thanks Elise. I had a question around the health IT certification rules particularly with regard to implementation and data collection around race/ethnicity data and sexual orientation and gender identity. My question or actually my concern is establishing the health IT standards is great, but the real issue is training for the collection of these data. Is ONC involved or developing requirements for training for the actual primary data collection of this information? If it is not, I really think that is the next frontier developing appropriate training and health care provider organizations to collect these data. I will stop there.
MS. ANTHONY: Thank you for the feedback. Generally, our certification program focuses on the technological requirements in terms of data quality, which might be part of what you are getting to. That is something I can definitely bring back to the team and see if some of the other teams are engaging in activities. We can follow up with you regarding more and providing additional information on what training you think might be helpful. We can see whether there are existing things happening in the private sector or that is something that is ongoing at ONC. But I can get back to you on that.
DR. SUAREZ: Vickie, I think, has a question.
DR. MAYS: Thank you. Can you tell me whether or not any of the mHealth data is going to be included under you for any reason? Part of what some providers are wanting to do is to use some of that data in terms of Fitbit, et cetera. I am wondering if that is on the horizon.
MS. ANTHONY: One of the things in the 2015 edition that we have included is some requirements – read only access through API. Those are application programming interfaces. This is something that is also mirrored in CMS’ rule regarding what was previously viewed download transmit options and that often is implemented through portals. It does not necessarily have to be. But CMS includes requirements to make information available – download transmit options as well as through API. That – in terms of technical requirements that we are calling for in our rule. We do think that allows more opportunity for the information that is included in an EHR, for example, or a health IT system that could be used in creative ways to benefit patient care.
DR. CHANDERRAJ: There was a recent meeting in February about ONC and abandonment of MUC, meaningful use. I was wondering if the new policy enacted by ONC has been vetted and if so, what are the new changes that are going to be?
MS. ANTHONY: New changes to Stage 3? Is that what you are asking?
DR. CHANDERRAJ: Certification of the IT regarding Meaningful Use or alternates for the Meaningful Use.
MS. ANTHONY: There are two components there. One I can speak to and one I cannot unfortunately. Stage 3 and the EHR incentive program is managed by CMS. There might be a CMS component to that question. Generally, ONC and CMS work very closely together in EHR incentive programs as well as through MACRA as we are moving forward. But CMS handles the behavior requirements of the providers and ONC handles the technological and technical requirements attached to the health information technology that those providers have to have when they are executing the behavior requirements that CMS calls for, if that makes sense. We have the technical component and CMS has the behavioral component.
In terms of Stage 3 and EHR incentive programs, that would be a question more for CMS. The current certification – the 2015 edition is the foundation for the October – the rule that CMS released in October. I think it is October 2015. That provided Stage 3 and what is called the Stage 2 modification. It will also be the foundation for MACRA as has been proposed in the rule. Obviously, when the comments come in from MACRA to CMS, we will look at the comments. At that point, further discussion can be held. But as proposed, the 2015 edition is the foundation for MACRA.
DR. SUAREZ: I think those were all the questions that we had. Thank you, Elise, for joining us and thanks for sticking with us. We are running a little behind and I know you had another commitment, but thank you so much for joining us.
MS. ANTHONY: No problem. My pleasure. Thank you so much for inviting us.
DR. SUAREZ: We are about 20 minutes or so behind, which is a record for us. We are going to take a break now. It is 11:09. We are going to come back at 11:25.
DR. SUAREZ: This next agenda item is one of the action items or the main action item for the national meeting in this meeting is a report from the Standards Subcommittee they are going to be presenting two letters of recommendations from the February 2016 hearing, one on attachments and one on operating rules.
They also have an update on the Review Committee report. We might not get to it until tomorrow. I will turn it back to Alix and Terri.
MS. GOSS: Thank you so very much. Thank you everyone for joining us today. Terri and I are going to be tag teaming through today’s two letters and possibly depending on how quickly we get through this the Review Committee outline. My co-chair, Ob Soonthornsima, is on the phone. He is going to supplement anything we missed or add any additional context or value that may be critical to our discussion.
The way we are going to run today’s discussion is that we are going to do some set up on the – we are going to start with the attachment letter. Terri is going to frame up, give some background perspective on it. Then I will give a brief general summary and then walk through each of the recommendations by title, each of the sub-committee members who have had numerous opportunities to review the draft letters. We then took the draft letters to the executive committee for discussion and then we also distributed them in the eBook. You probably have multiple copies.
We have a few little tweaks that we will also incorporate today based upon some feedback we got. We will walk you through those. Once we, as a committee, have an opportunity to review the attachment letters, we will then move on to the Phase IV letter using a similar approach.
Without further ado, I am going to turn it over to Terri.
MS. DEUTSCH: Good morning everyone. I am going to start out with just giving you a little bit of an understanding of the past of what has happened with the attachment letter. Some of you are well aware of it. And to some of you this is new. Please bear with me as I go through this.
The health care claim attachment is stipulated in HIPAA and it is one of the transactions for which electronic standards were to be adopted. Attachments are one of the first major opportunities for convergence of clinical and administrative health care data and information exchange through standards.
The exchange of attachments between providers and payers has and do occur today with the majority being solicited or requested by the health plan or the payer. Most occur by a paper, email, fax, or phone and sometimes through the health plan’s website.
In 2005, HHS published a proposed rule, proposing the adoption of the health care claim attachment. A final rule was not published in part due to the concerns with the maturity of the standard and the ability for it to be implemented.
The attachment standard has been addressed in four NCVHS hearings over the past five years in 2011, 2013, 2014, and the last one in February of 2016. Subsequent to the hearings, letters of recommendations were sent to the secretary of HHS. Initially, the recommendations indicated that it was too early to make formal recommendations regarding adaption of any attachment standard, implementation specification, and operating rule.
A later letter indicated that because the consolidated CDA Release 2 and implementation specifications for attachments had not been finalized, NCVHS felt it was premature to recommend the new standard that was being developed. However all of the letters acknowledged strong support from the health care industry on the adoption of a health care attachment.
Earlier recommendations have been included in the letter that we are going to be discussing today. Some examples of those earlier recommendations are expanding the definition of the attachment standard from a health care claim to supplemental documentation needed about a patient to support a specific health care related event such as prior authorization, referrals, et cetera using a standard format.
Attachment standards were to include three types of transactions of query, response, and acknowledgement. The standards were to support structured and unstructured data and the process should support solicited and unsolicited attachments and that they should all have application of minimum necessary requirements. On Thursday, you are going to hear about minimum necessary requirements. The proposed letter today is a compilation of the past recommendations to the secretary and the result of the February 2016 hearing.
The letter has an overview, having a summary of the transaction and the standard, the benefits of the transaction, the objectives and the purpose of the February hearing and the criterion of the factors that we use to evaluate the proposed standard to derive the recommendations. The proposed attachment standard represents a collaboration between two standard development organizations, the accredited standards committee or ASC X12 and the Health Level Seven International or the HL7.
ASC X12 provides the electronic data interchange, also known as EDI communication channel and HL7 provides the clinical data architecture or CDA, which is the vehicle to submit additional information required to adjudicate a claim.
Additionally, the American National Standard and the American Dental Association Standard Number 1079 for Standard Content of Electronic Attachments for Dental Claims describes the content for electronic claims attachment to payers for dental claims. And the National Council for Prescription Drug Programs or NCPDP recommended adoption for prior authorization attachment related transactions for the pharmacy industry known as the NCPDP script standard version 2016071.
Based on the review of all the testimony that we received and past recommendations, we believe that the standard meets the industry’s business needs, improves administrative efficiency, reduces administrative burden, is flexible and agile to meet future technological developments and health system changes are mature, are adoptable, are enforceable, and meet the other criteria that we use such as completeness, efficiency, complexity, consistency, effectiveness and ambiguity.
I am not going to turn it over to Alix to go through the recommendations.
MS. GOSS: Thank you for that set up, Terri. From a general summary perspective on the attachments, we believe that we need to adopt one standard definition of an attachment and establish the scope of the transaction. We need to adopt a set of mature implementable standards for the attachment of functionality. We needed to find a series of transaction process requirements, take an incremental flexible implementation approach no less than five years. We need to broaden the testing education outreach and compliance efforts related to attachments and ensure alignment of attachment standards with regulatory requirements for EHRs adopting certification program in MACRA and MIPS. In a general perspective, that is a summary of our recommendations.
I am going to work my way through each one of the recommendations. Terri will make any – if you have comments as we go through each recommendation one at a time, I am just going to – Recommendation 1, read the title. And then we are going to see if there are any comments. If there are, Terri is going to take those notes at the end of our discussion on the attachment letter. She will recap any of the changes that we have discussed and agreed to. Then our intention is if there is clear consensus, we will likely proceed to a vote. If not, then we will make any corrections and bring them back at tomorrow’s subcommittee session at 9:45.
Recommendation 1. Adopt one standard definition of attachments. Any comments?
MS. HINES: For those in the room, you will need to look at the version that you got from the E agenda book, the copies that were page 40 of the E agenda book. The hard copies that were provided today unfortunately only have the odd number of pages.
MS. GOSS: Recommendation 1. Last call.
Recommendation 2. Adopt the standards for attachment query, response, and acknowledgement. Any feedback on Recommendation 2?
Recommendation 3. Adopt the recommended standards. This is related to query, response, and acknowledgments. It also includes the value set for the LOINC codes.
Recommendation 4. Any feedback? Adopt additional standard for attachment response.
Recommendation 5. Identify standards for rooting/envelope of an attachment.
Recommendation 6. Define structured and unstructured data.
Recommendation 7. Adopt attachment standards. Pharmacy industry. We do not have the edited version on here. We have received a correction from NCPDP just to make sure we are referencing the right one. This is a small revision just to make sure that we are technically accurate with the current version. We would name NCPDP script standard version 2016071 for prior auth transaction as Terri noted earlier.
Recommendation 8. Adopt attachment standards for the dental industry also referenced by Terri earlier.
I see no hands up. I want to see if there are any comments on the letter of recommendations.
Moving on to the subsequent ones. Recommendation 9. Transaction process to support solicited and unsolicited transactions.
Recommendation 10. Avoid duplication and chain requests.
DR. CHANDERRAJ: I have a question for ten. Most of the information requested by the billing authorization wants to delay the payment rather than see the attachments. Is there a penalty for asking that information when it has already been submitted?
MS. GOSS: I think what we are trying to do with this recommendation is ensure that if data elements exist in the HIPAA adopted transaction sets that should be the mechanism for getting that and there should not be a request for information that was already in the base transaction. This should be for the supplemental data that would be needed.
DR. CHANDERRAJ: But is there a penalty associated with requesting that information?
MS. GOSS: The question of a penalty would not be something that NCVHS would set. It would be part of the rule promulgation. I think we have to wait to see what they would recommend in that. An industry comment could suggest that would be, but that would be only so enforceable beyond the feds.
MS. DEUTSCH: There is a requirement that if there is an adoption of standard and operating rules that they have to be compliant. Since we have not adopted yet the adoption of the attachment standard whatever happens now is something prior to it. We are recommending in this letter of recommendation that there not be this request for additional information. Does that answer your question?
MS. GOSS: Anything else on Recommendation 10?
Moving onto Recommendation 11. Conform to privacy and minimum necessary requirements.
Recommendation 12. Additional guidance for implementing the attachment standard.
Moving into Section 4. Taking an incremental and flexible implementation approach. Recommendation 13. Utilized flexible adoption and implementation approach. Any comments?
Walter, we have an audience member. Do you want to hold?
DR. SUAREZ: We normally do not take questions from the public. We do have a public comment period at the end at 5:30. If you could hold the comments until then, we will appreciate that.
MS. GOSS: If it is a clarifying question, I would be happy to talk to you afterwards.
Recommendation 14. Provide testing, communication, education, and outreach.
15. Publication of a notice of proposed rule.
Recommendation 16. Ensure alignment with EHR incentive programs MACRA and MIPS.
I am seeing no committee member comments or changes for the attachment letter.
MR. SCANLON: You are proposing that HHS consider a proposed rule —
MS. GOSS: We think that NPRM is an essential piece of activity to engage the stakeholders, make them aware of it, solicit for some further comments especially from any policy framework that HHS would put around or in the NPRM and then to give us a glide path for implementation. It was pretty clear that the use of HL7 standards is foreign to some in the health care industry. There is a need for an appropriate time for people to get their arms around that and put it in their own development cycles.
DR. SUAREZ: Just a point of clarification or two points. One is CMS published an NPRM on attachments already. It is almost seven or eight years old. There was a question as to whether that would be the NPRM and now this would be the final rule. I think we are recommending that there should be another NPRM.
And the second clarification – really the Affordable Care Act provision calls for the secretary to publish final rules – it is a point of clarification we are making by recommending that the secretary should consider following the normal NPRM on this process.
MS. GOSS: Walter, it appears to me that we have no issues with the letter. We do have one public comment, but I am not sure if it is a clarification or some concern.
DR. SUAREZ: We could potentially give the consensus and no comment on the letter take action right now. We have scheduled to take action tomorrow. We wanted to certainly take the opportunity later this afternoon when we have the public comment process to receive any public comments and then consider those as we add some whatever additional edits and then present back the letter to the Full Committee. Given the fact that there is at least one public comment and maybe there could be more, we will just not take action right now. We are going to then listen to the public comment this afternoon and get additional feedback and then finalize the letter and present it tomorrow for action.
MS. GOSS: That is wonderful. At this point, what we will do is go ahead and move on with the committee’s support to the next letter. We are going to now talk about the Phase IV Operating Rules.
MS. DEUTSCH: I am going to provide a little bit of an overview about the CORE and the operating rules to give a little bit of an understanding of what the operating rules are, what the authority is in CORE. And I know if I misquote anything, Gwen will correct me during the public comment period. I hope she will bear with me as I go through trying to summarize what it is that they do.
Basically what I am going to do is talk about what operating rules are, a little bit about the CAQH CORE process, what the currently adopted operating rules are and then which ones are being proposed. The question is what is an operating rule. And the operating rules are business rules that supplement other standards and define the business elements in the execution of administrative transactions. There are two types of operating rules generally. One is the infrastructure. An example of the infrastructure would be system availability and response time and the business process operating rules. An example would be the response actions based on specified situations.
The authority resides in Section 1104 of ACA, which is the Affordable Care Act, that requires the secretary of DHHS to adopt operating rules for the electronic exchange and use of health information for financial and administrative transactions and that there be an appointment of a nonprofit entity to develop these operating rules.
The Act also requires that the operating rule authoring entity focuses on administrative simplification, has a multi-stakeholder and consensus-based process for developing the operating rules, have a public set of guiding principles that ensure the operating rules and process are open and transparent, build on the HIPAA transaction standards, and allow for public review and update.
CAQH is the Council for Affordable Quality Healthcare committee on the operating rules for information exchange, which is CORE. That is how we get CAQH CORE – met the specified requirements and NCVHS recommended that it be the operating rule authoring entity.
The process that is used by CAQH CORE is that a subgroup gathers information to draft the rule and the language and requirements and that these then have a workgroup of experts to approve this or disapprove it. And then the full CAQH CORE voting membership either approves or disapproves and that final approval is through the CAQH board of approval.
The current adopted operating rules are health plan eligibility, benefit inquiry and response, claim status, electronic remittance advice, and electronic fund transfer. The proposed operating rules that we are going to discuss in the letter that you have in front of you are the health plan enrollment and disenrollment, premium payment, prior authorization and health care claim or equivalent encounter information.
The next slide has a graphic description of how this all works into the business process of health care. If you look at the left hand side, it has the different processes such as pricing and adjudication, payment and remittance advice. And then on the right hand side, it shows the relationship between these different processes and the transactions that are involved, the standards and the operating rule.
For example, if a provider makes a claim status inquiry to the payer, they use what we call the 276. The payer responds using the 278. The operating rules provide the infrastructure and the business process for the transaction such as the system availability for the transaction, which would be the inquiry and response to occur. The response time for the payer to respond to the inquiry is also included. And then the end product is the adjudication of the claim.
The letter to the secretary on the Phase IV Operating Rules provides the background information. We had two hearings on the Phase IV Operating Rules, one in February of 2015 and one in February of 2016, and the summary of the three main issues with the proposed operating rules.
The criteria that were used for this letter of recommendation are the same ones that I went over when I talked about the attachment standard. There also is in this letter the findings and recommendations. Alix will now go over the recommendations.
MS. GOSS: Thank you, Terri. I believe everybody has also received the advanced versions of the letter and had opportunity to comment. We have received some other feedbacks. We will also be talking about that today.
For the Phase IV Operating Rules, we are strongly supporting voluntary adoption of the operating rules by the industry. We are not recommending at this time to promulgate adoption. There have been three issues that we have identified as a Standard Subcommittee based upon the testimony that we received. The first one is required use of standards that have not been adopted, specifically, the use of the acknowledgment transactions, which have not been formally promulgated by HHS at this point.
Second, the testimony that we received and the multiple hearings that we held indicated that there were limited benefits at this point in time.
Thirdly, there were three of the four transactions for operating rules that have very limited use across the industry.
I am going to go through each one of the recommendations. I believe you also have a hard copy of this in front of you as well as received it in the E book. Rebecca used which version?
PARTICIPANT: Use the one in the E agenda book.
MS. GOSS: We are going to use the same approach and walk through each one of the recommendations that was put forth in the letter. Recommendation 1. We strongly support a voluntary adoption of Phase IV Operating Rules. Are there any committee member feedback on Recommendation 1?
Recommendation 2. Address the inconsistencies in the connectivity requirements.
Recommendation 3. Adoption of the acknowledgement standard.
Recommendation 4. Authentication.
Moving into Part 3 for the next recommendation, which is number five. We have suggested modifications that we have included within Recommendation 5.
In Recommendation 6, convergence of enrollment standard for health – it is five.
PARTICIPANT: — Recommendation 6, there are edits that are not showing up on your handout.
MS. GOSS: In Recommendation 6 actually in leading up to that section, we realized that there was an important component that we had omitted in the initial draft, which ties to the Center for Consumer Information and Insurance Oversight Companion Guides, CCIIO. We realized that in order for us to bring convergence in the industry related around the 834 transaction for enrollment, we needed to actually blend the HIPAA transaction that was already adopted with the 834 HIX implementation guide from X12, but needs to bring along the CCIIO companion guides for us to normalize these three pieces of the puzzle into one cohesive converged standard. We wanted to make sure that we appropriately included the CCIIO companion guides and have made a small tweak in the language just to make sure that that piece is brought along.
What you are seeing on the website or what you are seeing me present does not have this markup. I have it here because it was a late day catch.
MS. DEUTSCH: If you look at after Recommendation 5, it has proposed enrollment, disenrollment, and health plan premium payment operating rule. In the first paragraph at the end, it starts out by saying – this is the correction. I am reading the correction. In addition, health plans participating in the insurance market places (HIX) are having to work with the 834 standard adopted for HIPAA-covered entities, the 834 HIX standard, and the Center for Consumer Information and Insurance Oversight (CCIIO) companion guides. And the word mandated is crossed out.
And then if you go down to Recommendation 6, it says as indicated in the NCVHS February 26, 2016 letter to you, we recommend that HHS explore ways to bring to full convergence the 834 HIX (used by the insurance market places). We added and the CCIIO companion guides and the current 834 used by HIPAA covered entities. It really is just the addition about the CCIIO guide that we added in here.
MS. GOSS: I am going to move on from Recommendation 6, seeing no further.
Recommendation 7 relates to the 278. NCVHS reiterates its recommendations in the February 29 letter to the Review Committee. There were several specific ones related to the 278 transaction in that group of recommendations. Any comments on Recommendation 7?
We had that one technical correction or addition that Terri reviewed. We also received late yesterday some requests from CAQH CORE. There are several requests. The hard copy of the letter, I believe, was placed at each one of your seats. Hopefully, everybody has a copy of that letter, which actually is two letters in one. They sent us a letter a month ago regarding their support for the Phase IV Operating Rules and then followed up and included that letter as an attachment to the one we received late last night.
The first request is regarding our recommendation for voluntary support that they are asking us to consider what specific actions HHS could take to provide its voluntary support and outline such in our letter. It would be disappointing to lose momentum in improving efficiency, effectiveness and security of the various transactions and they would like us to make some further recommendations as to what HHS should do to support the voluntary recommended adoption.
Are there any committee member questions or comments on that request number one, which is located at the top of page 2 of their letter?
DR. CORNELIUS: I just have a quick process question with regard to this letter. Is it that we would vet with this adjusting, revisit the letter – something that would take more than this week to do or are we expected to incorporate the suggestions into the letter?
MS. GOSS: I think from a process perspective, first we need to make sure everyone is aware of what they have asked. We need to allow the committee to discuss it. I think the Standards Subcommittee has some extensive background on the topics and has established a pretty strong consensus as to these recommendations. I would like to hear from the Standards Subcommittee to weigh in to inform the larger committee’s conversation as to whether we want to entertain any further revisions or proceed with finalizing the letter and voting on it tomorrow.
DR. SUAREZ: One of the methods that we have now established as part of our transparency process is for any letter that is going to be taking as an action item during a meeting. We post the draft of the letter at least a week before on the website so everybody can see it. We can receive additional comments or any corrections or any issues as we have already heard a few of them here.
While in the meeting like we are doing here, we add any comments from the committee. We add to those comments. The comment that we received from the external resources based on the draft that we published. That is what we are bringing in and we consider all of those including the comments from the outside of the committee in finalizing the letter and then making a final decision. That is the step where we are right now is considering that input that we received as we publish the draft on our website.
MR. COUSSOULE: Just one point about that initial recommendation. I think we clearly heard during the hearings and discussed the encouragement to the continuation of the development of standards without asking the secretary actually to make some specific rules about it. I think the message is consistent with what we talked about. Whether the letter states that or not I have to leave that to other people who have a lot more experience with the policy side than I do to say whether that is adequately represented in there or not, but we clearly heard that during hearings and had that discussion. The point is well made. Whether there is anything different we can comment in the letter or not I am not sure.
MS. GOSS: I question also whether it is our role to tell HHS what they should be doing especially when they are not promulgating something and what role they have. I think that my experience is that CAQH CORE has done a phenomenal job with promoting webinars, their reference material, et cetera and would encourage them to continue to do that. Are there other committee members that may have comments on request one?
MR. LANDEN: I think the issue there and their issue number one is germane, but I do not think that it is something that we are in a position to put in the letter right now. I would suggest we work with CAQH CORE and others later to figure out what would benefit the industry in terms of what we might expect out of HHS to help along a voluntary adoption.
MS. KLOSS: I think in the past we certainly have made recommendations for HHS doing education outreach, other kinds of initiatives to move something along. I think that this is consistent with that. I do think we could probably bring that notion in the last sentence where we say NCVHS strongly supports access and evaluates. That HHS support – be engaged in supporting voluntary adoption.
DR. GOSS: I am struggling with HHS’ ability to do that and put resources to it if they are not promulgating something. In ICD-9 in minimum necessary and 5010. Those were things that they were actually endorsing and adopting. To me, that is the fine line here. I think we have all felt the level of collaboration in trying to make these operating rules something that is going to bring value to the market and there has been a tremendous amount of process there and promotion of it. But we did not hear that industry consensus that there was the value to add these. If we are not taking it to the secretary to adopt then how can they put resources behind it? That is probably a question for the feds.
MS. KLOSS: Something about technology and their availability on administrative simplification websites, other kinds of things, seems to me to be a reasonable collaborative step. There is a lot of evidence of public/private collaboration around topics that are to the public good, but not a standard. I think a sentence could bring that spirit into the recommendation.
MS. GOSS: I am hearing Ob make a request.
MR. SOONTHORNSIMA: I am trying to get Linda to clarify a little bit. Are you suggesting that we augment Recommendation 1 to expand on collaboration and look for —
MS. KLOSS: — that HHS uses its resources to – we will have to write that sentence. I do not think it would be inconsistent with what we have done.
DR. CHANDERRAJ: The right thing rather than delayed.
MS. KLOSS: Or just make sure that when – in the public presence of this, it is clear that there are voluntary operating rules available and where to access them.
MS. GOSS: Other comments on request one?
DR. COHEN: I agree with Linda. I do not think HHS needs to be promulgating a rule to actively support the voluntary implementation.
MS. GOSS: I think what I am hearing is a couple of people are in agreement that we want to add a sentence to just strength that one. I just want to see if anybody disagrees.
MR. SOONTHORNSIMA: It does not change the context of the recommendation.
MS. GOSS: It does not. But it certainly goes along with the subcommittee’s intent to strongly support voluntary adoption. Terri and I will take a stab at that. I have no concept of time. Walter, keep me on track.
Request two is to clarify if we are asking the secretary to revisit the intent of ACA regarding standards and operating rules including mandated standards that currently have a low adoption. I am not going to read the background on that. Hopefully, you have all had a chance to look at that letter. Are there any comments on request two?
MS. HINES: The clarification is are we asking the secretary.
MS. GOSS: We are not.
MR. SOONTHORNSIMA: We are only focusing on the task at hand and operating rules here.
MS. HINES: Where in the letter could that be made more crystal clear?
MS. GOSS: To be honest, I was not sure where they were getting that issue because I am not sure what in the letter prompted that personally.
DR. COHEN: It is your sense this is sort of out of scope for this letter.
DR. SUAREZ: I think the question has come based on the comments. The question is really whether by virtue of making the recommendations we are making we are sending a message in some way that the secretary should revisit the intent of ACA regarding standard operating rules. We are not doing that. What we need to do is go back in the letter to see if there is a message in the letter that gives the impression that we are recommending that the secretary revisit the ACA. We are not doing that. We are not asking the secretary to revisit ACA. Certainly, the secretary would have to go to Congress to have it revised. It would take an act of Congress to do it. What we would need to do is go back to the letter and make sure that it is not giving that impression.
MS. GOSS: This conversation actually ties to request two and three. We can take a look at that and see if there is anything that would be giving that impression. If there is, we will track changes on that.
I am going to move on to request four that they ask NCVHS to specify the type of testing that would be consistent for foundational expectations of existing standards such as those in Phase IV versus testing for new data content standards. Please address whether real world use or testing of that use by an external body would be considered sufficient. Moreover, we ask NCVHS to define the other non-statutory criteria of which authors should be aware.
In the letter, we have a reference to testing in the beginning. It is a very brief section. Actually, this was something that we tweaked as a result of our executive committee discussion of one of the initial versions of the letter. This is a pretty limited section in regards to testing.
We strongly supported the adoption of Phase IV Operating Rules and that this would provide CAQH CORE and the industry time to test, assess, evaluate, improve scalability, cause benefit and value to the industry. Testing is another way of getting good operational feedback that can make subsequent versions of a rule or a standard more robust. We felt that that was an opportunity to bring more value. I am not sure if there is committee commentary on request four.
MR. SOONTHORNSIMA: I think this is why we have to leave it to CAQH CORE and the industry to determine how they can improve if there is any opportunity at all to improve because we are not close enough to the actual operating rules development and how this would be implemented because we are evaluating the recommendation from CAQH, the practicality or readiness of the industry based on feedback. That is as far as we went. I do not think we can add anymore to that request. We are not close enough.
MS. DEUTSCH: Based on the comments that were made, it appears that for request number one we will look at the language that we have included about HHS and ensure that it reflects the supports of voluntary adoption or implementation of the Phase IV Operating Rules.
For request two and three, we are going to see if the letter implies that we are requesting revisiting the requirements in ACA for the operating rules.
MS. HINES: Terri, on this one, I would invite during the public comment period for any specific suggestions around that if there is something that we are not seeing and the public would like to comment on, it would be a perfect opportunity.
MS. DEUTSCH: That is a great idea. We will ask for that at that time. Absolutely.
In request number four that we are not anticipating taking any action or revising any language for that.
MS. GOSS: Have we correctly summarized the feedback on the Phase IV letter?
At this point, I believe we can move on. Do we have time left in our agenda item, Walter?
DR. SUAREZ: I think we have actually just did something we have never done, which is come back to our regular time.
MS. GOSS: What we plan to do tomorrow is cover the Review Committee table of contents outline, which everybody else has received in the E book. With that, I turn it back over to you.
DR. SUAREZ: Thank you so much. Thank you very much to the Subcommittee.
MS. GOSS: We just got a question on the first letter. He has a clarification on including eligibility in the definition of attachments. We will discuss this during the public comment period. Can you come back, Durwin, at 5:30? Is that when we are doing that?
DR. SUAREZ: Both at 5:30 as well as during tomorrow if there is anything else. But ideally at 5:30. I also received an email where there is some comments as well.
MS. GOSS: I will hold onto this one and we will hold it for 5:30.
DR. SUAREZ: We will review them at 5:30 as we open it up for comments.
We are going to take a lunch break. It is 12:20. Hopefully, we can come back at 1:15 to get started. Let’s just schedule it to come back at 1:15.
DR. SUAREZ: I think we are going to go ahead and get started again. I do not know if we have any new members that have not introduced themselves yet. We are going to go ahead and start with our next session.
We should first do actually the dinner count. Every meeting, as you all know, we have a special dinner social event. Tonight we are going to have our group dinner. As you see here, it is at PJ Clarke restaurant, not too far away. We needed a count of how many people will be present. If you could raise your hand if you are planning to be at the dinner tonight. Thank you.
This next session is really very special and important. We have always talked about how we need to go back to our roots, as they say, and go back into the kind of work and the kind of topics that we had originally been focusing on from the very beginning when the National Committee was created, the National Committee on Vital and Health Statistics. We are very fortunate and pleased to have the opportunity to begin that process of focusing again back into the topics of public health, the data and vital and health statistics. We thought of doing this by starting in this process that we are going to go through over the next several months by bringing in a briefing on public health data, the public health data landscape and hearing from NCHS. We heard from Charlie this morning, more from a larger policy perspective. We will hear some more details about NCHS’ work on data, public health data.
We will also hear from the Census. And then we will close with some presentations and discussions about the electronic public health case reporting. That is our initial set of topics. I know in September, we are already beginning to plan for additional discussion of some of these topics. Again, through the leadership from our Population Health Subcommittee and co-chairs and our Standard Subcommittee co-chairs, we are going to continue to build on this and talk more about vital and health statistics, some of the data policies around data quality, quality issues and all those things.
I am very pleased to welcome our presenters today. We are going to down the list on the agenda. We will start with Susan Queen from NCHS. We will hear an overview and then hear about the National Health Interview Survey and other health care surveys. And then we will go to the Census Bureau on an update on race and ethnicity collection. As I mentioned, we will wrap up this portion of the briefing with E-case reporting.
Susan, I will turn it to you.
DR. QUEEN: Hello. Good afternoon. I hope you will keep me on time because I am not sure how long and I do not want to go over anyone’s time because I know you have a lot to accomplish in a short amount of time.
I was asked to give an overview on NCHS, but you are going to be hearing for the National Health Interview Survey detailed information. You are going to be hearing on the National Health Care Surveys and the provider surveys in detail. You will be hearing in the fall significant detail on vitals.
I am going to a little rogue in what was requested. I will be touching on the NCHS data, but I think in terms of the committee and what perhaps where the committee might be of help or might be interested in knowing what is going on at NCHS, there are some other issues that have been percolating for a long time that are important to consider. I would just like to make you aware of them.
More importantly, I would like to talk about what NCHS’ role is as a statistical agency. This is the NCHS mission. Principal health statistics agency. I do not need to read this for you, but if you read it, you can see the focus here is not on a program, not on policy, not on regulation, not on enforcement. It is a purely statistical mission. Collection, production, dissemination of information.
CDC’s mission. NCHS is located within CDC. It was not always the case, but has been since about 1987. As CDC’s mission – health protection agency. If you look at the mission and how CDC accomplishes its mission, it is very clear. CDC has to take positions. It has to make recommendations. It has to have programs. It has to take actions. It has to have policy.
Since NCHS and CDC are located in a cabinet-level department, we are looking at a department who not only has a programmatic mission, but it would also have regulatory actions that are required, policy actions. Each of these entities accomplishes their mission in different ways. NCHS is housed within CDC, which is housed within HHS. Somehow you have to find a way that these missions are compatible, how you work together, how you collaborate and coordinate. It is certainly something that has been done for many years and is continuing to be done. But there can arise certain natural tensions I would say from this relationship, which is actually common for all statistical agencies in the US.
This is OMB’s definition of a federal statistical agency. Why I wanted to put this up is HHS like many other federal agencies has a lot of statistical programs, statistical activities, and statistical surveys. But HHS has one and only one designated principal statistical agency and that is NCHS. HHS also does have a unit or component that has since been designated as a statistical component and that is within SAMHSA. It used to be the National Survey on Drug Use and Health was in the Office of Applied Studies. It is now the Center for Behavioral Health Statistics and Quality. That component is a designated unit. But as an entire agency, NCHS is the only one. Obviously, there are other statistical surveys like the Medical Expenditure Panel Survey. At one point, there was some consideration within the Agency for Healthcare Research and Quality. Would they want to have a designation like SAMHSA does for the National Survey on Drug Use and Health? Would they want to have this designation as a statistical component for the MEPS?
And the issue would then be they would not be able to then publish in that office or program their policy briefs. There are benefits and then there are limitations. One of the things a statistical agency must do is to provide objective, credible, relevant, timely statistics that are free from any bias or perception of bias.
In brief, there are five major areas with NCHS regarding data collection. It is the National Vital Statistics System, which is our vital records, the National Health Interview Survey, a population survey that is a personal interview conducted in the home, National Survey of Family Growth, which is also a personal interview conducted in the home, National Health and Nutrition Examination Survey, which involves a personal interview conducted in the home followed by a physical examination and laboratory tests, and then the National Health Care Surveys, which are quite a range of surveys that collect information from both medical records, but from providers, from a variety of facilities, electronic health records. All of these together make up the primary and complete set of the NCHS data collection activities.
I am briefly going to mention the National Survey of Family Growth because you will be hearing about the other surveys today and this is one that is perhaps a smaller and not as well-known survey that NCHS conducts on men and women aged 15 to 49 now, looking at fertility, family formation, and reproductive health. It is one of the surveys that is used in more ways than many of us realize. Across the department, it is used for reporting on performance measures. Within CDC, it is used for reporting on progress and meeting the HHS strategic plan. It is a lesser known survey and it is probably the survey that is now to the greatest extent reliant on working with outside entities and co-sponsors and other agencies that support the survey. This is common throughout all the surveys, but particularly for the NSFG, there is a much greater reliance on external funding to be able to conduct the survey.
NHANES, as you know, has the trailers that travel throughout the country. It is now continuous. It used to be – I think it was a six-year or three-year cycle. It has been continuous for some time now. They are exploring all kinds of additional data collections that are affiliated with the family of NHANES, including some potential longitudinal components. This survey is the one that does all the growth charts, hypertension, estimates of prevalence throughout the United States of a number of measures that are reliant upon having both a physical exam, laboratory measure in conjunction with the survey.
That was my brief foray into data. Back to NCHS as a statistical agency. This is the slide that everyone always shows. It shows the various agencies in the federal statistical system. Coordinated in the center, you have the chief statistician at OMB, OIRA. All of these agencies that are listed on here have, as their primary mission, collecting of data, producing official federal statistics.
The US has a decentralized system. All of these agencies are housed within other departments unlike Australia or Canada, the Netherlands where you have an official Office of Statistics that is for the entire – all the range of statistics for their countries. You have Labor over here and Census over here and NCHS over here. All of them have to function within these different departments.
These agencies also vary to a great extent in the staffing levels and budget. The big three are going to be Census, Labor, and Education are the top three. I think Bureau of Transportation Statistics is one of the smallest. Bureau of Justice Statistics also has a pretty small office. If you think about crime reporting, they are responsible for some very important statistics for the nation.
Our statistical system was set up in the Constitution. I know I am preaching to the choir here. All the representation within Congress, the apportionment among the states is according to the numbers in the states. And the enumeration itself is in – you can probably quote it better than I – the manner as they shall by law direct and they meaning Congress. Here you have set in our Constitution the federal statistical system, but it is oversight and how it is run at a political level. Here you are trying to operate in an objective, credible, transparent manner, but you are underneath this mantle of politics. All of these federal agencies are subject to these rules, but they have to figure out a way to navigate and to behave and to operate in a manner which provides credible, relevant statistics that can be trusted. Ultimately, you will not get information from people, from facilities or from establishments if you cannot be trusted.
What is it that guides the federal statistical agencies? OMB provides coordination. The statistical policy office provides guidance. They have published statistical directives. They have published standards for surveys. I think all agencies like to describe themselves as special and unique and we are different. Program agency has to operate under certain strict program requirements. You have grants. You have grant policy. Statistical agencies have their statistical policy. They often are citing this to other components within their own department and certainly outside of their department about why they can or cannot do something. To some extent, this can be perceived I think sometimes as you are not being responsive to me. I think that statistical agencies can improve and do a better job on how they do respond while still maintaining their integrity, finding ways to be responsive and helpful. And providing access to data is a key function.
I think a lot of components within HHS have heard NCHS. I have only been back at NCHS for a year. You are special. Whenever I start talking about we are a statistical agency, which is why we cannot give you what you want the way you want it when you want it. Let’s figure out what we can give you, what we can provide, how we can answer the questions that you want answered.
This fourth bullet here, the Committee on National Statistics, which was established in 1972, was set up to advise the federal statistical agencies. They observed I think fairly quickly that many statistical agencies are routinely faced with situations that can be taxing in terms of standards and professional behavior. Examples would be policymakers, regulators or enforcing officials that want to get their hands on the data of individual respondents or individual facilities or establishments. That is more common than you would ever think for purposes other than statistical.
The other quick example that comes to mind is when a policy interpretation has been added or is trying to be added to either a statistical product or a press release that is a statistical product. Let’s add some text about our program and how we are helping to fix this problem or what does this mean for our program where you need to keep that nice functional separation very clear between the statistical information and the programmatic information.
CNSTAT published – they started in ’92, this principles and practices for a federal statistical agency. Since then they are on their fifth edition working on their sixth. But one of the first things that came out of that 1992 principles and practices were the statistical agencies needed a strong measure of independence. This is something that has been reiterated throughout the years. Instead of going away or being addressed or being taken care of, it is a continuing, ongoing, recurrent theme and issue such that in the last edition they revised the strong measure of independent to be independence from political and other undue external influences. There is constant pressure that the statistical agencies face.
Speaking specifically on this principle, these are either practices or components that CNSTAT has identified as being part of this measure of independence and having the functional separation from program and regulatory activities is one of them. I think NCHS despite its physical location or structural organizational location within CDC is functionally separated. Maintaining that or explaining that within CDC and HHS is actually the harder challenge.
It is not always well understood within a given agency that there is a distinction between what you are doing over here and what somebody in a program is doing over here particularly when you do have – within CDC, you do have statistical surveys where the program is rigorous and they know what they are doing. We are a conducting a statistical activity and you are too. Why can’t you share this embargoed information well in advance or why can’t you include this programmatic function and a policy within your statistical product? It is a challenge to explain and to continue in part because you have turnover of people. You have turnover of politicals. I think it is something that we can find better ways to clarify and articulate and perhaps publish even on our website in a better way that would inform within CDC and perhaps even within NCHS.
I would not be surprised to learn that there are individuals who come to work for any statistical agency who may not understand what their role is in a statistical agency. And why can’t I include in this my publication about why we should be doing more on such and such? Well, that is the program that talks about how to prevent deaths. We are the ones that count deaths. It is something that has to be continually articulated.
Charlie mentioned this morning – he said our independence is being tested. I think that is coming under increasing pressure. Some of that is tied very much to something he mentioned this morning, which is the timeliness of the data. As we release data more quickly and particularly it has become the case with the vitals and with the National Health Interview Survey as they pushed up. The federal release program – I know Jim mentioned ASPE and CHS collaboration. NCHS actually was doing the early release program for years. But what the collaboration more recently has been to even push up the time even further to a shorter period of time after the end of the data collection a quarter. As that has happened, the tension for access and being able to analyze and being able to put a policy spin is even more intense.
The organizational placement of NCHS. Actually, I mentioned this already that we are – NCHS is organizationally deep. I changed that. I am a very transparent person, but I changed the way I had phrased that. Charlie said don’t say it that way. This is an internal slide from the CDC. It is just an org chart. Where is Waldo? NCHS – it is on there on the left, the second column just above the bottom one. But what is really interesting about this slide and this is on the CDC website was there aren’t any connections between the lines of the group of boxes on the bottom and the ones on the top. You cannot get there from here. Quite frankly I will say there has been a very positive working relationship with NCHS and CDC in the recent years. NCHS is now one of two centers under another office in CDC. It has been a good working relationship.
But the director was designated of NCHS to be the senior health advisor on health statistics to the secretary, direct report to the secretary, as well as a direct report to CDC director. That designation still stands.
I want to give a budget update. For 2016, this is what was enacted. CDC had $6.2 billion. They got $272 million increase over FY15. NCHS, $160 million. We do what we do with that and then the help of the cosponsors that support the surveys. We did get the $5 million increase. It was targeted specifically to vitals. We did not get PPHF funds. We have stopped asking for those funds because it does not appear that we would get them. In some ways if you read the law, I can see how they are doing a very strict interpretation of where those funds should go. We had had funds and CDC has provided funds that help support the sample increase for the HIS to expand content and also be able to provide health insurance estimates for all 50 states. That has been for the past four or five years between the PPHF funds and then CDC’s support. The continuation of that is unknown and at this time not expected.
For 2017, the bottom line is – pardon me. At the bottom, I said Senate markup. This is the committee report. I made a mistake there. But they did increase what was requested by CDC. They cut NCHS by almost $4.4 million.
While saying at the same time in the report that CDC will maintain and continue to help and support states to improve the electronic death reporting, which is what the $5 million that had been given previously was targeted for.
At the same time that all of this is happening, there are a number of initiatives that have been put forth between OMB for all the federal statistical agencies, for executive departments. The new commission on evidence-based policymaking is but one. It does specifically call out all the statistical agencies that have to work together with Census and the lead really with no appropriation of funding to help explore possibilities of a clearinghouse, administrative data, with survey data. There is the OMB’s memo on sharing administrative data for statistical purposes, which relies heavily on the stat agencies to be able to help with linking data, looking at the quality of administrative records, where administrative records could supplement survey data, et cetera.
Departmental implementation of OMB Directive 1 so one of the new OMB directives that were published in December of 2014 wanted to know how are departments actually supporting their statistical agencies. This issue of your organizational placement, your independence, your ability to do things is one that is of a concern across the stat agencies. At that time, HHS did include in the departmental implementation – what they responded back to OMB they reiterated the designation of the NCHS director as the senior advisor for health statistics to the secretary. They did maintain that support in their response to OMB.
And then there is a new annual budget requirement that started a couple of years ago that wants to see how you are using administrative data in your budgets. How are you identifying high-value data sets and using them for greater purposes and the stat agencies again have an important role to play?
Everything else is – the future role for NCHS. Electronic health records, maintaining survey response rates, continuing to improve our timeliness, et cetera. All of these are things that are important for a stat agency to do. Any stat agency would be doing many of these things. They are to some extent resource dependent, but I think NCHS is doing as much as it can and is exploring ways and avenue and expertise across the federal system for ideas and guidance when you do not have a clear way forward.
Independence as a statistical agency is constant pressure and constant tension and ways to better inform how we operate within our operating division of CDC within HHS. I am going to let it go at that.
DR. SUAREZ: One or two questions.
DR. MAYS: Thanks for a great presentation. One of the things about NCHS in the past as a statistical – seeing itself as a lead in terms of being a statistical agency. It has provided innovation and training. There was actually a program where it was involved in trying to develop statistics particularly things like small area statistics, et cetera, and then the funds went away. I am wondering where the agency stands on seeing itself as still responsible for doing that because we are desperately in need of innovation and methods.
DR. QUEEN: It is now the Division of Research and Methodology. It is still doing innovative projects and working. As an office or division, that is their role, but they are also working with the various people on the survey components and across federal statistical system to look at – one of the projects they are looking at that started with a project that was on the HIS related to – it started as a possibility of a web panel and now it is exploring what is feasible for NCHS to be collecting via the web. There is no way you could do the HIS survey. They are doing on small seed funds some work in this area. It does make it much more challenging.
DR. MAYS: But it was much more public. It was funding others and bringing others in to the meetings. This is internal, but you had fellowships and what have you.
DR. QUEEN: I think Charlie mentioned. There is right now one small grant program, but some of that is funding dependent. Having the money to be able to even do something that cannot be completely externally funded, but that helps bring in external funds.
DR. O’GRADY: There is something that I would disagree with in your presentation and I think it is an important distinction between – that got kind of blurred. There is the notion of avoiding political interference. I understand that you have been under a lot of pressure. Certainly with the roll out of the ACA, there is a lot of political pressure, but that was certainly in the Bush administration as well. It is not just the Obama people. There is always that pressure. One of the dangers, as someone with the kind of training with people around this – is that it is not from some K Street lobbyist, as we sit on K Street. It is normally from well-trained PhDs from top programs who have forgotten where the firewall is and they are very good at eviscerating a statistical agency or other agencies for political means because they have all the tools to do that.
DR. QUEEN: I never meant to imply it was.
DR. O’GRADY: Let me finish. That distinction is certainly important and your vigilance is very important.
The other thing that I found more problematic is you have no entitlement here. Congress judges people, both parties, combined parties on the return on investment. They do not quite put it that way. And I can see if I am the director of CDC and I have to spend money on Zika or you, sorry, Zika is getting it. You have to be relevant to the questions they need answered, to the things they need to do. And to declare yourself we are the real statistical agency and we are the purest, that is just not the reality of the HHS I work in. You earn your seat at the table every day, every meeting. If AHRQ comes in with a better way to collect the data, answer the pressing questions that either party or the Hill has to do, they are going to get the funding. There is a reason CMS has their own survey of their own beneficiaries. The administrator of CMS wants to be able to pick up that phone and say I need a question on nursing home care and it happens. It does not involve a conference with OMB. It happens.
It sounded a little like you were feeling sorry for yourself. You are in a competitive environment with budget deficits and limited funds. You have to bring your A game. Don’t cross those political firewalls. If you are not responsive to the issues they are trying to – those hard questions they are trying to answer, they are going to go someone else.
And the fact that OMB is – those are not the people who pay the bills. It is the Congress who writes those checks that go in your appropriations. Just be a little careful to not wrap yourself in the statistical community so much that you end up just with – the actual needs of the people who do fund you.
DR. QUEEN: What you pointed out to me is that I need to work on my presentation. Actually, I have just finished reading your chapter and the vulnerability of the federal statistical system. I never meant it for it to come across that way. But I will say this. NCHS has not been able to have a seat at the table. NCHS has not been able to go through CDC to make connections to even make certain requests, budgetary or otherwise. The NCHS director as far as I can tell has never really served in the role of senior advisor to the secretary on health statistics.
DR. O’GRADY: But they can do that indirectly and that is what – if you think about how this has worked in the past. CDC, yes. They are busy with Zika. We can certainly argue about the organizational placement of your agency. Another secretary may come in and that will all change in about 48 hours. It just takes the secretary deciding. I think there is any legislation. There might be a little.
But the other thing is you have been very effective in the past and there are examples as you have done it through indirectly through other parties. Certainly through your relationship, your fairly close relationship with ASPE, you make sure that certain data gets forward. I know personally there are times when they come down to those internal budget decisions and ASPE says don’t cut them. We need the sample size to not be in the next round. The fact that you do not get the tent card at the table, that is much different from having real influence in these discussions. I think that your agency is in a position to do more in that. But if you wait for the phone call from the secretary, it is not going to come. You have to call and say we see that this is an important problem. Were you aware that we will have some data coming in? If they say give it to me early – the weights are all still screwy and stuff like that. Definitely draw the line.
But if it is something where – there is a tendency. All statistical agencies – not statistical agencies – analytic and data collection agencies within HHS to lose a little bit of who the focus is and who the client is. It is a little co-op – by academics, but that is somehow – they are your colleagues in this process, not your clients. The Congress is your client. The secretary is your client. It does not sound like – you and I are both old feds. It is sort of like that is not how we think about things. But that certainly determines whether you get $50 million next year or you get $5 million next year. How many HHS agencies have said we are going to take the money we were going to give to you and we are going to give it to NIH because we get a better outcome from NIH? We get cancer cured from NIH. We are not getting what we need from you.
All I am saying is just be careful how you approach these things and how you think about how you continue to play an important role and not be sidelined. Sideline means budget gets frozen if not cut.
DR. QUEEN: We are trying to work more actually with CDC. There have been relationships that have been newly established that we also hope to build upon and have more fruitful collaborations. There is a significant amount of that. There have been some strange instances. It is not unique to NCHS where don’t publish this table. It does not make our program look good. This is not just an NCHS issue. It is a broader statistical issue. I think the Interagency Council on Statistical Policy is trying to address. Your point is very well taken. I need to brush up on my presentation.
DR. O’GRADY: The point when you get a call from the secretary’s office and they say the secretary has to make a decision by five. That is different from we are Democrats. We want to beat up Republicans. We are Republicans. We want to beat up Democrats. It is a distinction that needs to be kept in line. It may not be perfect.
When I used to work on legislation, our friends at the National Academy of Sciences came out with great recommendations for preventive benefits for Medicare beneficiaries two weeks after the bill was signed. We would have taken a draft report. There is the pressure of that as well.
MS. LOVE: My comment is really a rhetorical question, but I am reflecting on the conversations at the state level too. I think we are heading into an era where we have to ask a question of who pays for the data and are the data a public good and if so then who pays. Nobody wants to pay. We are – the same thing with the states. I do not have an answer for that, but I think the conversation should be geared to – we tried to at the state and we will hear more on Friday on claims that they serve the public good and therefore it is a public utility and should be funded as such. But that does not sell to lawmakers. I do not have an answer. That is coming up at the state level too. Nobody wants to pay for it. Everybody wants the data. Private data banks are great. But when you are making public policy, you probably need the public and transparency –
DR. COHEN: Maybe Google can buy NCHS and solve the world’s problems.
DR. QUEEN: I definitely need to make sure — I need to work on my style. There is no two ways about it.
DR. COHEN: I want to pick up on something that Mike was moving towards, but from a slightly different perspective. I too think that NCHS is organizationally misplaced. Within the bowels of CDC, it does not have the visibility that it needs to have broader influence. I certainly agree. That would be wonderful if there was some consideration given to thinking about how a statistical agency of the quality and purview that NCHS has could be elevated organizationally to achieve the kind of prominence it needs to have some of the influence that Mike was describing.
I was wondering if there are opportunities to achieve that through the – you talked about other countries having centralized statistical enterprises, the notion of the clearinghouse for federal data, moving forward, the commission, the interagency council. There are a variety of emerging trends with respect to statistical data throughout the federal government that might provide NCHS with opportunities to be more involved at these higher level discussions around data access, data release, and data policy.
DR. QUEEN: I think NCHS will be particularly with commission because the statistical agencies have been specifically identified in legislation. I actually did not mean to suggest that we should not be in CDC because that is really not my call.
DR. COHEN: That is my comment. I am not quoting you. That is my firm belief.
DR. QUEEN: And because we have forged some really positive relationships. The clearing house also – I think you probably will be hearing about this more as a committee. It is not just statistical. It is all these other administrative and all kinds of other data how it can all be used effectively for evidence building. It is beyond just the statistical data. The visibility is an issue.
But I think one thing that I am working on right now is identifying for colleagues at CDC where they may be using our data to report on their own progress and meeting certain agency goals for CDC and for the department because HHS has agency priority goals. It turns out that data from NCHS are being used. We may not always even recognize it. How I can help identify that to CDC colleagues will help then say we need to maintain these because we are using them to report on our own progress for our program. There are a lot of interconnected ways to work through this.
DR. O’GRADY: Can I make a friendly comment on Bruce’s comment? At the end of the Clinton administration after a stumble on their attempt at health care reform, there were plans that when the Republicans came in they came across and for the most part agreed with of thinking about those sorts of reorganizations. The Clinton people were mostly responding to the idea of which – that they had four different surveys with four different counts of the uninsured. How were they going to make any kind of smart policy when the numbers were anywhere from 19 million to 45 million. They are used to uncertainty, but that is a little much.
But their idea was taking it in the other direction. Tommy Thompson at least in the Bush administration concurred. That was to bring NCHS and AHRQ in under ASPE to make them much more directly feeding into answering these policy questions, thinking about new policy, how it should develop, what data you need. It was not the European model. It was actually in the other direction, which you could tell from the comments, there is an unease – where is it policy analysis and where does it become political analysis and where is that line and are reticent to be viewed as taking a political point of view. It was actually attempted in the Bush administration. It was blocked mostly by AHRQ in terms of going to the Hill and then to the secretary and getting it squashed.
But that was the last real, viable I had ever – that was designed by the Democrats on their way out as a response to the stumble over health care reform and the lack of good data to make those policies. And the Republicans looked at it and went – it makes sense to me, but they were unable to pull it off. But it was that same direction of policymakers saying I need this data. I need it in the form I can use. It is not an academic exercise. It is a day-to-day decision making exercise.
DR. SUAREZ: Thank you so much. I appreciate it. We are going to go next to Marcie. Marcie is going to talk about the National Health Interview Survey.
MS. CYNAMON: Thanks for having me here. I want to talk to you about a couple of things, having to do with the Health Interview Survey. Susan talked a little bit about data dissemination. I want to tell you a little more. Apparently, you heard some other things this morning. I will not repeat that. But I want to talk to you also about one of the most encompassing activities that we have engaged in for a very long time and that is the redesign of the Health Interview Survey. I want to tell you about what is coming up and what the impact of these changes will be on agencies and estimates and how we are reaching out to our stakeholders for input.
With regard to dissemination, we have, as you heard, stepped things up a bit. We have a really robust early release program that faces – for anyone who does not know what the Health Interview Survey is, it has been around since 1957. The purpose of it is to monitor the health as a nation and we do that through the collection analysis and dissemination of data on the broadest possible array of health topics that can be obtained from individuals.
Our sample is address based and we have primary sampling units in every state. We are able to make state estimates on quite a number of factors. We do in-person interviewing, as Susan mentioned. Our data collection agency is the Census Bureau and has been since 1957. We are almost completely in person. The follow ups are allowed by telephone after initial contacts. Our target sample size is 35,000 households. In the past few years, we have been able to augment that up into the 42,000 to 44,000 range, which has enabled us to do more estimates at the state level.
Data dissemination. We release quarterly data files on our preliminary data and annual data files. These quarterly files are used to produce early release estimates of 15 key health indicators and also report on health insurance. The health insurance reports were really critical to monitoring the impact of the Affordable Care Act on uninsured.
The HIS figures are used as benchmarks for some of the other quicker turnaround surveys even though ours is fast. Some of the polling and other private survey research companies were able to do faster surveys, but they looked to us to benchmark their data against.
New in 2015, we were able to step up our production schedule by four weeks, enabling us to really compress the time of the release of data. We are very proud of that.
We disseminate data in a lot of different ways. We have webinars. We have classes. We do presentations at professional meetings and we publish NCHS and CDC publications as well as through peer reviewed journals. We have a staff of 44 professional researchers who all engage in publications and presentations.
Most important though because I know that you use our data, a lot of you use our data and know about how public it is. I want to talk about some of the really big changes that are happening. The last time the questionnaire was redesigned was 1997. As you can imagine, the world of health systems have changed somewhat. We are finding that some of our questions could use a little sprucing up, a little tightening up and a little modernization. We looked at what goals we should accomplish for our 2018 implementation of a redesign survey. We thought about how we could improve the relevance of the covered topics and how we could better meet the needs of the department and other stakeholders.
We wanted to focus on the leading causes of morbidity and mortality and their protective factors and the risk behaviors. We wanted to look more at access utilization and barriers to care. Given that more people are insured, now what are the barriers to care?
Many of you may be aware that the department is especially led by ASPE, working on alignment of shared content or overlapping content on federal surveys. And NCHS is definitely at the table there. We have been working with our sister agencies to come up with batteries of question that can be used on all health surveys. As you mentioned the problems with health insurance estimates varying so much across surveys. And identifying which survey is the primary. It should be used a primary source for which surveys these various topics should be covariates.
One of the biggest problems that we have had and it is not HIS or NCHS surveys, but all surveys have been seeing a decline in response rates over time. For us, we know that we need to shorten the questionnaire. My next three slides will demonstrate that. But we also need to eliminate and reduce some of the redundancy and focus more on what is HIS the best survey for and what is the most appropriate for us to do and what is more appropriate for other surveys. We do not want to duplicate what other surveys do better.
Talking about the length of the survey. It has gone up and up and up. We have added questions and not taken out enough questions each year to get back to a length that is workable for our public and for our interviewers. Many of you know that HIS is only partially funded by NCHS and we look to other agencies for additional funding. About 40 percent of our funding comes from other agencies and we add questions for those agencies to monitor their programmatic needs for data. You can see here there are some peaks and valleys. Those peaks are years that we have done larger supplements for other agencies.
There are three sections in the way the questionnaire is currently configured. A family section. One sample adult that is randomly selected from within the family and one sample child also randomly selected from within the family. All of these components had declining response rates. Those places where we have the valleys are those years where we have longer supplements.
Here are breakoff rates. The higher peaks are again those years that we have the longer supplements. Clearly, this 100-minute interview is causing some problems. Interviewers cannot complete more than one a night. We have a larger workforce. It costs us more to implement the survey. It is harder to get people to complete them. We hear a lot of feedback from the field about the exhaustion of respondents and how the quality of the data toward the end interview suffers. This is especially true for most vulnerable respondents, the elderly and people with more health issues. We know that we need to take a step back and think about how to cut back to what is the most essential part of the HIS while still selling time to other agencies to keep ourselves afloat.
These are the areas that we have identified as key content areas for the HIS. There is not anything too surprising here. Functioning and disability, health status and conditions, health insurance, health care access and utilization, risk behaviors, demographics, and a little bit of what we have not done so much of lately is social and economic determinants of health and health care access.
As we think about how much we can do on each of these, everything is a balancing act and everything is if you put something in, what comes out. There have been a lot of deliberations about how much of anything can get in the survey.
The format that we are proposing is a little different from what we have been doing. This is what we call the quilt or the crop rotation. The top section will be what has been called the core in the past. It will be again core questions that appear every year and go to everyone. That will stay the same for the life of the redesign, which we hope will be about ten years. It will have the information, the questions that will help us do our key indicators and demographics.
The next tier of questions will also be considered core, but they will be on topics that will rotate in and out on a fixed periodicity. For example, we know that it does not make sense for us to do a general health survey without asking about tobacco. We might have a very short battery of questions about tobacco use on the annual core. But every few years we might want to collect additional information on a broader array of issues related to non-tobacco products or something else related to tobacco use such as quit attempts. Not stepping on the toes of the tobacco use surveys, but something that would make the entirety of our questionnaire more valuable to users as they look across the covariates. These types of topics would rotate in and out on some fixed periodicity probably every two or four years. And those we call the rotating core.
And then finally, we will have our sponsored content. That can be as short as one question or it can be as much as five minutes. In the past, we allowed up to 20 minutes for supplement time and we will not be doing that anymore. We only had a few agencies that conducted such long supplements: the Cancer Institute, the National Center for Complementary and Integrative Medicine and National Center for Deafness and other communication disorders.
Those agencies – we are telling that we cannot do these 20-minute supplements because of the impact on the response rates and the quality of the data. We want them to think about how they can rotate content in and out, how they can break up the amount of content that they have. For example, the Cancer Institute is now looking at doing five minutes every year and what topics go together. I am looking across our ten-year plan. It can make sense to them when they might ask questions about – to be obvious, lung cancer and when we are asking more questions about tobacco use, for example.
They have really gotten on board with this. I was a little concerned because they are one of our primary supporters. They have people who work on cancer at CDC and those at NIH have teamed up to come up with a ten-year plan.
The National Center for Complementary and Integrative Health that had hundreds of questions on a few years ago has boiled down near critical elements to 40 questions. It can be done with real deep thinking about what is necessary and what is appropriate for the HIS.
One of the most important things that we are trying to get away from is using the HIS to study rare events. Mike, you mentioned thinking about who our stakeholders are, who our users are and who our funders are. We oftentimes get interest from those who would like us to add a question about an event or a condition that affects .01 percent of the population. We are trying not to do that in the future because once you add that question and then detail about that issue, you end up with few interviews that take two to three hours. We understand that there are very few resources or sources out there for this type of study. It does not appear that HIS is really the best source for that type of detailed information on very rare events. This is causing some concern for some of our data users who are not sure what source they will be using.
The structure. This is causing the most controversy. We are proposing to eliminate the family questionnaire on the Health Interview Survey. The family questionnaire is where a lot of demographics are collected and a lot of health information about all members of the family. We are proposing to have a sample adult and sample child questionnaire with very little information collected at the very beginning of the interview on the family. A couple of reasons for why we are proposing this change. Quite a few reasons really. One is that most of the publications that we reviewed in the literature and we did quite an extensive literature review used the sample child and sample adult, mostly the sample adult questions.
The family section is answered by a proxy person now who reports on all members of the household and that person may not be the best person to answer those questions. They are the person whose home when the interviewer shows up at the house. We do that rather than trying to find the best person in order to maximize the completion rates.
The proxy information is probably not as accurate as self-report, which will be the focus of our sample adult. And for sample child, it will be most knowledgeable person for that child. It is almost always the mother in about 75 percent of the cases. Almost 20 percent is fathers and there are some grandparents and non-parental guardians who complete the remaining percentage.
The concern in our user community is that we will lose so much of the family context of the survey that is unique to HIS. Listening to the concerns of our stakeholders and users, we have thought about what variables we can add back into the sample adult and sample child questionnaires that will provide not everything that is being lost, but a considerable amount with regard to demographics and social and economic determinants of health. It would have looked like this. It would have said sample adult, sample child.
The impact of the changes. We think with shorter interviews – and also cutting back of the family component. It took interviewers ten minutes before they got into the health component of the survey. By cutting out a lot of that upfront who is related to whom and marital status and all kinds of things like that that we will get some later. We can get into the health questions faster and make the interview more salient and capture the interests of the respondent faster. We believe that our data quality will be improved.
Another benefit of the redesign will be the updated focus. There were a lot of places in the questionnaire where there was duplication of questions of topics at least if not the exact question because things had gotten grandfathered in and we were afraid to remove context and change the order of things. We had done a little bit of that over the course of this past redesign and we saw estimates bounce around a little and we would take questions in and out from the context of various questions. We got very skittish about removing things and that is one of the additional reasons in addition to selling time on the survey why the interview got so long.
We will increase our person-level information in lieu of the family – all members of the family information so we will have more detail on individuals. Fewer rare event questions and rare conditions.
By laying out our ten-year plan, we think that we can address the issues of covariates and funding and knowing what is coming up so that appropriate agencies can think about their budget process with more advanced notice.
We are not making these changes in a vacuum. We have, as I mentioned, done extensive literature review. We have looked carefully at our reporting requirements, what the department and the what some agencies do with our data, how the data are used, how it is used for policy and program evaluation.
In addition to all those things, we have done an extensive and continue to do an extensive outreach to our user community. We have attended many meetings, professional association meetings, agency meetings, internal meetings, meetings of just groups who use our data extensively. We have released to the public three invitations to comment as we know more about how the questionnaire content is evolving. We engaged our LISTERV. The LISTERV has generated hundreds and hundreds of emails in response to what we have put out. Our first one was just a general plan. We are redesigning. Tell us how you use HIS. Tell us how often you use it, what you use it for, the impact of it on your program or your reporting system or whatever it is. We got a very good overview of how the data are used. That was in the fall of 2015.
And then in February of this year, we discussed the issue of the possibility of eliminating the family component. We raised the hackles of a lot of data users who responded to us and we have been responding back to them. We have rethought some of the decisions that we had proposed, our proposed decisions, not final decisions. Nothing is absolutely final.
And then just last week we released through the LISTSERV yet another request for comments. That one has a more detailed list of content and for the first time the children’s content. We had been focusing mostly on the adult content.
Next will be a release through the Federal Register. That is in late summer to fall of this year. Then there will be two more Federal Register notices, each one being more and more complete.
I would like to say also that we have engaged in what we call technical expert panels. These are not formal arrangements, but we have had panels on income, on child health, on program participation.
We have listened to our users. We have listened to the academic community. We have tried to incorporate the widest possible input into the modifications that we are making and thinking broadly about how to address the future needs for data. I invite members of the committee to submit recommendations to us as well. Thank you.
DR. SUAREZ: Thank you. Thanks so much. We have three questions. We are going to start with Vickie then Bruce and then Raj. Bruce is first.
DR. COHEN: I have lots of questions, but I will stick to two. I agree with the decision to eliminate the family component as difficult as it may be. I think that evaluation seems appropriate. I am concerned about some of the rare event issues. Unless NCHS or NHIS finds a home for some of these questions through the Census, ACS or otherwise, these are data that will not be collected. You said, however, you are going to be expanding some of the personal information. Are you planning to include gender identity in HIS? An increasing demand for gender identity, good population-based gender identity information I think is a real concern for many folks. I understand the prevalence is – I do not know how you define rare prevalence. Prevalence is around 2 percent.
MS. CYNAMON: That is how we define it.
DR. COHEN: Two percent. That is probably the threshold for —
MS. CYNAMON: Once you start trying to analyze that, you cannot do much with it.
DR. COHEN: The solution there is to collect over multiple years to build up the sample size to get more stable estimates. That seems to make more sense than discarding rare events. That is a longer discussion.
MS. CYNAMON: I would like to also say that in 2016 we implemented the new address-based sample. It is a lot harder to combine years. I met with my sampling statistician yesterday. We were trying to work out strategies for how to do that because the geography can change from one year to the next, making it more difficult.
DR. COHEN: That would be a severe technical limitation on the value of HIS because one of its strengths has been the ability to go across years and aggregate quarters for analysis. It is a longer conversation, but I would like to hear some more thoughts about that.
The other has to do with the mode. Have there been any considerations of moving away from a totally CAPI-focused survey to use multimode as a way to do follow ups or reduce the immediate burden of initial interview?
MS. CYNAMON: We have done some experimentation and we have been thinking about this. I still think that in-person interviewing has the highest quality data. We have done some testing of following up our respondents via the web. We had really abysmal response rates and ended up having to implement phone interviews to get the response rate up to 50 percent.
The interview will be shorter, but it will still be long. It is still an hour. That is a long interview to ask anyone to self-administer. We have looked at mini versions of the HIS to look at what could we do in a short mailed out survey or web survey and then add it into our in-person or combine it with our in-person data. But everyone has a different idea of what should be on that mini component. We are still exploring it. I do not think it will be a part of the first couple of years of the redesign, but it could be further into the decade.
DR. COHEN: The answer is gender identity will not be part of the redesign.
MS. CYNAMON: Right.
DR. COHEN: Do you know of any national data source that is planning on collecting gender?
PARTICIPANT: NSFG does.
DR. COHEN: States implement it differently. But the sample size for NSFG is much smaller and subject to more variance.
PARTICIPANT: (off mic)
DR. COHEN: BRFSS is the only state available data that I know of that states implemented in a variety of different ways so there is no consistent data.
PARTICIPANT: ASPE funded a special sexual orientation gender identity module, a special supplement. We are just now starting to get some of the data. If this is of interest to you, hopefully at an upcoming meeting then we can talk a little bit about that.
DR. COHEN: That would be great. It would be great to see that effort carried on. I am really sorry that NCHS is not picking up the mantle to do gender identity questions. It seems like a logical place for it.
MS. CYNAMON: With the maximum adult sample size of 35,000, we really cut back. That is the one of the losses going from the family to the sample adult focus.
DR. MAYS: I want to build on some of the things that were just asked, but from a different perspective. Susan Queen is going to know that I have brought this up before. One of the things that concerns me is when one survey gets rid of something, we do not have this plan to look across and consider the NCHS surveys as like a family of surveys. What happens is that if you decide to let go of something, there is no consideration from what I can tell and another survey picking it up. What happens is that as a federal surveillance group, we are missing something. I am just going to put that out there for the committee to think about whether or not that is something we need to be recommending or pursuing.
MS. CYNAMON: I would like just comment. If it was a topic that was that critical, we would have kept it.
DR. MAYS: I will give you an example. That is the way in which mental health happens.
MS. CYNAMON: We are expanding mental health content.
DR. MAYS: But I am just saying that in the past, what has happened is it could get dropped and then it would get cut back and then it isn’t until somebody comes back and fusses about it that it is back. Yes, I know now in terms of well-being, but I am saying I think as a process that has not been the case. You are speaking for just NIHS. I am talking about NIHS relative to the whole family of surveys and things that you are dropping. I agree very strongly with Bruce about the issue of gender identity disorder. It is like I do not quite understand even though the sample is small. What you do is you build the sample up. There are just some approaches that I think and I know you have probably heard it. Let me just get to my other question.
Is there going to be a significant amount of time and resources spent in terms of crosswalks so that as people want to do continue from the past, going to the future? Will we actually have the kind of either formulas or whatever we need to make sure that we can continue to do that?
MS. CYNAMON: We have absolutely no budget to simultaneous run the old and new questionnaires. It will not happen that way. You can look at – that kind of crosswalk, no. That was one of the initial requests that we run both simultaneously to be able to compare estimates from both. Perhaps the most responsive approach we cannot do, but other approaches we can do and that is looking at the previous quarter and delving into the data and seeing how it changed and if it changed. We think that there will be breaks in trends because we know that when you change the context at all, you see fluctuations in the data.
DR. MAYS: I guess what I am asking is if I wanted to do a paper and it is from 2000 to 2020, let’s just say we are in 2020, and I have the old data I have to use and then I have the new data I have to use. You have changed the question a bit.
MS. CYNAMON: You would have one of those little things in the middle of your table. That is what happened in 1997 too.
DR. MAYS: Let me ask my final one and that is is there going to be some kind of monitoring or some kind of piloting to know. There are things that you are saying that you think will be better. I might question some. And the question is whether or not when you make these changes, they have to stay for ten years or whether you have the capacity to pilot and to either return back or make adjustments.
MS. CYNAMON: Well, I am not sure what you are referring to because when the questions are good, we are keeping those questions. Really the biggest change is going from this family structure straight to the sample adult/sample child and that is not an easy thing to go back to.
DR. MAYS: I guess this is not just about the questions, but as a researcher what happens is that if I do not have enough context and cannot answer a question as well, if the sample is smaller and I end up with less power that my ability to contribute in the literature is actually – it is like advancing science is actually what makes me concerned.
MS. CYNAMON: That might be true if you had used the family section and used some of the estimates from there, but most people did not use that section. Most people used the sample adult. It might affect these rare events. If anyone in the family had eczema or whatever then yes. Your power will be a lot lower.
DR. MAYS: By race ethnicity, it is going to be a lot lower.
DR. SUAREZ: We have three more questions, four more speakers and 45 minutes to go. We are going to have to move very quickly.
DR. CHANDERRAJ: I have novel suggestion for data collection. I do not know if this is going to fly. Several organizations like health care, automobile industry, auto insurance, home, HUD, and everything. Can we make these organizations collect data, incentivize them rather than pay the staffers to interview the people going to the homes?
MS. CYNAMON: I am sorry. I do not understand.
DR. CHANDERRAJ: Collecting the data.
MS. CYNAMON: Not from individuals.
DR. CHANDERRAJ: The individuals come to the health care providers. They come to auto insurance. They come to the home insurance. There is a need for them to get the services. If each health care provider in the country collects on sample a day then the amount of people that are participating in the survey will tremendously increase.
MS. CYNAMON: I have a hard time controlling the people who work for me. I cannot imagine being able to control those people who do not work for me like health insurance companies and auto insurance companies. I do not know how you would coordinate and standardize.
One of the things that NCHS is known for is the quality and stability of its estimates. I know that when we talk about big data and we talk about alternative modes and putting all these little pieces of the puzzle together to come up with a bigger picture that we have data quality that we are not sure of. We know a lot about the quality of the data that we are producing now. Perhaps we are spending too much time and money to produce quality data and good enough would be good enough. For now, we are trying to do it the best way we can. I think that this is probably one of the best ways.
DR. O’GRADY: I have not really a question, but thinking about your rare event problem. More and more I do clinical trial work. There is I think methodologically disturbing kind of progression where they move from clinical trial. It is 300 people whose main reason for being part of this group is that they are willing to participate in the trial. But that is not a representative sample of anything other than willingness to participate and some clinical characteristics.
Then they have tended to move to registries, which again have very limited statistical properties. It is just again people who you have the records of. Then they try to say – maybe if you are talking about the Medicare records and you have 80 percent of the claims, you are close to a census. You do not have to worry about sample weights.
I guess what I am wondering is as we move more and more into electronic medical records and what not, whether not so much the collecting of the data would come from these other administrative sources, but the identification of this rare event of 3000 people who have cystic fibrosis or diabetic macular edema, some of these things that are important to your NIH colleagues and what not, but are never going to really hit that threshold to show up on either your survey or NHANES or any of these others.
As a building of that initial pull, it has a mixed method. You have very strong methodologies. That notion of thinking about could you do something that would at least approach a representative sample of one of these rare event populations. Then you could think about how you want to reach out to them, whether you want to do that phone or person or whatever. But you would use these other data sources and take them with the grain of salt they need to be taken with. Physicians are not great about coding various diagnoses at times, but use that to get you to what you think is a reasonable estimate of this rare event population and then think about what the appropriate data collection.
MS. CYNAMON: I guess your last phrase is key. Is that what the HIS should be? When we go back to thinking about what are the goals of the National Health Interview Survey, is that what we should be doing? To me, that sounds more NIH than NCHS.
DR. O’GRADY: I guess I was trying to think of – when you have – NIDDK comes in and says we really want to know about – I will stick with this diabetic. It is a sub-disease of diabetic eye disease, but it is important to them. You say I cannot put it on the survey. That just does not work. You have 20,000 people nationwide. But I have this alternative methodology for collecting the data and maybe NCHS does that method, but it is sort of a micro-survey that you would do of special populations and offer that to these funders as an alternative to this. What their traditional this is let me buy a question on your survey so I can answer the question I need to answer. It does not work for us and we do not really think it works for me. But we have an alternative methodology that is better than this. Ten clinical centers around the country have sent their data in. It is a 50,000-person registry. It is better than that. It is a real sample with real weights, but it is not – but you cannot get on the —
MS. CYNAMON: If your goal is not pure population prevalence, if you really want to know just more about the people who have this condition, yes, that would be a great way to do it. That has not been what HIS has been and I guess the question again is is that what —
DR. O’GRADY: I am thinking broader. Is that something when you say you need to talk to my colleagues in the – target population survey division who have a tool that is appropriate for your needs. We are not really that tool. And then you get to stay with your core mission. Still the other agencies are interested in sending funds to NCHS because you are answering the questions they need answered. But you guys are the only guys I know that could figure out that. How do I move from registry stuff and record stuff to have a real rigorous sample that somebody could actually hang their hat on and –
MS. CYNAMON: Didn’t Apple just do that?
DR. O’GRADY: I would not trust Apple. I would trust you guys first. Apple might say they do that.
DR. SUAREZ: We are going to move to Bob.
DR. PHILLIPS: My question has to do with MEPS. Is there a crosswalk with folks at AHRQ to understand how this is going to affect MEPS generally? And specifically, does the household component of MEPS stay?
MS. CYNAMON: Yes. The household component of MEPS will stay. Yes, we will still be their sampling frame. We have be in contact with them along the way so they know what is going on and have been thinking about – it really does not affect how to use our sampling frame. The linked files might look a little different, but otherwise no.
DR. SUAREZ: Thank you so much, Marcie, for the overview. We are going to go to Denys on health care surveys. We will have to adjust a little bit of the time at the back end. I really want to give as much time as possible for this.
DR. LAU: Good afternoon. I am neither Susan Queen nor Clarice Brown just for the record although I wear many hats. Good afternoon. My presentation may tackle some of the issues of rare conditions with the health care surveys. Hopefully, this may give rise to some additional conversations about that.
Good afternoon. I am Denys Lau, the deputy director of the Division of Health Care Statistics at the National Center for Health Statistics. I am standing in today for Clarice Brown, the division director, who has prepared these slides. But unfortunately, due to medical reasons, she sends her regrets for not being able to present at this meeting today.
I am honored to be here to describe how the National Health Care Surveys are leveraging the CMS Electronic Health Record Incentive Program, also known as meaningful use to move from manual medical record abstraction to electronic data collection using electronic health records.
I will begin with a brief overview of the National Health Care Surveys. The National Center for Health Statistics is the federal statistics agency responsible for monitoring the nation’s health, as we all know. The mission of the Division of Health Care Statistics is to produce accurate, objective statistics on health care to inform health care policy and serve a variety of research needs. The division carries out its mission by collecting, analyzing, and disseminating data on the use, access, and quality of health care provided in the United States and the health care organizations and professionals who deliver that care.
To carry out its mission, the Division of Health Care Statistics administers the National Health Care Surveys, which is a family of surveys of health care providers in in-patient, out-patient, ambulatory, and long-term care settings.
NCHS gathers data on care provided across the spectrum of health care delivery systems from ambulatory and hospital care settings to long-term care providers. The National Ambulatory Medical Care Survey, NAMCS, collects data from physician offices and community centers. The National Hospital Ambulatory Medical Care Survey, NHAMCS, collects data from hospital out-patient departments, emergency departments, and ambulatory surgery centers. The National Hospital Care Survey collects data from hospital in-patient departments, emergency departments, and out-patient departments.
NCHS places the National Hospital Discharge Survey, which for 45 years collected data in-patient discharges from hospitals. The new National Hospital Care Survey is the first of the health care surveys to move from electronic data collection beginning in 2010 with claims data and now moving forward to electronic health record data collection.
The National Study of Long-Term Care Providers collects survey data from residential care facilities and adult day care centers and uses administrative data for care provided in nursing homes, home health, and hospice agencies. This presentation will not cover our long-term surveys, but I have mentioned them here for completion.
The surveys share certain design features. Each is nationally representative, samples health care providers, and collects data from each sampled provider. Some surveys such as NAMCS, NHAMCS, and National Hospital Care Surveys collect data about encounters with health care providers, for example, physician visits in NAMCS and community health centers and patients’ discharges from hospital settings.
Some surveys such as those in the National Study of Long-Term Care Providers collect residents and participant data in aggregate form at the facility level where care is delivered.
These are some of the data elements we collect. Patient demographics, insurance status, zip code, medical conditions and in some of our surveys, patient identifying information such as name, address, social security number so that we can link the survey data to other data sources such as CMS data or the National Death Index.
Data on the encounter with the sampled health care provider include the reason for visit, diagnoses, procedures, and medications ordered and provided, diagnostic tests and types of providers seen during that visit. These data have traditionally been abstracted from medical records, originally, manually abstracting data from patient medical records using paper form.
In 2012, we used from abstracting medical record to the use of a computerized tool. Today, we are now talking about cutting out the middle step and moving to all electronic data submission.
Electronic health record is a longitudinal computerized record of patient health information generated by one or more encounters in any care delivery setting. EHR is designed to automate and streamline clinical workflow. EHR systems can pull data from many sources. In health care, data can be entered at the patient’s bedside, pulled from the pharmacy or come from labs and radiology reports. There are many commercial EHR vendors including Epic, Cerner, and Allscripts, which are just a few of the large hospital vendors.
EHR systems can contain various types of data such as patient demographics, progress reports, problems, medications, vital signs, past medical history, laboratory data and radiology information.
EHR data present a tremendous potential for the National Health Care Surveys. Currently, medical record abstraction for a sample of patient visits for NAMCS is conducted by the census field representative. They work for several works in the physician’s office to abstract just a sample of their visit. We believe that this burden on provider can be lessened with electronic submission of data with EHR.
Currently, abstracted data are entered into a laptop computer. While the laptop computer is kept secured at all times by field representatives, we think that electronic submission would be more secure. Compared to manual data collection, we believe that collection and processing of HER data would be completed more timely once the processing systems are developed.
EHR data would definitely be more clinically detailed because we will not be limited to a maximum number of diagnoses or medications that we do in our current surveys. There is the potential for more data in terms of volume. For NAMCS, we are moving from collecting data on a sample of visits to a physician office in a week to collecting data to all the visits to a physician during a reporting week. For the National Hospital Care Survey, we are collecting data on all in-patient ED visits and out-patient visits during the entire year.
Why now? We believe that this is a tipping point. The CMS electronic health records incentive program have greatly accelerated the adoptions of EHR by hospitals as well as physicians.
What would it take for us to move to electronic data collection? First is research. Just as survey methodologists took painstaking care in the design and development of manual data collection operations, the move to electronic health record data collections needs to take with great amount of thoughtfulness and care. To that end, we conducted research to ensure that data we abstracted from EHRs will be comparable to the data that we have been abstracting from medical records for over 20 years.
Next are data standards. Not all EHR systems and the data they contain are the same. We need to provide a standardized form for submission of National Health Care Surveys data to NCHS to enable automated extraction from the EHR systems. We also enlisted the help of our NCHS classification and public health data standards group to help develop data standards and interoperability standards for the National Health Care Surveys.
We also need incentives. The response rates for our surveys have been declining over the years. Moving from manual to electronic data collection will be a big ask of providers during a time when they are asked to do so much already. Therefore, we need to incentive providers to participate in our surveys.
With funding from the Assistant Secretary for Planning and Evaluation, we conducted two pilot studies at a large medical record to compare different ways of collecting NAMCS data. Traditional, medical record abstraction versus data abstraction from the EHR system and traditional record abstraction versus continuity of care document coming directly from EHR.
Overall, we found that abstracted and CCD were highly comparable to traditionally abstracted data for many surveys we collect in NAMCS. We also found that CCD data include rich progress notes that NAMCS can benefit from having. These findings give us confidence that we should expect that extracted CCD data will be comparable to much of the data we have traditionally abstracted in NAMCS.
Regarding data standards, the Division of Health Care Statistics worked closely with NCHS classification and public health data standards group, especially Michelle Williamson, who is sitting here with us, to develop the HL7 Implementation Guide for the National Health Care Surveys. This will provide a standardized format for implementers of the IG to submit data to fulfill requirements of National Health Care Surveys. This guide will support extraction of data from the providers, electronic health record systems or data repositories.
In January 2015, the HL7 Implementation Guide was published as a draft standard for trial use and then in March it was described in the 2015 Interoperability Standards Advisory published by ONC for health IT as the best available standard for clinical content and structure.
Regarding incentives, the CMS EHR Incentive Program also known as meaningful use program provides incentives to providers and hospital to achieve a set of objectives on the adoption and use of certified EHR technology to qualify for receiving incentive payments. The National Health Care Surveys are now officially named as an option available to meet the specialized registry public health reporting objectives per the final report on modifications to meaningful use in 2015 through 2017. This means that eligible providers, eligible hospitals and critical care hospitals can now choose to submit data to the National Health Care Surveys as one of the measures to meet the public health objective requirements under incentive programs.
Required by CMS and ONC immediately after the final rule was released in October 2015, we put up on our website our declaration of readiness for public health reporting. Ever since, we have been very busy working to answer phones to register eligible hospitals and providers. By registering, the provider agrees that if invited, they will submit data to NCHS electronically. For hospitals to participate in the National Hospital Care Survey, they will send all in-patient discharges as well as out-patient and emergency department visits. For physicians, they agree to submit all patient encounters for one week.
The ONC 2015 edition final rules specifies that should eligible providers and eligible hospitals choose to report their to the National Health Care Surveys between 2015 and 2017 as part of the fulfillment of the meaningful use public health objective, they are not required to use a specific standard or format other than the options specified by NCHS to do so. For Stage 3, however, they are required to submit National Health Care Surveys data from certified EHR technologies, guided by the HL7 IG.
The very busy part was referenced to this. Since October 2015 when the declaration of readiness for public health reporting was put up, we registered close to 128,000 eligible providers of which 535 were sampled providers into 2016 NAMCS. That is about 18 percent of the 3000 physicians we have sampled for NAMCS.
We are particularly excited that over 21,000 Kaiser physicians have registered with us and 85 of them are in the 2016 NAMCS. We have Dr. Walter Suarez to thank for that success.
We have registered 811 eligible hospitals and critical access hospital of which 43 were sampled hospitals in the 2016 NHAMCS and 93 were sampled hospitals in the 2016 National Hospital Care Survey. That is about 8 percent of the NAMCS hospitals and about 18 percent of the National Health Care Survey hospitals.
Because we are collecting clinical data from electronic health record directly without the need for medical record abstraction, we will have more clinical richness and depth. We will have the medications prescribed and lab and imaging tests ordered along with their results, information that we had not had before.
The sky is the limit as far as the volume of data. For example, in the National Hospital Care Survey, we now have more traumatic brain injury cases, rare conditions that we did not have sufficient cases to report reliable estimates previously.
For surveys that we do collect on the PII, we are able to link our data to other data sources such as the National Death Index to give 30, 60, 90-day mortality, which we could not do before either.
The amount of data we are expecting will increase dramatically. For instance, in 2014, the numbers of in-patient discharged records we receive have gone up ten times from over 120,000 records sampled from 200 hospitals to over 1.6 million electronic records coming from just 94 hospitals. The number of ED electronic data increased 190 times more than sampled records from close to 24,000 to 4.5 million records. And the number out-patient department electronic data increased about 720 times greater than sampled records from over 26,000 to 19 million records.
What are the challenges? What kind of challenges can we possibly face with such a wealth of data that we get? I am going to sum it up with three only. We are now faced with first file sizes. Where and how do we store this large volume of data?
Second is processing. How do we prioritize data cleaning, given that data are coming directly from the clinical records?
Third, public use files. How do we release such an amount of data to the public without compromising confidentiality of the respondents?
Moving forward, NCHS is working with electronic health record vendors to test and improve the HL7 IG. EHR vendors are building an interface for the certified EHR systems based on the HL7 IG. We have moved to testing and validation for the National Hospital Care Survey this week. We are planning to outreach to the sampled hospitals and encourage them to register.
For NAMCS, we are planning to start testing validation in a week or so and we are going to work with physicians to encourage them to register.
We are also developing onboarding systems, web portals for hospitals and providers to register with us. We are also going to continue with hospitals who submit their EHR data. We are developing processing instructions to integrate data from different sources such as claims and electronic health records. For NAMCS, we will continue to register physicians and check that the sample for NAMCS 2016 to see if they are eligible to report to NAMCS. We will obtain EHR data from NAMCS sampled providers and prepare the integration of harmonization of abstracted data and EHR data.
I think before I end, Clarice and I must acknowledge the many individuals at NCHS, ASPE, CDC, and ONC who helped navigate the waters of meaningful use, EHRs, and standards developments. This does not include the countless individuals including the meaningful use coordinators and others in hospital care organizations and individual provider offices who have worked with us to make this possible. Thank you.
MS. HINES: Thank you, Denys. Questions from the committee?
DR. STEAD: I appreciate the ambitiousness of the effort and the clarity of the update. I can appreciate that you will be able to get much better coverage of the parts of these records that are relatively simple. That probably is at least the admission counts and the encounter counts. To a degree, it may be the diagnoses. Beyond that, my guess is that you are going to need a level of cleansing that is difficult to imagine. In our work with these kinds of records in other settings, we in essence have to go from the raw data to clinical concepts where multiple different sources can feed into a clinical concept and it is the clinical concept that could be counted the way you counted your historic abstracted data.
In the abstraction system, the human does some of that. They do it with high variability across practices. The abstracted data is not clean either. But it is less heterogeneous through the abstraction process.
I think long-term you can get to where you want to go, but I think there is a robust concept identification and mapping exercise that will be more like natural language processing than transaction processing when all is said and done. I am just curious whether there is that kind of thinking. Are you thinking about bifurcating the data that already is countable, can plug into your statistics now? The ones that can be harvested and gotten to where you can put them into some process where we develop this. How are you thinking about parsing this? Otherwise I am not sure I see a path from here to there.
DR. LAU: I think it is a really good question. Just two prefaces. Clarice has really been much more involved in this process than I have, given just the diversity of work we have in division. But in terms of to my knowledge, we are definitely working with different sources to figure out the natural language issues with clinical notes when we get to clinical notes. To get there, there are so many steps from identifying what data are we going to collect. I think because just working with different providers and vendors, there are so many different systems and updates out there. Once we identify what we are actually going to get then we need to figure out how to harmonize it and to pool the data and standardize it.
I think at this point, we do not really know what is in the EHR data until they come. I think right now we have just a handful of hospitals who have submitted EHR data to us. I think we are working on more the feasibility studies to make sure that the data is coming in terms of what we need. Even an example is laboratory tests. Do we get the first laboratory test or do we get the last laboratory test? Especially at this point, we are looking at the entire episode of care. In the past, we have only collected discharge. With the amount of information for the entire episode of care that can range quite a bit as well as different services can be offered. We are still struggling with that.
That is a very good question, but it is definitely a multi-step process for us to get there. We are working with ASPE and different organizations to help us.
DR. SUAREZ: Thanks.
DR. PHILLIPS: Thank you, Denys. This is really exciting. Something I was emailing about just a couple of weeks ago is we launched a national primary care registry in February. I think the richness of the data, the depth of the data you are going to have are going to be much better. This will get better over time. I completely agree with Bill’s concerns. I think it will get better over time.
One of my concerns is that process that you are having eligible providers provide you the data. A weeklong extract of CCDs I can tell you out of the gate you are a missing a whole bunch of small, rural practices who use EHRs at the different end of the spectrum than the large players and that lack technical capacity to get you a CCD much less a week’s worth. That is a concern is that the sample you are going to get will not be reflective of what is happening across the United States.
One of the potential solutions for that is to work with some of the registries that are certified EHR technologies that are able to pull data and actually do some of the cleaning up front and verification with practices and that are doing reporting for PQRS that are pulling from a much wider variety of EHRs and solving that CCD export problem practices. It is a rich opportunity and I hope we can talk about that later.
DR. LAU: I do want to emphasize that at this point even though we are registering all eligible providers and eligible hospitals, the data we will be collecting from them is only among those who are sampled in the NAMCS. For those who have the capacity, they will be giving information to us electronically. But for those that are not registered with us, we will continue our manual abstraction for the next handful of years I would say. For those rural providers that do not have these EHRs that have the capacity to send us data electronically, we will continue our manual —
DR. PHILLIPS: Just really quick as an example, we actually incorporated the NAMCS meaningful use questions into our annual cross-sectional Census. And what we found is that while we have almost 90 percent of our family physicians have an EHR, only about 20 percent of those that have an EHR are able to connect to a registry. You lose a big bandwidth of people there who have EHRs, but cannot connect to you.
MR. LANDEN: I am very impressed at the progress being made. I see this as the initial development process, but it exceeds my expectations about how much it looks like you are going to be able to do leveraging the meaningful use objectives. I had been wondering before your presentation whether this was going to go into the new API concepts where you could actually pull data out, using the common clinical data set, but you did not go that way. You are using the public health reporting objective registries, which is more a push. I know that it has been a requirement for all certified EHRs since Stage 2 to be able to generate the CCDA, but the connectivity once that is generated is I think what the problem was being discussed.
Also, the CCDA contains both structured data fields and unstructured. You are aware of that. You are able to cope with both of them. It sounds like one of the few actual values that I have been hearing coming out of the meaningful use mandate. I hope to see this as a pilot because these objectives are – this is one of several that can be chosen to satisfy that particular objective. I am pleased to see that you are getting such good responses for that. I think, as we go down the road, success here will enable – I guess we will sell the value to the meaningful use stakeholders including the EHR developers and be able to build on the HL7 standards and implementation guides.
I guess I do not really have any questions other than have you looked at the impact of MIPS on the eligible providers. For the Medicare eligible providers, they leave meaningful use and go into something called advanced care information. I am not sure that carries the same requirement for the public health registries. I could be wrong. We may hear more about that tomorrow.
MS. KLOSS: Thank you. I had asked the question this morning of how far along this road we were. It seems like it is going quickly and that you could quickly be overwhelmed by a lot of data that is perhaps not very interpretable.
I had a couple of practical questions and I know this will be a learning thing. When I worked with the data aggregator companies, it just seemed like the process for acquiring, cleansing, normalizing data that Bill referred to were pretty sophisticated to be able to do anything with it. I just did not know what kind of resource or team you have and resources to deploy in that direction or whether it is sufficient or whether this is something that you are going to have to mobilize resources from other places to do it. That is one question. I have seen what it takes to do this. It is enormous. It is multi tens of millions of dollars of resources to do this on this scale.
The second question. These records I presume are identifiable. Then what provision is being made to de-identify or protect privacy and security from breaches or hacking or other kinds of mischief?
DR. LAU: I know for the first question we are going to leverage within the government as well as externally. There are a lot of companies out there that do these kinds of harmonizations and making sense of EHR data. We did an RFI I want to say October two years ago where we had different vendors coming to us to actually talk about how they would go about doing it and what we should expect. We are educating ourselves in that process.
In terms of the second question that was regarding security, the PII.
MS. WILLIAMSON: We have done some restrictions within the standard for de-identification on particular pieces of information. That was something that we walked through the implementation guide. In some areas, we would not be able to get the information. It would not be submitted to NCHS. We would not be able to receive it.
MS. KLOSS: Once you have it, is it de-identified?
DR. LAU: Yes, it will be de-identified. Do you mean for future – are you asking —
MS. KLOSS: When the data comes in, could you tell it is John Smith?
MS. WILLIAMSON: Certain pieces of information will not come to us as part of the standard. The standard is set so that they will not send that piece of information to us. We have identified, gone through the specification to say this is PII. Do not send.
MS. GOSS: Thank you. I support my colleagues’ comments about the level of progress and excitement about your update. I had a couple of questions. I will start with the connectivity aspect and the push kind of thought process that Rich was speaking to. I am curious about some of the transports. Do you have a very defined method for how entities already give you the file? I am not talking about the file content of HL7, but I am talking about how you receive it. Do you already have a very well-defined transport submission push method that you are using? Are you looking to expand that to some of the tools and the toolbox that have been rolled out under HITECH?
MS. WILLIAMSON: We did not define specifics in terms of how they will send the information to us. NCHS is exploring multiple mechanisms for how we will receive the information.
DR. LAU: We have various options we have given them up until 2018 when they have to follow the IG. We give them the option of sending us CCD for this year and next year. We also give them a custom abstract for also the next two years, but by 2018, they have to follow the implementation guide.
MS. GOSS: Michelle, the implementation guide expects that this is going to be an EDI-type transaction. It is going to be a – it is a CCD so how it gets from point A to point B up to the users.
MS. WILLIANSON: We have not defined that.
MS. GOSS: One of the things I would offer when you mentioned about incentives is to think about the role that health information exchange organizations are playing in the providers’ clinical document sharing world. Not only do you have a query-based functionality through health information exchanges that can be leveraged, but you have push as well and you also have the direct protocol. Those with the incentives that are already out there that might be something that you could piggyback on to further promote value especially look at some of the activities that are being funded under the Medicaid incentive program.
MS. WILLIAMSON: I do not think at this point we have looked much in terms of health information exchanges. I know you had also mentioned that for the rural community at this point we have really been focusing on the future of MU and trying to get it directly from the EPs and the EHs, but I do not think we are ruling out the potential for the future to explore this.
MS. GOSS: It is a value add for the health information organizations and their provider relationship. I think that there is a natural synergy to align incentives and try to get more submission rates especially when you look at some of the complexities that states are experiencing in their Medicaid program.
MS. WILLIAMSON: I think it is something that the center has to grapple with in terms of getting data from HIEs versus directly from the providers. We talked about this in vital records as well.
MS. GOSS: It is not that the data should come from the HIE, it is that the HIE should facilitate the provider’s submission of the data. They can help with normalizing it. They can help with getting it to the try right mechanism. You could create a pipe with one HIE and they can get you a lot of providers.
MS. WILLIAMSON: I am with you. I think it is something that we have to internally grapple with and see how to move forward.
MS. GOSS: In specific to slide 15, I think I heard you say between slide 15 and 16 that the data is CCD, but also you have UBO4 on that so you are getting two forms of data.
DR. LAU: This is 2014 when we were collecting data. At that point, they were only giving us claims data. Right now with EHR, we are moving to that direction. Just to show you what we are getting just from claims data.
MS. WILLIAMSON: They tried for a while to use the 837. Because we were missing so much clinical content, they thought let’s try the CCD route and see if we will get more of what we need.
MS. GOSS: The 837 is administrative financial. It is not intended to be clinical data. Thank you for clarifying that.
DR. SUAREZ: Thank you so much for the presentation. We are going to move very quickly to our next presenter from Census. They have agreed to move rather fast in their presentation. We are a little behind. I know, Laura, if you are on the phone, we are going to go right after the presentation from the Census to you. Thank you for your patience.
MS. SHIN: I will introduce myself. I am Hyon Shin, chief of – Census Bureau here today with my colleague Nicholas Jones – director of race and ethnicity outreach and research –
In the interest of time – we are going to go right into it. We will just go over it really quickly some background, what we are doing for our national content tests and our research for race and ethnicity – testing as we prepare for 2020. We will go quickly into our race and ethnicity study plan. Where we are in our testing and also the enormous amount of outreach and engagement that we have had with our stakeholders and our advisors over the last several months as well.
MR. JONES: Our research on race and Hispanic origins are really prompted by a number of factors. We are going to go quickly through those – essentially leading up to the 2010 Census. We have received a lot of feedback from communities across the country. We looked at data from our recent American Community Surveys to see how people are reporting. We are finding generally that people are not answering the questions as readily as they did in the past, not finding themselves in the questions or not seeing categories for themselves and they do not feel that they get the data that they would like to see tabulated about their individual communities.
We also had a very large over the last several decades increase in the population that we call “some other race”. It intended to be a small residual population of less than 1 percent. It is going to be the third largest category for race over the past several decades. By 2020, we would predict that it might be the second largest category.
This is primarily made up of individuals of Hispanic origin who when answering the race question say I am some other race. That race is Hispanic. We are concerned that if we do not make changes that these have impacts not only for the Census, but also for the federal statistical system in terms of the base that we provide as most other federal agencies do not collect or tabulate data with the sum of the race category.
We also had concerns leading up to the 2010 Census from communities, from organizations, and from Congress about the ability to have these changes made at the time that we are preparing to print the 2010 Census. While this was not possible to change the Census, these concerns did lead us to do more research.
The research that we are going to talk about very briefly is part of a long history of research at the Census Bureau to improve the way that we collect and tabulate data on race and ethnicity. You can see from the slide, over several decades we have been experimenting with ways to improve the questions so that they reflect a way in which our society is changing, becoming more diverse and also recognizing the complexity of our population’s changes in terms of race and ethnicity.
The last point on this slide is abbreviated as the 2010 AQE. I am going to give you a brief recap of this research if you are not familiar with it, but it stands for the 2010 Census Alternative Questionnaire Experiment research on race and Hispanic origin, AQE because it is a mouth full.
This was the base and it was our largest, most comprehensive research effort ever undertaken on race and ethnicity. Find ways that we can improve the question format, to improve the instructions, to improve the ways that respondents see themselves and therefore to improve our data.
If you look to the next slide, this is just a high-level overview of what we talk about as one of the main research dimensions of this, AQE, which is looking at the question format, testing the two traditional separate questions on Hispanic origin and on race, which you see on the left behind me and also testing a more exploratory combined question where the concepts are put into one race and ethnicity.
We had, as I said, a very large scale effort to look at this during the 2010 Census. It was a three-pronged approach with a sample of half a million households in the United States. We had a survey with both a re-interview component with one in five of those households and also a series of about 70 focus groups across the country in different racial and ethnic communities to learn more about how people responded to the different types of questions that we presented and also why they reported the way that they did given the different options that we presented to them. This basic conceptual comparison is what we would like you to keep in mind as we talk about the current research that we are doing and leading towards 2020.
Essentially the AQE was very successful in that it showed us that we had good strategies to move forward with, but we did not intend in 2010 to make a decision ten years ahead of time for the 2020 Census. We wanted to do more research to improve upon what we found and that is the research that we are going to talk about next.
In terms of just high-level findings, I am going to speed through this, but the combined question that I presented to you was shown to be very successful. It did not reduce the proportions of people who reported as Hispanic or black or American Indian or Alaskan Native or Asian or Native Hawaiian Pacific Islander. It did reduce the reporting of some of the race and also the reporting of white. These are ways in which we saw the data in comparison to the separate questions. You might think about the ways that you tabulate data from a separate question approach. We look at a non-Hispanic white population. That is what essentially the white population reflected in these results.
The other benefit of the combined question was that we had much lower non-response rates to the question itself. We also got detailed reporting increasing for many groups including for the first time the reporting among whites and blacks who have never in a race question been asked to report details origins like German or Jamaican.
Finally, we find that the results with the re-interview reflected better self-identification. People found themselves more easily and they were more accurate in their reporting compared to the survey and the re-interview response.
Finally, one common theme, but this is pretty important. People liked the combined question. We learned this from the focus groups that they said these are good ways to approach the question. I did not like what you did before. I like this approach. That is a good success.
All these dialogues that we had with the NCT have helped us to get to where we are today. We have been talking about this research since 2020 when we released it with communities, with advisors like yourselves, with Congress, with stakeholders, and letting them know that this was the first step in us moving towards 2020 with more research to come.
I am going to give you a quick overview of a research study that we fielded last year and this is the 2015 National Content Test or the NCT. This is our primary vehicle for determining the content of the 2020 Census, not only for race and Hispanic origin, but also for the concept of relationship and looking at ways to optimize self-response. It was conducted last summer and last fall. The Census day was September 1 and we went – collection in late August through the end of December.
We had a large nationally representative sample. When I talked about the AQE being the largest effort undertaken with half a million households, this was even bigger. 1.2 million households across the country with an oversample of key population groups including small populations like the Middle Eastern or North African population, the American Indian and Alaskan Native population and Native Hawaiians and Pacific Islanders. We can talk about some of those strata and the sampling effects if you are interested later.
But we are able to over sample to enable us to get wide and deep in terms of depth and also the ability to look at different options that we are testing. These content areas are being tested in the 2015 NCT because by 2017 April, which is less than a year away, the topics need to be provided to Congress for the 2020 Census. By April of the following year, 2018, the final question wording needs to be provided to Congress. We are not very far from having those deadlines upon us. Results are needed and documentation and communication is needed once we move forward.
The NCT like the AQE also included a re-interview operation to further to assess the accuracy of the research that we are doing and again it is our primary vehicle for making decisions to move forward.
MS. SHIN: The methodology that we used for conducting the NCT and the evaluation that we are going to be doing with the data that we collected – what we wanted to present is as Nicholas mentioned, the NCT had a nationally representative sample of 1.2 million addresses. We also had the re-interview. The methodology and the study plan – we discussed not only how we process the data, but we code the data, edit the data obviously to have a final data set in order to make our recommendations, as Nicholas said, for the 2017 content and for the results later this year.
The goals of the 2015 NCT race and ethnicity research. We are still evaluating our key components and the dimensions to explore. The first is the separate question and the combined question, as Nicholas presented, of what we did in the AQE. We are continuing that research as well in the NCT. This is also the first time that we are actually examining a response category by exploring how to collect data for Middle Eastern or North African populations.
The additional dimension pertaining to the wording of instructions and the question terminology, we also want to optimize detailed reporting for the groups for all respondents and to understand that they are able to report more than one race or more than one ethnic group as well. The overarching theme over all of the data collection that we have done is our web-based technology.
What we have developed obviously in order to evaluate the NCT data is we have a comprehensive study plan. It is about 100 pages long. We have additional 100 pages of appendices, help text, the re-interview questions, the survey questions themselves. We have been told that the study plan is actually pretty easy to read despite the depth of it. And a large part of the study plan is dedicated to introducing the plans for our data analysis and for our decision criteria because of our transparency about how our data from the NCT will inform the potential content for the 2020 Census.
Just really quickly I just want to show you. Some of the question format that we tested on the paper form is on the left hand side. There are handouts floating around, but I apologize. I do not think I brought enough for everybody. We are continuing to test, as we said, the separate question where we ask about Hispanic origin first and then we ask the race question. In the middle, you can see where we have it combined with write ins where we have now combined the Hispanic origin and the race response categories into one question format with a dedicated write in line for people to express to us via write in their detailed information.
And on the right hand side, we also have what we call combined with check boxes. Not only do we have the major OMB groups of white, black, Asian, American Indian, Alaskan Natives, but then we also have the detailed check boxes for people to report that they are German or Jamaican or Korean, for instance.
As I mentioned, the overarching experiment/dimension that we are looking at is our web-based design as well. We are continuing of course because we are collecting it on paper. We have to collect it on the Internet as well. Our separate question for the Hispanic origin question.
In this format what we see is you can check that you are of Hispanic origin and then you get your separate question for race in our traditional format.
Then for the combined questions for the race and ethnicity and the write in areas, what you have is then the combined question for race and ethnicity. As you check through your major groups, what we have are subsequent screens that come up that then allows people to have the ability to write in their details. This format is to help distinguish the check box options on an Internet self-response versus the write in that you have. That is also based on prior research that we have had.
In this example, you can see that for every OMB group check box that you select, you have the corresponding detailed check box or the detailed collection that comes up. In this case, we have what is your white and then also for your Asian origin so you are able to give us more details.
Then the combined question for race and ethnicity with the detailed check box write in areas, we have the same question that asks about the race. And then once you check your race self-identifying characteristics then you have the ability to then check your detailed information as opposed to just having the write in area. You are able to check. In this example, for instance, for somebody who selected that they are Hispanic, Latino or Spanish origin then we do have the ability for people to check that they are Cuban, but also have the ability to write in even more details that is not part of the first six detail check box information as well. And also subsequently this is for the detailed information for the black or African American population.
MR. JONES: You might imagine with this dimension, we are looking to see what did we learn. We are exploring all these different aspects of ways to tweak the question format to find out what approach yields us the best, more accurate results. This is a list of the research questions that we are exploring in the NCT. We have some examples that I am just going to highlight. We are looking, for example, to see which approach the separate question or the combined question with write ins or the combined question with detailed check boxes, which approach yields the most accurate responses as we measure through the re-interview.
We are also looking to see which format has low or overall item non-response rates, looking again across the question formats. We are looking to see what is the effective reporting of detailed groups. Do any of these approaches yield better detailed reporting when we are looking at check boxes or write ins or different questions being separated. We are also looking to see, as I mentioned earlier, with some of the race reporting issues. Which of these provide more accurate reporting among Hispanics in terms of reporting their race?
When you look to the next dimension as Hyon mentioned, we are testing in the NCT a Middle Eastern or North African category as a dedicated category. For those of you who are not familiar with the OMB standards, Middle Eastern or North African is a category that is defined under the white population. White is defined as people with origins – people of Europe, the Middle East or North Africa. This goes back years to when OMB looked at the first reiteration of these standards in the mid-1990s and a category near eastern or Arab was not recommended to move forward from the federal statistical system. More research was suggested. More research has been done. This research is primarily looking at whether or not to include that category moving forward and also how to operationalize that category.
As you can see here with a simplistic slide, we are just showing you where the Middle Eastern or North African category would be situated in some of our experimental forms. What we have done with all the research is list the categories in order of population size in the United States and also the examples listed in the population size from largest to smallest both vertically as well as horizontally.
We have listed some questions also regarding the Middle Eastern category of what we are testing and what we are exploring in the study plan. You can see here, for example, we have a number of different questions that relate to the accuracy of reporting Middle Eastern or North African responses. These are being used to help us understand the different ways in which we can improve the accuracy of data for respondents of Middle Eastern or North African descent. Essentially in the survey, half of the designs that we are testing across all these dimensions contain a Middle Eastern or North African category. Half of them do not. And then we are making comparisons across those both at the survey and the re-interview.
Additionally, as Hyon mentioned, we have other dimensions of the NCT and all these are crossed fully across the different dimensions. We are looking at the evaluation of the use of different terminology and different instructions. It may seem very simplistic at first when we are asking people to report what they are, but we know that the ways that we ask these questions have impacts on the ways that people report. We are looking at terminology. One of particular importance is testing the old instructions to mark with an X one or more boxes. This is first brought in with the 2000 Census for multiple race reporting. But then we are looking at that in comparison to an instruction that says mark all boxes that apply. What have we learned from research? Many respondents do not read the instructions, but they also stopped at that visual X and they mark one box. They did not read beyond it. We are moving the box. We are changing the instructions and we are noting they may report more than one group in hopes of having people understand that they can do reporting that they would self-identify with.
We are also looking at the clarity of the question. We know that these concepts mean different things to different people. What is your race? What is your origin? What is your ethnicity? To some people it means all the same. To others, they mean very different things. We are experimenting in this research with using race or ethnicity, race or origin, and also no categories at all to just ask people which of the following categories do you identify with and then to see again how they respond.
There are some examples of the different instructions and term issues that we are experimenting with and trying to understand. We want to see how does this impact multiple race reporting. How does it impact detailed reporting? What is the accuracy of the reporting depending on the instructions and the terms that we utilize?
MS. SHIN: Where are we now with our research? We conducted the NCT. We are right now in the process of finalizing all the data processing including coding and editing. We are doing feedback and outreach, as I mentioned, with our myriad of stakeholders, who have been following our research for years. As we try to finalize the study plan and hopefully working on the data analysis over the next couple of months in order to have the result that hopefully we will be able to present to all of our stakeholders and our advisors later this year.
Part of that is we recognize that there is a great interest in what we are doing and therefore we are doing this ongoing engagement with everybody. We have been talking with our communities, our organizations and leaders of the myriad of race and ethnic groups. We have also talked to academic scholars and expert researchers. We engage with the civil rights communities. We have also had an extensive federal register notice that went out before we started our NCT research and got public feedback as well.
We shared the NCT study plan with stakeholders and organizations. We actually held a national webinar on April 27. All of this is designed to help the interested groups familiarize themselves with the research that we are doing.
Where are we in the next steps? I am just going to skip this slide other than to basically highlight the last two, which is, as Nicholas mentioned in the beginning, all of this is we are mandated to provide to Congress by April of 2017 the topics that are going to be on the 2020 Census with the final wording in April of 2018. We will leave it there for any questions.
DR. CORNELIUS: A quick question. I have appreciated the Census Bureau publishing the methodological studies over the years. I am curious as to what will be the plan with regards to publishing the methodological studies that you are doing.
MR. JONES: At the high level, if you look back at the 2010 as some of the most recent experiments and evaluations that we have done, we posted all those to the Census websites and made the results available to the public. What we included for the 2010 AQE is a comparison. It was not only the report. This is a similar product to what we did in 2010, but also extensive tabulation tables that were provided in the appendices to help the public understand what did we do to help researchers understand what did we find and then to give us critical feedback. It was an important feedback loop that we had from the 2010 AQE. It is part of what helped us move forward to the 2015 work and we plan to do the same. We will be publishing the results and also preparing them to be available online.
The other challenging part of this though is that unlike you guys, most people do not want to read a long report. We also bottled this down into a very small explanation. A lot of extensive studies, extensive research, but people would like to know what is the big picture. We produce that too. We produce infographics. We produce presentations. We plan to do it again to help people understand. The bottle line is this. This is what we are doing. This is what we recommend and this is why. That is the way that we have been able to communicate with all levels of interested parties whether they be extensive researchers, Congress or just the person on the street who is reading the news.
DR. SUAREZ: We are going to go to Bruce and then Nick.
DR. COHEN: In Massachusetts, we have been using the single question combined format for over ten years for collection of vital statistics, program data, and hospital discharge data. But rather than including sub-race categories, we include a separate ethnicity question so you collect ethnicity separate from race. From all of our pilot testing and feedback, we find that cognitively it is a much more appropriate task for all respondents. Certainly we find it virtually eliminates the issue of Hispanic choosing other race.
A couple of quick observations. When we look at ethnicity among whites, we find that around 70 percent of whites choose their ethnicity as American. If that is not one of your check box categories, I think you are going to find a huge difference in responses where you have write ins versus check box. There was enormous amount of cognitive dissonance among whites if you do not allow them to say that they are American as their ethnicity, an unexpected finding for us.
When OMB in 1997 revised infamous directive 15 from 1977, they really screwed up. I hope you are going to give them the appropriate feedback that they need to rethink how they promote throughout the federal government. The implications for numerator data collection after we change the denominators are going to be an enormous task for all executive agencies.
Finally, I am sure you are going to find a dramatic increase in multiple races whether you just include the check box or the X box or I think your approach is better, making it clear that people can choose more than one. Hispanics do like to indicate in addition to their Hispanic race other aspects. We have used two modes. One that has Hispanic/white, Hispanic/black, and Hispanic. We find that many Hispanics like to represent other parts of what we call ethnicity. You would call race. Having to deal with producing statistics where you have a dramatic increase in multiple race respondents is going to be an onerous task not only for the feds, but states as it filters down.
MS. SHIN: Absolutely. To address the three parts, we actually currently see in the American Community Survey a lot of people who report that they are American. We do have a coding and classification system where we do the coding. As we have presented – the vast majority of the data that we get comes from people self-identifying for a check box. However, we do have hundreds of thousands of write ins. And a lot of what does come in is American. People express themselves that way.
One of the dimensions that we are looking at especially with the NCT and because for the first time we are allowing write ins for the white and black populations that we have not done previously is that if somebody checks that they are white and they write that they are American, we will classify them as white American, much in the same way that if somebody writes that they are Asian or checks that they are Asian or writes that they are Korean and then writes that they are an American then they would be counted towards that population as well.
To go to your question about OMB, we have been engaged in an ongoing discussion with OMB for the last couple of years. At this point, we have monthly meetings with OMB and the other federal statistical agencies that this will impact. Absolutely. We know that that is a very big component of what we are doing and we recognize that right from the beginning right after actually probably the AQE where we are engaging with them.
Finally, for the multiple reporting, because we allow multiple reporting now, we would be most likely tabulating the way we have done it traditionally, which is we allow for the alone populations so people who only identify as being one. I am Asian alone. That is how we would tabulate out.
For people who do multiple reporting, we would do it the same way now as well where we allow for Asian and any combination. We call it AOIC in our tabulations that we do. We will continue that approach where we would tabulate and produce data for those who report one race, those who report multiple races and many different combinations.
MR. JONES: I just want to add to what Hyon said. This is really built on the research that we did with 2010 AQE. We had these encouraging findings that you are talking about, more accurate multiple reporting, more accurate detail that is being provided and the ability to show that we can produce the data to move forward. This research will help to really cement that and move towards 2020.
DR. SUAREZ: Thank you. I think we were going to go to Nick.
MR. COUSSOULE: A question and a comment at the same time. I can appreciate the complexity of what you all are trying to do. It is pretty interesting to try to figure out how to get the data more accurately, but I also come back to the question of to what end question. When I look and say how do you determine whether to distinguish the Middle East or North Africa at all. Why? How is it going to be used? I know that is not a simple question at all. Part of when I think about this being accurate is good, but why am I trying to create different subsets and what am I going to do, which may lead me into doing a very different way to collect it. I am trying to understand. What is driving you towards saying this is the way I need to try to collect and be accurate?
MR. JONES: The big driver is what the American public is looking for in terms of data that represents their communities. We have done a really interesting infographic. I talked about how we use some pictures to tell the story. If you look from 1790 to the present across every decennial Census, the categories, the concepts or the instructions have changed every single decade. We certainly do not have the community that we had in 1790 here in the United States now much less the community that we had in 1960. But each decade more categories are added to really articulate the ways in which we have been changing. As I mentioned in 1997, this category was brought up as a potential topic.
And what we are hearing from the community is this is really a community-driven effort, not a federal government driven effort. The communities are saying we do not identify ourselves as white. We are not seen as white. We are not having tabulations of data that help us see who we are and how we are changing. We need that diversity. We need that data to be presented.
It also goes within the categories. There is more heterogeneity within groups within the white population, the black population, Asian, Hispanic. Any and all of those groups have a lot of heterogeneity within them and that diversity is not currently captured through the race question for all categories.
We have three different concepts that we have used at the Census. There has been a question on origin since 1970. There has been a question on ancestry that was in place for many years on the Census long form. It is now in the American Community Survey. There has always been a question on race, but they collected different things. We are trying to pull all that together into a way that would really provide more accurate data for all communities through questions that work to both collect and elicit the results that we can then provide back. That is the work that we are looking to do and it is really driven on the context of who we are and how we are changing and becoming more diverse.
DR. RIPPEN: Again, thank you. Very thoughtful. I know it is a lot of work and it is very important for a variety of reasons especially as we grapple with health equity issues and things like that. I guess just a nuanced to build on what Nick had implied or kind of suggested and that is sometimes if I am asking you to identify and I do not see who I identify with, I am going to want to have something that I can identify with. It is an important thing whereas if you don’t ask. I think it is a chicken and egg thing.
The other nuance is how people feel. My husband likes to tell a story. He is Puerto Rican. He has a lot of friends that are Puerto Rican. They filled out a form and the gentleman put in Puerto Rican and white. His skin was a beautiful ebony color. The person who got the form was really upset that the person did not put black because he put white. Again, this whole question of how do you identify what does it mean to be in this category white versus black. I know the history of it, but it is just some nuance with regards to again ultimately how do we want to leverage and how do we actually make it meaningful as opposed to separate.
MR. JONES: That is also part of the work that we did both with the 2010 AQE as well as a lot of qualitative work that led to the refining of the 2015 NCT work. We are understanding how people are self-identifying, given the different questions that we ask, how they are reacting to the categories and the definitions that the federal government uses. A lot of that is what is being challenged. In its entirety, the federal system defines race and ethnicity for the world here in the United States in terms of how we classify people. But the entire world is not considered in the definitions. South America is not in there for race. The Caribbean is not included. Brazilians are not defined as being Hispanic or a particular race group. All these topics and categories are ones that we have been wrestling with for decades to try and understand how can we provide more meaningful data for those communities. This research is helping us to understand how the individuals from those communities response given the different options that we present and that is how we are looking to move forward with recommendations as to how to continue to improve our data as opposed to rooting it in what we thought or knew 10, 50, 100 years ago.
This research will have a lot of answers that are going to be provided to help us understand how to move forward. But I want to say this also. There is not a perfect question on race and ethnicity, but we can do better and that is with the work that we are looking to undertake.
DR. MAYS: Thanks for your research. I think it is going to help us grapple better with some of the things that we are grappling with. I have two questions. One is that – I am in California so we do the California Health Interview Survey so we are used to multiple race kind of questions. But one of the things that happen is that the CHIS had to make a change. We had a question that would say but which race do you identify with the most. We were forced to do that change because of OMB. I guess I am wondering, for example, whether or not you are going to be allowed to have a question that would say which race do you identify with the most. From a data analysis standpoint, a person can have five things, but it is like the lived experience is what impacts their health and well-being. That is question one.
And question two. We also found on CHIS that self-identity starts to get problematic when you hit about the age of 70. There are certain questions that – it is not just the race question, but some of the other questions that when you hit about the age 70, you have to really think about what results you are getting and also if you have limited English proficiency. I am wondering if you are also in your research looking at those two things.
MR. JONES: Thanks for your questions. I will go with the first one. We recognize and we did this research back in 2010 with the AQE results to try and follow up and see how the individuals who report multiple races do they identify with a primary race. That question got a lot of negative feedback in saying that multiple race respondents were saying these are my primary races. I do not have one. My primary race is white and Asian, white and black, et cetera. We moved away from that concept. We do not tabulate backwards. We are trying to think about how we can tabulate data to move it forward and make those bridges. We develop bridges with NCHS as part of the 2000 work when we first tabulated multiple race data through the Census. We are really looking at that projection that is very important in terms of age. This is one of the fastest growing demographics and we are looking to see how can we improve the collection of the detailed data especially for respondents who may not even know that they can report more than one race now even almost 20 years later.
Your other comment about age or language. This is also part of the work that we are doing in our mid-decade testing. As part of the research for the NCT, we are looking at age cohort differences to see whether or not this reporting is different across age groups, which we expect it might be. That is how we understand through the re-interview whether or not we are getting more accurate results for the ways that people understand the concepts and understand the questions.
We are also designing with this research that it was both a Spanish language form as well as English in terms of design that was utilized for all the work. But we have done a lot of research to improve the ways that we will then feel this with other multiple languages. We had language support through the 2015 NCT for about six primary language groups. But then with our 2016 and 2017 research, we are expanding that even further. We have now questionnaires designed in Korean, simplified Chinese. I forget the other one. Vietnamese. Knowing that these populations are growing and they need to have the ability to answer in their native language. We are doing work to make sure that the concepts also translate in those native languages so to be very sensitive to that to ensure that what we are asking translates well for the respondents. I would be curious to see from the work you are doing in California, which kind of leads us into where we going. What have you learned? I would be happy to talk with you about your work.
MS. SHIN: And the other part I will say is what we have presented is all of our quantitative work that we are doing. 1.2 million households. However, all throughout this, we are also doing cognitive interviews, usability testing as well in other languages. We have a really robust way of doing translations at the Census Bureau. We also have a translation office that I think just recently started as well. And what we are trying to do with that – that is a like a separate – we are all interconnected. Research is to show that in other languages, the way we conceptualize race is very different. Who would be responding to a Census form in Tagalog or Vietnamese or Russian, for instance, their concept of race is different. We are trying to be very culturally responsive and culturally appropriate in the way we ask questions and not just have a literal translation of how we are asking it in English and just having that exact same terminology wording in those other languages. We are very much aware of that and doing continuing research in our qualitative work.
DR. SUAREZ: Thank you so very much for your participation. We really appreciate it. I am sure we will have more discussions about this at some point and potentially even some framing of observations and recommendations.
MS. SHIN: All of this research has been building obviously since the 2010. We are hoping that we have results later this year. We are going to be doing the outreach and the presentations and the very transparent way of showing what our research actually found. We are all really anxious to see what the NCT data are showing us.
DR. SUAREZ: We look forward to hearing about that.
We are going to go to our last speaker, Laura Conn, who is patiently waiting on the phone. She is going to talk about E-public health reporting and then we are going to go to our much deserved break. Thank you for your patience.
Laura, are you on the phone?
MS. CONN: I am, Walter. Can you hear me?
DR. SUAREZ: Yes.
MS. CONN: Great. Thank you. Good afternoon everyone and thank you for the time to talk about electronic case reporting today. As Walter said, my name is Laura Conn. I am the director of Health Information Strategy in the Center for Surveillance, Epidemiology, and Laboratory Services at the CDC.
Electronic case reporting or eCR that we finally call it. The reporting is legally mandated reportable – by health care providers to state and local public health agencies electronically. There is value in eCR for both public health and health care providers. Public health can receive more complete accurate and timely data to allow for earlier interventions, to reduce – transmission and identify and control outbreaks more efficiently.
It has often been difficult for health care providers to know the what, when, where and how of reporting. They have typically not been very good reporters to public health – automate the process of case reports, the burden on health care providers can be diminished in their obligation to report these conditions to public health fulfilled. This will also allow for more communication between public health and health care to advance population health goals.
At a previous NCVHS meeting, you heard about the CDC Surveillance Strategy from Dr. Chesley Richards. But I show you this slide – review for the four pillars of this strategy. I hope you will see how advancing electronic case reporting can contribute to each of these goals.
While electronic case reporting is a public health community wide effort, the Center for Surveillance, Epidemiology and Laboratory Services at CDC where I work fill that electronic patient report and is the future of surveillance and that the time is now for action. Electronic case reporting is one of the public health objectives in the meaningful use Stage 3 incentive program. We need to work together with health care during this time of transformation.
Our priority focus is getting initial case report while implementing a scalable architecture for electronic case reports that shows incremental, but tangible success.
What do we need to do? Public health needs to build stronger relationships with EHR developers and the health IT implementers of these – we need a standard for the data public health needs in the initial case report. We need to facilitate electronic case reporting by utilizing shared platforms and tools using secure and scalable standards and technologies. I will talk about the progress that we have made in these areas. We also need to support state and local public health agencies to receive – electronic case reports.
This picture shows the entities involved – and the flow and some of the components needed to accomplish eCR –- starting in the upper left hand corner in the column that says healthcare, a provider in its normal practice of care would record information about an encounter with a patient in the EHR. There would be a set of trigger codes, which I will talk more about coming up in the electronic health records – information recorded in the EHR – trigger codes. It would be the generation of an initial case message sent to public health and to a service to adjudicate – report ability. If the case is reportable, it would determine which jurisdiction would need to receive that initial case report and route that to the public health jurisdiction for inclusion in their public health system.
A notice of report ability or information that can be provided back to the – care provider would be developed and sent back to say yes, this is reportable or this may be reportable – additional information may be needed.
In the future, we also hope to be able to add the ability to request supplemental data on these conditions, which may include data that is not currently in an EHR, but would be needed by public health to complete a full case report. That is what you see there on the bottom half of the slide.
The next slide is a very similar version to this picture, but focuses on the top half – initial steps in the previous diagram – just to show that there are – many efforts underway by the different public health partners. I am going to skip on just in the interest of time because I am going to talk about – going forward.
The triggers. A match to a trigger code would initiate – and the sending of an initial electronic case report. There will be one set of trigger codes for all conditions for all jurisdictions. This is critically important because EHR vendors could not possibly maintain a separate set of codes for each public health jurisdiction. But this does, however, require a second level of public health decision support to confirm report ability to a specific public health jurisdiction.
All trigger codes are indicative of a possible reportable condition that are using codes that are used in EHR – LOINC, SNOMED, ICD-10 that represent diagnoses, laboratory orders when a condition is reportable upon condition and laboratory sets and results.
The triggers are being authored in a tool called VSAC, which is hosted at NLM. And the code set itself is hosted for retrievable by implementers on PHIN VADS. There is a draft set of conditions that is available now or a draft set of codes for five conditions: pertussis, chlamydia, salmonella, gonorrhea, and –
The next component – electronic initial case report. Last summer CSTE convened a taskforce to identify the data elements of interest to public health for initial reporting in an electronic case report. While I doubt you can see the details of these 40 such elements, they are grouped – information about the provider and health care facility, the patient, the identity, contact information and demographic and clinical information about the encounter including diagnosis, laboratory tests and results and symptoms, et cetera.
There was the recognized need for a standard structure to be able to get these data from clinical care. Some guiding principles that were used in order to determine the standards that we use – C-CDA, which needed to be implemented in the short term that we felt like were needed to use an existing standard, which is why the C-CDA was chosen. Also, because it was already in use by vendors. We worked to use already certified templates. We also were looking for something that public health would be able to accept and use and something that could be used in the exchange using existing trust networks.
The HL7 Public Health and Emergency Response work group took up the work in October of last year. They used the data elements identified by CSTE and drafted the implementation guide that was balloted in December. Comments from this guide were reviewed earlier this year. The next draft of the implementation guide was approved – for publication at HL7 in mid-May. We are expecting publication of this implementation guide for the initial case report any day now.
Public health agencies will need to support – receive and use these new initial case – CSTE is beginning work on a toolkit to assist public health agencies in what will be needed in their jurisdiction in order to implement electronic case reporting. They will need to assess requirements in their readiness. They will need to identify the resources of both people resources and financial resources in order to accomplish this work. CDC is working with existing cooperative agreement programs, the Public Health Emergency Preparedness and the Epidemiology and Laboratory Capacity cooperative agreements to help sort this. There may be other opportunities with programs from CMS like the 90/10 match program that can also help support this.
I have mentioned that reporters often find it difficult to know what is reportable and to where and the details of those reports. CDC and CSTE have working on building a public health knowledge base that standardizes reporting – jurisdictions into computable logic that will provide public health decision support for reporters to confirm the need to report to public health.
The public health decision support tool will be initially hosted on the public health community platform that uses the APHL’s AIMS infrastructure. It will receive the initial case report process, determine report ability and to what jurisdiction the report should be sent so it can be routed appropriately to public health.
The AIMS platform as the technical infrastructure for the public health community platform will provide a shared space for tools to use in support of electronic case reporting. The platform can receive the initial electronic case report – RCKMS – provide notice of report ability back to health care and route the initial case report to the appropriate jurisdiction or jurisdiction for action.
I just wanted to point out here that platforms do not store data, but for short periods of time while it is processing through the decision support tools and being routed and confirmed for delivery.
That is an update on some of the technical components that are underway to support case reporting. Equally importantly perhaps at times more important are the communication and collaboration aspects of this activity. Public health has been working to establish the business case or electronic case reporting in communicating it with a wide variety of partners both inside and outside the CDC. Discussions are underway about data use and other legal agreements that may need to be in place to accomplish these goals.
Art Davidson from Denver Public Health has reminded us often that in order to succeed, we must be one public health in working with health care.
Art’s advice is particularly cogent in the face of messages like this one from a prominent EHR – the public health is like a thousand gnats that consistently pester you, but they never help you do anything.
It is time to change this dynamic and build a strong working relationship with health care and our vendors. There were several meetings with EHR vendors – February this year with an agreement to work together to advance electronic case reporting.
In addition, there have been – of electronic case reporting at the Advisory Committee to the Director of CDC, another – like yourselves – last fall, they passed this recommendation that says by mid-2016, which is essentially now, CDC should convene appropriate partners to develop recommendations for a national strategy for eCR.
To that end, there have been considerable work with partners to discuss different perspectives to build and pilot different components to support eCR. There is a draft manuscript in the works that lays out a high-level approach technical framework. These meetings with vendors did result in an expressed interest to work together. In fact today and tomorrow one such meeting is occurring that includes public health, heath care and EHR vendors to discuss a vision, governance, and initial implementation of eCR. This meeting is being convened by RWJ with support from Deloitte – Health Informatics Institute. We are looking forward to the outcome of those discussions.
There are other upcoming discussions within the public health community around electronic case reporting that will occur at CSTE annual conference next week and then the Public Health Informatics Conference in August in Atlanta.
The draft technical framework for electronic case work has identified the shared vision and is intended to address these bullets here. Foster interoperability, limit the burden of reporting for EHR developers, address jurisdictional needs of health departments, and establish governance.
Somewhat of a summary. I am just going to actually skip over this. It covers a lot of the components that I have already spoken to.
In April of this year, the ACD made additional recommendations to the CDC director. The first is around governance of electronic case reporting. You can see here that there are very similar principles from the framework –
It identifies that public health programs need to adapt standards that are widely adopted in health care. That the governance group needs to have a hand in determining changes to the standards being used. There needs to be a consolidated approach to working with vendors and that public health people evolve to this new way of doing business.
To summarize, for success, we need to build a working relationship with EHR developers and health care, show incremental success, keep interoperability in the fronts of our minds, learn as we go, improve over time that we can achieve our long-term vision.
I just want to thank and acknowledge the many passionate public health colleagues that have gotten this effort to where it is today and continue to push to achieve our goal. Thanks for your time.
DR. SUAREZ: Thank you so much, Laura. This was very good. Do we have any questions from any members of the committee? I know we actually coincided this particular meeting with the one that you mentioned, Laura. We are scheduling or looking at scheduling some additional time this September to hear more about the meeting that is occurring this week in Chicago and learn more about the platform development and E case reporting and other areas around public health and vital health statistics reporting.
MR. LANDEN: Like the previous presentation we heard, I am very pleased to see it building on standards and the C-CDA and the connectivity and the interoperability that has already been introduced by meaningful use. I like that.
I do want to go on record of saying I am not the EHR developer that is the source of that gnat quote. Although in some context, I can identify up to a large extent with that comment.
DR. SUAREZ: Thank you again, Laura. Thanks so much for the patience. We are going to take a 13-minute break. We will start back at 4:45. We are going to be talking with the co-chairs of Population Health. I know we have some guests from our work group that will be presenting at 4:45.
DR. SUAREZ: I want to thank our chairs of our Population Health Subcommittee for agreeing to graciously move their block of time to tomorrow. We are going to do Population Health tomorrow. We are going to do it after lunch tomorrow. We are going to replace the committee’s strategic planning session, which we were going to talk more about this week’s hearings and all that. Hopefully, we will be able to do some of that as well as we go tomorrow through the agenda in the morning and see how much we are going to save if we save any time.
In any case, we are going to go directly to the work group on data access and use. I am going to turn things to Vickie.
DR. MAYS: Thank you. It is very hard being up here now and being mindful because when I am back there, I want my tent up to be able to ask the questions and stuff. I am going to try and be mindful of the time, keep my responses shaped up so that we actually can get out of here on time. Thank you to my colleagues who yielded their time.
Let me get started now. Part of what I want to do today is to tell you where it is that the work group is going, some of the changes that we have embraced, try and do it pretty quickly because I think that we will also be talking – I know we will be talking about some of these things in the work group. Since the meeting is continuing on, several of you will be in the work group. I am going to do this shorthand because I want to make sure that we have time for the presentations.
One of the things that has happened is that if you remember a couple of times I presented what the charge is to the work group and it has seven long pieces of things that we are to do. It really is pretty overwhelming. Part of what we have tried to do is to really figure out what is the best space for us to be working in right now where we can make some change occur.
In the past what we did was we worked pretty much as consultants, which is why we are pretty nimble. We would respond to requests from HHS, things that Jim wanted to see us do. And partially what has happened is that it has kept us busy fixing things and running after solutions. But each time what we are seeing is in doing that that we would probably do much better to have a big picture approach that really begins to develop some type of guidance for HHS in general. Thanks to Walter and Rebecca and Jim and the discussions that we had, we are going to turn the corner and actually look more like a subcommittee even though we are still a work group. You have this in your E-agenda book.
Part of what it is is we are going to start thinking about domains of guidance. In thinking about where is it that we want to focus such that we can give guidance to the federal agencies, it will allow us to probably integrate things that have been going on in standards, things in population health as well as the data stewardship principles. We really do this work as work that will be across the committee of the whole. That is number one.
Overall in terms of our approach, here we have set up for you a sense of what the purpose is in terms of our approach working with these federal agencies. We really wanted to develop a set of guiding principles. We will think about in terms of developing these guiding principles where it is that we want to focus. It could be a letter to the secretary. One of the things we will talk about is what are the products.
There will be guidance, but we are operating in a space that when we talk about products it may also entail, for example, putting a blurb in a professional organization that really fits the group of people who are data entrepreneurs. It may mean using a blog for people who really work in the space, using federal data and what they are often interested in is how do I fix something so that I could use something. Our goal is to increase access and use. Our goal is to expand the use cases that some of the federal data currently finds itself. It usually is really good in terms of data entrepreneurs and probably researchers. But we want to start thinking about – and pop health has been very good at this, thinking about the community and also getting down if it is possible to also think about the consumer.
We want to think about best practices, which will allow curation of the data better, which gets into discussions we have had previously. We want to think about the development of the data matrix, which will be most of the discussion that we are going to have both in the work group tomorrow as well as to introduce you to what that means today. We want to think about metadata standards. I never thought I would be crossing that standards boundary. I think it is being on that review group that even with great trepidation that I think about that. There are several things that I think in general you will see we will want to work with subcommittees to help us. We really do need a partnership as we think about the development of this guidance.
Process. We can leave this – great detail on that today because that is something the work group will talk about. The timeline. We want to redo our work plan. We want to get a sense of what bandwidth that we have. Remember, our appointments are like typically on a yearly basis. We also have to get staff in place to work with us. The bandwidth of how much we want to cover. I think that is something that is better talked about after we have been in our work group, but we also have some staff.
Let me just ask if there are any questions about big picture here before I turn the presentation over to Helga and Josh to really give you an idea of what we are talking about in terms of what we think this guidance will be like in terms of a matrix. You have an outline of that in the E-agenda book. Questions at this point?
We started with the idea of a matrix, which was to do a crosswalk with some of the work that was being done in pop health. Helga and Josh have been some of the primary leaders on this. But what we want to start with is giving you some background as to what is this about, where are we headed, and what some of the issues are.
Josh, I am going to let you —
DR. ROSENTHAL: As Vickie said, I am only going to take ten minutes. I am going to be very quick about this. It is just my name up there. If there are any misspellings, lack of spell check or any other silliness, that is me and not Helga. You guys should have all received an outline of the data matrix. I just want to walk through ten minutes just a couple of ideas that show you what we are thinking and then ask for your feedback on this.
First, what is the data matrix, data access and use matrix? Essentially, we are putting together a working draft, cataloging the major components of, and how to describe information necessary for people to increase access and use. Necessary to support, increase, and extend of different users to find, access and use the data. Essentially, we want to answer questions like this.
How can a data consumer find data? Things around speed, ease, accuracy. How can a data consumer quickly tell if the data will be meaningful to their abilities and goals? Do I have to download it and spend time and try to figure out what is going on or can I look at it and understand?
How can a data producer or distributor, I will talk about the difference between those in a second, ensure that the data is meaningful to different users? This touches on things like usability, utility, and others.
Finally, how can a data producer/distributor increase the reach of an ability for a user to find the data? I am talking about the pull versus the push model. This is particularly salient in light of health indicators and some other things going on.
Essentially, when we think about distribution outside of health care typically what we think is there are two models that have dominated for the past few years. This lecture I give at Harvard from time to time and how do you get data out to people.
The pull model is an early model of web and information and it is structured around the destination site, a place where you would ask users to go, to drop what they are doing and go to you and require users to find the site and then go to it, pulling specific users to a specific site. And the model here is to think build it and then they will come.
The push model is very different. It is a later model where information is pushed to users where they are through a variety of distribution channels. This is known as web 2.0 and the idea here is think of throwing seeds to the wind. Let a thousand flowers bloom.
Just to make this very specific, HHS asked us, Jim and others, from time to time, who is using our data. We can sort of answer it. We can look at the web logs and we can say there are 1500 people at AHRQ and they downloaded this. But when you start looking at broader distribution, there is literally hundreds of millions of users using HHS data in a push model of which we are not aware.
And the way to think about this a little bit is if you want to do a pull monitor, which was Health Indicators Warehouse, you have to have staff that is salient, expert and can do a top-down model of organizing all this stuff.
A push model is a little bit different. You put some stuff around it, not necessarily technical metadata, but some stuff and then you allow different people to come in and get it. I will show you what I mean by that.
We want to ensure that the data gains as wide of a reach as possible. By the way, if you want to do push or pull, if you still have to have pull, that is fine. It needs a number of the same elements, which we will address in here.
This access and matrix just focuses on various categories of information, things like data publisher, data description, currency, frequency, attributes. You can look through it at your leisure.
Just to be really clear about it. It is not designed to replace standards, your current work. It is not designed to be a technical system or architecture. RWJ has done several listening sessions and meetings they have been involved in where they have been recommending technical architectures. That is not what we are talking about.
It is designed to address the specific elements around getting the data out there to increase access and use beyond just general data practices. POD and OSTP has done a good job of putting general data practices out three, but not specific for health care.
When we look at this, we think what work do we need to do on this beyond the draft you are seeing. We need to extend it and expand it. We need to have it reviewed by a wide variety of users and groups. Its current form is kind of high level and it is done intentionally so specifically to keep everyone from getting stuck into the details, but to be granular enough to be able to be worked through and valuable to specific use cases.
And the questions we are thinking through as we speak is this is useful. Is it useful for data producers and distributors as a mechanism to inform production and distribution in order to facilitate use and access, as a guide to evaluate data assets and distribution mechanisms, and the degree to which they actually facilitate access and use? It is a different twist on it.
What is the best format or artifact, as Vickie spoke about? Does this work overall have the necessary elements to raise issues and inform strategy around policy and use of HHS data? What is the best name for it because we like naming things?
Before I close, I just want to throw this out there. If you want to see more extended detail, we can show this. But when we are talking about what we mean by some of these terms, if you are not familiar with this, this is available in your index. When we think of data access and distribution, there are at least four layers. This is some of the work we have done over the past year. Of course, there is the data. How useful and usable is it? There is the information architecture. How does someone actually find information within a site or delivery mechanism? There is the UI and user interface and user experience. How does someone use and experience a site or delivery mechanism?
Most importantly for what we are talking about, there is an ecosystem interaction. How does someone find a site or a delivery mechanism of information? That ecosystem potentially especially in a push model should cause us to reexamine some of our assumptions. When we think of this to walk through, there is broad applicability moving down to general security. They solve different audiences. You can think of enclaves as having the smallest reach. We are all familiar with those of being the most secure. Then you move up to destination sites, web 1.0 where you have to go a place. Then there are meta sites and repository. This would be data.gov. Most people are familiar with that. It is not really a site conveying data. It just pulls together metadata into a reasonable format.
And finally, there are things called data browsers. For those who are not familiar with what a data browser is, that is really important because here you have tens of thousands and hundreds of thousands of people using HHS data in this push model. It is only using certain sets because only certain sets of HHS data have been wrapped with things that are identified in this data matrix.
And then finally, there are additional, secondary, and tertiary delivery mechanisms, people taking HHS data and repackaging it where consumers are using it. To walk through a couple of examples. What is a data browser?
Here Google has pulled together HHS data along with a bunch of different data. They did what Health Indicators Warehouse was trying to do, but they did not do it from a top-down approach and they synthesized it. The metadata works pretty seamlessly from source to source. How did they do that? They did that with certain sources. They had enough of these data elements that we are defining in the data matrix. If you have not played with that, you should go into it.
The way you can think of it is you do not have to download data you just interact with it in a browser. When we teach kids how to use data or bring in health care , entrepreneurs and just for background, which is a little bit of what I do as well. They do not have to know how to code SAS or R. They can go to a browser and interact. There are 500,000 people that did the diabetes comorbidities challenge from ReadWriteWeb and TabWell where they are using HHS data. None of them could code and they come up with very interesting things. There are data browsers we should be aware of.
Then there are these secondary systems as well. Here is Yelp and ProPublica using HHS data, hundreds of thousands of users. ProPublica has actually taken HHS data and made it reasonable for users so they have a data store where they have repackaged it and wrapped it. You can download a set that has data integrity and full metadata behind it.
US News. Full disclosure. This is my little company working with US News. They are working – if you have ever seen best doctors and best hospitals, they are using HHS data. Anyone looking at that is actually using HHS data as well as well as hundreds of millions of payers and providers using it.
What I hope to close with here is that we are building this data matrix, which should be helpful in a push or a pull model. The things we are hoping to answer are how can a consumer figure out if data is worth using – a data consumer, I mean, And how can a data producer actually incorporate some basic things that not only allow them to sustain within a destination site, but allow their data to go to the wind and be used by hundreds of thousands of people instead of just a handful of people. That is our goal.
DR. MAYS: Can we do the matrix next? Then they will see it and then take questions. I realize and pardon me, Josh, for this that some of you may not know Josh. I am so used to Josh in our work group. Josh Rosenthal is a part of the work group. He is the CSO of RowdMap.
DR. ROSENTHAL: We are a little start up that was founded on a dare by Todd Park, who was a friend, and said no one is using this stuff. We are Instant Young Entrepreneurs of the Year. We work with payers and providers covering a hundred million Americans in 48 states, essentially just using HHS data to do things like design delivery networks, calculate risk, et cetera. The beautiful part now is to teach health care entrepreneurship courses at places like Harvard, Hopkins, and MIT.
In the olden days, there is always the data barrier to creating a value proposition, meaning if you want to create a company, you have to ask a payer or provider for data. And now all of a sudden, you can walk in and basically have – with these data releases, we filed a number of Freedom of Information Act requests, which were granted. We have data in 97 percent of the providers where we take old school data sets like BRFSS, Behavioral Risk Factor Surveillance System, and we publish in the British Medical Journal with Wennberg, who is part of RowdMap as well, showing how it out projects cost compared to claims and EHRs. You can say why is that interesting. It means the standard status quo health care delivery system is all of a sudden incredibly interested in old school sets like BRFSS.
The point is there is a lot of data out there, which we think of in terms of specific settings like population health and other things that are incredibly powerful to other sources and actually improve privacy as well at the same time.
DR. RIPPEN: I think to build on that, there is a lot of work that has been done, looking at what kind of data people are looking for from an entrepreneurial perspective and then also what are the characteristics that they may be interested in. Let’s just say a lot of people being reached out to datapalooza you were mentioning before.
I just want to put into context because I think that there is a lot of expertise at least with regards from some of the perspectives of the potential users. I am going to pull everyone back to more of the not so exciting.
DR. ROSENTHAL: This is incredibly exciting. When you sit down with the kids who are used to comp sci outside of health care, they basically say where is this stuff in health data? They say, what are you talking about? I have to go to an enclave and I cannot find any of this stuff. It seems crazy to them.
DR. RIPPEN: Actually, we are going to be doing a lot more cross walking. I think, Bill, you put out some recommendations of the different kind of matrix for data and what are the elements. All these things have to be included in here. Also, in an agonistic way, I know we all talk about health, but we all know from determinants of health while it is not really just what we consider what is happening in the health care system. As we start thinking about these groups, we have to step back and say how do we not bias it and how do we actually capture the category.
What would be an example? For people who are publishing data, what are the things that they should actually be using to describe it? People can do it in a consistent way so that you can actually make a determination if you think the data might be useful or not. What you do not want to do is spend a lot of upfront time before you realize that it is not going to work. It will not be on the nice maps because if it is only a one-year thing, you are not going to be able to track it for the next year. Again, what are the characteristics?
We would like some feedback and ideas to make sure that we are thinking about this balancing robustness with that you can actually do it. Again, where is that balance? Obviously, you always want the data owner for a lot of different of reasons and then the theme. Again, what we are really thinking here is that this would actually line up with the population health matrix because what you do not want to do is have multiple different ways of defining topics and categories. Since it is broader, how do we do that?
The purpose becomes really important because if you are collecting data as part of a health care delivery, the kind of data and the purpose becomes important versus research versus something else. Again, the context, the audience. Because the more you know about, the more you may understand some of the limitations of the information or not.
Geographic scope. We will get into a little bit more detail later. If we really do want to reach out and support communities, it has to reach out to that. The funding becomes important because it is mandated or un-mandated. It has some implications for a longer term. Again, I know Bill probably thinks this is a déjà vu club. And then the priority status. Is it important to the organization that it is actually capturing the data or not? How is the data collected?
There is the whole thing about the currency because it is important to know how often it is. What is the lag time? Then the data attributions also with availability because sometimes you can get data, but you have to actually do a formal request. That is a barrier to use. What are the limitations of the data, the file format, attributes and so on?
And then in the end some more details that actually go more to support things like the push model. What are the technical specifications that you are actually using and making available so that people can actually use it for other applications directly or they have to actually modify it?
And then obviously over time from the user perspective, are people using it? It goes to usability, but that is after the fact as it relates to something that would be shown.
Again, this is a high level. We really do need to think through and make sure that what we are thinking about is robust enough, but not too difficult for people to actually do because no one will do it.
I think, again, cross walking it to some of the earlier work just to make sure that we are not missing anything from the multiple groups become really important also.
And even some of the lessons learned because there was shared space for data that had challenges because standards and how do you maintain. Is there some lesson to be learned there too? That might be helpful too. That is the overall approach.
DR. ROSENTHAL: The key elements that are necessary to extend access and use even within a pull model, but especially to a push model.
DR. MAYS: What we would like to be able to do is to — let me just say a couple of things, Bruce, and then we are going to open it up for questions. Part of what we want to be able to do is to draft out a little bit more of this matrix and then send it out to you and ask you to think through these issues with us. Initially, it was we were going to send you this question just asking you what do you think and what do you need when you go for look for data. I think we are getting far enough along that what we want to do is actually have you look at the matrix and tell us particularly from the perspectives of pop and standard, et cetera. What kinds of things would you want to see, but at the same time because I thought we were going to have a presentation about the data warehouse? At the same time, we cannot design a process that gets so onerous that what happens is that it requires a lot of caretaking and maintenance, but something in which a person can go on, find out quickly. Is this data useful to me? Make a commitment because we put it in a format where they are more likely then to be able to see its value and then move from there and utilize the data. We want to make sure that the information is there, that it is in a format that makes it usable and that it is not onerous to the data owners.
DR. RIPPEN: And then just to add because once we have what we think is a fairly reasonable matrix or template or however we want to call it, the intent is to actually provide it to some of the HHS’ data providers to actually try to see if they fill it out. What are the challenges and how would they actually go about doing that and get feedback? Because ultimately if this becomes what we consider best practices for actually describing your data in a consistent way for data users at whatever level then I think we might have an interesting approach moving forward. I hope I got that right.
DR. MAYS: It sounds perfect. Let’s open it up and then I will make some comments a little later about RWJ and their space and where I think we might want to move to in terms of some collaborative efforts with them.
DR. COHEN: I think this is great. What would help would be use cases. I think in order to really understand how it all works I think some potential divergent use cases. Let’s say I am a community advocate and I am interested in smoking. What would the length be for how this matrix would help somebody like that if that is a target audience? Or is the target audience researchers? I need detailed multiple cause data by race and ethnicity. Policy makers. Think of potential audiences that you are targeting for the description of the data either in the push or pull model and work from that end and at least that will help me understand how the utility of this approach.
DR. MAYS: There are four use cases that we are going to try and do and there is a way in which once a data owner decides to do this that they can collect analytics in the background to help them determine who is continuing to use it and who are they missing. Data use cases would be data entrepreneurs, others who want to take this data and actually do something with it, ProPublica, companies. The second would be the researchers because we know that is bread and butter for many of these data sets. Third would be communities. It is either communities or community organizations. We have not quite settled on that. And then last because I think it is an interest of something the secretary has promoted is consumers. I get a diagnosis. I want to find out should I or shouldn’t I have surgery. Is there any place I can go to and look up what happens in terms of African American women who have radically mastectomies? What is the outcome or something like that?
One of the ways to attempt to help the data owner is for them to be able to collect analytics in the background to say who breaks off. If you do a little survey – we have all had that. There are usually two or three questions. How easy is this used, et cetera.
You can then determine whose needs you are meeting and whose needs you are not meeting. And then maybe one of the things to do over time is for those agencies that see that they are not meeting certain categories to figure out how to improve. Again, all this is something we – you have to see how onerous would it be to do these things.
DR. RIPPEN: The team actually had some really great use cases and ideas about, for example, if it was for a consumer or if it was a community. What are some of the ways you could actually share that information in a way that could be understandable. And the use cases are important because they can help you go back to the data elements to decide are you robust enough. You can always have a way to interpret it for different user groups. What you would present might be different. But the core data elements that really describe at a higher level the data is the big first nut to crack. And then the use case just allow you to make sure that you are not missing anything and that maybe you were not biased too much for one user group or another.
What this level will not do is actually address the user interface issues and how data is presented and download it to push or pull. But rather the first cut with regards to – one of the biggest challenges is trying to find is there data for me that might be relevant. It is not taking what a researcher needs which is the detailed thousand page kind of description that they actually have to go through to make a determination sometimes if they can use for certain things. Others they may not need quite as much the detailed data description.
DR. COHEN: How can they evaluate whether this data – how can they use this to evaluate whether this data is what they need or they need someplace else?
DR. RIPPEN: And again this is an important thing if this is valuable or not. If I just need to know who has what kind of data. Let’s say I am interested in diabetes, some data about diabetes. I want to know what is available and the limitations and how that data was made. If I can quickly peruse to make a determination of might this work or not that would be the first cut. I always have to then do more analysis to make a determination of whether or not really it is perfect for me. I would say this is your first cut and perhaps something that you could use as an example though, I am scared to use it, would be I need to do the abstract when you do a MEDLINE search. You read it. Does it seem like it is a good match or not? Hopefully a little bit more specific than a MEDLINE search because a lot of times they will not tell you what the results are nowadays and the methodology sometimes either. Enough to give you that is reasonable. Maybe the 80/20 rule.
I think if you are asking to provide everything, you will wipe out half of the users. This is more can I find something relatively quick. All of us that actually consume data know that it takes a lot more work no matter what to actually get to the next stage unless you are a distributor.
DR. ROSENTHAL: I think we are thinking along the lines of next step is to create use cases. For me, geography is huge. Is it a county zip, HSA. What is the thing actually at basically. That is one of my questions going in is this for research. I think use cases, greater detail, and extending and expanding. Before we did too much work, we want to say does this conceptually make sense. This is a different way of looking at it. They are not necessarily standards or technical format, but what are the pieces that need to be in place? How come only a couple sets achieved having a 100 million people look at them and the others did not. Why is that? When you start to drill into it, it is because they had these elements. That is the kind of question we are trying to drive at.
DR. MAYS: And it also might be that the data goes to information, meaning that you have a consumer. What you are able to tell the consumer is if you go here, there is a bunch of policy briefs that use this data. If you go here, here is where there are infographics. Part of also thinking about this is thinking about some of the alternative ways that a data owner may want to put their data up. It is like think of something like ICPSR at Michigan where the data sets are there and they have places where the results of what has been done with this data set is there and a common person goes and then they can just actually find, not the data, but the results or the information that has come from the data. It is also thinking that way.
DR. ROSENTHAL: Tracking some of that. Right now, we do it anecdotally. We organized Health Datapalooza and we had 1,500 people.
DR. COHEN: — analogy and that would really make the data a lot more useful.
DR. ROSENTHAL: Exactly. We know it anecdotally right now.
DR. CHANDERRAJ: Something to add on to Bruce is that I know it is a lot of development and infrastructure that needs to go into, but I think to add a comment on the validity of the data would be good enough for the consumer. And also at the end of it, a survey, if the consumer has looked at the data and has pursued the data and ended up with a positive or negative result, that would be useful to know.
DR. ROSENTHAL: That is kind of in that final bucket when we start moving up from data into information architecture and then that kind of sits at UI/UX level. That is the fifth bucket down there basically saying did a consumer look at it. Did they find it was useful? In different systems, sometimes you can do it automatically through stars and likes and things like that or you can do it through a survey or through a self-identification. Usually outside health care, there is a feedback mechanism that tends not to be at the data level, it tends to be at the distribution level. But then the results of those are sometimes taken down back through the data level.
DR. RIPPEN: And actually some of the members of the working group brought up some of the nuances for different types of users as far as how would someone know that it would be applicable to their use case versus not. Again, going back to what is the intent of the data, why was it collected. Sometimes we see data and we say how great, but then we did not realize it was in this context. You cannot actually apply it. That is a future opportunity because it is one thing to put out data so that people can convert it into knowledge and action. You have an opportunity to maybe educate people a little bit too on how to be effective users of information and data. You have different opportunities.
I think the big question for us is is there the 820 rule that actually can characterize data in sufficient detail so that people can actually make a pretty informed decision before they invest a lot of effort or apply it to things that they really shouldn’t.
DR. ROSENTHAL: And some of that just to use the flip side of that is here is the intended use of the data. The BRFSS example is so great. Right now, there are payers and providers literally deciding do I want to go into an exchange market. Do I want to go into a geography? How do I write a product? They are doing that based on BRFSS, which was never intended for that. It just so happens to work really nicely. There is a push to get out more data and that is great, but lots of data we have is also incredibly applicable to different things that they are not keeping track of and that is very valuable too.
DR. RIPPEN: And actually that is the evolution of knowledge. To your great point, how do you actually do that feedback loop? I think those things will actually accelerate the ability to leverage information.
MS. LOVE: I just want to say good work. We did a project for the California Health Care Foundation for community county supervisors. We came up with very similar criteria here. I think great minds think alike. We did put cost and linkage variables. The cost of the data because some of the counties did not have budgets for $8000 Medicare extracts. Geography was critical, cost and format because a PDF file is standard report. We just filtered that all out for them. There were a few others, but linkage variables and cost. I can share that report.
Then what we did is we catalogued them. Then we actually went in and did an evaluation. This gets into domains. And some domains could not be populated. But then we did use case. But use cases of how were these data used in a policy decision or a budget or a nursing home planning event in a county. You start building into the data the utility for a particular audience.
DR. RIPPEN: The nuance for us because if we are saying that maybe recommending this as a best practice for data providers from at least within HHS. I guess I would say what would be reasonable because, again, it is a level of effort. And then also what can be done versus what can be done by the user community feedback and without the bias because sometimes people are going to have bad experiences.
MS. LOVE: I do not think the use case upfront putting into that makes sense, but it would make sense if you want your data to be used to have somebody later plug some creative uses of it just to give an idea.
DR. ROSENTHAL: Typically, when you see this in other domains, the user community does that pretty effectively. If I am in a well-functioning ecosystem, the data producers are connected into that as well.
MS. LOVE: The whole thing that the California – county supervisors and a county in California does not have the budget or the time to metadata analysis – we built this little database saying these are the domains. These are the databases. Do not waste your time Googling.
DR. RIPPEN: And actually building on that and also some of the work of the population group. It would be wonderful if we could go beyond HHS and actually – I guess I would ask everyone to also think about how can the information be used to actually help us identify where their gaps might be or opportunities.
MS. LOVE: We will send you an analysis. There were some domains that just could not be populated. We categorized as severely inadequate. Data at a geo level and that even measures were not available.
DR. RIPPEN: I think Vickie had it as one of the aims with regards to being able to use it from a broader perspective too as we think about determinants of health. If it is very effective then obviously everybody can use it. Now we have a consistent way of thinking about how do we summarize important information and data so that at least it can be used or evaluated quickly.
MS. LOVE: It could be a census tract —
DR. SUAREZ: I think we are just about on time.
DR. MAYS: One more thing. I gave you two handouts from us. They are from the Robert Wood Johnson Foundation and where they are going in terms of their focus on data. Dave Ross, who is part of our group, is the co-chair of the group that they have. And part of what I really want people to look at is – on the one that says data for health – learning what works report, themes and recommendations, is if you look at the recommendations, what you are going to see is that many of those recommendations are somewhat in the space that we are in. They are in the space that pop is in. They are in the space that the work group is in.
One of the things that we are going to do is to actually sit down and talk with Dave since he is the co-chair, but with RWJ and see if we can bring them to the table with us. What we are doing is finding out how do we better implement what we come up with, where are the gaps because they have already had a listening session. They traveled to five different cities. Karen DeSalvo has been working with them. They have made recommendations as to what CDC and ONC should be doing. They are pretty far along in this. I do not think we should do this just to do this, but instead I think we should maximize the effort so that we build on what it is that they have already found and figure out where is the spot that we can best help HHS in terms of their data as opposed to just kind of being all over the place.
If you could look at that because in the September meeting, I am going to make sure that we are able to have more of a discussion.
DR. SUAREZ: Thank you so much. This is terrific work from the data access and user work group.
DR. MAYS: Thank you to Josh for coming in to the meeting today. We appreciate it.
MS. HINES: Before we go into public comment, just a quick note. We are going to be in federal room A tomorrow. Do take everything with you and the hotel will re-set us up in this other space. We walked passed it on our way here.
DR. SUAREZ: We are going to go to public comment. We are going to get public comment about any of the topics we covered today. I do want to acknowledge that some people have submitted comments via email and other messages to us specifically on some additional suggestions for corrections and slight modifications to the letters that we reviewed this morning from the Standards Subcommittee. We are going to open it now for public comments. I think we are limiting comments to about three minutes per person. Any public comments at this point?
MS. LOHSE: Gwen Lohse, CAQH CORE. Thanks everyone. I know you are having a very long day. I will try to stick to three minutes. I am reflecting comments that the CORE chair and the vice chair, one is from Christus Health System, which a huge health system in the Southeast, and the other one is the head of testing and implementation within Aetna. You have a health plan and you have a provider that are co-chairing. They represent a multi-stakeholder board with AMA representatives and others sit on and then the 140 participants.
We took that letter and they did and we have demonstrated some examples where we would suggest that you reconsider the draft letter on the operating rules. They were sent to Walter. I only had so many people’s emails. Walter, a few of you. I am going to read through those.
MS. GOSS: Let me ask something, Gwen. We did receive the letter. We did review the requests in that letter today. And then after that conversation that we had as a committee, we did talk and we gave you the opportunity to give us some very specific proposed language. I hope that you have also come with something in hand that we can look at since we need to turn this around.
MS. LOHSE: We did and it just arrived in hand. We have responded, but I do want to acknowledge that this was over a three-hour period per your request. We did the best job possible, but we obviously do not represent the board. We have reflected their comments and our comments. I do want to give the opportunity for our board to see if they have any other comments after we give you this. I am putting that out there.
MS. GOSS: I certainly understand that. We have also gotten some additional comments that we have tried to respond to and incorporate today. I do think that our goal is that those of us that are working on the letters plan to look at this tonight and incorporate what can be tomorrow before we bring it forward.
MS. LOHSE: Fantastic. Great.
PARTICIPANT: Are the comments highlighted?
MS. LOHSE: They are. Starting on page 2. Before you go into the details, they reflect around three main topics. One is either adhering to the statutory language as it is today either written by the Congress or the ACA or the existing federal regulations.
The second one is allowing a voluntary effort to move forward with voluntary work if there is an absence or a vacuum of a regulation like acknowledgement.
And the third one is just the awareness of the value that the 140 entities that worked on the rule saw. I know you heard comments from those that were in the hearing, but there were 140 other entities that did see value.
With this said, I do not know how you would like me to go through these comments.
DR. SUAREZ: We are going to take look at look at those. Thank you so much – editing and inserting some language that we can review. I think it will be the responsibility of the chairs of the subcommittee or staff and then presenting those tomorrow.
MS. GOSS: I would ask that the Standards Subcommittee members that were a part of this process look at this tonight and let me know and Terri because Terri and I are the ones who are going to the most of this – your perspective on that as soon as you can or maybe we can huddle first thing tomorrow morning and go over it so we can all do our independent thinking, bring it back collectively and see what —
MS. LOHSE: We did write our comments in your voice to help you with the process. When is the public comment period tomorrow because our chair happens to be in town and he may drop by?
PARTICIPANT: 1:50 p.m.
MS. LOHSE: Fantastic. Thank you.
DR. SUAREZ: Thank you. Any other public comments?
MS. MTUI: A quick message from Durwin Day via WebEx. He says the committee has done an outstanding job of addressing the attachment recommendations.
DR. SUAREZ: For those who do not know Durwin, he is part of the Blue Cross Blue Shield system. He worked for Blue Cross Blue Shield in Illinois. He is one of the co-chairs of HL7 attachments work group and also a member of X12. He is quite knowledgeable in understanding of the standards and worked on the development of the standards.
MS. GOSS: We also had a comment earlier from an individual. She was not able to stay, but she did meet with Terri. We went over her comment. She had a question where there was a conflict. I think we got that resolved. She did not feel the need to stay.
DR. SUAREZ: As I mentioned, we did receive a few more comments. I do not know that we need to mention the sources, but certainly they came from Mayo Clinic, from the X12 chair, from NCPDP, which we heard earlier during the discussion of the letter on attachments. We have received several of those comments and they are being reviewed and considered for incorporation into the letters by the chairs of the subcommittee.
MS. GOSS: We are thankful for those because they tend to be very –
DR. SUAREZ: Parson me, Maya?
MS. BERNSTEIN: The written comments that you have.
DR. SUAREZ: They are written comments. They are comments submitted in written form. As we review them tomorrow, I think we will just have a chance to mention them. We will go through the letter and the edits tomorrow.
MS. GOSS: What I wanted to note was that they were technical – we forgot an N behind X12 for the correct technical TR3 title. It was those kinds of corrections. Or in the case of Mayo, they wanted to make sure that how we represented it was just a slight tweak to make sure it was more in lined with how they wanted it represented.
MR. LANDEN: There was nothing in those that would require us who are going to be reading the core letter to have copies of those comments.
MS. GOSS: My take is that it is not. It is nothing substantial. It is qualitative for betterness.
DR. SUAREZ: Any more public comments? Hearing none, I think we are considering our first day of the meeting adjourned. Thank you so much.
(Whereupon, at 5:45 p.m., the meeting adjourned.)