[This Transcript is Unedited]
Department of Health and Human Services
National Committee on Vital and Health Statistics
Subcommittee on Population Health
February 20, 2014
Hubert H. Humphrey Building
200 Independence Avenue, SW
Washington, D.C. 20024
Proceedings by:
CASET Associates, Ltd.
caset@caset.net
P R O C E E D I N G S (4:15 p.m.)
DR. COHEN: So this is the meeting of the Subcommittee on Population Health, and our focus is on developing our work plan. The nature of the conversations today were remarkable and I think incredibly generative in terms of helping provide us focus and direction. I’d love to continue in that vein. Bill and I can briefly summarize what we’ve done and a tentative work plan agenda that we helped frame, which in point of fact can all be thrown out given what we heard today.
That option is not listed on the paper, but certainly one of the things that we will propose I think is an alternative to deal with the emerging exogen issue that we discussed this morning, and I think a real impetus around– Jim mentioned the mosaic workshop, building on that, focusing on granularity because that’s an issue that touches all the work we do as a potential additional option.
Let me just quickly frame what we’ve done and have Bill discuss part of what our options for the work plan are. What we really want to do, our goal for this session is to get feedback from all the committee members and all the staff, about what you think would make the most sense for us as a Population Health Subcommittee to tackle first. That’s the goal for today.
What you have in your book is a wonderful four page write up called, Community CLS for Health, NCVHS Plans, and this is a document that Susan Kanaan put together after numerous conversations and telephone calls and interactions, that essentially frames what we were thinking about in terms of the Population Health Committee, but the Population Health Committee as a facilitator for the broader committee activities.
Initially the four streams of work that we developed are something called the Framework Project, which Bill will talk to you about in a second, which has two broad work streams, and then the other two broad work streams are a continuation of our learnings from our previous community as learning system activity.
They focused on an astute observation that I think Len Nichols really clarified for me, is we need to meet communities where they are, and we need to be able to figure out where communities are. If we could do that we would go so far down the road in being able to provide useful technical assistance and resource and data. That’s the third stream.
The fourth stream sort of emerged as a consensus from a variety of our previous meetings and from our workshops, really reinvigorating what– and it was discussed here throughout the day– what’s the role of the federal government in reinvigorating these conversations and providing support and trying to come up with some discreet, actionable recommendations for the Secretary that would move the federal government, HHS, in the direction of being a more effective partner in providing community support. Those are the four general work streams.
The other option that really emerged today is I think that crosses that all of the activities is what– I initially called it actionable geography, but really it is granularity. How do you define granularity? What are the issues? What data are available? What other data can be brought, and as several people have mentioned, it’s a function of the use of the data, linking our needs for the data to the potential uses. That’s another I would put on– if we have four work streams, that’s a potential fifth work stream focusing on having a workshop, getting all the players together.
Tomorrow morning I’ll report on the Data Council. This really resonated with many of the HHS agencies. This is a conversation they’re very excited to have and I think would be a wonderful partner for us in that area.
Are there any questions? That’s my summary. Any thoughts, comments immediately?
DR. STEAD: Thank you Bruce, and I particularly want to thank Susan for the work she’s done in hooking pieces together, and then Nancy Breen was willing to join the phone calls and take on actually developing one of the appendices and has been very helpful to this process, so thank you for that.
As Bruce has described, and has written in the short document, we’re envisioning this as a set of activities that collectively support our effort to focus on the federal government’s role in helping communities learn at national scale. Although we describe them as work streams, they’re actually one set of interrelated work. That’s one key piece.
Second is we worked through this, and the framework work group is drawn from all of the subcommittees of NCVHS. As we went through this it was clear that this was work for the committee as a whole, that the Population Health Committee could facilitate it, because everything needs to sort of be led from somewhere, but this is work for the committee as a whole. At the end of the day, Larry would have to correct me if I’m wrong, but I think we actually want this basic short document probably adopted by the committee as a whole. I don’t know how we deal with that, not as something that’s up to anybody but as something that we think can guide not everything we’re doing but an important set of things we’re doing that we want to do as a committee as a whole.
It does have the four pieces that Bruce has described. The framework project, I’ll just draw your attention to the top of page two, and this reflects literally hours of discussion that Susan distilled out key sentences. It’s important almost to just read your way through a few of them. Whatever their domain of activity, today’s health data users need a parsimonious way to capture and organize high leverage data from multiple sources as well as understanding how to use and repurpose the data appropriately, one point.
Most users come to data and analysis from one perspective or level and may not have a way of seeing how they could draw on other data or levels. Diversity in the data is valuable. The framework project is designed to help communities and others work with that diversity to solve problems, catalogue and organize information about data into serviceable structures. It plays very well into the conversations we’ve had this morning and through the rest of the day because the objectives of the data continuum framework are to optimize measurement across levels, clarify the choices involved in assembling, and using data at various population levels and to assess the gap between existing data sets to the complement needed to guide health improvement.
Our discussion around granularity is an example of what we mean by both the continuum and the different methods one would need to think about with data at different levels of that continuum. The methods framework is to organize information about the characteristics of various data sources and methods so they can be repurposed for use at various levels and to optimize use across levels. Our discussion around things like timeliness and how you would have to characterize timeliness and how you would have to capture that in metadata are examples of the kind of methods that we’re talking about as are one vision of the kind of recipes and so forth that we’ve been talking about in this latter set of discussion.
If we made all this come together correctly, it would allow us to dock those recipes, if you will, to the right types of data sources or level in the population. At high level, that’s just pointing out some key pieces that are in the document, and then just draw your attention without really going through them to appendix one and appendix two. This has been so abstract. They were an attempt to give you a concrete example of– in both of them population level went across those examples and columns, and methods or characteristics of data types came down them in rows, and we just then put some of the examples in the cells and the simple example I did was one that I just distilled from a public panel at one of the IOM committee discussions on social and behavioral determinants around asthma and exacerbation of asthma and what kind of things in the electronic health record might help people–
(Pause find file)
DR. STEAD: As you see this is just a very small example, but one column for the individual level, then community/agency panels working with individual level data but from a different perspective and then a small area of variation, which is aggregate but related to the area in which that individual lives or set of individuals lives. Coming down, the things that need to be explicit and the methods relate to is what is the type of data source, what elements are relevant to this particular problem, what is the voice of the person who has captured the data or who has originated the data?
Then it is granularity, back to that example, and then whether it’s individually identified or not. Then what actions, you begin to get into how you actually use it, to get into that case, and then what are the stewardship requirements that you have to deal with as you deal with re-use? Appendix two is one that Nancy put together, with information from various data sources at those different levels from the cancer world, if you want to speak a second about that Nancy.
DR. BREEN: Bill actually developed this format and then it seemed like it would be useful to have a second example. I said I would try to develop one with cancer. I think as we were working through the different levels or continuum, because as I said to Susan we use “continuum” in a little bit different way at NCI and at some of the other institutes. It’s more healthy population, disease, diagnosis, so we’re still toying with language here. These might be levels of analysis, the individual, the community, and small area of variation, but we’re still struggling with all those things.
Also, when I put in surveys, Bill said, and that would be the top-left cell that is not bolded, which is not a– that actually has information in it– he said would that be at the individual level? I said only the micro data would be at the individual level, so a lot of this information is going to have multiple uses and I think it’s going to take a lot of thinking on the part of this committee, and then everybody here with all the different expertise’s around the table to try to sort this out and figure out how to use it.
Just to mention a couple more things, I also tried to bring in some of the tools that NCI has developed in order to look at some of these aspects. For example, there’s a planning tool called Cancer Control Planet, which when you go into Cancer Control Planet it’s got a lot of information. You go through steps and you find out for your locality– and it’s just down to the county-level. I think we want more granularity, but it’s a start– to see what the cancer-related descriptive data is like, what kind of surveillance, what is cancer incidence, what is screening so that you can see where are the areas where work is needed in your area.
Then you can go through and it provides evidence-based interventions that have worked for these things, which you can pull off the shelf and utilize in your communities. The idea is to be able to help you identify what issues there are in your community and then what interventions have an evidence-base that you can use to try to work on and make meaningful change to improve those issues. Then the Cancer Trends Progress Report is, like it says it provides a lot of information on different aspects related to cancer.
It’s easy to digest data. There’s a lot of tabular information and also graphs that are easy to read. Those go down to– usually it’s just the national data, and so you need to use that with Cancer Control Planet if you want to compare national with more local data. Then thinking about community agencies, I think that it was a little bit different, and I may not be quite right on this but when you get into the world of cancer beyond screening anyway, it’s a disease.
It’s pretty complicated for people to navigate, and it requires pretty intensive medical services. Navigators, primary care physicians, hospitalists, we’re not at the level of landlords and people in households anymore. It just seemed like these two examples would provide a lot of information that would help us thinking about how this might work on the ground in terms of thinking about this continuum and this framework.
DR. STEAD: So that is sort of as far as we got. As we thought about work plan, I think what I at least became confident in is that we weren’t going to get a lot further with email and short phone calls. We can come back to the work plan. What I would put on the table is that it would be useful for the framework work group or some such small set to actually spend a day in a face to face where we actually tried to populate enough use cases, not real work cases, that we could then design a workshop.
I think what we’ve got to do is get enough of an understanding of what we’re trying to do that we can then decide who are in fact the experts we need to bring to the table to have the right workshop or hearing to get to where we would then have something that could be turned into some form of a white paper that would describe this and that would allow both us as we go through the kind of conversations we’ve been going through to today and in recommendations to the secretary help people see how this kind of framework would guide people to what their choices were, to what the right methods were, et cetera.
DR. BREEN: I think one of the things that I find overwhelming that I hope that we might be able to make some headway with is those 908, or however many data sets that the federal government has available to it. That’s pretty overwhelming. Is there a way that we can help people navigate through that to understand what the purposes of those data are and how they might be able to use them together or independently and in ways that would allow them to answer the questions that they need. To me that’s a big part of the task.
DR. COHEN: Let me spend a couple more minutes describing what other activities we thought about, and then I really want to open it up to feedback from everybody to give us some direction. The other piece of activity that the Population Health Subcommittee is interested in working on is sort of a continuation of some of the communities of learning system activities.
In particular, the two areas of keen interest are– I guess keen interest initially– to have a better sense of how we understand what’s useful for communities, and we’re describing that as community readiness. How can we assess what– and the focus here is around data needs in particular, health data needs.
Susan Queen had done some initial work looking at what the federal government had been doing. She was surprised to find very little. The feds have lots of community-focused projects, community development projects but very little on providing support for community data use. Now that we’ve got Chris back and Pat Remington and a variety of folks who are very familiar with providing data and data support to communities, we wanted to expand our investigation to look at what’s going on in the non-federal sector around providing technical assistance and health data support at the local level.
The whole goal of this process would be essentially to come up with some guidance, so for data providers, whether it’s federal data providers, state data providers, private data providers, they would be able to understand better how to provide meaningful data to communities because they will be able to better assess where communities’ abilities are at with respect to using data. This is a conversation that would involve communities extensively. What emerged in these discussions at the same time we’re assessing community readiness– it’s an opportunity to assess where the federal government is at in terms of its resources for providing data support.
The projects you see involved in this dual track of population health work would be to gather more information over the next couple of months, maybe write some framing papers, and then hold a two-day workshop, and we’re targeting the fall, it could be sooner if we’re ready– to work with communities, work with data intermediaries, work with foundations, work with federal agencies and other partners to come up with a framework that will allow us to better help provide support for communities.
The outgrowth would be a report and recommendations and guidance as well as specific recommendations to the secretary on sort of what Walter was talking about before, reorienting or reinvigorating the federal workforce to provide data support. That was our other thought.
What emerged today, this morning, is a potential third train, which is to focus some of our efforts in addressing the issue of granularity, building on the work that’s been done and really the clear need to provide guidance as part of healthdata.gov, as part of directions for the secretary, as part of the emerging strategic plan. What is the role of the federal government in supporting communities at the local level in terms of data support? We were talking over lunch and the three tracks that came to mind to explore are the federal government can try to provide data down at much lower levels than they’re currently doing.
The federal government– other folks can provide those data. The federal government can teach communities how to collect the data at that level so communities can actually generate their own data. Exploring the federal roles and actually providing data and supporting efforts to collect data at lower levels of aggregation, that would be the nature of the workshop that would involve a variety of sectors of folks who are experts. I guess the other train is if we can’t collect data at the local level, can we provide small area estimates that are acceptable to communities?
If we can’t actually collect the data in actionable units, can we provide some estimates– if we can’t figure out what the smoking rate is in Brookline, are there ways to use existing data sources to estimate it? The federal government is really good at some of these statistical techniques and supporting services. I know I was just talking with Chris, and Community Commons is getting into the area of small area estimation. BRFSS is. It’s an emerging potential option for helping communities with the data they need. I’ll stop now.
One other thing, I’m sorry, Chris reminded me our focus today has been on data, but data are only one input into how communities make decisions. Although for me keeping my eyes on the prize, the immediate prize is helping support community data, but the ultimate goal is helping communities make decisions that they want to make. Ultimately whatever we do needs to be in the broader context of how communities use data to improve the quality of their life. I just wanted to remind folks of what our gestalt is. Sorry for talking too long. What do you want us to do?
DR. GREEN: So I have three points I want to make. One is it has been a great privilege for me to watch this group do what it has done to produce the three or four pages up there and this page. It’s been not just interesting, it’s been impressive. I wish to express appreciation for wrestling this to one page with a framework project and community readiness and technical assistance. I think that’s a laudable achievement that I want to celebrate.
That’s the second point I want to make. What you said with your additional comments about the granularity issue and the three, possibly four, streams, from my understanding of where you provide that is actually embedded in the framework project and the community readiness and technical assistance thing. I am not clear, including from today’s discussions about granularity, that we need to reach out and talk about the level of granularity of data as if it’s a separate methodological thing out there that we need to study. I would predict that it will emerge as a pivotal issue in the work of the framework project and assessing community readiness and this sort of assistance.
The third thing is I wanted to express appreciation for the staff support that got this to where it is. I’ve got one opinion here. I think this is beautiful. I heard your first comment about whether you might want to have some sort of approval process to forecast that. I’m sitting here triggered by you, Bruce, to say maybe in a few minutes we might want to hit the pause button, think back through what we’ve heard from these three groups with the work plans and just ask ourselves if we’re prepared to endorse them in content as what we’re up to. I see them as a coherent package deal that’s dangerously close to setting up into coherence.
(Laughter)
MS. MILAM: Along those lines the issues discussed with granularity in terms of geographies but that also gets into dates and looking at your enabling deliverables from other NCVHS sub-groups on page three, part of the facilitating CLS document. It would seem that this is an opportunity for privacy and population health to work very jointly to identify use cases and then figure out what use cases are really important for communities today at these different levels in the framework and then from that what stewardship privacy and security practices are needed to facilitate it?
DR. BLEWETT: You may have gone over this in other meetings, but what level of locality are we talking about, state, county, city?
DR. COHEN: I should have the definition of “community” tattooed on. You brought up an interesting issue. The feds think of small areas as states. States think a county is a small area. People think of small areas as their neighborhood. It’s all of those depending upon where you sit, and what is the obligation of the federal government to provide support for areas. I would think below county is the way I’d think of small areas. You’re right that NCHS is trying to generate state estimates. BRFSS is generating county estimates. Folks at the county levels are looking at towns and communities. We generate data by neighborhood. You’re looking troubled.
DR. BLEWETT: I guess I haven’t wrapped my head around is it the federal government’s responsibility to support neighborhood use of data. I guess the assumption is that yes, it is, of this group?
DR. COHEN: That would be a wonderful conversation to have. My answer would be yes.
DR. STEAD: Let me speak from a point of view of the thought of the framework. In the framework what we’re really trying to do is to help people think about data at a number of levels, and in that case we’re talking about from the individual to small area to whatever communities they think themselves as part of, which can be a number, to the population as a whole because you can measure– you measure things differently, and you interpret them differently at those different levels that are in fact summative. That’s one part of it. If in fact we can have a method for understanding and helping people think about those differences so they optimize how they use data from those different levels together to deal with a problem, then we would have a way to sit down and have a reasoned discussion about given that what is the proper role of people in different places in working together to give us the infrastructure to use data to improve health? That would then answer the question of what’s the role of the federal government. These things all connect, as I think Larry was saying, but they’re not monolithic answers. Does that help?
DR. BLEWETT: I think the use cases will be really important and should be real. When I look at the framework– I think about some of the work that I’ve done at states doing transformation of their Medicaid payment reform, which is really trying to figure out a way to do multi-payer analysis and link with communities and public health, and how do you do that? Cancer would be a perfect example. There I think it’s almost supporting local public health to connect with the healthcare system at the community level, which is generally the county– it’s generally the county– or the local units. That to me seems like a proper role for the government to support that intersection of public data with the private sector. That would be one use case that I could rally around, which is big gap.
DR. GREEN: I appreciate you bringing that back up again. It gives me a chance to harp on this. Susan did a great job for us when she was writing up some of our first reports in work here, getting our epistemology straight about what we meant by “community”. We also went by a functional notion of a community of solution, that we accepted the notion that the problems that people need to have solved have no idea where the county line is. They don’t know where Missouri ends and Arkansas begins.
I’m working my way around to what’s the role of the federal government. The role of federal government is to enable communities of solution even if they pass county and state lines. It’s a functional sort of way to think about it. That’s not based in the structures of whether they have a health department, what the agent of operation. They don’t have to work through a county health department. There are six counties in Colorado that don’t have one, so that’s not a very good idea for working with those six counties. I think we’ve already accepted our journey here that we will have this– we understand what we mean by “community” and we accept the fact that the diversity of data uses need to be able to fit into a framework project of some sort. We’re going to have to deal with some ambiguity of– how does that hit you?
DR. BLEWETT: I guess the use cases will be really important.
DR. GREEN: We have 14– we started it with 14. We’ve been riding those 14 horses for quite a while now. It’s time to get some updates and further ones to guide us. I think there’s widespread agreement with your point that use cases will drive it– Paul will make this case. When Paul goes off the committee we can just wave a thing and say, Paul Tang point and talk about use cases.
DR. FRANCIS: I just wanted to say that it is really hard, and I think part of what we’re recognizing is it’s really hard to parse apart what’s for which subcommittee to be taking leads on. One of the things that we should probably be doing– I mean, we certainly should be doing is making sure we don’t step on each other’s toes. As we– you heard some of the kinds of things that we’re looking at and as we hear the kinds of things that you’re looking at, we should kind of say well, you’re taking the lead on this and we’re taking the lead on that, and take it from there.
DR. STEAD: Absolutely, there is a sentence or so in this document where Susan says that this work goes back and forth between theory, research, and practical application. You can think of the frameworks, and there are information sources– ways to organize stuff. If we were to the way I would hope it could be done, any time we tackled a problem, and we needed to use data or we needed to help people talk about methods, much as we were having the recipe conversation, we would be able to look at this emerging structure.
It would start out as something that had a lot of ambiguity in it. The thinking is designed to be sensible. We can almost start with the examples that we cobbled together in this first thing. They’re initial examples. As we begin to have some of these conversations, we could look at those as version 0.001 of something. We could say how does that inform us? What of that can we use as we do this? Now let’s solve this problem. We’ll build out other things as we do that.
Let’s then look back and update the frameworks until they get to a point that they provide enough of a landscape that we can write a white paper that helps people that don’t live through this have some prayer of understanding what we’re talking about and that we can then begin to see gaps and make recommendations to the secretary.
I’ve spent a lot of time reading what NCVHS has done and the related things from IOM over the last two or three months. What I’ve discovered is there are lots of good examples about the data continuum at multiple levels. Interestingly enough there’s almost nothing that relates methods to that. The methods are all expert that are in people’s heads that are applied one problem, one case at a time. They’re not actually harnessed in a way that could then turn into these kinds of recipes that we’re talking about. I think part of this work we can maybe grab. Maybe that will become the use cases. Maybe we can take some of those examples that people have built. I don’t know. It’s a lot of ambiguity here.
DR. SCANLON: The other thing I think, just as a guiding principle, health data and statistics has one sort of approach to the geographic and national and so on. There are other places like the census and others where they have much more experience. I think they see as a responsibility because they collect it to begin with obviously, the collection, the dissemination, and use, of fairly local level data with all due protections of privacy. They have advisory committees to advise them, and they bet a fair amount of money on it.
Of course, they’re the only ones who have all this data, so obviously they should be the ones to do this, but there are models from the social science side, particularly Census, that could fit us here. There are models from the business side, the Bureau of Economic Analysis and so on, and there are models from– not so much education, transportation, or others, but there are other models to look at. In the case of the census I think they’ve taken a fairly active role. They have the data, and it’s a matter of getting it and under certain circumstances– in the case of health, nobody has all the data.
Someone said that NCHS is the National Center for Health Statistics. It’s not the Center for National Statistics, which would include state and then local. They could be. You could have a program there that just isn’t focused on their own surveys but actually works things, but it’s a little different, health. The good thing about health is we have a lot of administrative data that needs to be more– starting from vital statistics, claims data, payment data, and so on, that is available at the local level under certain circumstances.
I just think we could probably learn a lot from census and others, in fact pulling some of those folks in and how they do it. How do they provide technical assistance? How do they disseminate? I’m a little afraid that the theory that if you build it, a website or something, they will come isn’t the answer anymore. We have more stories of tools and websites that’s we’ve abandoned because we thought it was the answer and it went away. We just have to be very thoughtful.
DR. MAYS: I just want to follow up on that comment because even when we had our hearing and we had the census people in, they’ve got a ton of stuff that are actually excellent, but what they’ve found is that even though they have– because I sent them one of the reacts and what have you, they had to go to the point of having a community. They are called community cens.us– Despite all that stuff, that’s actually turned out for them to be very successful in terms of getting the data out of the door. We have one at UCLA that focuses on the native Hawaiian and other Pacific Islander populations. I think there are some potential models to think about.
DR. SUAREZ: I just wanted to comment on something that Leslie mentioned that made me think about this. As we look at this type of project, for example, the facility and community learning system, this entire framework, in reality the benefit and the uniqueness of our committee as a whole is that we attack, to use a word, attack every problem or try to attack every problem from a three dimensional perspective.
One is population health and in terms of content. Another one is privacy and security, the data stewardship frameworks, and another one is standards. In some respects I’m going to begin to think that in each and every one of the projects we should be thinking about the three pointed, triad approach to every one of these components. I would hope that this framework gets that triad perspective of saying there is definitely a component around all these aspects of methods and population health elements, but there’s also the stewardship dimension. There’s also the standards dimension. We’re seeing it in a number of other projects.
While there’s a lead, and that’s what Leslie was pointing out, there’s a lead in a particular initiative, I think the other two subcommittees should also be engaged. I have the end– to tell you the truth, our subcommittees are four or five people big, so it’s not like we have a subcommittee of 24 or 10 or even– that it’s really and truly– so I’m hoping that as we move forward with this project we really see them from this triad kind of a perspective.
DR. COHEN: That is a great segue. I’m from your public radio station, and it’s fundraising time. We want to engage everyone from all of the committees and the work group in all of these activities. The piece that we aren’t going to get to today is really actually asking for folks who are interested in these streams. You don’t get any preference if you already identified yourself as a population health subcommittee member. These are activities that cross, you’re right, everything we do. We want to make sure we have input from everyone. The data workgroup, the stuff we were talking about this morning fits perfectly– they all mesh. We want all of these different perspectives in these activities.
DR. STEAD: Let me just raise a question, which may be simply impractical or wrong to even raise. I’m still a novice on this committee.
(Laughter)
DR. STEAD: The reason we constitute as a framework workgroup was what this really is a framework to dock all of the stuff we’re doing as a committee. It happened to get docked in the population health, but it is logically a workgroup of the committee. That’s what it is. Because the– if you think of it that way, then the method of how you put the framework together and then having a focused workshop around what we can do now to provide help to communities this year around granularity would actually be a useful product of population health that would add information into the framework.
If you look at the kind– as we take the things we were talking about this morning around granularity and around timeliness from the perspective of how we get the right hooks for metadata that can work with the data use group, those are specific examples that would build the framework. One of the thoughts– since I think that one of the challenges that people have with the framework is because it’s so abstract.
I came out of the data workgroup discussion this morning and I said maybe what we should do is to focus on answering the– first have a focused– I think timeliness is simpler. So we can have a focus day on timeliness, the answer to that piece of the data workgroup, populate that card, if you will, in the methods framework, then go back and do the same thing on granularity.
At that point we would have an example of methods, granularity would force us to get at levels. I could see going back and forth like that, but I don’t know– I was coached that this thing needed to be docked somewhere. It made sense to be docked in population health as much as anything else. It’s a committee as a whole.
DR. GREEN: I have got two suggestions to try out on all of you. It’s hard with this group to sense when you’re ready. I think you’re ready to make a definition of the work we’re going to do together for a while, at least 2014. We’ve got vector support well beyond that.
I think this is a good point for us to as a group, endorse the reports we’ve heard this afternoon and today, from the three subcommittees this afternoon, and say the work plans they have presented here today make sense. They’re on track. They interdigitate. They’re what we want to do, and answer the question “what”. That’s the “what” question for NCVHS for at least 2014.
When someone says what are you doing, you whip out these charts and tables that we just saw and says that’s what we’re doing and that’s the answer.
The second notion I want to try out on you is that this is probably not– it’s not probably a productive enterprise to decide to take names and addresses of whose going to do what right now. What we ought to do is if we can agree on the what, we give Debbie Jackson a work assignment of preparing the agenda for the executive subcommittee meeting soon, for us to actually answer– take the timelines you’ve put out and put them on a grid and start dividing up the work.
To build off of Bill’s word doc, we don’t have– the full committee, I think should be seen as where this docks and that the subcommittees are basically our mechanisms and our means to advance on more than one front at a time. We’d be working in a coordinated fashion, but in smaller groups of five or six, but those people are not working in isolation and they’re not their own universe. They are tied by a canon. I’m glad to see Justine back down here at the corner because unless there’s some objection, I think we need to think of the Workgroup as a fourth player, a fourth deployable resource to help get this “what” done. How does that hit you? Are you in your comfort zone? Have we talked about this enough to say, let’s do this?
DR. TANG: I’m going to try to converge what you and Bill said. If the framework was the organizing principle for this two-day workshop, for example, the three use cases you started out with, timeliness, in order of difficulty, then granularity, and the one I’ll throw in to definitely get Privacy on board, is data segmentation. The output really going into the community readiness assessment is a community ready to consume the output product of the frameworks applied to these three use cases. There could actually be a workshop that applies literally the products in a way that’s truly coherent and ends up with a product.
MS. GOSS: Coming back to more of an administrative aspect to what you were saying, Larry, I think it makes a lot of sense, and it’s interesting– sort of parallels with some of the discussion I had with Bill after a lunch conversation back to the use cases and the parallels for the workshop. I think that it also means that the executive committee’s going to need to take a look at all the work that are on the very robust work plans of each of the subcommittees and make some very hard decisions to focus us– not that all of it isn’t good– but we’re going to have to really hone our work efforts and our thoughts based upon the framework of available time that we have for the committee.
I know I saw in the privacy executive committee minutes that there was a discussion about maybe going to five meetings a year. I know that wasn’t really realistic, but we have way more on our plate than we ever have time to do. That’s an ongoing concern and impacts the ability of us to deliver something of value.
DR. TANG: I don’t know that we need five meetings, but we need some combination of meeting and workshop. This was a very hybrid– we actually did work today. I think–
(Laughter)
DR. TANG: But you need a dedicated– what you described was dedicated heads down work. That’s a true workshop, but this was a really good working meeting that had some outputs that feed into a dedicated– you call on certain people to come in and do that work.
DR. FRANCIS: It struck me that data segmentation was kind of a different set of issues in certain ways.
DR. TANG: It is, but then it called to question in a very tangible way on how do you deal with stewardship in context of building a framework–
DR. FRANCIS: It is a critically important issue. I’m just not so sure it’s as germane to the immediate community use question.
DR. SUAREZ: I agree with that. I wanted to just go to a different idea. I think all our products, as I mentioned earlier, from a question that Bruce asked, all our products are really committee products. We don’t produce things by subcommittee. They come out to be the committee products. I think what you said about the structuring, in some respects that is what we have been doing less formally in that lately we have joined hearings, for example, or we have joined products like the HIPAA report, or we have joined letters coming out.
I mentioned the Population Health or Public Health Standards Letter– it’s not just about standards really, and it’s not just about the Standards Subcommittee. It’s really something that comes out from the three subcommittees. I’m just thinking that in reality that process has already started. Maybe yes we need to formalize it a little more. What we need to really think about to the point that Alex mentioned is as we formulate our own subcommittee– as we call it subcommittee agendas, what we haven’t recognized is that work items that work from all the subcommittees are going to impose on us.
We need to add those pieces into our own work items because our own subcommittee– we did actually include in some connections on potentially this year later activities on public health standards, but we had not brought in the possibility of developing some section from component aspects of this particular framework being presented today from Population Health. We really need to identify in columns the projects and in rows the subcommittees, going back to one of my main concerns that I’ve said before, and really highlight what are the activities at the end of the work and then how each of the subcommittees will participate in those activities.
DR. MAYS: I think what bothers me is that it’s an enormous amount of work that you really have to be clear about whether there are the resources to do it. I think the ideas today are exciting, but when we leave this room, it is like, do we have enough staff, do we have enough time? Some of the groups have an agenda. I guess the notion of what the executive committee has to do I think is a hard task. You’re talking about two hearings I guess and white papers, and I heard Standards already has hearings going. I heard the Workgroup has– from a management standpoint it just seems enormous to have each group doing what it’s doing and then to have this interdigitation that also has to go in order to get the quality of the products. Something’s going to have to give or– nothing has to give if there’s enough resources.
The other thing is I think you really need to be clear that the direction you’re going in is not necessarily what the community prioritized. To say that this is what the community– that the work of this is what the community wanted, and when you keep going out to the community, you’re going to get a bit of pushback, but if you feel that this is important to help the community, that’s what one should say as opposed to going out and stirring them up as we keep going and asking them questions. Then we don’t do what they asked us to do because as we said, we have to do this first, and we have to do that first. I don’t want the community starting to like– the more we push out until we have to also learn how to work with the community, they’re not the academics and the feds.
DR. COHEN: I want to build on what Paul said about the work we did today. I think the true value was we were all sitting around the table, and we all had very different perspectives. It made the product of today incredibly constructive and high quality because we can all blend our ideas together for consideration.
What I think we really need to do when we think about the structure organizationally is essentially we need three or four national committees that are subcommittees that represent the different points of view in each workgroup rather than being identified by population standards– rather than segmented and topic areas, essentially we need to pick people from each discipline to form workgroups to leave the discussions and initial work to address the problems. I think we’re evolving to that kind of structure. It makes sense organizationally to think about how to do that strategically.
DR. CARR: This is just an update. I want to thank everyone again for the input this morning. Lily and I have worked to meet the deadline that Debbie set by 5:00 so that we can post– well, we have to post tonight so that we could discuss, and potentially vote tomorrow.
I think we got major concepts. We actually reduced it down to two topics, and we can discuss further tomorrow whether we want to add something else besides timeliness and metadata.
Walter gave us a tremendous resource that came out of OMB on metadata that really strengthened what we were saying. As I’m rethinking what we did, there were a couple of edits I meant to make that I didn’t make. It will be posted tonight. I appreciate your continued input. I think if we can say two things and say them powerfully, that’s better than waiting six months to say three things.
MS. JACKSON: Just a point of clarity, what we have been doing for the last year, about a year and a half, is that items that are coming up for action are posted within a day or so of the meeting so that the public has a chance to see what the item is in time for the discussion. We didn’t have this in time for today, but I did ask for something that we could post for ourselves, for the public. It can be a very different document after you finish tomorrow, but at this point even if it’s conceptually organized for a framework, as you’re saying, it would have something out there. It would be on the usual placement, on the homepage and the discussion for the agenda item.
DR. WALKER: I just want to echo Vickie and call it out in this great one-pager. I think the critical success factor for us that we can say who we think our customer is, that we can develop whatever we develop in a way that the customer says thank you. It’s in the one-pager. It’s what Vickie said. I think we ought to be better than anybody else at understanding some customer– we can’t serve all the customers. I think that’s one of the clear messages here. We can’t serve them all in version one at any case. I think if we could say this is the customer, this is the need that we have gone out and validated that they need, here we’re going to road test the DCMF.
That’s what we’re talking about, going out there and saying when you actually watch what they’re doing and listen to them and do a cognitive walkthrough with them about what we think they might need and then they say yes or if they say no, which is more likely than not, then we go back, start over again, and we keep doing it until they say yes. If we do that, we will score, but that’s hard.
DR. GREEN: So I have a proposition. Since Jack Burke has said so much today, I wondered if I could entice him to making a formal motion that the committee approve the work plans as presented today.
MR. BURKE: Mr. Chairman, I would be happy to make such a motion. I would suggest that the committee consider the plans presented this afternoon as being adoptable in their current form and that we vote to approve the deliverables for ’14 and beyond based on those descriptions.
DR. BREEN: Second.
DR. GREEN: Are you ready to vote? All in favor? All opposed? All abstaining?
(Motion carries)
Good, tomorrow we’ll start with Standards at 8:00. Everyone’s invited. I remembered in the early days of my experience here I basically was just grateful I wasn’t on the Standards Subcommittee because I couldn’t understand what they were talking about. Now after getting through the last half of yesterdays — this is a hell of a good thing here. You might not want to miss it is my point. Come early tomorrow and join Walter and Ob, and then you guys do your best to get us started at 9:00 with the Full Committee.
We will open up with the famous Dr. Bruce Cohen making his presentation that has been postponed so far three times this year. Then I think we’ll do just fine to move on. We’ll go to Justine and we’ll proceed through that. We have a heart stopper at 10:00, and wherever we are we’re going to want to interrupt and do the 10:00 a.m. Social Determinants of Health with Carter Blakey. That will be our plan for tomorrow. I think we’ll be able to end on time. Thank you for your attention, your dedication today. See you at dinner.
(Whereupon, the subcommittee adjourned at 5:33 p.m.)