[This Transcript is Unedited]
Department of Health and Human Services
National Center for Health Statistics
National Committee on Vital and Health Statistics
September 14, 2017
Hubert H. Humphrey
200 Independence Avenue
Table of Contents
- Call to Order – William Stead, MD, Chair
- Commission on Evidence-Based Policymaking – Report and Recommendations – Lucas Hitt, CEP, Sherry Glied, CEP
- HHS Leadership Updates
- Chris Jones, ASPE
- Charlie Rothwell, NCHS
- Elsie Anthony, Office of the National Coordinator for Health IT
- Mona Siddiqui, Office of the Chief Technology Officer
- Health Terminologies and Vocabularies – Part 2 – Linda Kloss, Chair, Subcommittee on Privacy, Confidentiality & Security, Donna Pickett, Chief, Classifications and Public Health Data Standards, NCHS
- NCVHS Workplan Formulation for the Remainder of 2017 and 2018
P R O C E E D I N G S (8:30 a.m.)
DR. STEAD: Colleagues, welcome back for the second day of the National Committee and the fourth day of the NCVHS marathon. So, if we will start with our roll call, I am Bill Stead, Vanderbilt University, chair of the National Committee. No conflicts.
MR. THORPE: Roland Thorpe, Johns Hopkins University, member of the Full Committee and member of the Population Health Subcommittee. No conflicts.
DR. COHEN: Bruce Cohen. Members of the Full Committee. Co-chair of the Population Health Subcommittee. Massachusetts. No conflicts.
DR. RIPPEN: Good morning. Helga Rippen. Alertgy. Member of the Full Committee. Member of the Security and Population Health Subcommittee. I have no conflicts.
MS. MONSON: Good morning. Jacki Monson, Sutter Health. Member of the Full Committee. Member of the Privacy Subcommittee. No conflicts.
MR. COUSSOULE: Nick Coussoule, BlueCross BlueShield of Tennessee. Member of the Full Committee. Co-chair of the Standards Subcommittee and member of the Privacy, Security, and Confidentiality Subcommittee. I have no conflicts.
MS. GOSS: Good morning. Alix Goss of Imprado. I am a member of the Full Committee; co-chair of the Standards Subcommittee. I have no conflicts.
DR. ROSS: Good morning. Dave Ross, Taskforce for Global Health and Emory University. Member of the Full Committee. Member of the Population Health Subcommittee. I have no conflicts.
MS. KLOSS: Linda Kloss. Member of the Full Committee. Co-chair of the Privacy, Confidentiality, and Security Subcommittee. Member of the Standards Subcommittee. No conflicts.
DR. CORNELIUS: Lee Cornelius. Member of the Full Committee, Population Health Subcommittee, University of Georgia. No conflicts.
MS. HINES: Good morning. I am Rebecca Hines. I am with CDC’s National Center for Health Statistics. I am Executive Secretary and DFO. I want to also welcome – on the phone, we have Sherry Glied this morning. Good morning, Sherry.
DR. GLIED: Good morning, Committee.
DR. STEAD: We also have Lucas Hitt.
MR. HITT: Good morning. Lucas Hitt from the Commission on Evidence-Based Policymaking.
DR. STEAD: Welcome all. I think we can go on and start and turn the floor over to Lucas and to Sherry. Thank you so much for being with us.
MS. HINES: Sherry, can you get closer to the phone? We are having a little trouble hearing you. We are putting your slides up now. I had sent you the link to the WebEx. If you wanted, you could see that Lucas has got the slides going.
MS. GLIED: Let me just start and when it comes, I will do it. Thank you very much or the opportunity to speak with you about our new – the Commission’s Final Report on the Promise of Evidence-Based Policymaking. It was released on September 7th. We are very excited to talk about what we did.
I know that – I always feel a little badly because I know that we didn’t do everything that everybody wanted. The final recommendations are a series of overwhelming agreement within our committee and would substantially, I think, improve how we produce and use evidence.
Let me give you a little bit of background on the process, on what we did, and what the Commission’s findings and recommendations were.
The Commission was created because of a lot of decisions that had been taking place around how we could increase availability and use of data while protecting people’s privacy and confidentiality. Speaker Paul Ryan And Senator Patty Murray created the legislation that created the Commission and charge us, really, with evaluating how we could meet that mission –
DR. STEAD: Sherry, we are really having trouble hearing. It is coming through as if it were in an echo chamber. We are just having trouble.
DR. GLIED: Give me one minute. Lucas can keep going. I will get back in.
(Dr. Glied calls back in)
DR. GLIED: Here we go. So, we were charged with trying to figure out how to improve access to data and also protect privacy and confidentiality. We were specifically asked whether we ought to have a clearinghouse for program and survey data. That was one of the questions that we considered in our deliberations quite a lot.
I would note that although the impetus for creating the Commission was around gaining access to evidence, 5 of the 15 commissioners were appointed because of their expertise with respect to privacy. That has been a really strong theme of the work of the Commission. You will see it in the results.
I think as a data person, as an economist, as an evidence-based person, it is really encouraging that the government decision makers in congress, in the executive branch, across both parties, wanted the Commission to look at the design and how we could use evidence to improve the design and administration of federal programs. I think the basic impetus for this is really great. This is interesting because often evidence can make the work of policymakers more difficult. They may discover that a program that they really like doesn’t work as well as they would like it to be. It is, I think, important that we think about how to use evidence to improve learning and improve the administration and design of our programs in government.
The Commission held hearings beginning in the summer of 2016. We had an eight month long fact-finding process with a lot of feedback, including Charlie Roswell came from NCHS to talk and lots and lots of folks we spoke with. We had 7 public meetings with 49 invited witnesses. There was a survey of federal offices to get a sense of what people in the federal government felt was their – were their needs with respect to generating and using evidence. I think, actually, that was a really important part of the process, getting a sense of the heterogeneity across the government, both in the processes of collecting evidence and in the processes of using them.
After that, we had a long deliberative process. Lots of discussion, sometimes heated, about where we wanted to go. We listened to the evidence that we had heard and the comments that we had received. After all of the input, we came up with 22 recommendations. I want to say – and Lucas is probably going to shoot me for saying this – we started with 51. We were basically told that no one would read 51 recommendations. We had to get it down to 22. I would encourage all of you to look at the entire Commission report. There is a lot of material in there that isn’t specifically a recommendation, but that provides some guidance about how things might be done within and across government that I think is also quite useful.
Let’s go to the report and its major themes. The final report contains the findings and the recommendations. They really show up in three areas. The first one is improving access for approved evidence-building projects. Helping get access to data to researchers so that they can actually figure out the evidence to support programs. Second, maximizing and strengthening privacy protections for the data that are used in evidence-building. What can we do to make those data more private, more confidential? Third, and I think very importantly from my own perspective, strengthening the federal government’s ability – capacity to use that evidence – to use those data for evidence-building. I think that capacity is very strong in some parts of the government, but weaker in others. Again, a large theme of the report is to equalize the playing field at a high level across government so that protections for privacy and confidentiality, the use of evidence-building, the use of evaluation are more broadly used and at a very high level throughout government.
As you will notice when you read the report, many of the recommendations, especially in the early chapters, address issues of data access and privacy. These are pretty critical. One of the big questions, as you will recall, is whether or not we need a clearinghouse. The Commission was asked to decide whether there ought to be one. In the end, we decided that there shouldn’t be a single clearinghouse – a warehouse of data – that would have all government data in one place. There are places that do that. It is in some respects, an appealing idea, but we felt that it would be – it was not necessary. That it would raise more problems than it might solve. Instead, what we want to have is a national secure data service – a service that would facilitate access for agencies as well as non-government researchers, but that would not be a clearinghouse that contained all of the data.
The Commission also recognized the importance of having capacity within the federal agencies and making sure that statistical units across the government and their components are well coordinated.
Another piece of this that was very important and that we spent a lot of time discussing was use of administrative data. We – there is a very strong cultural protecting confidentiality within the statistical data services. That is what they have been charged with throughout their existence. We are increasingly able to use administrative data that can supplement the use of statistical data. It actually provides a lot of information on its own. Unfortunately, or fortunately or whatever, just as a matter of history, the agencies that collect administrative data don’t have the same history and culture of thinking about data confidentiality and protection and also about access to data. A lot of what the report is focusing on is how to improve access to and the privacy of administrative data and then trying to put all of that together to facilitate evidence-building.
Let me go through this particular aspect of this. Now, I want to talk about recommendations for data access. A lot of the people on the Commission were really focused on how we could make data more accessible to researchers. A lot of the comments we got were about that. There are a lot of barriers for people outside of government and I think, even within government, in getting access to the data.
The Commission recommends the establishment of a National Secure Data Service to facilitate access to data for evidence building while ensuring privacy and transparency. I think transparency is another idea here that there isn’t enough of yet in government, in how the data are used. Basically, the service would be created by bringing together and building on the existing resources that exist within government. It would facilitate secure access to government data by qualified researchers inside and outside of government and especially to datasets that are created by linking information from multiple agencies. So, there is great potential in building evidence using linked data, using data that link administrative data and survey data or different administrative datasets or different survey datasets. How are we going to do that well in an atmosphere of privacy and confidentiality?
So, we recommend requiring very stringent privacy qualifications for acquiring and combining data for statistical purposes at the NSDS, as we call it. Also, using this service as a locus within the government to develop and implement state of the art methods for combining data for evidence building and for providing cutting edge privacy protection and technology for data access. A piece of this idea is really to bring the best minds, the statistical minds around the government, the best practices around the government, together, to really establish the state of the art in data protection and data access.
This will also require reviewing and revising laws authorizing Federal data collection to ensure that limited access to administrative and survey data are possible under strict privacy controls. The idea here is to actually make data more available. At the same time, make sure that that data are treated with as much privacy protection as we possibly can.
Finally, and I think very importantly to external researchers, trying to develop a uniform process that would work across the government for researchers to apply and qualify for secure access to confidential government data for evidence-building purposes. One of the things that we heard a lot from researchers is that it is very confusing. Some places have well-established processes for getting access to the data. Other places really don’t. How can we build a process that allows people to have a streamlined way and a mature way that is consistent across government for getting access to private data? Particularly, that is a problem when people want to link data and the requirements of different agencies might exist.
Recommendations for modernizing privacy protections. So, you will notice, if you read the report, that there was a lot of mention of privacy. Somebody actually went through the report and discovered that we use the word privacy 408 times. It is not that long of a report. So, privacy was a real concern here.
One of the things that I think you will all be aware with the Equifax thing last week is that privacy is an evolving and continually challenging area. Our protections for privacy need to keep up with that. We need to be thinking about privacy in kind of a dynamic way.
One of the places that this shows up that I think NCHS has actually – again, it is a question of trying to coordinate and consolidate efforts across government – is thinking about administrative data and how those data are released, how de-identified data are allowed for private release – sorry, for public release, to make sure that as data are merged or combined, we don’t unwittingly expose people to breaches of privacy and confidentiality. We need to be constantly evolving our practices to make sure that we are taking advantage of emerging technology and that we are aware of new emerging threats to privacy and confidentiality.
The report lists four recommendations for focusing on this. Requiring comprehensive risk assessment on de-identified confidential data. The focus here is on de-identified data because of the concern that even de-identified data can eventually be traced back to respondents if it is not appropriately – if the de-identification process is not appropriate, is not sufficient. Second, adopting modern privacy-enhancing technologies for confidential data so that we constantly improve the way that we keep data private. Third – I think this is really important – assigning senior officials the responsibility for stewarding data within government departments. Making the data part of the responsibility of the senior officials in a more explicit way. Finally, codifying policies for maintaining the integrity and objectivity in Federal statistics to promote continued trust in the accuracy of the information being used. Again, a concern that people really – not only respondents of surveys, but users of them, are confident in the validity of the data.
Not too many slides left. We should get there soon.
We also made some recommendations for using the data, for strengthening the Federal capacity to make use of the evidence once it is produced. It is not enough to produce evidence. People need to be able to carry out policies that use that evidence and make use of it better. The report presents five recommendations related to increasing capacity.
The first one, which sort of already exists at HHS, is to establish a role of a Chief Evaluation Officer, who would be responsible for leading efforts to coordinate departments’ overall evidence building activities in partnership with the offices that may be developing statistics or administering programs or conducting policy analyses. Have somebody in charge of the evaluation process within each agency.
Closely related to that, develop learning agendas. Several of the departments already have learning agendas. The goal here is for the departments to identify where the need for evidence is greatest and then position their research dollars to pursue those priorities. Help researchers outside of government understand what the really key questions are and pursue and multiyear agenda to try to get there.
Third, improve coordination of government-wide evidence building. That is try to diminish the fragmentation of evidence-building across multiple offices so that it supports evidence-building in the way it should. This will also help to facilitate coordination across OMB, rather than the other way around. Quite a number of recommendations here are about streamlining processes that involve the departments interacting with OMB.
Fourth, align administrative processes with evidence-building activities, opportunities related to procurement and information collection that might make it more streamlined. This is very sort of inside government bureaucracy talk, but there are many things that make it hard for evidence building to take place within the government. There are simple processes that could be put in place to make them easier.
Finally, ensure that sufficient resources are available to support evidence building. We actually talk about how it is not enough to say evidence building is important. You have to put your money where your mouth is. We talk about a number of funding flexibilities, such as the potential for multiyear funding, funding set aside, transfer authorities, and the creation of what the Commission refers to as an evidence-incentive fund to recapture unobligated balances.
Next, we have some recommendations related to state-collected data. There are many concerns about not having access to the state data on programs that are Federally funded, but state administered. These were actually a lot of tough challenges. The Commissioners had a lot of discussion about what is the right balance to strike as it relates to increasing access to state-collected administrative data, such as earnings, TANF, SNAP, vital records, and a whole host of other administrative data sources.
Ultimately, we included two recommendations related to state data. It focuses entirely on state-collected data on quarterly earnings. It turns out that one of the most important sources of data that is very hard for researchers to access and could really be very valuable is information on earned income. There are a bunch of ways to access that data now. We talk a lot in the report about ways that we could – strengths and weaknesses of those and trying to think about how to improve access to quarterly earnings data. We don’t actually list a specific approach, but we talk about different ways it could be done.
The second recommendation related to state data essentially lumps together all other state-collected administrative data that would be beneficial for evidence building, noting that where there is a substantial Federal investment in a program, Federal departments should, consistent with the law, direct states to provide the data necessary to support evidence building. It is a broad kind of a recommendation.
When we began to look individually at each individual administrative dataset source collected by the states that might be valuable, it is pretty clear that there is no one-size-fits-all solution. Again, in this recommendation, we frame the recommendation broadly and focus on the outcome that we are seeking and ask the Federal departments to think – effectively, recognizing the role that support from the Federal government can play in enabling states to comply with the reporting requirement. The report suggests the Federal departments explore and adopt incentives to support either compliance with the requirement for states to make data available or to incentivize the voluntary action of making administrative data available. I should note here that the experience of NCVHS was actually very important in helping us to think through this issue.
Next is our NSDS, the National Secure Data Service. The idea is to really bring in a new actor in the world of evidence building. The Commission recommends that it should be a principle statistical agency, so that it has an exclusively statistical mission. In contrast to the existing PSAs, it would have a role that is functional rather than topical. While technical and analytic services would not be an initial priority, a basic level of services like that might be necessary to encourage collaboration and build capacity for evidence building. We might like – as happens in other PSAs, allow analysts at state, local, and tribal governments and external researchers to qualify as approved researchers and submit statistical research projects for approval to the NSDS.
To implement the NSDS, the Commission recommends building on the infrastructure and expertise already developed in government, including at the Census Bureau and at various other places, especially within the Commerce Department, requiring public input, guidance, and participation in the policies and procedures for NSDS activities through public and stakeholder representation on a steering committee. Establishing a new transparency and accountability portal so people will know how the data are actually being used, which is something that doesn’t always happen now. Ensuring the government has administrative flexibility to be able to innovate continuously on privacy and protected data access. And increasing efforts to make information available about the government’s current data inventories and related data documentation. So, there is a big section here on just creating an inventory of what information exists in the government and making that readily accessible to researchers.
So, with all of that, as you can see, we have released our report. We had a discussion of the recommendations. We have gotten a lot of very positive feedback. The report lays out the Commission’s vision for better use of data for evidence building across government. We understand that much of what we recommend will not happen right away. Implementations remain to be worked out. We think that if the Commission’s recommendations are taken up by congress and the president, a lot of good will happen. We very much appreciate the contributions of NCVHS and other agencies in the government to our process. Let me stop there and take your questions.
DR. STEAD: Sherry, thank you very much, both for the work the Commission has done and for your insightful comments. We will use our normal process. I will let Bruce start the questions.
DR. COHEN: Hi, Sherry. This is Bruce Cohen. Thanks. This is right in the wheelhouse of issues that the Population Health Subcommittee has been investigating and might be investigating in the future. This is very timely for, I think, the National Committee to consider as we move forward with our work plan. We look forward to working with you as we do so. That is great.
I had a couple quick questions. One has to do with vitals. The last Monday and Tuesday, we had extensive hearings around vital statistics. I was wondering, first, if you could provide more detail about your discussion related to vital statistics and creating – incentivizing or doing what to make sure that folks have more access to vitals.
The second question had to do with the data service. Will the data service be for analysis only? Will the secure data service be a gatekeeper for access to individual level, de-identified data that researchers might want to take away and use in their own space?
MS. GLIED: Let me take those questions in order. I want to be clear here, we did not have any specific recommendations around vitals. One of the things that we really learned from NCVHS is that it wouldn’t be enough to just tell states they should hand over data. I don’t want to make my colleagues seem naïve, but if you haven’t ever worked with state data authorities, you might imagine that just telling them that they have to hand over all of their data would be sufficient to actually make it happen. I think we were disabused of that notion by the tremendous efforts and the incentives that vital statistics have developed over the years to try and get the data. Actually, I think it was more of a model than something that we had particular recommendations around.
On the second question, the idea is that data may – would leave the NSDS. People wouldn’t – I don’t think that we envisioned it as being some sort of hyper-office, but that it would be either de-identified or accessible in research data centers, accessible in ways that maintain the privacy and confidentiality of the data to researchers who had been certified as having appropriate protections in place. Much like the current system that exists within NCHS, for example, which has its research data centers and allows different kinds of data to be released, the idea is not that it would be the primary place that de-identified data would be released throughout the federal government. That already happens through open data initiatives and so on. Rather that it would help develop methods for ensuring, for example, that those data are appropriately de-identified.
DR. COHEN: I am not clear. I didn’t quite hear you. Is it sort of an expansion of the RDC concept? Is it a broker to release individual data to researchers to take home and use in their secure environments?
MS. GLIED: It might be either of those. We aren’t very clear on exactly how researchers might get access to the data. I think that would be a function of the particular characteristics of the researchers and the environment at that moment. It will be doing the linkages, for example, so that by the time a researcher would get a dataset, they wouldn’t need to be – the data would have been de-identified to the extent possible after the linkage had been made. Does that help?
DR. COHEN: That is great. Thanks a lot. Great work.
MR. HITT: If I may add a little bit to Sherry’s explanation, one of the aspects the Commission recommends is sort of a tiered access model, depending on the sensitivity level of data. So, to Sherry’s point, NSDS may be providing access in a secure enclave. It may be providing it under a licensing agreement. The tools that access may be provided through would be dependent on the sensitivity level of the data in question.
DR. COHEN: Thanks.
DR. THORPE: Thank you for an informative presentation. I have a couple of questions. First – the first question is how is this – how are these recommendations with linkages different than what we already have in place. I guess I may have missed it in the presentation. I really didn’t see that. Could you elaborate on that?
The other thing is – well, let’s go with the first question.
MS. GLIED: What is the difference from what takes place right now? So, right now, imagine that somebody wanted to link – wanted to link data on mortality and income tax return. This is an enormous project with no obvious way of being undertaken beyond trying to get all of the people who own all of those different data resources to be willing to participate, to work with the tax people – with the IRS because those data are extremely sensitive. It is very challenging when these things spill across departments to make this happen right now. The idea here is that particularly for projects that spill across departments and need linked data, the NSDS would facilitate that kind of linkage. Does that help?
DR. THORPE: Yes. Kind of yes and no.
MS. GLIED: Do you want to amplify and I will try and answer it better?
DR. THORPE: Are you suggesting that what we have in place – the proposal that you have or the recommendation you have – I’m sorry – will be a clearinghouse place where all of this can be done in one place, where all of the players come together and agree to play together and make it easier for the researchers or the consumers of the data – I won’t use just researchers – for the consumers of the data? Is that what I am to understand?
I am really having a hard time trying to understand why we need this with what we have in place now.
MS. GLIED: What we have in place now is that different agencies release their own data. If you want data on vitals, you go to NCHS and they have data. You have to deal with whatever confidentiality requirements they have and get access to data. That is a pretty good system that we intend to maintain. The problems that researchers identified particularly were those that involved linking data across two federal departments or between a survey in one department and administrative data in another department. The notion here – one important function of the National Statistical Data Service – not the only function – would be to make those linkages. To go to commerce and get data from commerce. Go to HHS and get data from HHS. Put the data together into a combined dataset and release that combined dataset to a particular researcher, who has been approved, in a tiered access environment. That is actually very hard to do right now. That would be considerably facilitated by this.
Other functions of this would be to be a government-wide resource for improving privacy protections across all of the PSAs, across all data release, and to homogenize top processes that are used across government so people who come in and are not working exclusively with one department will find it easier to move among them. Did that help? Is that better?
DR. THORPE: That is a little bit better. The next question is how is the fee structure going to be determined? You have a lot of different players in here. Everybody has their own – there is a lot of different heterogeneity about how the fees – how you access fees now and to whom get access.
MS. GLIED: We did not actually get into the details of the fee structure. That has got to be something that is worked out by the committee – the steering committee that is developing the service. We didn’t propose a particular fee structure.
DR. THORPE: My final question is, so, did you all talk about once you have this resource, are you going to have remote resource data centers? I know Census has around the country where they have these remote resource data centers whereby the close proximity for researchers to go in and conduct their analyses or whatever they are trying to do. Was there any conversation around that?
MS. GLIED: Yes. So, this is the tiered access point that I was making earlier. The idea here is that the NSDS would take – for any particular dataset that had been developed and linked, it would essentially assign it a level of security kind of tier. Some kinds of data, if they are sufficiently deidentified and the researchers is sufficiently authorized – so, say a university researcher who is able to operate a secure environment within their office might be able to access certain kinds of data right on her desktop. In other circumstances where the data are more risky, you would have to go to a research data center, either a Census research data center, ones that are geographically all over the country now, or another Federal research data center, depending on what the project was, to access the data. It would really depend on the level of privacy required for that dataset and the nature of the person who was getting access to it.
DR. THORPE: Thank you.
MS. KLOSS: Sherry, thank you. This is Linda Kloss and I co-chair the Privacy, Confidentiality, and Security Subcommittee of NCVHS. I really appreciate the emphasis of the report on all things privacy.
A couple of questions. Did you envision that – there is a lot of reference to ensuring that state of the art tools are used and risk assessment is used in de-identification and other mechanisms. Did you envision that this center could actually be a way of actually helping to advance those methods?
MS. GLIED: Very much so.
MS. KLOSS: Could you comment on that some? Our subcommittee just did a review of de-identification practices under HIPAA. The whole ability to do risk assessment surfaced as a very big deficit. There is a lot of talk about methods for privacy, but there is certainly not in widespread use in private and probably not also in government. Some comment.
MS. GLIED: Our conversation completely mirrored that kind of conversation. In fact, there is a great need for these protections. There is not enough research on it. Particularly, there is a lot of conversation, actually, interestingly, about to what extent can we learn from the private sector and so on and then some discussion of how the needs, particularly of government data, are actually sometimes quite different from those of private data. There needed to be the advancement of methods for ensuring privacy and for developing the de-identification processes that allowed release of data. That also seems very important for people to be able to understand what their government is doing. And that that was a priority.
We are a little bit hamstrung by a limit on how much we could ask for funding. We really were not supposed to be doing – asking for much money. You will see in the report references to the idea that this is an area that we do believe the government needs to develop and needs to be developing on a continuing basis because this is something where – you know, hackers develop new methods every day. We have to be developing those methods, too.
MS. KLOSS: It just seems like, historically, government has been kind of a laboratory for advancing some of these techniques.
MS. GLIED: Yes. We were very much – that is something that we discussed.
DR. MAYS: Good morning, Sherry. This is Vickie Mays. Thank you for everything you have done on this committee. We know you have been great at looking at a lot of the issues that we are interested in.
I have a couple of questions. The first is, in terms of what goes on in this center, one of the things that sometimes we worry about is making sure that small populations or specific populations are going to be able to have a fair share at getting analyzed. Often, we have suppression mechanisms. I am wondering if there is – we have all been asking for kind of the state of the art analytic approaches, which would allow us to do the small populations. One of the things that I know is that Census is sometimes good at doing this. Is there any resources that would be connected with this Center to assist with making sure that we are able to do these – usually, the small populations are racial and ethnic minorities of specific subgroups. Is there anything that is going to be able to push the envelope to ensure that those populations will be able to get their fair share at people analyzing the data for them?
MS. GLIED: I want Lucas to jump in but let me give you a couple of ideas. You can fill in if I am misspeaking.
We didn’t, I think, spend much time addressing this very particular question. I think there are two aspects of the report that actually do have some positive implications for this problem. One is this whole notion of tiered data access and privacy and confidentiality and access as being something where the standards need to differ depending on the characteristics of the project, the researcher, the environment, and so on. It is very contextual.
I think, actually, it offers opportunities to look at small populations in ways that haven’t been authorized before in circumstances that – how should I say this? You would be in a different – you would have to be a different kind of a person in a different kind of environment to deal with certain kinds of data for small populations than you would have to be to deal with large populations. We envision that all kinds of analysis could be done and that the particular claiming would depend on that specific context. It is a little bit – like a confidentiality process, it is a little more bespoke, a little more tailored to the particulars of the analysis. I think that would definitely help on this.
The second piece that I think is really important is the big emphasis on linking survey and administrative data. Administrative data really helps get beyond some of those small population problems. The better we can do with linking the survey and administrative data, I think, the better we can serve analyses of those populations.
DR. MAYS: I love those interpretations. My second question has to do with the committee that put this together was kind of on both sides. It was nonpartisan. The question will be, as we go forward, and there are questions that people want to use data to answer, is this a situation where any questions can really be answered? Is this really geared towards certain kinds of questions and not others?
MS. GLIED: First of all, I would say that although this was, in principle, a commission that was bipartisan, in fact, I would say it was not partisan at all. It was very hard to tell during the meetings who was coming from where. I think the questions of data and evidence building are not partisan issues. There was some concern about what questions might be asked. This is really a studiously neutral report on that. The idea is that – I think that there is an element of it. What questions would be asked would be – these learning agendas, for example, will surely reflect the political priorities of the departments. That is inevitable. That is always the case. The NSDS, itself, is intended to evaluate projects simply on the basis of privacy, transparency, confidentiality, and so on, and not on the basis of partisan politics.
It is a political system. That is inevitable. I don’t think there is anything in here that – in fact, I think, on the contrary, it is really about trying to make the data more accessible for everyone. I think very important, in that respect, the transparency elements of the report that really are about letting the public know what the data are being used for could be a very important tool in ensuring that data are not being used in a form that is not scientifically valid or that is partisan in certain kinds of ways.
We also, I should say, in a quiet way, made a plea for making sure that all kinds of data are available. You will know that there are kinds of data that the federal government has been, over time, reluctant to collect or to make available. This is really about set the data free to the extent that you can do that with privacy and confidentiality.
DR. MAYS: Great. Thank you very much.
DR. STEAD: I believe the order of play is Denise, Alix, Nick, and the Roland.
MR. HITT: May I just add additional detail that I think will help address your question, as well as Dr. Thorpe’s question. One of the things we don’t really dig into in the presentation, but it is very much embedded in the report, is when we talk about NSDS and creating NSDS and so forth, there is an underlying legal framework that goes along with that. As the Commission went through its work, one of the things it identified is a lack of clarity and consistency in many of the statutes that govern access – the collection and access to data and the protection of data.
I would categorize it as – there are some statutes that are very specific. Generally, what we would look for is specificity as to the who can access the data and what the data can be used for. There are statutes that are very specific. There are statutes that may speak to either the who or the what, but not both. Then, quite frankly, many of the statutes just don’t address it directly at all. In those cases, it tends to – it kind of falls to agency councils to determine how to interpret that. Most agency councils tend to interpret that more on the conservative side. If they don’t see an exact authority, they will interpret that as not having that authority.
We also encountered a number of examples where one agency’s council would interpret it one way, another agency’s council would interpret it a different way. Ultimately, that creates a roadblock. Within the Commission’s work, NSDS is envisioned as having a very broad authority to acquire data. We refer to this as sort of a yes and less. The default position in statutes should be that data is available for statistical purposes, unless there is an explicit prohibition in law that is put in place by the Congress. The Commission then goes further to say once this new legal and privacy framework is set up, congress should reconsider any of those exemptions, given the fact that some of the privacy issues have been resolved.
To your question about access – many of the issues that might – particularly when you are talking about suppression and other reasons why things need to be protected, it goes back to the confidentiality requirements in a statute that are either blatant or implied. The ability for NSDS – again, to the question previously about sort of linking and so forth, right now, part of the challenge is that one agency’s data is protected by its statute. The other agency – maybe, you know, you are trying to connect something from NCHS and from Census. You have two different statutes that don’t have the same sort of who and what. Negotiating that reconciliation can be very challenging and in some cases, not possible. However, if both agencies were able to acquire administrative data, protect and disclose it under a consistent statute, it now means they are operating on the same plane and can make choices in a more standard way.
MR. COUSSOULE: That is exactly where I was going. Just to follow onto that just a little bit, the NSDS, itself, wouldn’t necessarily have the authority, but you envision more some statutory changes being made that would effectively give it that authority to be able to harmonize those rules.
MR. HITT: Yes. Very specifically, the way the Commission presents this is there is an existing statute called the Confidential Information Protection and Statistical Efficiency Act, CIPSEA. It is a statute that currently applies to the 13 principle statistical agencies and then three other agencies. It is essentially the sort of gold standard for the acquisition, protection, and disclosure of data that has been acquired for statistical purposes. Ultimately, what the Commission is suggesting is that NSDS should be created. It should be given the authorities vested in CIPSEA. There should be a series of other statutory changes to ensure that not only does NSDS have the authority to acquire that data, but other agencies have the authority to give it. What we found is usually you need – in theory, you could have just one of those two parts of the equation. The best way to ensure there is not a hiccup is if both sides of the equation have that authority.
MS. LOVE: I really have two questions. I should introduce myself. I was late. I apologize. Denise Love. National Association of Health Data Organizations. Member of the Full Committee and Subcommittees Population Health and Standards.
My first question is really simple. This does not duplicate the linkage through CARRA through the Census because we are having some state pilots right now where they are doing some fantastic linkages with disability, household, foreclosure, the SES data, IRS tax, with some of the public health databases in a couple of states, including hospital discharge, including vital records and some of the public state data resources. So, they will link them in Census and give it back to the states. The MOUs are being negotiated right now. Would that duplicate anything?
MS. GLIED: No. I think CARRA was kind of a model for what we were thinking about here as something that – take that same idea and goes beyond it, makes it more general.
MS. LOVE: So, just so we don’t duplicate it.
MR. HITT: Not duplicate, but expand upon and improve. One of the great limitation of CARRA is it operates under a Title XIII authority, which means the who and the what are very specifically limited. The who basically has to be a sworn Census agent. The what is, in theory, to improve the Census mission. Admittedly, they are as creative as they can be to find ways to make projects fit that – those requirements. Those are a good example of the existing limitations on the who and the what that the Commission would like to see broadened to statistical actors and statistical purposes.
MS. LOVE: I got the impression they are really expanding their focus. I just would avoid duplication. In my world of state data, we have huge data gaps that I don’t know how to fix. We have OPM data, FEHB claims data that is sort of locked up by the payers. We have the VA data that every state wants and needs. It varies how states work out those arrangements. We have IHS and tribal data that states really can’t put into their mainstream utilization datasets. I could go on. I won’t even start with part two.
Could this center be a source for some of these huge and glaring data gaps that states could draw down appropriately with the appropriate linkages? I don’t think we are going to convince payors. Even though OPM has no problem with it, but payors are worried about liability, et cetera, et cetera. The Federal agencies could play a huge role in filling these data gaps appropriately. Would this model be something to look at?
MS. GLIED: I think there are two elements of this. On the one hand, I think, to the extent that – this is about filling data gaps for analytic purposes. I think that is very much within the scope of what NSDS would do. If there are particular analytic research projects, evaluation projects to assess a particular program at a point in time, that would be appropriate. I think it can’t be see, I think, as kind of an ongoing source of programmatic data that would be used, for example, to alter payment rates or something very specific. It is intended for use – for statistical purposes – that is for research and evaluation purposes that don’t target individual actors.
I think it goes part of the way, but not, perhaps, as far as a state might want, for example, in the health arena, where you would want to know which hospitals are misbehaving across all of their billings. That would be a difficult thing to envision.
MS. LOVE: I think it could be very helpful with denominators and some global data. I don’t think – it is just that these are huge gaps in the state data systems. It could conceivably be a statistical source of these data so at least they could know what the gaps are. They don’t even know –
MS. GLIED: Remember the data will not reside there. There would have to be a project that involved getting everybody to agree – now, also, the authority of this thing will not stretch into private and state records. I don’t want to make it sound too – I am a little bit nervous about sounding too optimistic about the potential for bringing private and state records in there because there is not authority.
MR. HITT: As Sherry mentioned earlier, the Commission was very – it was informed by some of the experiences, including around vital statistics with respect to sort of the federal/state partnerships such as they are. I think the Commission was thinking very much about trying to create sort of an incentive model concept here where NSDS could provide resources – analytic and data resources to states in return for states providing analytic resources – data resources to NSDS. Of course, your list of here are the things that the states don’t have, there is sort of an equal list of state level data that the Federal government doesn’t have. I think the idea was hopefully that there would be – both sides would recognize the value proposition.
MS. LOVE: There needs to be something because when you have states like Alaska, Hawaii, and New Mexico building datasets across their population, but when half of the tribal, VA, or Federal data are locked up, you don’t even know- you can’t even benchmark. Medicaid can’t even benchmark. It really is a Medicaid business case at the very heart of this, not just the states. I think it would benefit Medicaid. We can talk more. Just in concept, I wanted to know if this was something we should be thinking about.
DR. STEAD: Rebecca.
MS. HINES: Given your incredible journey over the last year plus and what you know about this committee, do you have any thoughts about given the work you have done, your recommendations are – they are specific, but not detailed. What could this committee do or focus on that you think would be helpful moving forward?
MS. GLIED: That is a great question. I think there is a couple of things that strike me. One is this point about privacy protection and the technology around it and how the Federal government might be a player in that world. What form would that take? I think that is one set of questions around this.
A second, I think, is really leveraging the expertise of this committee in getting access to state data and really providing some – even more detail about how that works, how incentives can be used so that we could get more state data. That was a real need that the Commission identified.
A third is to think about how this committee might want to work with this. It might think about working with the NSDS, in terms of what kinds of – how you might think about aligning existing confidentiality and privacy requirements with new ones and what differences there might be there.
I think a fourth one is also the connection between data and evaluation is not always as clean as it might be. Thinking through that as well. What kinds of new data will be needed for evaluation purposes and how can we be sure that we are actually producing data in the service of assessing our program? Let me ask Lucas, as well.
MR. HITT: I think all of those are great. I would particularly emphasize the state data. As Sherry mentioned, in the recommendations, we have two specific recommendations that address state-related data. One is specific to earned income data, which interestingly, HHS is one of the few agencies that actually does receive that through the National Directory of New Hires. The second recommendation related to state data really directs departments to sort of look across their program inventory and identify where are there existing engagements with states. Is there an opportunity to maybe if the department is receiving a sample, does authority exist to increase that to a universe? Are there opportunities for departments to expand their inventory of state collected data within existing authorities and in the case of where additional authority is needed, to identify that and propose that through proper channels, generally in the budget process? I think that would be one that is really very specific to each department. Obviously, you all would be in a good position to sort of think that through.
The other, I think, that is really very – you know, the statistical agencies and the Federal statistical system play a key role in how this vision would be realized. Obviously, the department here does house one of the 13 principle statistical agencies. I know there was a question earlier about, you know, capacity to do things such as disclosure review, risk analysis, and things like that. One of the things the Commission found is that there is a great deal of expertise on how to do these kinds of things that exists within the federal statistical system. It doesn’t necessarily exist outside of that system. I think this is one of the areas where the Commission believes that some of that existing capacity can be better utilized in a more broad way.
Part of the reason why the Commission was not more explicit in exactly how that should be done is that, quite frankly, the arrangement of every department is very different. Sort of the organizational structure and location and scope of principle statistical agencies are different in every department. Again, I think having to sort of think through here is the basic idea, but how do we best realize that within HHS versus in different departments I think would be a very important thought process that this committee may be well suited to.
DR. MAYS: I just want to get a sense of it is a foregone conclusion that this is actually going to exist. Is it going through a process and it is going to have a budget, a home? Do you know anything about those things?
MR. HITT: I can give you a little bit of insight as to sort of where those discussions are right now. Obviously, the Commission, most unfortunately, doesn’t get to wave a magic wand and make everything true. It does have a great advantage in the fact that the sponsors of the Commission include the Speaker of the House and a senior member of Senate leadership.
Since about early August, when the recommendations were largely agreed upon by our members, the other members of the staff and I have been in negotiations with various congressional committees. There is legislation that will probably be introduced before the end of the month that will be what the Speaker refers to as down payment legislation. It will do a fair number of the things we have talked about.
It will create – Sherry actually referred to this – a steering committee to think through the exact details of exactly how NSDS should be stood up. As we have sort of alluded to, a lot of the elements of NSDS probably exist elsewhere in government. CARRA, for example, could be an idea that would be moved over. The RDC network could become part of NSDS as opposed to being specific to the Census Bureau. There is a fair amount of thinking about what exactly would you move, what would be the implications of that, what needs to be new. I think you will see that in this initial legislation. You will see a lot of the privacy and legal framework in the new legislation. You will see a lot of the capacity dynamics around evaluation, chief evaluation officers, learning agendas in the first package.
The second package is also being worked. It will be introduced probably the first of the year. Whether or not it makes it all the way through the rather arduous process that is legislation becoming law, I can’t say for sure. I can say that when you have the Speaker pushing it that is about as good of a hand to play as any. We are very hopeful. I can also tell you that in our conversations with the Hill and with OMB, we have gotten – to be blunt, delighted, but surprised – we have gotten – no one has said, oh, that is a terrible idea. Even some areas where I expected people to go, well, these budget times we are not going to be able to so on and so forth, those reactions really haven’t come. Like I said, somewhat to my surprise.
DR. RIPPEN: When you were talking about the states and either increasing the amount of data or making it more streamlined, is there any reason why you couldn’t have the same concept of the federated. You know you have each agency. They have their own data. Because it not always a good thing to bring it all together. I know agencies may need specific state data to do their kind of core mission. For those, they probably already have agreements in place. Would there be any – I know 57 is a big number. As far as, you know, you’ve got your agencies, it is a smaller number. Is there any reason why you couldn’t use the same philosophy, as far as having common – you know, kind of a set of standards or approaches and allowable uses? Would there be any reason why that wouldn’t be possible from a state perspective, as opposed to just saying states give me the data and we will take care of it?
MS. GLIED: I don’t think our intention is to say please give us the data and we will take care of it with the possible exception of the quarterly income data, which is a different box. The basic idea is if there is a project that comes along and that requires the use of these data, then – if there is a reason for having the data, then the data could be requested. The datasets that we mention there are ones where it is pretty clear that there would be such analyses and such evaluations that would be necessary to be done. They are large Federal program in which the states play an important role, where it would be important to assess the performance of those programs in conjunction with other federal programs, for example, not only for the administration of that program, but to assess how these programs interact, for example, in the lives of Americans.
It is a little bit different from the way a department currently might access data for its own administrative purposes.
DR. RIPPEN: There are some things that you just have to do. Again, it strengthens kind of this relationship between federal and state. Just something to think about.
MS. GOSS: This conversation has been really interesting. I look forward to having the opportunity to read the entire report. Pardon my questions if they are a little bit short-sighted because I haven’t read it yet.
The conversation, as I hear it, is about organizing the federal agencies at this point to garner advancements and how we approach linking data and collaborating in a protective manner with the data, but still harnessing the value from it. That conversation cut short of going down to the state level. I realize we have to start someplace, but there is a huge – so, at this point, there is nothing coming out that is going to change really what the states are doing or how the states are playing other than potentially – I am thinking about how much the Feds, like in a Medicare arena, provide datasets to states or researchers for a variety of reasons that are supportive of evidence-based medicine. Is it reasonable for me to expect that right now, it is just a sphere around the federal environment and that it will be a long while until there is a bleeding out to the states’ impacts or even the products of the federal government that states or researchers use?
MS. GLIED: I think we envision states as being one of the customers of the NSDS. So, states might very well be approaching the NSDS to get certain kinds of federal linkages made and linked to their own state data for their own evaluation purposes. I don’t think states are in any way cut out of this.
It doesn’t envision – and I think a second aspect of this that is important to state is that hopefully, the NSDS will actually serve as a model for states that might want to do very similar things with their own data and make those data available to researchers who want to investigate state programs. At the same time, the federal government’s ability to tell the states what to do is limited. The report is really focused on what the federal government could do.
MS. GOSS: Thank you.
DR. STEAD: This is Bill Stead. From my perch, you have done a remarkable job at trying to put the framework together. You really have done well at flipping it around to, in essence, building up from a confidentiality framework. Then if I am hearing right, providing a linkage functions, not a database, that allows the data to be brought together for a particular purpose, released for a particular purpose, under the protection of the framework. That makes great sense to me. Am I playing that tape correctly?
MS. GLIED: You are completely correct. That is exactly what we envisioned.
DR. RIPPEN: Quick comment. Did you actually look at what South Carolina has done with their state data and how they have gone across the different agencies?
MS. GLIED: I am trying to remember. I don’t have the report in front of me. Lucas, did you want to do this one? We did look at some states.
MR. HITT: We looked at a number of states. I am smiling because I am actually from South Carolina and very familiar with and know some of the people involved in that project.
We did not hear from them formally. We invited them to testify. The gentleman who was running the program at the time was unable to, scheduling-wise. We did interact with them off the record, in terms of not being at a hearing specifically. We are aware of that. We are also very aware of the program that South Carolina is participating in, where they are one of five states that are working with CREC to develop a – sort of a broader state data model. I think the hope is that that same kind of model can be more broadly implemented.
MS. GOSS: CREC?
MR. HITT: CREC is the Center for Regional Economic Competitiveness.
DR. COHEN: So, I am sure you mentioned this earlier. Did you clearly define the purposes to be research and evaluation and exclude administrative or commercial uses?
MS. GLIED: Yes.
DR. COHEN: So, these data would not be available for commercial uses?
MS. GLIED: Correct. They could only be available for statistical uses. I don’t think that necessarily rules out certain kinds of commercial uses, for example, a commercial use that involves trying to figure out what the overall relationship between purchases of X and some disease or whatever by a vaccine maker would be fine, but not something that said here are the 27,000 people you want to target for your next marketing campaign.
DR. COHEN: So, commercial users focusing on statistical analyses –
MS. GLIED: Could be – I mean, part of the reason for that, if you think about it, is that there are a great number of contractors who conduct a lot of the evaluation for the federal government. We wouldn’t want to close them out from being able to do this. At the same time, the real focus here is on statistical analyses. I think there are actually relatively few other situations where commercial users would need or want to conduct evaluation that improve government programs, after all, that is what the purpose of this is and are only statistical.
DR. COHEN: So, the goal is to improve government programs.
MS. GLIED: Right.
DR. COHEN: Using statistical analysis and that is clearly defined.
MS. GLIED: Right.
DR. COHEN: Thanks.
DR. STEAD: Can I ask one more question from my perch? How readily available did the Commission think the methods were or the best practices were related to the rigorous risk assessment of de-identification.
MS. GLIED: Insufficiently. I think a focus is improving those methods. Making the ones that exist better available, but also improving those methods.
MR. HITT: There was a question earlier about the role of NSDS and its role in sort of leading the development and implementation of privacy enabling techniques. That is one-third of it. Another third of it would be sort of data related sciences, data linkage sciences, and related. Right now, a lot of that work happens, but it is not any particular entity’s specific mission.
DR. STEAD: In our last dive into de-identification, it is very clear that although covered entities are required to, in essence, state that they have done that, largely, it is not done. Not only do they not – so, it is a huge –
MS. GLIED: Correct.
DR. STEAD: That would be a place where leadership would be awesome.
MS. KLOSS: I may be done, but it is done in a very formulaic fashion without risk analysis that would – or problem how data are to be used and how to take extra steps to guard against reuse that weren’t appropriate.
DR. STEAD: Thank you.
MR. HITT: Just two administrative things for you all to know. One is last week, we were very pleased to see an announcement from the Bipartisan Policy Center. They have received funding and are going to continue the – essentially, carry the Commission forward. The chair, Catherine Abraham, and co-chair, Ron Haskins, have agreed to continue on in leading the project there. BPC also hired the policy director from the staff. He will continue to get up every morning and think about these things.
MS. HINES: Who is that?
MR. HITT: His name is Nick Hart. We are very excited about – you know, to the earlier question about how do we make sure this doesn’t just sort of sit on a shelf somewhere, we have worked to try to create some sustainability there. Also, in that vein, specifically to the statistical operations, BPC has partnered with COPAFS, the Coalition of Professional Federal Associations on Statistics. That is now led by John Thompson, former director of the Census Bureau. They will be meeting tomorrow and focusing on these topics further. So, we’ve got two large professional organizations are going to be focused on this specifically.
The other thing that may be of use – Sherry mentioned all of the public input we got, the survey of federal agencies and so on and so forth. All of that material will be made public. I think it probably will be published later this week in a second volume of the report, which will live only online in several hundred pages at this point. Particularly, if any of that data may be useful to any of your work, it will be available at cep.gov. Ultimately, BPC will launch a new website where all of that material will live as well. In the federal world, we have to figure out how to shut down our websites when we go out of existence. I am sure the archives will have it, but I’m not sure how easy that is to access. That material will remain available.
DR. STEAD: Thank you very much. We are on time for a 15-minute break. We will reconvene at 10.
DR. STEAD: Let’s call ourselves back to order. Vickie, would you like to read yourself in?
DR. MAYS: Vickie Mays. University of California Los Angeles. Member of the Full Committee, Pop, Privacy. I have no conflicts. I apologize for being late.
DR. STEAD: Welcome back. I want to thank the – did somebody –
MS. STRICKLAND: This is Deb Strickland. I had joined about 8:40. I just wanted to announce myself.
DR. STEAD: Formally read yourself in, please, for our FACA colleagues.
MS. STRICKLAND: Sure. Deb Strickland. Member of the Full Committee and the Standards Subcommittee. No conflicts.
DR. STEAD: Thank you. We are looking forward to the next block of time, where we are going to hear new things that are happening from key perspectives that might – and how NCVHS might respond, in terms of our workplan. Charlie Rothwell is going to lead off.
DR. ROTHWELL: I am Charlie Rothwell. I have the honor of being – I think it is the seventh director for the National Center for Health Statistics. I have known all but one. I have been around a long time.
I think the presentation that you got this morning was absolutely spot on to where the National Committee should be looking at and how it could maybe facilitate us working together not only within this department, but across departments.
When you take a look – first of all, when you look at the report, I would first go – not read what normally you all do and that is read the recommendations or the executive summary. I would go and first of all, look at who is on that commission. A question was asked whether it was partisan. Yes, it was. But it was partisan in a different way. It had one group of folks who are privacy advocates and it had another group who were data mongers, if you will, data manipulators. That is where the partisanship was. Frankly – and they are extremely competent.
When I looked at the makeup of that commission, I was concerned that they probably would kill one another and nothing would come out. There couldn’t be anything further from the truth. I was completely wrong. What they have come out with is much more than I ever anticipated and I think is spot on on what we should be doing within the Department and where we should be going as a governmental agency.
Let me start by an example. I think this goes to one of the questions that was raised during that presentation. Why is it that we don’t link information? Well, is it always because of legislation? Many times, it is, but many times, it is not a natural act. It took us I forget how long – four years, maybe five years, to link our survey data within the National Center for Health Statistics with HUD data. Was it because legally we couldn’t do it? No. But it was an unnatural act. When the legal folks on both sides looked at, they are trying to protect their risk. If it is not something that has ever been done before, why should we move ahead?
The other is the usual question is why we, at HUD, are looking at health information? That is not what we are about. And why are we, in HHS, looking at HUD information because that is not what we are about? Frankly, what we are about is to provide the appropriate services to people that are in our communities. They are receiving a variety of services. Health data is not all about collecting data in health. It is about the smorgasbord or services and places where people receive services in their communities. We need to be able to link that information up.
In fact, finally, we were successful. It showed that people who were receiving various types of HUD services, in fact, were much – had much better health outcomes in a variety of areas. We will need to look at that over time to see if that really holds up. It also might allow us to focus certain activities better in the future for both departments. That is one thing.
The other is a question about, well, why are – don’t you have expertise in the National Center for Health Statistics to do disclosure review? Why is it going over to Commerce? Why is it going someplace else? Well, I think the capability in the federal statistical system are great, but they are not even. Bureau of Labor Statistics has expertise that NCHS doesn’t have. NCHS has expertise that BLS doesn’t have. Census has expertise that neither one of us has. Together, in fact, I think in a serve perspective that they were talking about, we can really make sure that we are not being too restrictive and in some cases, not restrictive enough because I think there is great variability in how our disclosure reviews are being done within even the Federal statistical system, much less outside of it.
Then, as you know, as programs have been developed in this country in all departments, legislation to allow them to collect information usually was protective for a variety of good reasons. Well, sometimes – well, not sometimes, but many times, that legislation is quite old. The environment didn’t exist that would allow such linkages. It wasn’t that people didn’t want this to happen. It was that they wanted to ensure that the privacy of the individuals or the institutions that were providing this information are, indeed, protected. We really do need to be able to create an environment, a protected environment, where people and congress feels assured that nothing untoward can happen or at least the risk is minimized as we link and move datasets together.
I see this as strengthening both the Federal statistical system, the 13 agencies that make that up, as well as the administrative data that we would like to use more that we haven’t done in the past and put in, in statute, protections for that administrative data that has never been there before that would allow it to be linked and still be protected, which is what the initial statutes were that allows those program data to be collected.
Finally – I won’t say finally – this creates an environment, both from a congressional perspective and, hopefully, from – I hope from a congressional perspective and then from an executive perspective that this is expected for us to do. We are expected to do business with one another. We are expected to link information together. We are expected to analyze it. Period. It is not NCHS. It is not Census. It is not CMS. If we are looking at issues of health, if we are looking at issues of employment, whatever it is, we need to be able to look across this government and use the appropriate data that we have to make the best decisions possible to either change tax code, for example, or to provide better education to – money for educational purposes, you name it where federal government and state government, by the way, has to make some very difficult decisions, sometimes without the information they need.
I really think that what this commission has put in place is a good down payment. It is not the end. So, what we need to do is figure out how to begin. Hopefully, congress will initiate some of the statutes that need to take place, i.e. take one of the OMB directive one, which basically protects the credibility and the confidentiality – a variety of other things – and CIPSEA that statistical agencies have and make it available – put it into law and then make it available – broaden to other data activities so that when we bring things together they are all protected in an equal way. So, in my opinion, the Federal statistical system – my counterparts should be running towards this and not running away.
Now, for this Department, what does this mean? Well, I can only tell you from my little seat. Just take it or leave it. We have now – Sherry Glied was one of our – in my opinion, one of our better ASPEs. We have had a variety of great ASPEs over the years, who have guided our evaluation activities. One of the recommendations of the Commission was to have something similar like that in each department to really crystallize on how we should be looking at our data.
The Commission also said something else. It said, you know, what we need to do is to – instead of looking at it from after the fact, we need to look at it before we are looking at linkages. We need to have somebody senior within each department that will look at data possibilities with existing systems and systems that are going to be developed.
Let me explain a little bit about that. Over the years, as various agencies have developed their information systems, whether it was for their own programmatic purposes or for data dissemination purposes like a statistical agency, those systems were put in place and probably stayed for years and years and years and years. Technology is changing so fast, both from a platform perspective, from a software perspective, and from a capability/storage perspective that you can no longer do that. What we are going to be seeing, I believe, over the next several years is a major change in many of those base systems. That is the time that we need to look at making sure that there are links in those systems that will allow better linkages when needed. It doesn’t mean that they have to be done all the time, but at least when you are going to try to link that you have some hooks in there that make that possible and also that that is built into the system, itself, and that that is a responsibility. Even though I may be a program that is very narrow in my area, I need to make sure that this data can be pulled out appropriately and easily. We have to have the technology base to do that. We have to make sure that we have the best equipment to do it and the best software to do it, which means we have to bring together both the technology, the Office of – the technology officers. We need to bring together the chief information officers and the security officers. We need to bring also together the data person, who understands the underpinnings of the data, itself, and how – what should be linked, what should be made available, and frankly, what shouldn’t.
Right now, our department is probably closer to that than we think. We do have a Data Council. The Data Council, though, hasn’t actively been involved with the chief technology officer and their staff. We are trying to do that. I am a co-chair of that council. The chief information officer and their organization has a representative in the group, but we have really not been, frankly, working together as a team. This is my opinion. I think that can change. This is one of those things – what the Commission is requesting, I think, is for us to be not in the same organization necessarily, but working together in order to make what the Commission wants to take place.
I think, unlike a lot of other departments, we have the offices available and the director of NCHS – it has been named the statistical advisor to the Secretary. There is in place some things that could help make this happen in this Department quicker than in other departments. That being said, there are some things that we can’t do in this department. I think that is why this data service, if you will, is really important. We may need to have a safe zone where we do the data linkage, especially between departments. Why should Commerce trust us? Why should we trust Census? So, let’s have a safe zone where everybody has had input to this, that we are convinced that this is a place where we can play cards together and get going. I really feel that we need this overarching capability.
The other thing when you read their report, you will see that they are asking OMB to reorganize. They weren’t real specific there, but I think it had to do with, again, the Chief Technology Officer. I think it had to do with the Chief Statistician of the U.S. I think it had to do with the Chief Information Officer role in OMB. In other words, they want these folks to work together more than they have in the past. It is not to say that there is warfare going on. That is not the case. It is just that we have had our well-defined roles. We stick to them. It goes right back to – this is a function area. It goes right back to my comment about what is health data. We have been very siloed on how we have looked at data. Health data is my data. That is how I am going to look at it. Anything else is really not all that useful. Well, that is not the case. In order for us to really do better, we need to, in fact, broaden our scope of datasets as well as we need to bring the statistical agencies closer to the Chief Technology activities of the departments, as well as bring – and with the CIOs, in order to make this happen. Without either – the three activities going on and without something like ASPE making sure that what we wind up with is something that is really needed from an evaluation perspective, then it is not going to happen.
I will stop right there.
DR. STEAD: That is an excellent sort of set up for this whole set of leaders. If you look around you, with ONC, with the Office of the CTO, and with ASPE and with the statistical piece within the Department, we are trying to get that cross-talk this morning, in terms of implications with our committee. That is a very nice launch.
Questions for Charlie? Do we move ahead?
MS. HINES: Let us move ahead because Elise has to leave in half an hour.
MS. ANTHONY: I hate to be the reason that we don’t get to ask questions. I apologize. I do have to leave a little bit early for another meeting that I couldn’t get out of.
Good morning, everyone. My name is Elise Sweeney Anthony. I am the Director of Policy at ONC. I have had the pleasure of presenting before many of you in the past. It is usually either myself or Dr. John White. We work in tandem.
Today, we are going to do an update on our work on Cures – 21st Century Cures Act. I would be surprised if you guys haven’t heard that at all. I am sure – it might be easier for me just to jump into Title IV.
So, most of the Health IT provisions are in Title IV of the Cures Act. The Cures Act deals with a number of different pieces from FDA to SAMHSA to ONC. So, we have been hard at work at ONC thinking about how best to implement what congress has asked us to do, making sure that our focus is on interoperability and exchange, patient access, and really bringing together from the outside, stakeholders, in terms of networks being able to connect and move patient information appropriately.
So, I am going to go over a little bit of the key sections, but I am going to spend a lot of time talking about 4003. There is some upcoming work that we are going to be doing. We want to make sure that you guys are aware and engaged.
So, first, 4001. One of the key pieces of 4001 is around Burden Reduction. I think congress recognized from a number of hearings and then ultimately, through Cures that there is a significant amount of burden that sits on providers for a number of reasons, some around documentation, some around how the systems are constructed to move that information. They have asked HHS and ONC is working with our federal partners on this, to look at ways to address the burden that exists for providers and to put together a strategy to address that. We are calling that the Burden Reduction Strategy. So, we are working across our federal partners now on that piece.
There is also another piece in 4001 around specialty certification. I think that is in recognition that providers obviously have different sets of information that might be helpful to their practice. Congress has asked us to look a little bit into that and to see what we might be able to do by way of our program to support that.
4002 deals with a couple of pieces. One is around EHR Significant Hardship. There might be situations where a provider is using a technology and for some reason, the technology isn’t working properly or what have you. It could be a number of scenarios. The impact that that could have on a provider who is attempting to report under one of CMS’s programs, most notably probably QPP, Quality Payment Program. So, we are working with CMS as they put together language to support that provision.
One of the main pieces I did want to spend some time on is the Conditions of Certification. So, for those who know our program, we have focused very much on the technical criteria that should exist in systems to support the movement of information or support the proper recording of that information. Sometimes, that is functionality requirements. Very often, it is standards-based requirements. Ultimately, it is around technical components. Then the conditions of certification component of Cures, Congress has asked us to think a little bit broader. They have given us some authorities to look at things like information blocking, whether a developer is blocking the flow of information across systems, for example. They have also given us the ability to look at real-world testing to make sure that the product is operating as it does in the field. There is a number of different conditions. We are looking at the best way to implement that. We are hard at work on that provision.
One thing I would note there is we have done some work prior to Cures, thinking about how to get at what is happening in the real-world environment. So, when the technology is actually released, after it is certified, a product has been purchased and implemented by a provider, we have surveillance system in place with our accredited certification bodies. We also have a system for surveilling the products on a reactive basis so someone can come to us and say, hey, there is a problem with the product I purchased and we can look into it. We are excited about these provisions that Congress has added, too, to help us further address some of these issues that we have seen and also that the community has seen at large.
There is also a component around the EHR reporting program, generally. That is kind of looking at how EHRs are operating and to give more information to providers who might be purchasing that technology.
I am going to skip 4003 because that is where I am going to spend most of my time. I am going to go to the next one and come back to that.
So, 4004 is around Information Blocking. As many of you know, some years now ago – I can’t believe it has been so long – we released an information blocking report. That was in response to a request from congress to look at information blocking and the ways that that is happening. The report was – really helped to spur conversation around information blocking not only from the view of a developer who might be doing it, but also others. There could be some health systems that are doing the same thing. Congress recognized that as part of 4004 because they note that there could be a situation where a health provider is information blocking or where a developer or other type of exchange is information blocking.
This provision is one that we are not working on by ourselves. It is in conjunction with OIG, Office of the Inspector General, who has the enforcement capabilities here around potential fines that could exist for developers – not for providers, but for certain developer types around information blocking. We have been holding a series of meetings with those who have come to us and asked to share their stories or their experiences around information blocking with OIG. That has been helping to inform our thinking. If you folks have any awareness or interest in reaching out, please let me know. We are happy to refer you to OIG, who is organizing these meetings.
The goal is really to think about information blocking as it impedes the flow of information or as Dr. Rucker would say, data liquidity. So, it is something that is very important to us. You will see that weaved across our work going forward.
4005 is a dual provision. It focuses on registries and the interest of congress for us to look at standards associated with registries and for developers to incorporate that into the systems they have. There is also another provision around patient safety activities and the treatment of developers for that purpose, when looking at patient safety activities.
4006 is around patients and the work to make sure that patients have access to and information about the access of their records. We are working on that with OCR, of course, the Office of Civil Rights.
4007 and 4008 are around GAO Studies. That, obviously, is not something that we would be doing. Obviously, we would be working with GAO as they request on the formation of that information.
Let me bounce back to 4003. First, the Health IT Advisory Committee. That is probably one that you guys are pretty interested in. As many of you know, in the past, we had two advisory committees. We had the Health IT Policy Committee and the Health IT Standards Committee. Both of those were very instrumental in development of policies not only on the standards side, in terms of what standards we should adopt and move forward with, also on things like privacy and security, consumer information, as well as the policy around some of the CMS work, not only from stage one, stage two, stage three, and MU, but also more recently, around QPP.
Now, as we have kind of – we have worked on closing out those two committees, as per Cures, we are now in the process of winding up and setting up the next committee, which congress has called for, which is the Health Information Technology Advisory Committee. That committee would serve dual purpose. Not only looking at policies, but also looking at standards. It is putting some of those functions into the same committee.
There is a pretty significant standards section there, looking at standards and making sure that the committee is involved with that. There is also somewhat of a priority list that Congress has identified for things that they would like us to look at first. Those are things around exchange, privacy and security, access, et cetera. From there, there were opportunities for us to expand on looking at other areas as well.
So, this is something that we are hard at work on. As many of you know, the GAO appointees have been set in place. Most of the congressional appointees for this committee have also been set in place. HHS released a Federal Register Notice a little bit ago announcing a request for nominations. We received those. We are reviewing the list now. Once all of that is set into place and we have all of our committee members in place, we would go ahead and stand up and start the committee. Our goal is to do that in the winter. So, it wouldn’t be too far from now.
We are excited about that opportunity not only in terms of looking at some of these new issues that Congress has asked us to explore around interoperability, et cetera, but also because it still calls for us to engage with NCVHS as we always have. That is specifically noted in the statute as well. We look forward to engaging with you as we move forward there.
The other piece that we are going to spend time talking about, finally moving to the next slide, is around the Common Agreement. So, the 4003 specifically calls for ONC to think about a trusted exchange framework and whether we develop or support a Common Agreement in that vein. The idea is that Congress specifically said that networks that are moving information, are they communicating? Is the information – health information able to move appropriately between or over networks in an interoperable way? They have asked us to engage in that and to help that happen externally.
So, we are in the process of making sure that we are working with the external community and our internal partners in that process. What we don’t want is to do anything in a vacuum. We always try to make sure that that doesn’t happen. In the policy team, we specifically make sure we are engaging with our stakeholders.
What we have done is we had a kickoff meeting in June, end of June. The purpose of the meeting was to, one, share some of what we have heard about what the existing networks are. Many of those networks presented at the meeting, everyone from Cara Quality to Direct Trust, a number of different folks who move information across networks or across systems for purposes of exchange, however you define what that is. It was exciting because we were able to hear the different perspectives and the progress that has been made. We also heard about some of the areas where there are still shortfalls, where things are not moving as they should and still progress needs to occur. That was the purpose of the meeting, to make sure we are hearing both sides, the good and the bad, and not to necessarily look at just the progress that has been made. I think all of those who participated in the panel were aware of that and open to that discussion, which was important.
We also heard a lot of public comment around this issue as well at the meeting. That was everything from the mechanisms for exchange, the ease of exchange, access, patient access to that information when it is moving, to patient matching. A number of different issues came up.
After that, we opened up a public comment period. The public comment period was an opportunity for those – for folks to submit actual written comments to us about their thoughts about what a trusted exchange framework and potential Common Agreement should look like. We provided a couple of categories. We tried to make it as easy as possible. So, we made it a web portal so folks could submit comments on any or all of the categories we provided or just provide general comments, everything from standardization to transparency, cooperation and non-discrimination, security, and patient safety, access and data-driven choice. All of this is now being examined by my team and others at ONC to look at the comments we received as we move into our next step.
Our next step – well, before I talk about that, I should talk about some of the things that Dr. Rucker noted at that meeting that he thought were critical. We kind of captured this in a blog that we released right after. Just to note kind of the three areas that he wanted us to think about and wanted the community to think about, in thinking about a framework and a common agreement.
One was patient access to health information without special effort. The second was providers and organizations being accountable to working – that are accountable to working with populations so that they can receive information as a group or information on a group of individuals, as opposed to having to do one-off queries or inquiries on a patient, which can obviously be time consuming. The last was open and accessible APIs, which is another component of Cures that is called out in the conditions of certification, where congress has asked us to think about APIs and that they should be available without special effort, as the term is noted.
Following receiving the comment, we are reviewing the comment now. We then are planning at September 29th webinar. This will be our second meeting to engage with the public on the Common Agreement. The goal of that meeting will be to share some of the feedback we have received through the public comment process and also to share an analysis that we are working with contractors on on what the existing frameworks or networks that are out there do and what they don’t do and how they are set up. We wanted to provide kind of a foundational look at that information.
Following that, we plan to release a draft coming up in early 2018 at the latest is our thought, on what the Trusted Exchange Framework and Common Agreement would look like. Then, again, we are going to have another public comment process. We want to take in feedback from folks on what they think about the draft we put together. The goal would be to release the final in 2018.
So, as you can see, there are a number of opportunities that we have built in to engage the public. The reason for that is that we can’t do this alone. Even when we were working on prior documents, working on the roadmap, et cetera, we have always said we can’t do this alone. This involves private stakeholders to be involved and at the table as we try to get information to move to the benefit of the patient and to help providers do what they want to do best, which is care for the patients. We want to make sure that we have a process that supports that going forward.
So, the webinar, I noted a little bit about what the webinar is going to entail. I would encourage folks interested to RSVP as soon as possible. As I understand it, we are almost at capacity for the webinar already. Please RSVP as soon as possible. I will say that on our website – you can capture this through the blog, if you click on the link to the blog – there is a full page of information on our website around what we heard – what was said at the Common Agreement because we have the recorded presentations available, as well as the agenda. At the end of the second webinar, we will do the same. So, for those who are not able to get in, please know that we will make that information available on our website, as well.
With that, I will stop and take any questions.
DR. STEAD: She is going to have to leave. If we want to probe, we should do it now.
MS. GOSS: That is phenomenal. I have been tracking with this pretty closely. I think she gave a really good overview. I actually am just excited for the next step in the process and to be able to be involved. I think there is a tremendous linkage between what they are doing and what we are doing, especially because I think Linda is about to make some very dramatic comments.
MS. KLOSS: Actually, I was going to ask a question. Not on the Trust Framework, but on in what direction is your work on patient access taking you? I know we had the new guidance from the Office for Civil Rights in 2016 on patient access. It has certainly ramped up. I would like to know what your thinking is and what the process is for taking that further.
MS. ANTHONY: So, this is where I probably have to be a little bit of a lawyer and say that, you know, some of these things we are still working on internally in terms of the path forward and how we are going to capture that. So, we don’t have any announcements I can share now, in terms of the direction and so forth. I will say the commitment to the I would say number of guidance pieces that were released around 2016 around patient access and seeing HIPAA as more of a vehicle for health information access and not an impediment, as has been the case sometimes externally, I think that that commitment still exists. I know that that commitment still exists, not only with me, but also with all of our leadership at ONC. We continue to work with OCR as we think about how what we have done fits in with 4006 and what needs to continue to happen under that.
MS. KLOSS: I think that is an area where there is some opportunity for collaboration with the Committee. Certainly, we know that 2016 guidance has made a big difference in practice – real practice changes in the field. I think we would like to contribute in any way we can.
MS. ANTHONY: Wonderful. Thank you.
DR. STEAD: Thank you. If no more questions, let’s move onto Mona Siddiqui.
MS. SIDDIQUI: First, thank you so much for allowing the Office to come and give an update. I really don’t have a formal presentation, but just wanted to talk a little bit about the work moving forward.
You all, of course, know that healthdata.gov has historically lived in the CTOs office and been an external facing – primarily, I think, the open data movement, been an external facing project, really trying to make data open for entrepreneurs, for researchers, externally. I think more and more where our Office is really looking to right now is how do we leverage this data for greater insights within HHS. Rather than having – I know this happens sort of every cycle, a new administration – that we want to understand all of the data assets and how we use them, I think we want to be a bit more deliberative about it and take some of the priority areas that have been identified by the leadership and really start to understand what the data assets are across HHS and across the different agencies around those domain interests and trying to understand what some of the barriers are to data sharing, whether they are cultural, whether they are from the technology perspective, or whether they are from a legal perspective.
What I would say is sort of the next iteration or healthdata 3.0 – really want to understand those sort of challenges, the data assets, and use data for insights in decision making so that the policies and the programs, whether it is a funding mechanism, those are grounded in evidence. This is going to require, obviously, many different steps. Part of it is going to be to reach a data governance process that is HHS-wide. It is also going to be about drafting data use agreements that are a starting point across the different agencies so that – somebody in one agency told me recently they have 200 – at least 200 data sharing agreements with another agency right now in place. That should not happen. There should be an agreed upon starting point at least. We also need to look at other barriers, whether they might be financial and data sharing across agencies. So, really trying to understand what some of those are and bit by bit trying to address those through an enterprise-wide data management strategy. I think that is where we are really looking to right now.
DR. STEAD: What do you see as the intersection of this work and what we just heard from the Commission on Evidence Based Policy?
MS. SIDDIQUI: I think it links really, really well. I will admit that I haven’t actually read the entire report yet, but I think the foundation is the same. As background, I am a data scientist and researcher by background. When I see decisions getting made without the right evidence and research behind it, whether it is in how we allocate dollars or whether it is in how we manage a program or how responsive we are to a particular crisis, I think it all, to me, at least, points back to our data and really grounding those in evidence.
I think we will have to see some specific recommendations, how they match up to some of the work that might come out of Reimagine and some of the specific priorities that get highlighted. I think the foundation is correct.
DR. RIPPEN: We did a lot of work on data and the question of how do you summarize it and organize it. If you are actually doing kind of a review of the datasets and actually asking for information about them, you may want to pull up some of the work that we had done on how do you summarize. We actually got a lot of input from our committee – I guess it was a working group that actually included external players, too, as far as what might be helpful and how might you do it in a consistent way so you don’t have to do it again. Just something to consider.
MS. SIDDIQUI: This is more around how to attack the data?
DR. RIPPEN: Yes. It is more like how do you summarize it so that everything is – so that you can actually do a search so you can just kind of get a sense before you dig in.
MS. SIDDIQUI: Thank you.
DR. ROSS: Just a question more about the span of responsibility for your office. Do you get involved with or even observe activities where a Federal agency is involved in data exchange, but with healthcare entities? So, it is not just a discussion about – I would say across public health service agencies here, in HHS, but with an agency like CDC interacting with healthcare? You mentioned governance.
MS. SIDDIQUI: Right now, we are working on a project where we are highlighting some – both challenges as well as the mechanisms of how we operate between, for instance, a CDC and a state and the data that the states have. It is not a comprehensive look right now by any means. In those interactions, we are starting to understand what some of the state level, again, differences are and what has worked and what hasn’t in data sharing. I think it goes to a larger question around multilevel stakeholder data sharing, whether it is within HHS or other federal departments.
Again, in interactions with some other federal departments, the pathway is easier in some than in others. I think the same goes for states. I think trying to find what has worked and being able to highlight that and being able to share some of those practices is going to be helpful moving forward.
DR. ROSS: Are you involved at all with observing the Digital Bridge?
MS. SIDDIQUI: A little bit, again, but we haven’t done a deep dive. I should say we welcome – you have been steeped in this space for a long time – resources that you can point us to and work that you have done in the past. We certainly don’t want to duplicate efforts. We want to build upon the work that you all have done in the past.
DR. ROSS: I point that out just because it is an exercise in trying to understand if a governance discussion can be established that helps move – helps healthcare work with public health, public health work with healthcare, for the exchange of data. It is more about governance than it is about technology, honestly. It is something that you might want to observe.
MS. SIDDIQUI: Absolutely. I think our office firmly believes that this is not a technology issue at all. This is really more about having institutional processes in place and working through some of those bit by bit.
DR. STEAD: We will send you the links to probably a couple of things if you have not seen them. The first is the letter that we recently wrote to the Secretary that attempted to provide really actionable, short-term guidance –
MS. SIDDIQUI: I actually have seen that.
DR. STEAD: Around healthdata.gov. So, you’ve got that. We really did try to set aside boiling the ocean and say some things that could be done in 6 to 12 months.
The other thing is much more complex, which is our Health Data Framework, which, in essence, is a white paper talking about how we could apply metadata to datasets to make it easier to discover and to know how to use them. That might also be helpful, although it is much longer term.
MS. HINES: I will send that to you right now.
DR. MAYS: I guess I just wanted to follow up to get a sense of whether those are the kinds of things that are in your ballpark like when we start talking about like healthdata.gov or we start talking about ways in which they make data more easily accessible, usable. Can you talk about where you all are on that?
MS. SIDDIQUI: Yes. We definitely want to continue that focus for external users, but I would say that one of the challenges with healthdata.gov, as you all are familiar with, is that it has been hard to sort of track exactly how it has been getting used by researchers and by entrepreneurs, which really was the initial purpose to really open up that data. Because of some just website functionalities, we aren’t able to completely track how that data is getting leveraged. We want to continue to do that – continue to open data.
We also realize that a lot of the insights that can be generated for us and how we run our programs should happen behind a firewall, where there is no concern around PII getting compromised. That should happen with agencies sharing data. As a first step, we really need to figure out that process internally. Not that we want to close data, but how can some of those data assets across agencies lead to better understanding of whether a program is working, of whether the right resources are getting deployed, whether we are achieving the outcomes that we want.
I think that is sort of the other sort of flipside of this. We also need to turn it internally, which we haven’t done in a systematic way.
DR. STEAD: Chris, thank you for jiggling your schedule around so that you could be with us to think with us from the perspective of ASPE.
MR. JONES: Sure. Thanks for having me here today. I think all of the comments – actually, I think it fits well into some of the stuff I was going to talk about. I think there are sort of specific examples and then broader activities that are happening.
Just a couple things I wanted to update on around some of the Department’s priorities. Some of these are more baked than others, but I wanted to talk about those.
So, when Dr. Price came in to HHS, he identified addressing the opioid epidemic, serious mental illness, and childhood obesity as his top three clinical priorities. Due to I think just the pressing nature of the opioid issue, that is the one area where we have really rolled out a strategy where on childhood obesity and serious mental illness, we are in the process of developing those strategies. ASPE is engaged in all three of those across our different offices. Our Human Services Policy Office is heading up the childhood obesity side. Our Office of Disability, Aging, and Long-Term Care Policy is heading up some of the serious mental illness activities. In Science and Data Policy, we are leading the opioid. That is a role that we have been playing for the last several years in the previous administration.
As many of you know, the President also stood up a commission that Governor Chris Christie is chairing, on substance abuse and the opioid addiction epidemic, in March. That sort of was somewhat of an impetus for us to figure out what our strategy is that we wanted to be ahead of the curve a little bit, knowing that they would have recommendations out. We, very early on, met with Dr. Price and presented him with a five-point strategy that tries to address the opioid issue in a comprehensive way and largely building on work that was already happening within the Department, I think making some important pivots to certain areas – I will talk about those in a minute – and then taking from what congress had signaled to us, both through the Comprehensive Addiction Recovery Act that was passed in the middle of 2016 and then more recently in 21st Century Cures, which had a substantial investment to states in the first part of the Cures Act around addressing the opioid issue.
So, the five points of the strategy – the first area is improving access to prevention, treatment, and recovery services for people who have opioid use disorders, opioid addiction. It is the area where we are making our most significant investments. The Cures Act, as I mentioned, included $1 billion over two years for – primarily focusing on treatment. We have evidence-based treatments that are highly under-utilized for opioid use disorder. There is quite a gap between access to those services versus the people who need them. We put out $485 million in May – May 1st – to states. All states received funding to expand their treatment capacity, build out recovery support services, and then a small proportion of the money is used more on the upstream side of prevention and more primary drug use initiation prevention.
We also have other funding that is soon to be announced from SAMHSA that is specific in areas of prevention, treatment, and recovery, to continue to build out those specific areas within the opioid’s context.
The second area of our strategy is targeting the availability and distribution of overdose reversing drugs, so naloxone for reversing overdose. Again, this is a scenario where there has been quite a lot of change at the state level. The legal environment exists in almost all states where you could go into CVS and get naloxone like you can a flu shot, without needing a patient’s specific prescription. What we are finding from the research is that is sort of a passive approach. We are missing critical opportunities to equip high-risk populations with naloxone.
Our previous opioid initiative in the last administration was more generally focused on naloxone. We felt very strongly that it really should be using data to help drive where do we need to deploy to the high-risk populations, who are the high-risk populations. CDC has done a couple of Epi-Aids with states to look at this. There are touchpoints like leaving the hospital, leaving the criminal justice system, leaving residential treatment that are substantially – those populations are at substantially greater risk for overdose. Thinking about how do we look at our funding announcements for naloxone? Are we targeting those areas?
The third area of the strategy certainly is relevant to this group. It is improving public health data and reporting or surveillance as I like to shorthand it. This is an area that is of particular interest to the Secretary. I am sure that you guys are getting quite a lot of pressure. He would like that we have absolute near real-time data for everything. Obviously, that is not possible. NCHS, in particular, is making great strides in the mortality data, in particular, and trying to get that information out as quickly as possible. This was an area where we felt like we needed to be more explicit in our strategy. It was sort of behind the scenes in the previous opioid initiative. It really is a fundamental part of our work. If we are going to make policy in 2017, we need to have as up-to-date information as we can.
CDC is a primary funder in this space. They are actually creating a new surveillance system through the Injury Center at CDC. They have currently funded 32 states and DC to sort of draw from the bioterrorism space in syndromic surveillance and using emergency department data to look at nonfatal overdose that are showing up in a more real-time way and in a consistent way. The medical examiner/coroner system has its problems. There is not a lot of consistency across states. NCHS, again, is trying to work through some of that. This is really trying to create something from the ground up that everybody is using, although there are obviously going to be coding issues if you are looking at ICD data, but that everybody is using the same set of definitions and they are reporting it in the same way. That is certainly an exciting areas of work that I think the Secretary will continue to press us across the Department to deliver more real-time data.
The fourth area of this strategy is supporting cutting edge research. Again, this is another area that was not explicit in the previous opioid initiative. This really runs the gamut of sort of the basic science research around pain and addiction all the way through how do we use that information to develop better products, bring them to the market faster, and then how do we translate what we know from what is happening in clinical practice or innovative policies and scale those up in communities. So, sort of the research enterprise and being explicit that that is a critical component of the strategy. Secretary Price is very much interested in ensuring that our funding dollars are going towards evidence-based practices. I think for many areas, opioids included, we would like to have more evidence for the policies that we are putting forward dollars to implement.
Then the last area of the strategy is improving the practice of pain management, which I think is a fundamental issue of how we got to where we are. There were changes in how opioids were being prescribed. We can’t – although illicit fentanyl and fentanyl analogs are sort of an acute issue that is very concerning, we have to address the underlying drivers of the issue.
The Department had largely focused areas on abuse and overdose over the last couple years and then separately working on pain-related issues like the National Pain Strategy. We wanted those to be very much intertwined because they are a part of the same issue. There are millions of people in the U.S. who have pain and who need access to high quality evidence-based pain care. They should not be harmed as a result of our work on reducing abuse and overdose. We have to pursue policies that are balanced.
That is the opioid strategy in a very brief clip. The second area, I just want to talk on 21st Century Cures. As you heard some specifics from ONC on their section, the Cures Act is a large piece of legislation. There are over 200 sections in the Act. Largely, at the end of the last administration, since it was passed, it was sort of a decentralized approach to implementation. When new political leadership came here, they tasked us and ASPE to really think about how are we looking at this across the Department. Congress will come calling and saying where are you on Section 1003 or whatever it may be.
So, we went through a process to not only look at overall implementation, but identify high priority sections, where we felt like the Office of the Secretary, the Secretary’s councilors, in particular, would want to have visibility and early input. NIH has their own section. FDA has its own part of the bill, SAMHSA, CMS. Largely, they would be carrying that work forward and we would see it at some point, but there were certain things like the Innovation work – there is large funding streams for both FDA and NIH innovations. There is a funding stream for opioids. Those are things where we need to see it from a broader perspective of the Department to see how all of those things are fitting together. For those particular sections – it is a little bit more than 50 sections – we actually went through and said who are the players in the Department who need to be engaged on this. While it may sit in NIH clearly as an implementer, the policy decisions they may make have implications for FDA or CDC or CMS and we need to make sure that we are coordinated on our policy work.
So, we are using our Strategic Planning System, which also is housed within Science and Data Policy in ASPE, to monitor implementation. Now, we are just starting that process of actually creating what are the action steps deliverables under the sections. Many of them are spelled out in statute. Like FDA has to have a public meeting in a year, issue a guidance within 18 months, issue a final guidance within three years, et cetera. We want to be able to keep close tabs on that to see, one, where are we being successful, but also where are there challenges that are coming up.
It is a complex piece of legislation. Lots of things are already moving. Some of them were relatively self-implementing, but we continue to see – I think NIH put their information out around certificates of confidentiality last week – maybe it was earlier this week. All the dates blur together. That is an area where it is an NIH policy, but it has implications for actually even outside of HHS. There has been an interagency working group who has been engaged on those things. We are trying to look at it from an enterprise perspective. This is – congress gave us a lot to do. We want to make sure we are doing it thoughtfully.
It is also in the period of transition, which has made it even more complex. There is stuff that is happening and had to happen because congress gave us timelines, but political leadership wasn’t in place in some agencies. There has been some starting and stopping trying to move that forward. We felt like we need to have some way to very quickly assess our progress on that.
I think I will leave it there and certainly, entertain any questions.
DR. STEAD: Is that overview of who is doing what in the Department available in a form that we could browse it as we think about our work plans and how to – what the right connection points are for us in ways we might be helpful?
MR. JONES: Not really. I think if there are particular sections of the bill or thematic areas that you are interested in, we can connect with our Data Policy colleagues to get the right folks here for a discussion on that. It is pretty much an internal document.
DR. STEAD: In many ways, we are trying to ask you where we can be most helpful. We bring a unique lens in that this is a very interdisciplinary group. Many groups like this are less interdisciplinary. You can get the cross-talk between public health, healthcare, data, standards, privacy, confidentiality, in a way it is very hard to get any other place. If we knew which parts – for us to guess from the outside where that is most needed makes us sort of a blunt weapon. If there is any way you could let us – just keep in mind, we obviously want to be as helpful as we can.
MR. JONES: Sure. And as we are continuing to build out the specifics of what people are looking to do within each section, I think we can provide that lens to think through where there are places where we would want to engage in that way.
MS. LOVE: Thank you for that excellent overview. I am especially interested in the opioid strategies that you laid out. I am working – work with state health data organizations that are aggregating claims and hospital and emergency department databases. One of the things that every state is making a priority, of course, just like the Federal agencies, is the opioid crisis.
They are drawing on these data from the emergency rooms and monthly feeds from payors for commercial, Medicaid, and Medicare Advantage, which the Feds apparently don’t have. They are finding interesting things like nurse anesthetists in some states are a big provider of chronic pain services and the variation in prescribing.
I guess my question is as you go down these strategies, can we figure out how states and the federal agencies can work together? Setting up the CDC Injury Surveillance System is fine, but these are claims databases that are coming into states right now on a monthly basis. The emergency department – we have over 40 states with monthly feeds from emergency department, all payors, 100 percent of the patients in acute care hospitals. How can we sort of link these together?
The second issue is, at the same time that they are doing this, we come up with 42 CFR Part 2 that says the states and others cannot have substance use disorder data, which is a collision course that is beyond my comprehension to understand because I don’t know who is getting that data and what they are doing with it. We are meeting October 2nd in a workshop with our SAMHSA friends and states to noodle this out. So, I guess the bottom line question is can we work better together as states doing similar work with the Federal agencies?
MR. JONES: I think that is absolutely a goal. One of the things that Secretary Price has been doing is actually traveling around to hear from folks on the ground, often working with the governor’s office to set up roundtables of key players to hear where there are successes and where there are challenges, in particular, where there are federal activities that are getting in the way. HIPAA – some of it is misinformation around HIPAA and Part 2. Some of it is actual barriers. Those things are clearly coming back to help inform our policy here. Both of those areas are being looked at.
I think through both the SAMHSA and CDC funding, there is a desire to leverage what is actually happening. Not necessarily create new stuff, but how do you put it together? So, CDC has another funding stream through the Injury Center. It is called their Prevention for States Program. At least half of the states get funding. Part of that is to like link different data sources together, to give the state the ability to do that, and then use it to identify where are there hotspots in the state, where are there areas of need, and how is epidemiology there potentially different than other places. SAMHSA, through the large funding stream through Cures, does support some of that work. It also requires states to do an assessment of what is going on in your state. You really can’t move forward with what policies are you pursuing with this funding without doing that.
MS. LOVE: What makes this a challenge is they don’t get any of these federal funding. They are 100 percent state funded or almost 100 percent state funded. I think sometimes that gap is just because CDC is funding a different kind of data. They may not look to other existing data.
MR. JONES: Oftentimes, CDC’s funding is going to the health department.
MS. LOVE: These are in the health departments, too.
DR. ROTHWELL: Denise, if I could interrupt, if you want me to contact Deb Houry, who heads up the Injury Center –
MS. LOVE: Yes. That would be helpful. I am just trying to make a point. Thank you for listening.
DR. COHEN: Charlie, I have a question for you. You are very familiar with what this committee does. It sounds like you are incredibly engaged with the Commission’s report. What would you suggest would be a sweet spot for us to help HHS move forward implementing the Commission’s recommendations?
DR. ROTHWELL: Now what I would say is I would go to the – their set of recommendations, which are really only on two pages, and that is a good way to start. I hate to tell you which ones to look at first because it will tell you my biases.
DR. COHEN: That is what we want to hear.
DR. ROTHWELL: I do think it is broken down in improving private and confidential data access, modernizing privacy protections for evidence building, that one. The recommendations – it is 3.31, .32, and .33 – probably are important. The recommendations with 5.1, 5.2, 5.3, where they are really talking about OMB reorganizing. As OMB reorganizes their activities of how they look at data and how they support the collection of data from an chief information officer perspective, a data science perspective, and a statistical perspective, that will, in fact, drive a lot of what is going on in the individual departments, frankly.
I think that you will see here, when you look at these recommendations, how your committee, as an outside FACA committee, could relate directly to the Data Council and also the Office of the Chief Technology Officer in the Department and also ASPE and also the Chief Information Officer at ONC, to look at where and how we can work together and make recommendations.
I have been talking a lot about data linkage. You know, there is a lot of times you don’t need to link data. The last thing we need to do is spend a lot of time trying to link stuff that is not going to be used. So, there really has to be a concerted effort to look at where the sweet spot really is and where we should be linking within the Department and outside of the Department. You might want to think about what information you think we should be looking at. That probably – first of all, you will need to – if you don’t have the data warehouse, if you will, of information that you would like to see, then that is something that you could ask the Department for. Then take a look at where you think there could be possibilities of data linkage and analysis across the Department.
The other is what I was trying to talk about in the – for the future. I will just use CMS as an example. CMS has done a fantastic job, in my opinion, in trying to take their data, which was used for the largest programs in government and trying to make that data available for analysis purposes for ASPE and others. That was a very expensive, time-consuming process because it had to be done based on existing systems. As those systems change, we need to make sure that if there are areas where we want to pull data out of, we need to make it so that it is easier for them to do, instead of trying to do data curation, which we call it, after the fact. That slows everything down.
The other is if you feel – it is up to you to see whether you feel – maybe you will look at this and say, well, the Department is already doing this. Don’t worry about it. I don’t think you will.
The other is – I will just take my agency on. NCHS has to be – we have to be hit over the head. Yes, we have been flat funded for a long time. What does that do? That means that you hunker down and you stay with your existing systems and try to keep them going. Well, you know, maybe it is time – just the opposite – to look where there are other opportunities. This, I think, would give us a framework to do that. That is my concern, frankly, of my counterparts in other departments. Some of them are in much, much worse shape than NCHS is. I don’t want them to run away from this because, what are we going to do. They should be running toward it because this might just give them the key to their future.
DR. STEAD: Thank you.
DR. ROSS: Given the time, I will just make a quick comment. Chris, you mentioned on the opioid challenge, under key strategy of improving public health data and report, you mentioned NCHS. We just finished earlier this week, a two-day hearing about vital record data. Just one comment to make and make sure that you have the Secretary really actually understand, there is broad federal benefit – I think I am going to synthesize one of the things that I think we heard at this hearing – broad federal government benefit and also private industry benefit from birth and death data. You think about opioids and death data. We lean heavily on the states to just produce it. We lean very heavily on the states to modernize their infrastructure, give them virtually no resources. NCHS has a little bit of money that they work with the states, but it is probably in the order of a tenth of what is needed.
Not to argue Charlie’s case, but the reality is more real-time death data would help inform this problem in a big way. A lot of progress could be made. But they are going to have to put some money – I think the federal government, HHS, is going to have to put some money to that – very pointedly to vitals infrastructure, not just splattering money on public health generally. I think probably, as tough a pill as it is to swallow, to aim something so narrowly categorically, if we don’t – if money were to come and we don’t aim it that way, it is not going to give you the result you want.
From what we heard, it is huge potential, huge benefit, small amount of investment, but investment is needed.
MR. JONES: I appreciate that comment. I think some of the fundamental issues with mortality data are not the purview of CDC, per se. They get the data. They turn it around pretty quickly. It is what is happening at the state and local level, whether it is specificity of coding and documentation or even just the process of turning that around from somebody signing a death certificate all the way up.
There is a separate group that is chartered under OSTP and National Science Technology Council that I co-chair on medical legal death investigation. We have been looking at this more broadly. We make funding decisions on causes of death – so, cancer, heart disease – because they are large numbers, we give them a lot of money. We know for other areas that some of those heart disease deaths, maybe they are overdoses. It is just a 65-year-old who didn’t wake up and they classified it as cardiovascular or whatever. We are potentially not even aligning our funding with the realities of what is happening. So, I think we do appreciate that. DOJ has a little bit of funding that also goes out to support medical legal death investigation. We are trying to figure out, in the budget environment that we are in, how do we craft an ask that resonates. I think opioids is the potential lever to pull because it is such a pressing issue and it is so complex. With the emergence of many, many synthetic opioids, our system is just not well equipped to capture that.
DR. ROSS: I just worry we are piling on to the state vital registrars to modernize, do all sorts of stuff. I think a willingness is there. I think if we are going to play fair, I think the feds have a share.
MR. JONES: I think we recognize that funding is the largest role for us at the federal level. It is the approach to getting it.
DR. ROTHWELL: It is just not the state vital registrars. It is also the medical examiner system. They need that support, too.
I would like to make one other comment on this. I will just take our role as a federal statistical agency. We happen to be located in CDC. If we weren’t in CDC – we have been in a variety of places over the years – my guess is that we wouldn’t be as fast reporting right now as we are because CDC doesn’t really care about vital statistics or our surveys. They want data right away for surveillance purposes. That has been the pressure on us. On the opioid issue, we would be perfectly comfortable, as statisticians, to stick with the ICD. That has been our business. Well, we need to do natural language searches on the drugs that are being reported and be much more specific. That is the surveillance perspective.
What I am getting at is as we look more holistically in the Department, I think people like my program and other programs in the Department have to think broader about their systems. So, again, getting back to this Commission activity, I think if we play it right, it will improve NCHS systems and it will improve a variety of other systems within the Department and other departments and then allow folks like in ASPE to be able to really guide the Department more appropriately.
DR. STEAD: Thank you. We need to bring this to closure. It has clearly sparked common interest across the board. I really am grateful for you taking the time to come and help us think. Thank you. We look forward to further interactions.
MS. KLOSS: We thought it would be useful to move, as we guide this next part of our discussion. What an interesting segue we had just talking about code and classification and vocabulary support for the opioid initiative. A very good transition.
This next hour we have labelled as part two of our work on health terminologies and vocabularies. With me is Donna Pickett, Chief Classifications and Public Health Data Standards Chief, and also the NCHS representative to the World Health Organization family of international classifications. You heard Donna on the phone at our June 21st meeting, but she couldn’t be with us in person. So, I am very grateful that she is with us in person today because she will nudge me when I go off track. As Bill says, I am a generalist. I am a generalist in this terminology and vocabulary world. I have been interested in it for many decades, but I am not a deep expert.
We also, I believe, have on the phone our colleagues from the National Library of Medicine. Vivian, are you there? Okay, well, they may be dialing in. I had an email. They had intended to be here and then decided that they would have to dial in.
I am – the agenda shows me as co-chair of the Privacy, Confidentiality Subcommittee. I am very much wearing my hat as a member of the Standards Subcommittee in this discussion because this is not a privacy issue. This is a standards issue. A little factoid that will be of interest, I think, to everybody, Donna did some good background work for me in preparing this slide deck. One of the things that she found is that the Standards Subcommittee used to be called the Medical Classification Subcommittee of the National Committee on Vital and Health Statistics. It got changed to be the Standards Subcommittee – it got broadened when HIPAA was enacted.
Part of our goal here as we look at kind of our overall NCVHS charge, certainly anchors in uniform data standards for patient medical record information and advising the Department on health data collection needs and strategies, which is where health terminologies and vocabularies fall. As we undertook this, we said that what we were going to set out to do is do a contemporary look at health terminology and vocabulary landscape. We are still in the environmental scan. What this slide deck does is kind of add some additional information to the scan, but by no means completes it.
I do want to say that this slide deck is kind of as of September 1st. I look at it as an evolving document. I will say that I had some corrections made to the slide deck through Donna’s review. The colleagues from the American Medical Association reviewed the section on CPT and made some corrections. Vivian and her colleagues at NLM emailed some corrections to me that are not reflected in this document, but we will get back with them and make sure we are framing all of these systems accurately. Look at this as very much a document that is a work in progress.
We thought we didn’t need a big narrative report right now, although our goal will be out of the environment scan, to produce an environmental scan report. I thought for prompting dialogue that this would be the best format.
So, you know our goals here are to look at the changing environment and implications for timing and approach to health terminology and vocabulary standards; needs, opportunities, and problems with development, dissemination, maintenance, and adoption, and actions that HHS might take to improve development, dissemination, maintenance, and adoption. You notice that none of those say deep dive into the content of the code sets. We are looking at governance and coordination and timeliness and actually all of the things that we have talked about through this week, whether we were talking about vitals or predictability roadmap beyond HIPAA. There is a theme that is running through our work for sure.
The purpose of going through this slide deck is to tee up discussion. So, I am just going to expect you will flip up your tent cards on any slide where you have questions or comments to make and then we will tee up a little more formal discussion. So, I am going to go through this first part fairly rapidly.
This – there is a lot on this slide and I think as we do a narrative report, we will draw on the content of these reports. This is a 20-year view, a two-decade view of what NCVHS has done around this topic and those two points of that charge. I had a recollection from the 1990s that the Committee had articulated some principles for guiding choices for the standards in this area. As I said, Donna went back in the archives and learned a lot, actually. Right?
MS. PICKETT: It was very helpful.
MS. KLOSS: Yes. To the point of Bill’s earlier work in reviewing past annual reports, it was very helpful.
Now, I am not going to bring those principles forward for discussion today. I don’t think we are quite at that point in our environmental scan study. We have kind of accumulated the body of knowledge from which to use as a launching point. The Committee did a Uniform Data Standards for Patient Medical Record Information work in 2000 and a report on Information for Health: A Strategy for Building the NHII Infrastructure in 2001. I will say that this report in 2000 is really the report that made recommendations, for example, that led to the development of UMLS. Really important recommendations that we are benefitting from today. That was 2000. This is 2017.
In 2003, the Committee formally recommended that HHS should initiate the regulatory for ICD-10. There was another letter for adoption of PMRI standards. We did a series of letters supporting ICD-10 implementation dates through 2014. The last really major, transformative kind of set of recommendations in this space from our committee was 2003.
Any questions on that history? All of the report links are there. Anybody who, you know, wants to do a deep -dive, you may. This is our – this is right in our wheelhouse. It is in our charge. I hope that through this exploration, we will look at some directional strategies. Any questions? Any comments? I am just going to keep going because I want to get to the point where we tee up the discussion.
I thought it was helpful to just define these terms. I think there is a lot of use of terminology versus vocabulary versus code set and classification set. I won’t read these, but, in fact, we didn’t have to define them anew because the NCVHS did this before. I don’t see any reason to need to update them. These were part of that seminal 2000 report that defined these terms. This whole world – for this project, we have labeled it as health terminologies and vocabularies, understanding that a terminology kind of subsumes codes and classifications and vocabulary is a broad set of specialized terms that facilitates precise communication and that is is a synonym of nomenclature.
Did someone just dial in?
MS. AULD: This is Vivian.
MS. KLOSS: Hi, Vivian. I indicated to the Committee that you would be joining us from NLM. I indicated that you had sent some corrections that will get into the formal report after this meeting. Thanks for your help.
MS. AULD: Excellent. Thank you.
MS. KLOSS: So, when we had our meeting in June, we deliberately started by looking at terminologies and vocabularies that are developed, maintained, or disseminated by the federal government. Specifically, we looked at the CDC, NCHS, the ICD-10 – Donna provided that report to us. We heard from CMS, ICD-10. We didn’t really talk about HCPCS, but we will talk a little bit more about where that fits here. We heard from National Library of Medicine on SNOMED CT, LOINC, and RxNorm, and the UMLS, in general, as the source – the United States source for these systems.
That was why we designed June as we did, but we knew that wasn’t the whole picture. So, in figuring out how to frame that whole picture, we are just presenting now a series of tables that kind of begins to round out the whole picture.
I will say that I had kind of a struggle in how to group these. There are a couple of slides that really reflect those that are named in regulation and then a couple that are – you will see the structure as we go through as a way of framing. I want to say that one of the points that Vivian made was that I didn’t make clear throughout these that all of these standards are – all of these code sets are available through UMLS. UMLS has them available. There is a licensing process for individuals or agencies or groups to be able to download them. This doesn’t explicitly say, but I just want to underscore that.
So, this first page then are the systems that we heard from from NLM – SNOMED-CT, LOINC, RxNorm. We just kind of organized the input that we got with regard to their purpose, ownership/governance, maintenance and dissemination and release schedules. That format is what we have used for all of them. We are not going to walk through this because this is background. There won’t be a test. But I think it is useful to kind of see how – what the logic is.
Then we move onto the systems that we looked at and are in use in the U.S. in the ICD world – ICD-10, which we talked about earlier in the week a lot, the International Classification of Diseases, the native ICD-10 that is used in this country. And there is a statutory basis under which it is used. We just didn’t research what the statute was at this point. This is what we use for cause of death coding on death certificates. We don’t use ICD-10-CM. We use the native ICD-10. Interesting little wrinkle. The ICD-10 in the U.S. has been used for mortality coding since 1999. We were using ICD-10 in this country long before we implemented ICD-10-CM. What we use for morbidity is ICD-10-CM and ICD-10-PCS. As you heard in June, ICD-10-CM is the purview of NCHS. PCS was developed for use in the United States, brand new procedural coding system, new design. CMS maintains that.
Any questions about that? Just to make all of these relationships clear.
Then we move on and other named standards are the Code on Dental Procedures and Nomenclature (CDT), Current Procedural Terminology, which is – both of those are named HIPAA standards. For CPT, it is a named HIPAA standard as HCPCS Level I. Then there is also the named HCPCS-Level II, which are codes that identify products, supplies, services, ambulance, other things that need to – other data that needs to be captured, again, purpose, ownership, development, and maintenance and release schedule.
MS. LOVE: This may not be the best place and time so I will respectively withdrawal, but I get a few calls on occasion about the CPT. We have states with transparency laws to publish shoppable outpatient conditions pricing. The AMA has pulled the – that is a proprietary code. You have to pay us to do this public reporting that is being done under law. I think the AMA is trying to work this out, but it is on a state by state sort of licensing agreement. What troubles me is it is named as a HIPAA standard, but it is a proprietary dataset. I don’t know what you do about it. You don’t have to really – it just kind of confuses me. We are trying to work with AMA. It really puts a monkey wrench into public reporting.
DR. RIPPEN Just a quick comment to follow up on that. I don’t know how to summarize it, but if there are any other cases, just a flag as far as that there might be a nuance that relates to ownership or –
MS. KLOSS: I tried to be pretty specific on ownership.
MS. LOVE: We have only had CPT as an issue.
DR. RIPPEN: Just some notation as far as cost – associated cost. Just their business model.
MS. KLOSS: As an additional column?
DR. RIPPEN: Yes, or somewhere flagged out because there are implications from a policy perspective and a diffusion perspective.
MS. KLOSS: Finishing out the named standards, the National Drug Code, which Vivian reminded me actually is part of RxNorm. RxNorm includes the NDC. I don’t think my description of the structure of this code is quite accurate. I will improve that.
You can see there is quite a list of named standards. I think for the most part, they are public domain or available under some kind of licensing arrangement, such as SNOMED-CT, there is a relationship through NLM that makes that available for use in the U.S.
MS. PICKETT: To Denise’s point earlier about proprietary code sets, in the NPRM, the first NPRM related to HIPAA standards and code sets, there were some principles, as Linda mentioned earlier. One of those principles was that it be publicly, freely available. However, in the write up in the Final Rule, CMS did discuss at length those code sets that had implications of use because of the proprietary nature of the code sets. Again, initially, it was a de facto standard. If many of you recall, for the initial standards under HIPAA, they were truly looking at de facto as opposed to identifying something that had not been well-tested, well-used, or had any familiarity with possible use in a larger environment.
MS. LOVE: That perspective is helpful. I do think AMA at least has tried on a couple of states to be responsive.
MS. PICKETT: If anybody would like this citation, we could also probably put that as a footnote for the slides.
MS. KLOSS: Okay, I just think it is important to also look beyond what is named standards. We heard, in June, about the International Classification of Primary Care. This is part of, as it says, the World Health Organization family of International Classifications. It is considered a related classification, developed by another organization, but considered related to the family. A family relationship, not a child.
The International Classification of Oncology, 3rd Edition, this is actually used in the United States, though it is not a named standard, but it is used in tumor or cancer registries. This is considered a derived classification because it is based on ICD. We do use that one in the U.S.
The International Classification of Functioning, Disability, and Health, again, that is a reference classification, according to the World Health Organization. It is a separate classification with its own vocabulary.
Donna indicates to me that that is not really used much in our country. Some of the functional disability codes now exist in ICD-10, some of them.
Then there is International Classification of Health Interventions, which is like a procedure coding system that World Health Organization wants to develop that would replace the old procedure coding system that in our country, we have replaced with our own PCS. This is not likely to ever be used in the United States, but it is under development for use in other countries that otherwise may not have a procedure coding system alternative.
That kind of finishes out our view of the International Classifications part of the World Health Organization. As I said, Donna is our representative to that, WHOFIC, The World Health Organization Family of International Classifications, to unbundle the acronym.
Also in use in the United States is the DSM. As we know, that is the Diagnostic and Statistical Manual of Mental Disorders. It is not a named standard, but it certainly is used by psychologists and psychiatrists and mental health professionals for their purposes. International Classification of Nursing Practice – I guess that is also World Health Organization. Medical Dictionary of Regulatory Activities – these are – I just went to UMLS and I pulled some others out that I thought were of interest in covering areas that we hadn’t seen covered in the other categories – DSM most certainly, Nursing Classification, MedDRA, regulatory activities. Beyond this, there are another hundred systems in UMLS when you look at all of – there is multiple nursing classifications, for example. A lot of these systems are available through UMLS in other languages. It wasn’t possible or probably necessary to go through all hundreds, but these were just ones that I thought were of interest and perhaps should be highlighted.
The next slide is the same. The Gene Ontology work – actually, that was just added fairly recently, 2016. You can see that there are new ontologies being developed around emerging areas. That was why I pulled that one out. That was just six that were of interest to me. Just showing you, if you will, the breadth. So, we are still in our standard work, primarily focused on those that are named standards, but this is a big environment.
With that and without going into any other detail, if there are no more questions, I would like to kind of move us with our remaining time into the discussion.
It seems like, based on what we are setting out to do, that the variability that occurs and perhaps out of that, the opportunity, comes in three chunks. There are things relating to governance and coordination. It is not possible on the tables to kind of describe all the levels of coordination, but, for example, Donna just came from two days of the coordination and maintenance group that oversees changes to ICD-10-CM and ICD-10-PCS. That group meets twice a year. Yesterday, there were 40 proposals with regard to ICD-10-CM. Coordination.
In my conversations with the American Medical Association, they explained to me in more detail the kind of coordination that exists between their work and CMS for purposes of maintaining CPT as a named standard. There are a lot of levels and layers of coordination. I guess – we will look at some of the questions I have raised around this. There probably are other questions. I think this will kind of take us where we are going next.
There are – there is great variability in development processes, maintenance processes, supporting technologies, and I think if you read any of the detail in those lines, you will sort of start seeing how that comes out. There are some best practices. There are some areas of consistency. But there is a lot of variation. We also see variation in how these systems are used and in the research and evaluation that does or does not support their peruse testing or their ongoing evaluation.
I think those are kind of the issues around our – what I had kind of hoped to do was just lay this out and throw out some discussion. With regard to coordinating and governing bodies, these relationships are all kind of different. As I look at these relationships, three questions came to mind. There may be more. Is the current level of directional and operational coordination across named standards adequate? Are the paths for sustainability adequate? Kind of what I mean by that is there is some differences in how things have evolved. I served on an AMA CPT-5 advisory committee many years ago. Since then, what they have shifted from was version numbers to the 2018 version. It is just an annual process. The system doesn’t get thrown out and revised. It gets updated. What we saw in ICD from 9 to 10 was a complete redesign. Certainly, it was based on 9, but it was a significant enlargement. The biggest change has been seen in procedure coding in this country. We know procedure coding is the basis for DRG’s and payment still. That was a very big change. That is what I kind of meant by going forward. How do they evolve? Do they evolve by earthquake or do they kind of evolve – are we going to have another earthquake in ICD-11 is one of the questions that comes to mind? Is PCS being evaluated so we know how it needs to evolve?
Then research – is there research on how best to measure validity and fitness for use? At what points in the process is that being applied? Is that research adequate?
Are those the right questions to advance our thinking? Are there other questions? Are there questions about the current coordination in governance?
DR. RIPPEN I was going to add maybe a different topic.
MS. KLOSS: Add it.
DR. RIPPEN: The other is when do you know that there may be another standard that you might want to consider as a named standard for inclusion? For example, since we are focusing so much on genomics, when does it become core because it ties to treatment and all other kinds of things? Just kind of a nuance as far as how do you know.
DR. COHEN: I think understanding the interaction of uses and the dynamic nature of change in use is a key question.
MR. COUSSOULE: As I look to the way you framed the questions, is the current level of directional and operational coordination across named standards adequate, follow that up to say adequate for what. To me – again, I am always a think with and in mind kind of guy. I think framing up the either problems or concerns that are trying to be solved is really necessary before you can answer that question. Not just problems and concerns, but opportunities as well. I don’t want to make it sound like it is a response to negative, but also a potential for positive.
MS. KLOSS: As you think about our environmental scan, this would be – this whole governance and coordination is a worthy area to do further exploration around based on uses – evolving uses and needs that aren’t being met by current coordination and governance approaches.
DR. RIPPEN: And maybe, going back to the positive, when things are working. So, we can maybe apply it to other areas, too.
DR. STEAD: Another thing and I am not sure where in your sequence of issues this will come in, but we need to pay attention to maintaining the archival history and relationships of some semantic meanings. I am not yet sure the degree to which this group really has gotten a gut level understanding of what the UMLS meta-thesaurus actually does. I really think to get to where we have time to really understand and ask the questions – because it now has close to 200 vocabularies. It, in essence, keeps each of those vocabularies in their native form with their native definition. What it does is link related terms to a base concept. It isn’t another standard. What it, in fact, does is turn the set of vocabularies into a knowledge resources where you can easily see what the links and relationships are. Then I begin to show my ignorance. I believe it does, in fact, handle longitudinal versioning, but I can’t promise that.
There is a piece of this that is around governance and coordination of process. Then there is another piece of it that is around the technical informatics approach to how we preserve meaning in a time when meaning is changing. I think to the degree we can figure that one out and I think that is where UMLS has maybe gone the furthest – it actually backs – it has been evolving since the early 80s. It was preceded the 2000 pieces that were part of the HIPAA relationships and Meaningful Use and so forth.
I think that is a part of this. It probably deserves its own slide at some point because it is a different concept, I think. Yet, it makes this a lot easier or it could make this a lot easier.
DR. ROSS: Do you know the answer? Is Bill right that it actually does versioning over time?
MS. KLOSS: I don’t know. We will ask Vivian.
MS. AULD: Yes, it does to a certain extent. In fact, that is something that we are very careful of. We tell people once the vocabulary is in there, we will not take it out unless we are forced legally to do so, simply because it has all of that historic information. It is intended as a resource so that as you go back in time, you can find out what a code meant at a certain time. If you go to the ICSU(?) website, you will see that the antecedent version of SNOMED are no longer available, but they are still available through the UMLS simply for that historic and research perspective. We have made it clear all along that that is what we intended to do. Everything was set up so that we could continue to do that.
One other comment that I wanted to make is that we probably want to add something in the chart, add a column where we are indicating whether or not the terminology is regularly updated in the meta-thesaurus. There are a couple of them there that we do not update. The two that I know are old versions are ICD-10 and ICF. In both cases, we talked with Donna and her staff and came to a decision that it wasn’t worth the resources that would be necessary to get a copy and the negotiations that would be necessary with WHO, et cetera, in order to keep them updated. With ICD-10, it is a closed group of people who need them. They won’t necessarily come to the UMLS. With ICF, as Donna said, nobody is really using it. I would be useful to have input from the committee as to whether or not we are making the right decisions there.
MS. KLOSS: Interesting. Just to take that comment back to the use we talked about earlier, how are mortality coders kept current with the latest version of ICD-10?
MS. AULD: They go directly to NCHS to get the ICD-10 update.
MS. KLOSS: They get the files through NCHS. Okay.
MS. AULD: Yes.
DR. RIPPEN: I would just suggest that – given that ICD-10 codes can cause blips in diagnoses, it may be important to be able to go back historically.
MS. KLOSS: The next set of discussion provoking questions are around the maintenance processes and the development processes, that second chunk. They generally have – all of the maintenance processes sort of share this theme of having a steering committee and having clinical advisory panelists in the field, and a process for public input. I certainly appreciated, as I went through this, that things have become more transparent and open over the last decade since I last looked at this more closely. There has certainly been an evolution in that regard. The guideline development process, as well as some sort of communication and training and then a Q&A process to get field questions answered. They kind of share those common characteristics, although it seems like each works quit differently in some ways. Is that true?
MS. PICKETT: That is true.
MS. KLOSS: So, the questions it prompts are are maintenance processes regularly reviewed for improvement? Aggressively enough? What technologies are used for this? How manual is it versus – you know, are we using the kinds of tools that could be used in that regard, really true data governance tools? You know there are a lot more robust technologies now for managing definitions over time of code sets. Are there any steps missing in these processes for best practices? Anything else in this that comes to mind?
Then development. I think ICD-11 is on path. It is a good example for us. It is one of those things where I am going to suggest going forward, we take a little deeper dive also. Is it being designed so that it is kind of a major redesign? Are we going to get to this rolling version rather than the explosion? And the timing and we need to get, I think – through the Standards Committee, get predictability roadmap at some point to consider recommendations for the evolution from 10 to 11. Doing that sooner rather than later.
DR. RIPPEN: Just a quick question, this is the development and maintenance of the standards, themselves. Do we touch upon or is it important at all to touch upon the diffusion of those standards and the updates within the technologies?
MS. KLOSS: Yes. We do. When we look at dissemination, we have both the UMLS as a unified source and then the various agencies and organizations with their own distribution. A bigger circle for all of the commercial sources for many of them, where these versions are repurposed. I guess one of the questions that comes to mind is really do we need to disseminate through all of these? Could it all be off of one platform? I don’t know. That is a dumb question. There is cost associated with everybody maintaining these sources. There may be some version variability.
DR. ROSS: Is there a documented problem because of that? Is there a versioning problem because of that – multiple platforms?
MS. PICKETT: I will just speak from our perspective as the maintainer for 10-CM and previously, 9-CM. We have not had any problems. If there have been, they haven’t been with the official files. It sometimes has to do with a publisher adding bells and whistles to their printed version that sometimes leads to confusion from the users. An example, they will see a definition in one of the various published books. When there are questions about it, they will call us. However, we don’t include definitions in the official version. We refer them back to the publisher. Other times – this is probably in the past, but could very well be happening now. Occasionally, when we have released the official files, there may be some typos in it. Some of the publishers thought it would be nice to help us. They made the corrections in their version. Well, of course, now, you have a disconnect between the official files, which are updated once a year, and a published version. Again, the official version is the official version unless we issue an ARATA. You do have those quirks.
Sometimes, it is made more complex by a particular payor’s interpretation of a code and how they use it, again, beyond scope for the official version. We will get questions if a publisher has included information about DRGs or that this is something that you wouldn’t be admitted for. Again, that is outside what we do in terms of updating and maintaining the classification. Again, it is a question that will get asked. The typical response is you have to talk to the publisher of the book because we have no knowledge of what we did. We are kind of like not the police in terms of what bells and whistles have been added. So, we are aware of those things, usually anecdotally, if somebody calls us because they know there is a problem. As part of our official maintenance or publicizing the fact that someone may have inadvertently introduced errors, that is not something that is kind of within our scope.
MS. KLOSS: It is interesting in our increasingly digital healthcare system that a lot of the dissemination method is books, published books. In terms of update, if you look at RxNorm, they are updating some codes daily. That is a very dynamic – whereas, in some ways, because of the process and books, we are otherwise issuing official updates yearly and locking down those official versions, let’s say with the case of CPT, in the summer in order to give all of the developers and everybody a chance to do this and for the books to get published for something to be effective in January 1.
So, it does seem like, looking forward, we might look at how dissemination takes fuller advantage of technology.
DR. STEAD: Two pieces, around dissemination, the fundamental shift that I cannot imagine will not take place at some time is from the idea of dissemination and incorporation into everybody’s systems, which then are on a certain version, to the idea that there will be a service that, in fact, can unpack any version and tell you what it was. We will not get to predictability as envisioned in the workshop if that isn’t the end game for dissemination. That was a key piece of PCAST 2010. This is not apple pie and motherhood. It is, in fact, doable in today’s world. It is a fundamental shift in thinking about it. I think where in the continuum that would become mainstream I think would be when you would really get a change.
I think for those of you that don’t live within the RxNorm and Daily Med world, the key breakthrough there was getting to where – when the FDA issued a new product structured label, in essence, when there was a new manufactured product, which is a brand of combination, if you will. At that very time that was released and that then is that day or week mapped into the vocabulary so that as you can imagine, the same thing would take place when a new testing device – a new result became possible because there was a new testing device. You would actually begin to get these things that are, in fact, manufactured data. The drug is a manufactured, the drug test – you would actually get to where those releases were occurring at the source. Therefore, you never got in the situation where something is not available when you are in the system beginning to use something because it started out at the source.
Those are two examples of a complete shift in the thinking that is a little bit analogous to what the Commission on Evidence-Based Policy was proposing, where we shift to this new framework, confidentiality framework, and everything else flows from that. Fundamental shifts in thinking that we will have to debate, which are and which are not ready for primetime when.
MS. KLOSS: And what research could be done to advance that.
DR. STEAD: We are going to run out of time. The one thing I wanted to know is whether we have heard the answer – I guess do we know the degree to which ICD – do we have any instinct about the degree to which ICD-11 is going to be an explosion and ICD-12 is going to be an explosion or whether they are going to be more iterative?
MS. PICKETT: That is a very detailed answer. Is the structure of 11 different from 10? Yes, by design. As Linda was mentioning earlier, ICD-11 – what WHO is trying to do is to look at real-world application, how things are being used now, “what is fit for purpose”, and how do you – what is your pathway forward.
Basically, what is happening is there is clearly an understanding that the previous update every ten years isn’t going to work. With 11, it has been built so that it is much more easily expandable. So, you won’t have these disconnect.
Now, that being said, if there is a scientific reason to change something because of a definitional change or something along those lines, that will be incorporated, too. It will affect your trends. But that is something that is a necessary change. How do you evaluate that?
Again, the ten-year tipping point for when you get a whole new version and complete chapters have changed, you may not see quite as much of that based on how it is being built. If the knowledge base changes significantly, clinically, scientifically, you are likely to still see some changes, but not in the same way where you have this set of three books and all of a sudden, they are thousands of pages larger than they used to be just because there was a different way of thinking about how a disease should be classified.
MS. KLOSS: I had the same question, Bill, and when I teed up where I thought we needed more study – if we could just focus quickly on that, we will be ready for lunch.
I identified three things that I didn’t think we knew enough about. I have added since a fourth, that is the UMLS. I thought if we did – for part three, we could do a little bit more drill down into the design of 11, a little deeper dive into what formally we learned about the transition from – to CM and PCS, and what is the status. What are we learning about how well they are working? Enough time has lapsed now. We should have some data impacts and so forth. Fourth, I would just like to take a little survey look at governance models from other countries for code sets and vocabularies. I thought three and then I added Bill’s fourth, the UMLS. I put that question to you. Do you think those would be productive areas for, again, the next chunk of shared learning?
DR. COHEN: Yes.
MS. KLOSS: Going forward, just from – we could probably save this. I was trying to think of, okay, what does this suggest for our work plan? Maybe we can take this up this afternoon. I thought in the first half of 2018, we would finalize our environmental scan report and use that as input for a stakeholder roundtable. That we could, in 2018, host a stakeholder roundtable, a working session to frame models and recommendations, kind of like predictability roadmap model. Then the Standards Subcommittee could give some thought to considering how this integrates into the predictability roadmap. Then I saw the second half of 2018 and the first part of 2019 being reserved for perhaps one additional hearing, a draft report, preliminary discussion of final recommendations to the committee, and then approval in June of 2019. I was just kind of looking out. I think, in terms of activities, I think we are going to need to do a roll up your sleeve roundtable with experts to help us understand the technology and governance and other dimensions of this as we tease out some specific recommendations. Just reserve a possibility of another hearing before we wrap this up.
DR. RIPPEN: Just a really quick question. You went from the development all the way to use and kind of assessing. It sounds like you are going to use ICD-10 as kind of an educating kind of use case of how things were done it sounds like.
MS. KLOSS: You know, I think we – because it is a core named standard –
DR. RIPPEN: That is okay.
MS. KLOSS: It may not be the best future model.
DR. RIPPEN: That is okay. The one thing you may want to consider as you talk about use is also the impact on not only healthcare, but researchers, public health. That is all.
MS. KLOSS: For sure.
MS. PICKET: That is a very interesting comment, Helga, because much of what you just mentioned certainly went into discussions at WHO in the development of ICD-11. A lot of that happened before we actually sat down to start looking at content. They did a landscape to see, well, how is 10 currently being used, readily understanding that it really isn’t only used for mortality. When you say, well, it is also used for morbidity, not every country uses the word morbidity in the same way. Other countries may have their own national modification of an ICD. Again, they may be using it for in-patient only with some headway into outpatient or physician office. There are countries like us who use it across the board. Our use case for morbidity is very different than the use case for France or Australia or Canada. Again, a lot of that discussion occurred at the front end for a change, as opposed to halfway through or at the end. They came to a realization that countries, even though they were using even native 10, weren’t all using it in the same way.
MS. AULD: We see a lot of similar parallels with SNOMED-CT and LOINC, as well. Both of them, at an international level, variability in how people are using it. The one thing that NLM is trying to encourage is now that SNOMED is being used for clinical encounters, how can we also promote its use for research and public health, as well, but especially research.
DR. COHEN: It would helpful in January or before to expand on this conversation to better understand the variation in uses of these basic systems. I would really appreciate that.
MS. KLOSS: Yes. Very good.
DR. STEAD: Thank you, Linda and Donna and our colleagues at the NLM. This is a – we have made a great progress in really turning this landscape into an organized view. I think you have got a good plan for our next steps. Thank you very much.
Now, we will adjourn for lunch. We have a shorter lunch so that we can get out of here at the time people want to. Please be back by 1. Thank you.
(Luncheon recess at 12:20 p.m.)
A F T E R N O O N S E S S I O N
DR. STEAD: We are now going to work on our work plans. And just to remind everybody of the process that we’ve laid out, we’ve used the time over the last month to get the input on the predictability roadmap, the vitals hearing, the TNV scan, and the first part of the Beyond HIPPA scan. We’ve heard from a subset of HHS leadership and the CEP. We’re treating all of that as input into our thinking about what our work plan ought to be doing forward.
And what we want to do this afternoon is take a block of time for each of the subcommittees to think with the committee as a whole about what the key options or alternatives might be in thinking at this juncture. This afternoon is not about making decisions about the work plan, although I think we do need to make decisions about how we’re going to use our time in November.
And our time in November we’re not thinking of as a face to face meeting, and I think we’re probably not thinking of it as hearings, but we are thinking of it as two days that have been held on our calendars that could in fact give us the opportunity to work together on key topics, not necessarily with all of us having to be engaged throughout all of the two days. But since the calendar is the biggest barrier to progress, that’s just an opportunity that we can take advantage of, but we would need to figure that out as we go through today.
So you might have in your mind which things would you like to barter for blocks of time in that period of time. Then we’ve asked that informed by this conversation the subcommittees will do the sausage making in their subcommittee meetings, and we’ve asked that each of the subcommittees submit draft work plans to Rebecca and me by October 15, and they will be distributed to the executive committee, which will use our October 24 executive committee call to try to synthesize a line, if you will match them up as a set against the strategic plan and the criteria that we worked on yesterday.
And then after that executive committee that will give us time to have the work plan reviewed by the different key stakeholders, ASPI, OCR, CMS, whoever else, the National Center for Health Statistics, to get feedback to us, and then we would finalize the work plan on our December 7 Executive Committee call. So that’s the journey that we’re sort of in the middle of, and this is the next step. So are there questions about the overarching process before we begin to hear from subcommittees?
MS. KLOSS: I do have a question. If in our discussions we think about how to use the November meeting time, isn’t that, could that not be part of this sequence?
DR. STEAD: Yes. We will be, I would think about that as being more doing actual work than working on the work plan. So I would think of, we might want to do a chunk of TNV background so that we didn’t have to do it in January for example. So what we might do now is sort of get the key chunks. As always the exact way we would use them would undoubtedly morph as we go through the process of finalizing the plan, just like the agenda for this meeting has morphed right up until today in fact. Flexibility is our middle name. Is that good? Then Bruce and pop health.
DR. ROSS: One quick question. ASPI’s work plan review, is this a formality or do they really pass it around, do they really chew on it and give us substantive comment back?
MS. HINES: They do. They will. One example to remind us all, Alix is not here, she and I have been belly aching over this, that the review committees required activity through the ACA every other year, and that has been on ice per ASPI. They’ve basically said don’t do anything related to the Affordable Care Act because there’s too many unknowns. So that’s an example where ASPI has been very clear, don’t bother with that right now, just to remind ourselves of that.
DR. STEAD: There is a natural interaction between our committee as an external FACA and our fiduciary responsibility to for lack of a better word say what we think is important to say. And aligning what we’re working on where possible within that frame with the things which will be most actionable to the department, and be most helpful, because the chances of getting uptake are simply much greater. And so it’s a back and forth process.
DR. COHEN: Okay. Population Health Subcommittee. We’ve learned a lot this week, up through this morning. So I would say there are two big buckets for us moving forward, and then one potentially smaller bucket. The first big bucket is moving forward with vitals. I’m not going to go through all the details of the themes that emerged in the vitals hearing, but clearly we have a lot of work to do. I’d say in terms of products that we’ll be working on, there will be a report summarizing the workshop.
Other potential products discussed include use cases to make apparent the criticality of the data and the current challenges, and we talked about substance abuse opioids, Zika for births or changes in early elective deliveries, and catastrophic events leading to mortalities, potential use cases as a product.
We talked about a scan and a diagram of vital statistics uses, really understanding who uses the data, for what purpose, where they get the data. And parallel to that is a diagram or a table or a short report that focuses on the funds flow, where the money goes and who pays it.
And then on the short-term, will a letter to the Secretary emerge as part of these ongoing deliberations. In the intermediate and longer term visioning certainly will be a key, re-imagining vitals, not to necessarily replicate paper records. Will our explorations over the short course lead us to want to hold more hearings and workshops to delve into some more detail about some of these issues? Will there be additional sets of recommendations from our future learnings so that we write additional letters to the secretary?
And will other products emerge? The Monday and Tuesday was such a rich conversation, should we be doing a white paper or vitals primer for folks who weren’t able to come to the hearing but who might benefit from something like that? So certainly vitals will occupy a significant part of the population health subcommittee work plan. I’ll stop there to get suggestions and feedback from folks about that aspect of our work plan.
DR. STEAD: I have two thoughts that have emerged over the course of this morning. One is given the comments and information that Chris provided to us this morning, is there a way, we had previously talked about opioids as one of the use cases.
Given the fact that it sounds like the point for the opioid effort within HHS is in ASPI and is reasonably close to us, could we figure out how to align, are there recommendations that we could make that would be helpful to the vitals effort in the short-term and helpful to the opioid effort? So that’s a little different from opioid, as you look for holes in the line it would seem to me that would be one thing that might be an opportunity that I hadn’t recognized.
MR. COUSSOULE: Just asking for clarification, Bill, are you thinking that some of the topic questions of the opioid epidemic topic this morning becomes the use case to push some of the short term needs or things that got identified in the vitals?
DR. STEAD: Yes. If we are going to make a difference with opioids in a way that you could maybe break open modest additional funding for example? If we could somehow, they’re doing things like the ED based surveillance. If there is a meaningful opportunity from short-term enhancements to in the death record, because I suspect that’s where we would be talking about this case, could we use that as a short-term step that would be beneficial to the department, to the secretary given their priorities, while also maybe getting first things done that would make the death reporting more robust.
DR. COHEN: It would be great if we could. I would frame the question understanding the opioid data ecosystem as really the issue, and it involves a lot more than vitals and it’s a lot more complex, and I don’t know how rapidly we can turn that around, because it has to do with changes in reporting pathways from coroners and medical examiners, from emergency departments, from hospitals, from public safety, as well as incorporating that all into more rapid death reporting.
It would be wonderful if we could commission that kind of study to help ASPI and increase the visibility of some of our activity. It touches the death system, but I don’t think it — So that would be my response. Part of it would be related to it, but it’s a bigger tent question.
DR. RIPPEN: I would agree with Bruce on that. I think though there could be a middle ground, which is as we do the kind of summary of what we learned and the process as it relates to birth and death, the use case actually is flagged.
And the reason is because then it’s based on what we heard as far as the limitations, and it could be more as we know the limitations, and this does have an impact on getting the information about opioid. So I think that is building then on what we heard, because as Bruce suggested making it a focus only on that is a very different thing, and based on what I heard that really wasn’t of interest to them.
DR. STEAD: That is fine. My only other thing, I’ve tried to turn around in my head what we learned from the commission, and I’ve wondered whether that, whether we need to think about aligning whatever we’re thinking about with vitals with their strategy.
And I haven’t been able to find a good match in the end because of the fact they’re narrowing the scope to an analytics and statistics. But I think that’s the other thing, as we carry forward the thinking from the first two days we want to think if there’s a way to either pilot or leverage what’s going on as a response to CEP in the vitals work.
DR. RIPPEN: Don’t forget about the birth though. Just don’t forget about the birth because addicted kids.
DR. COHEN: As we discussed the second bucket which has to do with data access, maybe we can talk about that a little more, but I’d like to finish vitals first. One thing I can think of is they talked about the federal-state interaction. Thinking about federal leverage that could help support state registries going to their legislatures to require them to report more rapidly as a really nice general fix rather than having to go to 50 different state legislatures if the feds require that we do blah blah blah in order to maintain our funding or our accreditation or something.
So is there a federal lever that we could use that might emanate from some of the recommendations of CEP to help support efforts to free up states to report data more rapidly, that would be the nexus that I see for that. But I’d like to hear other comments about the scope laid out around vitals.
DR. THORPE: I am more pragmatic, I’d just like to talk, when we talk about work plans I like to talk about products and get the end in mind so we can get a flow process, that’s just my thinking. So the question I have for you Bruce is you mention a white paper in your presentation, and then the word report was tossed around yesterday. So my question then is how different would that material be in the two, how would we distinguish the two?
DR. COHEN: I guess the report summarizes everything we heard. And that would certainly I think be our initial product because it was such a rich experience. I don’t know how long, to summarize everything that we heard in two days would be I think a really incredibly long report, so I don’t know how deep we can be in a report, we might want to use down the road taking a deeper dive into some of the issues to create more focused white papers.
That’s just a thought, if we feel we want the report to be something that people read and is not overwhelming, and if we find that there are issues that we want to explore in more detail we can consider them later. But I hadn’t really planned, the report is the basic product, but in my mind structuring that report is going to be challenging because we did hear so much.
DR. THORPE: Can you lay out the short-term products, mid-term products, long-term products? I think that would be helpful for me organizing my thoughts.
DR. COHEN: Sure. The short-term products I envision is a report summarizing the workshop. I would like to see if we can get the scans and the diagrams about uses of vitals and funds flow, and a letter to the secretary as the short-term products, and perhaps the use cases. I don’t know how long it will take to develop those. That might be more of an intermediate.
DR. ROSS: A couple of thoughts. The report from the workshop, we ought to set to have done by no later than the end of October, but probably in the next couple weeks so it will be assembled and the committee can chew on it a little bit.
DR. COHEN: It is a little longer process, Suzan will need to get the transcripts back and develop it, I think the subcommittee will need to meet and try to figure out how to organize it, so I think it will probably take a little longer.
MS. HINES: Just a quick comment that in the past because I’ve been with her for two or three reports, and she generally proposes an outline, and then we go back and forth on that, and then she fills it out. So I think the next few weeks in terms of conceptualizing the outlines is going to be important for the subcommittee.
DR. ROSS: So the report would be, to Roland’s question, it gets done when?
MS. HINES: I think as soon Susan can do it, which I think the end of the October, it’s a push, but I think she can do it life permitting.
DR. ROSS: Then in parallel I think we agreed we will move forward with an analysis of the uses of vitals data and the funds analysis kind of in parallel, that would happen November-ish so we can have something in our hands.
MS. HINES: That could start now.
DR. ROSS: We can produce it and have it in our hands come November-ish.
MS. HINES: Just so the committee knows, Gib Parrish has expressed interest in doing that, and we’re going to try to work that out.
DR. ROSS: And that’s kind of end of ear funding I gather that could help make that happen. Similarly, I think on the white paper idea my thought there is that yes, there was an awful lot heard, but I have to ask is that the responsibility of the committee, it would be great to have our name on such a thing, but to really develop a really good white paper given all that got said.
It occurs to me that the amount of energy the NAPHSIS folks had and the vital records folks that were in the room had around the idea of forming a community of practice to improve state practice, that maybe CDC’s cooperative agreement with NAPHSA said they might want to urge them to build the white paper and then training materials around it, because what we saw was such a broad education not just for us, but frankly most everybody in the room was learning something, that they could produce some educational outputs that would inform many people. And that goes well beyond the mission of this committee. So I’m just trying to understand where we’re going —
DR. COHEN: You’re right, we have limited bandwidth, and we’re not an implementation.
MS. HINES: We are not an analyzing group.
DR. ROSS: I would urge that NCHS talk with NAPHSIS or somebody and move that in as a work product that they could produce, I think that would be a great outcome of this hearing, that it really propelled that kind of thing to happen. It’s in their interest to do it I think.
DR. COHEN: Bill is giving me a soft hook. A couple more things about vitals, then we’ll move onto our second big bucket which I think really is sort of welding the next three bullets on that slide. Good, let’s move on. The second big bucket for me, and I’d love to hear the population health subcommittee folks and the committee as a whole, there is a space relating to health data. Let me just summarize what these bullets are and why I think the next three fit together.
Understanding the health data ecosystem grew out of our APCD work, and our discussions talked about understanding how all the data fit together in the federal data system, particularly within HHS, the content interactions, the geographic level, how the data are accessed, what the training is needed to use the data, what the restrictions are for release, the return on investment of the data, that was our thinking about the ecosystem. We had had a bullet in place at Bob Phillips’ urging about really thinking about following up on the commission of evidence based policy making, and after this morning session that seems like a no brainer for us.
And the fourth bullet there resulted from conversations we had about concerns about public access to research data. All of these are in the HHS data sandbox as a bucket, I don’t know whether there are sub buckets we’re going to want to break out or try to figure out how to scope an activity, a primary activity related to data. So I’d love to hear from other —
MS. HINES: Can I add some information to your point? So to the third point about maintaining public access, we received a draft letter that’s going to come informally from a group, basically saying they can no longer access key data that they need. Then I informally from a colleague that I worked with on developing the community health status indicators has noted two CDC systems have gone offline, gone dark, the Health Indicators Warehouse and the Health Data Interactive.
And there seem to be these decisions being made that are affecting a lot of people but somehow there isn’t a direct line of communication, and so one idea we’ve discussed in the hallways is using that time in November to actually find out what data that have gone away is really impacting people’s ability, especially this whole issue of community learning systems, because communities need access to data, and all of a sudden things that they’ve regularly been getting they’re not getting, all these sites are going down.
So I wanted to make you aware that we’ve seen a draft letter, it’s going to be coming in officially to the committee about specifically the BRFS being much less accessible to researchers than the Health Indicators Warehouse, and then the Health Data Interactive. So that’s three systems, and the Community Health Status Indicator such as it is today is also I think just been taken down or is about to be taken down. That’s four community health systems going offline in a very short period of time.
And I’m not sure it was systematic, it just happened, and just because that’s the space I’ve worked in my whole career, I’m aware of each of these four things, and it has just been brought to our attention willy-nilly and maybe we need to do something systematic around that. So I just wanted to expound a little on that. Sorry, Bruce.
DR. CORNELIUS: There is quite a bit of discussions over the last few days about the RNDCs. And somehow we need to get on top of that with the problems with accessing data at the research datacenters.
DR. COHEN: Thanks, Lee. Other concerns around RDCs is certainly related to accessing data.
MS. KLOSS: It may just be last half of four days. The second bullet, that just seems like a huge undertaking, in a whole new direction I guess. Does that work exist anywhere? I guess I’m just wondering where it came from, it just seems huge.
DR. STEAD: It came out of the discussion from the June meeting about possible next steps to HPCD.
DR. RIPPEN: There may be something more manageable that I think many of us know, which is really where are the buckets of specific types of information. So for example when people are talking about the vital statistics or vital records and adding all these other components that are part of an EHR, that may be in a higher level.
I mean that’s nothing new, but at least what are the sources, not the details of databases, but just where are the different sources and where are they going as it relates to potentially standards or nomenclature, things like that, just so that if you think about alternatives to understand if it’s available or not, more future. But not as detailed, I think that anything that lists details of databases and stuff is not —
MS. KLOSS: I don’t even read it as detailed databases. I just read it as trying to map the universe.
DR. STEAD: My own gut, because given the magnitude that Vitals has evolved into, and Predictability Roadmap and beyond HIPPA and TNV, that I would almost think of this as something Pop Health might want to consider putting on a cue list that might be surfaced when you’ve really got Vitals behind you.
Because my sense is Vitals may involve you for maybe the next two years. If you think back to the work that led up to the measurement framework and where we are, is vitals as big as that, are can we keep it smaller, I don’t know, but it seems to me that we might want to have this on a list that we reconsider when we consider the 2019 work plan.
DR. COHEN: I want to hear from Nick and then I want to respond to my former population health cochair.
MR. COUSSOULE: Pretend I don’t know who else that is. I’m going to spend less time on the three bullets and more on the way you framed it up. I think the way you framed it up really makes a lot of sense, and it fits right within our charter, in our wheelhouse. I think that the danger, I’m a little bit with Linda, I think it’s an elephant. So I think the danger or the opportunity would be to frame it in such a way that it would be complimentary to other things happening, but leveraging the uniqueness of what we can bring forth in the committee to kind of go and do that.
And I have a few things in my head, and some of it may tie into some of the standards work and roadmap work, some of it may tie into the other work we talked about the last couple days. So I think there’s a real opportunity to move some of that forward, I’m just not sure what the scope should be to make it both doable and unique enough that we can add some value to it.
DR. COHEN: That is a perfect segue for me to respond to Bill. My wise former cochair of population health always taught me that having two buckets is better than one, even if one bucket is bigger than the other. It’s true that Vitals might be a large focus for us, but I think it would behoove us to have another project simultaneously.
And I think Nick hit the nail on the head, if we are going to be in the data space related to HHS, is there a project of less scope that would address specific questions or issues that could connect to other activities in the committee, and try to move forward on some of the stuff we heard this morning around the commission on evidence based policy, and some of Charlie’s suggestions so that we can be moving forward not with a huge project but something manageable.
DR. STEAD: To be clear I was only advocating bullet two as part of my opinion. My sense from what I hear is we do need to do something in the public access space, and I think we do need to do something in how HHS can most meaningfully respond to the commission. So I’m not suggesting we opt out, I think both of those, they could be examples of medium sized buckets, and I think the Pop Health ecosystem is probably bigger than vitals.
DR. COHEN: Okay. I think we have achieved consensus that we’re going to not focus broadly on the health data ecosystem, but we’re going to try to come up with moderate scopes of work relating to CEP and data access to move forward. We’re running out of time.
MS. KLOSS: One quick thought with regard to the health data ecosystem. An early question might be to research what has already been done. I just can’t believe this hasn’t been described through the Data Council and through other means.
DR. COHEN: We haven’t seen it. You are right, it might exist. So any suggestions? I want us to be able to move on. Any suggestions about scoping for CEP activities? Or Rebecca had some suggestions about moving forward on some of the data access issues. I really appreciate it, and the subcommittee is really going to have to figure out how to scope those two issues.
The last one, the measurement framework, I don’t think we need to think about this right now, but we’ve created a lot of momentum from our past work. I’ve talked a little bit about what Shoma(phonetic) is doing, and there are other activities related to it. At some future time, it might make sense for us to review what’s going on and see if it makes sense for us to take another bite of the apple from our perch.
DR. STEAD: Given this, do you want to carve out any of the November time for Pop Health work?
DR. COHEN: Rebecca, if you think we can get input, get conversations going from folks who have data access, I think it will be valuable for the subcommittee and the full committee to hear that input for us to figure out moving in that space. Vicky, do you have any thoughts?
DR. MAYS: We had a discussion about ways to think about November. I don’t think that we are finished in terms of vitals. So when you talk about access, there’s stuff if you’re going to write a decent vitals report and decisions that we have to make that also still need to be done.
DR. COHEN: So you think we might want to use some of that time to talk more about vitals?
DR. MAYS: I am not sure you appreciate the workload that you put out in terms of how many gaps are in it and how much work needs to be done if we’re really going to make it.
DR. COHEN: I am sure I haven’t appreciated it fully.
DR. MAYS: There is still a lot of work just to have a report be other than Dear Secretary, we’ve got some problems here, but instead a real coherent thing, Dear Secretary, we have some problems but we also have some solutions.
DR. COHEN: Let me write that down.
DR. STEAD: Dear Secretary, we never realized that Vitals could be so complicated.
DR. COHEN: So two possibilities for us to consider for November from Pop Health related activities would be spending time continuing the discussion given where we are at the vitals report development and the product development, and the other has to do with understanding the access issue more. And if we don’t do the access issue in November it’s not going to go away, we might want to think about building that into our January meeting.
DR. MAYS: I was actually thinking about having people use that time to actually give us comment, things like we talked about. I don’t think we’re going to do on marriage, enforce, and all that kind of stuff. So there’s things we haven’t heard about, and we said we wanted to deal with that. It’s not about just us talking, but should we not see if we can have in any of that time people presenting to us?
MS. HINES: I don’t think there is time to organize that kind of thing between now and two months. Do you think we can?
MS. BERNSTEIN: Maybe we can.
DR. MAYS: Some of us know the people that work in that space. I thought we were going to do it online, I didn’t think we were coming in. We’re talking about a virtual meeting in November, right?
MS. HINES: I don’t know what we’re talking about.
DR. MAYS: I didn’t know we can even consider face to face.
MS. HINES: You can’t have four Full Committee meetings, but we certainly are still able, just like we had the standards meeting on August 21st, and the other meeting, individual subcommittees can do what they need to do, within limits.
DR. COHEN: I would suggest pretty much everyone around this table was at the hearings on Monday and Tuesday. If you have specific suggestions of how you think we could use the November time to explore vitals issues, just email us so we can think about the best way to use that time.
DR. STEAD: Thank you. We will now move to Nick and standards. And we’re trying to time each of these blocks for about 30 minutes.
MR. COUSSOULE: I will try to speak for Alix. I know she is usually shy and unwilling to share her opinions. The priorities that we see, and I apologize I don’t have a slide for you, but obviously the predictability roadmap is a big issue for us. We’re kind of in the, we’re not even sure how to say in the middle of it, we’re between the middle and the end, I’m not quite sure where we are in there, and there’s still a good bit of work to do, we’re still trying to reconcile and get our hands around all of the output out of the session as we communicated yesterday in our update.
If I look at what else is out there that we believe would be some priorities, the first one that comes to mind is the prior auth, and then hearings on the prior auth and trying to understand that challenge. A second one really would be the patient matching question. Now it was raised in the letter, but I think it’s been something that’s been spiraling around for a long time. I know there are some different challenges within the federal side of that versus what we might try to accomplish.
But I have to believe this is, nick personally I haven’t talked to Alex about this one in any level other than we both agree this is in the top three priorities of what we would view. There is so much tied into that with a lot of other things that we’re doing that I believe that would be important for us to start framing up.
And then the third one, if I were picking a number three on this, which is also listed in the parking lot items, would be something akin to a Chief Innovation Officer form. And the reason we like this idea is that as we start talking about some of the longer term challenges and issues, whether it be in privacy or data access or security or standards, trying to get some feedback, and maybe this is an environmental scan, but trying to get some feedback on what are the creative people thinking about doing differently that we would need to consider as part of our continuing efforts. So that’s the way at least Alex and I would view the top three.
DR. ROSS: Sort of a workshop idea? Interview them, or what?
MR. COUSSOULE: It could be a workshop. In my head it would be more sitting around a table, because I think we’re going to want discussion. It could be a workshop instead of a hearing like this, but it’s really about how do we gather input into some of the more creative folks trying to push the envelope a little bit so that we can be considering that information in the medium to long term in what we’re doing.
Sometimes we tend to solve today’s problems, oftentimes, and I think it’s an opportunity for us to get some feedback to make sure we’re looking forward a little bit more aggressively too. So those are kind of the top three that we have, again I don’t want to speak for the other committee members, this was not done as a kind of committee exercise, it’s really just Alex and I talking about things that we believed were on the priority list from our perspective. Comments, questions, concerns?
MS. KLOSS: From the Predictability Roadmap perspective, I’m just adding a little bit to Alex, because we had a chance to talk about it at dinner last night, we did think that maybe taking half a day of that time in November for standards to kind of do some strawmen roadmaps, just take the reports of the results of the workshop and just use that as a working session to lay out some possible, what this might look like, because that will be important to engaging anybody, getting some reaction.
MR. COUSSOULE: I am in agreement with that. Alix and I had talked about that. I’m thinking more of a long-term, 2018-2019 as opposed to just November. But actually I do believe personally that that would also be a very good use of sometime in November.
MS. KLOSS: Then getting a handle on when we think that roadmap can actually be done. Is that an end of ’18?
MR. COUSSOULE: Done is an interesting choice of words there.
MS. KLOSS: Something that becomes a working tool.
MR. COUSSOULE: Absolutely. And I think that part of the challenge we have frankly as all of us do is enough time to get into framing up those kinds of things. And I think setting up some dedicated time for us to do that as a group would be very helpful.
We can certainly do some work in the subcommittee to prepare that much better, to give people something to respond to, as opposed to let’s sit back and try to whiteboard everything, which counterproductive might be the wrong choice of words, but it wouldn’t be as productive as it could be, and then we could come in with here’s two or three options with impacts and gaps and holes and opportunities and then respond to that, I think that would be a very useful way, in my mind it would help us a lot, I think it would be a very useful way to also get the rest of the committee input into that.
One other thing I will add, as you notice I did not talk about a couple of the other initiatives that are going on that would likely need standards input at some point in time depending on where they go since there’s the thing we just talked about where the terminology and vocabulary and even potentially somewhat on HIPPA things. But the good news about that from my perspective is that it does show some of the dovetail between a lot of the activities that we have going on here, that it will involve lots of the committee and not just one portion of us.
DR. STEAD: stead: I think as the Standards Subcommittee works on the work plan over the next few weeks, we want you to include, at this juncture we’ve viewing TNV as homed in standards led by Linda. And so you can take credit for that as part of the work plan.
MR. COUSSOULE: Interesting use of words.
DR. STEAD: It is going to eat bandwidth, and it will provide deliverables. So realistically it limits what else we can do.
MR. COUSSOULE: Recognizing when I’m talking about those top three items they were going out a ways.
DR. STEAD: Then we have the wildcard of whether the review committee will reappear.
MR. COUSSOULE: 70/30 might come up, or there are a number of issues that are outside of our control directly, obviously.
DR. STEAD: Other comments or suggestions around what standards should be thinking about?
MS. HINES: I have a logistical question, given the conversation about November. Are you interested in trying to use that time to get work done virtually, I guess using some kind of web based tool, is that what you’re thinking?
MS. KLOSS: That is what I thought.
MS. HINES: That is helpful.
MR. STEAD: I think that we proved that we can get a fair amount of work done that way leading up to the predictability roadmap workshop. And so I would think that we would maybe think about that as a model if practical.
MS. HINES: Knowing that the public could call in and listen to us.
MR. COUSSOULE: I think the important part would be to have enough pre-work done to make it a reaction discussion as opposed to a —
DR. STEAD: I am beginning to hear that as a way to move forward we might sort of think of it as three four hour blocks, and let each subcommittee manage a block, but have as many people of the full committee join for as much of it as we can, because sort of what I’m hearing is rather than trying to micromanage it at this time from this level, that might be how we would want to do it. You could think about the feasibility of that.
MS. HINES: So for instance, on whatever that date is, November 28th, there would be three to four hours in the morning, three to four hours in the afternoon to take what we’re doing right now and expand it out, and then the next day, and then finish in the middle of the day the second day.
DR. STEAD: Yes. Really I was thinking largely a meeting with three four hour blocks, but it would be a day and a half meeting in essence. Let me ask, Nick, let’s assume we begin to get some alternatives for the predictability roadmap, and say as we come out of the November time, we come in and — Is there a way that we could then rethink prior auth as a test of the predictability, of the ideas that surface in the predictability roadmap. I’m just wondering if we begin —
MR. COUSSOULE: I am not sure I can know the answer to that one, you might want to run that by Alix.
DR. STEAD: The question would be we will take the Predictability Roadmap so far in abstract, and then it will end up with recommendations. If there are actually some ways of working together — I just think about whether, because that could actually be a very neat way of beginning to kill two birds with one stone.
MR. COUSSOULE: And I think just from the standpoint of I know there’s potential activity going on in regards to attachments. That might all tied together as somewhat of a use case that solves a couple of different problems in different groups.
DR. STEAD: And during that process we would come out with some recommendations about prior author attachments, and we’d come out with version two of the predictability roadmap reflecting what we learned. Okay, then are we ready to switch gears to the privacy, confidentiality, and security, a little bit early, so we’ve got a little bit of —
MS. KLOSS: I don’t have a slide deck either, but I’ve just been identifying some opportunities and will throw them out for discussion, because I haven’t talked about these with any of the subcommittee members. So with regard to beyond HIPAA, I think we’ve got four things that kind of got identified that we need to do. We need to summarize the hearing, thank you Rachel, we’ve already got your notes, and that was amazing.
And we need to then review that combined summarization and make some decisions about additional areas for study, and some of those suggestions came up. So I would think that perhaps we can get that done on a call before November. We need to work as a subcommittee to define the scope of and outline the report we want out of the environmental scan. I think it will help us to move that along. Well it will certainly help Bob Gilman to begin his work productively. So I think the sooner we can brainstorm that and do that the better.
Third, I think we need to understand the recommendations that came out of other relevant reports. We teed them all up, but we haven’t really just looked at them and done any cross-thinking about other recommendations, and it would seem to me that some of that may end up getting footnoted or brought in or quoted in the environmental scan report. So I think it’s the kind of research side of the scan report that we can step up. And I would think that would be something we could talk to Bob about doing as part of his work.
MS. HINES: That is listed as one of his things.
MS. KLOSS: So I think if we just have a list of what those documents —
DR. STEAD: If we include that in the outline of the environmental scan.
MS. HINES: Exactly, there’s what we heard, and then there’s what we need to find out outside of yesterday, and then what’s relevant.
MS. KLOSS: We may actually want to follow up on some of the suggestions we heard yesterday on other people that we may want to hear from. There were a couple of really great prospective panelists that weren’t able to be with us, and we actually could do that virtually, I mean we could think about doing that in November.
It would be good to circle back with Simpson Garfinkel and have him bring us up to date on NIST. That John Brownstein, we just had a couple of other. So we may need to try to do that in November. So as far as that gets me, this work probably gets us to the January meeting. And so the question will be, we won’t have the environmental scan report done by the January meeting.
MS. HINES: No?
MS. KLOSS: We’re hoping close. Then we had thought that we would have some preliminary discussion about that report or that draft in January perhaps if we really hustle. That’s the plan.
MS. HINES: That was the plan to have a draft by January.
DR. STEAD: Is your thinking, at one point we had talked about the environmental scan, and then we would build a framework of levers of something, we were thinking, and then at which would be more of a vision piece, and then we would have recommendations, and where in the process do we begin to draft our understanding of what the levers look like, or is that something that actually —
MS. KLOSS: That comes after the scan. So we have the scan, then we have a creative session of some sort.
DR. STEAD: One question, I don’t know, if we play back to how we built the measurement framework, truth be known, we went back and forth, we had a first version of the environmental scan, and then we began to iterate —
MS. HINES: Actually, no, we had a meeting, and discovered there was a whole lot we didn’t know. Then we commissioned the environmental scan for the non-health sector to find out what they understood, combine that with the health sector, then had a first draft, then had another meeting, then had another draft.
DR. STEAD: But Gib was very helpful as we went from the environmental scan to the framework.
MS. HINES: Super helpful.
DR. STEAD: And I didn’t know did we want to, as we framed the charge for the environmental scan for this, I don’t know whether we want to suggest that it includes structuring what’s learned as a draft set of leavers. Because that would be, then that would set you up. You would then need a meeting. But that would set you up and that might be a bridge too far or might not.
MS. HINES: I don’t know but I would say these guys can noodle on that and figure it out.
MS. KLOSS: Well we had said in the expanded scope of work we said phase two would develop models and illustrative future scenarios laying out assumptions and identifying areas of uncertainty. Levers is another good word for that. And then we were going to have an invitational roundtable to explore those.
And then phase three we would be prepared to make some recommendations to the secretary, and that was going to be in the form of a letter and then phase four, if you’ll recall we were going to try to do something for the broader community, some kind of stewardship guidance or some kind of framework or toolkit or something. So we do have that creative piece, and we’ll certainly consider whether we can tee up out of the environmental scan the drivers of —
DR. STEAD: A version zero that would actually be the basis for framing the discussion of the meeting.
MS. KLOSS: That was our plan. That the purpose of this report was not to publish it widely but to support the invitational —
MS. HINES: You might be surprised that the Environmental scan turned out to be a really nice resource that we do have on our website, it could be useful. What do you think Rachel, does that make sense? So we really press between now and January to move that scan, and that’s consistent with the work contract we have with Bob Gilman.
MS. SEEGER: It will be in very good shape by January.
MS. BERNSTEIN: I think that is right. Bob expects that sometime in December, I think the committee can help him by posing, deliberating, suggesting any kind of resources that we find interesting as he’s asked us to do yesterday. That will help him gather both by getting a better sense of what the parts of the department that you are particularly interested in and things that you find interesting that might be helpful to others of your colleagues that he could incorporate.
MS. KLOSS: We thought our first task would be to summarize the hearing findings and then try to nail at least a working outline for the report, so there are no surprises. Although he has literary freedom to rearrange things, but I think we need to know what we want covered, that we’re okay before going. So that’s that report. Now other privacy, I just have jotted three, and again discussion with the subcommittee or our staff, our experts.
This whole issue that we heard about this morning with the access, patient access issues under Cures, I’d be interested in hearing from Rachel a little bit more about that, and is this project all baked at ONC, is there some role we can play there, it just seems like such a nice fit to the work that we’ve done or had been contemplating doing with accounting for disclosures and access and disclosure management. So, question, is that all a technology play, or is there a policy play, and is it pretty far down the road, or is there something we can do to contribute to that. There’s a lot of emphasis on access in that.
MS. SEEGER: Yesterday I passed out for everyone our brand new brochure that was launched last week with some seed money from our friends Steph Cavanaugh and friends at All of Us. I can see one of them is sitting out in front of Dr. Stead. So this is phase two of our Information is Powerful Medicine Campaign, it’s really geared towards women in their role as caregivers, mothers, aunts, friends, sisters, and oftentimes the primary decision maker in health care.
And also as patients. And speaking to them in plain language about their right to access under HIPAA and the importance of regularly using their health information to track their wellness and to be more involved in their care, to manage their prescriptions, their lab results, et cetera. We’re really proud of it.
And also reminds people about how to get involved in a research program of course, and share that information with a third party like a researcher. This particular initiative was done in tandem with another one of those Medscape modules that we’ve created in the past. This is for healthcare providers providing free CME and CE credit, again on the patient right to access.
So these are two small initiatives. We don’t have a tremendous ability to push it out. So we’ve been talking about a second phase of this in terms of a dissemination plan, doing some digital media buys, some marketing, trying to expand our reach, working with associations and others consumer facing groups to try to gain greater awareness.
I think where NCVHS comes in is linking to organizations and others within the state and local community health space who look to you all for guidance. And it’s very important for us I think to try to continue pushing this patient centered care via the patient right to access under HIPPA. It opens doors to better health.
DR. COHEN: So Rachel, if I went on Get It Check It Use It I could follow that and I can get all of my medical records?
MS. SEEGER: No, all you would get is more information on your right to access. But that really is the challenge, because your health records are in so many different places.
DR . COHEN: Again, I would love a pathway to help me figure out how to get all of my information. I run into problems because I’ve called up a hospital and the medical records department never responds, nobody is there, and in order to continue my care elsewhere they won’t proceed without those records from the hospital where I had knee surgery 47 years ago. What do I do?
MS. KLOSS: You have a right to it.
DR. COHEN: Thank you.
MS. SEEGER: I do think that there have been many discussions about a longitudinal health record. Many healthcare providers I know feel passionately that there’s a lot of superfluous information in there that would not be useful to a patient. But I do think that if you think about, over the course of an individual’s lifetime, if a child has been diagnosed with a murmur or a VST when they’re borne, that when they’re 40 or when they’re 50 and they might be having cardiac issues, it’s so important to be able to go back and look at that information. Or the importance of a family health record. Many of us know that as we’re getting older we’re learning that maybe a grandparent years ago had an issue, and the information is lost over generations. Genomics is helping us get to more information.
MS. KLOSS: I think my other suggestion to you Bruce, is that you should get your care at Sutter Health, because Jackie and I were talking this morning about how they’re amping up their access resources.
DR. COHEN: That would be wonderful, if we could figure out how to —
MS. KLOSS: So it is happening.
MS. BERNSTEIN: That would help from this day forward for him.
DR. THORPE: I have a question. I think this is great, but I don’t see how it recruits men, and why was it only for women. Because if I saw this I would think this is just only for women and not to do this. And men have pretty bad health profile, live shorter lives than women, so I’m just trying to understand in the all of us picture, and what we saw we’re recruiting yesterday and the great information she gave us yesterday, just trying to think about particularly minority men, is there any campaign made about access with regard to that.
MS. SEEGER: Phase I was focused on men, specifically young black men who have sex with men who are at risk for HIV. Our money is like miniscule for these activities. So what we try to do is hyper focus our reach so that we get the biggest bang for the buck. And that is why this particular effort is focused on women. But we have been posed a question, if there was additional money coming in FY18, what would we do for a phase three.
And the challenge, and this is not an all of us initiative, it’s an OCR initiative, even though we got some seed money, the challenge is reaching this one million cohort, that NIH is challenged to do, or you heard Steph saying more than one million. Because people get information in different ways. But this is a step in the direction to try to educate all Americans about the importance of accessing and using their health information and being part of their medical team.
MS. KLOSS: I think that is an area of some interest to our privacy subcommittee access, and whether there’s a way we might contribute to the cures work in access, and a related issue on my short list for other privacy was accounting for disclosures, that’s kind of hanging out there as an operational issue that needs to get resolved under HIPPA. And we had some discussion with OCR about whether they would think that as a worthy project for the subcommittee, and I’m just not quite sure where that stands with all the shifting priorities.
MS. SEEGER: Continues to be on OCR’s long-term regulatory agenda.
MS. KLOSS: It would be helpful if the agency had some recommendations from NCVHS.
MS. SEEGER: I think it is always useful to have recommendations from NCVHS.
MS. KLOSS: That is something the subcommittee will chew on when we work on our long-term report before the next exec committee meeting.
DR. STEAD: Let me ask one question while we have Rachel here. As you went through summarizing yesterday, are there additional levers in the space we were talking about yesterday that OCR feels you need in the environment in which we’re operating in? It would just be helpful to know.
MS. SEEGER: I think it is challenging because much of what this activity is focused on is beyond HIPPA, so it’s beyond HIPPA covered entities and their business associates. So as I was ruminating yesterday a lot of it is outside of the scope and jurisdiction of OCR. The recommendations may be for the department. Some of it may fall within the wheelhouse of FTC as we heard.
I think that there are other emerging technologies that we didn’t hear about yesterday except for one small mention, Jeremy mentioned machine learning. So we have spent a lot of time on big data as part of our de-identification slash re-identification two day hearing, and a minimum necessary quite honestly. And as we sit here technologies are just changing by the minute in the way organizations are delivering healthcare using data to do that, diagnosing patients, being able to manage quality, care delivery.
So we didn’t hear a lot about that yesterday, and we also did not hear a lot about applications, there was a little bit of talk about internet of things, but it really wasn’t getting into medical devices, and there was a little discussion about more patient centered wearables and internet of things, but there wasn’t a lot of discussion about things like infusion pumps that are outside of the scope of HIPPA, but where we often see privacy and security not being built in on the front end with research and development of these new technologies, and only when they’re about to come to market do folks realize that they needed to be thinking about HIPPA, FTC, FDA, and others.
DR. STEAD: As I listen to this, one thing you might consider as a subcommittee is we’ve got a bullet on Pop Health about implications of the CEP Report for HHS. it seems to me that given the pieces of that report around privacy enhancing technologies and techniques, and also around risk assessment for DID, we ought to think from a privacy, security, and confidentiality view, if we can help HHS think how to move forward in that space. My sense is we’d have different recommendations there than will come out of Pop Health.
MS. MONSON: Perhaps there is an opportunity, the FDA is doing a lot of work on the biomedical devices, Suzanne Schwartz and her team, and perhaps it’s an opportunity in November to talk to her, because I think they’re doing a lot of great stuff.
MS. BERNSTEIN: They are. Dr. Schwartz has spoken to the HHS privacy committee. They do not see privacy, as far as I understand from her, they don’t see privacy as sort of in their bailiwick, unless it’s specifically related to patient safety. So we heard a little bit about that back and forth yesterday. And so I think one of the things that maybe is worth considering is whether outside of HIPPA there are other authorities, whether they’re FDA authorities or other authorities that the Secretary already has, even if they’re outside of HIPPA, that would be worth exploring.
MS. SEEGER: We talked to FDA, NIST, the Cybersecurity Center of Excellence, FTC, all these partners all the time. I think one of the challenges is that small providers don’t. And even midsized providers don’t. And when you go to, I pick on infusion pumps again because they’re on my brain, they’re particularly vulnerable.
When many of us go to even a midsized provider, you’re hired an out of date notice of privacy practices, you just wonder what’s going on on the backend of that provider’s office with their infusion pumps and all of the other types of devices that are connecting onto the network, and their vulnerabilities with respect to cyber.
So the taskforce has done great work in setting out a number of objectives for the Department and for the healthcare ecosystem, and cyber is really important, but it’s not the only challenge out there for covered entities, business associates. I still think that the notice of privacy practices is one of the biggest problems out there. Rendering a compliant notice, it’s not happening.
MS. BERNSTEIN: Do you think what we heard from Dr. Devaney about her work on consent forms is maybe applicable to some of what we do with the notice of privacy practices? Can we learn from their experience about how to tell people what’s happening with their information or that data? Or are we already doing that? We’re probably already doing that, but I’m not aware of it.
MS. SEEGER: HHS, ONC, and OCR have developed a model notice of privacy practices in English and in Spanish, we’re working on some updates to that that need to happen, so we’ll be re-issuing those model notices. They are being used and downloaded quite a bit, but not nearly as often as they should be. I think everyone here in this room has probably received recently an out of date notice that has the date of 2003 on it. So not compliant under this rule.
MS. BERNSTEIN: I have a doctor I go to regularly that is very high tech, and he gives me a tablet he makes me sign every time I go the same notice, and the same notice I can electronically mark it all up, and I do every time, most people don’t do that, but every time I sort of give them notice that I’m not signing this part, this part isn’t necessary, and so forth. but most people aren’t that sophisticated and don’t know what the rights actually are because that’s what’s supposed to tell them —
DR. ROSS: They ignore all your markups?
MS. BERNSTEIN: It’s all about all the disclosures they can make.
MS. KLOSS: I think that is why I always feel some obligation to have a HIPPA focus for at least one of the projects for our committee. The Beyond HIPPA is terrific, but I also like to think we’re impacting and moving the —
DR. STEAD: I think you have gotten good input. So I think we’ll bring this block to closure. We’ve got a game plan around November, we’ll have to work with the cochairs around sequencing the three blocks and work on technology to support an agenda, but I think we’ve got a game plan there. So if everybody is okay we’ll pause for public comment.
MS. HINES: Yes, I will start off. Steve Lazarus sent us in a comment. I’ll just read it, it’s not very long.
As a Medicare beneficiary and former WEDI Chair, I commend the NCVHS for taking prompt and direct advisory action to the Secretary regarding the need for industry-wide training partner testing and workflow redesign prep for the SSNRI change.
In addition to the letter, the one that we have been drafting, I hope that all of the nuances discussed to the committee are communicated ASAP to the decision makers at Medicare. I also suggest consideration of issuing cards to new beneficiaries only for the first 90 days, and then evaluate the impact on the industry monthly as a way to prepare the industry and the public for the significant change.
DR. STEAD: Is there other public comment. I don’t see in the room. On the phone?
MS. HINES: Is there anyone on the WebEx?
MS. CHRISTIANI-MTUI: No.
DR. STEAD: Before we adjourn I want to pause a second and thank the NCHS staff, who have done what it takes to get us here and to keep things moving. In particular Marietta Squire, Debbie Jackson, and Katherine Jones.
And I also want to recognize the staff to the subcommittees, Maya, Rachel, Susie, Lorrain Doo, Kate Brett. Their dedication has enabled us to be productive. We have been doing an unbelievable amount this summer, and the way in which the staff have reached out to presenters and to organized panels such as what we did around the vitals hearing and beyond HIPPA is just remarkable, and we are extraordinarily grateful.
Today we’re also saying goodbye to a major player who supported the committee for years, so we want to give special thanks to Jeannine Christiani-Mtui. I remember early in my tenure when I was trying to get to the next stop, she said we try to never lose committee members. She meant to physically losing them. And she has lived up to that during her 12 years of service to the committee, and this will be her last meeting. And so the support behind the scenes often not seen has simply been wonderful, and we are grateful.
DR. STEAD: We will have a virtual meeting in November of subcommittees. Subcommittee blocks but with as broad participation from committee members as possible. With that we’ll adjourn. Thank you.
(Whereupon, the meeting adjourned at 2:30 p.m.)