[This Transcript is Unedited]
National Committee on Vital and Health Statistics
Subcommittee on Population Health
June 19, 2013
National Center for Health Statistics
Hyattsville, Maryland
Proceedings by:
CASET Associates, Ltd.
Fairfax, Virginia
caset@caset.net
P R O C E E D I N G S
MS. MILAM: Our first topic on the agenda — well, it’s our second bullet point, but products from the roundtable and dissemination. I think Marjorie’s idea was fantastic. Is there other dissemination we should do? Do we want to try and get on agendas?
DR. COHEN: If folks haven’t realized, we’ve just switched from stewardship and privacy to population as outgrowths. I just wanted to set the context straight. A lot of the conversation applies both to stewardship and establishing, obviously, the framework for moving forward with the Population Health Subcommittee as well. I’m sorry to interrupt, but I wanted to make clear, a turn like– not 90 degrees, maybe 33 degrees.
MS. GREENBERG: Can I just clarify something? I know this was discussed, but I guess this is still in play. Clearly we came out with a single summary of the meeting. It was an integrative meeting, and it was the right thing to do. We left open what kind of products might come out of it. We’ve got on the table a stewardship product, but it will be informed by the Population Health and Standards and Quality, those who want to weigh in on the Quality. That would be a separate product of some type.
What I was proposing was for the joint product, which was the summary, as a way to kind of reach out, which could be both on agencies that wanted to get engaged with the stewardship issues, as well as with the technical assistance issues that are on the agenda of Population Health. It sounds like we’re definitely talking about several products — it might not be short-term products.
MS. MILAM: We’ll talk about– a large part of the purpose of our discussion here is to talk about next steps, but I think we’re going to find we’re not ready to say we know exactly what our next product is. In terms of sharing the summary, do we have other ideas of presenting the summary or other dissemination that we know we want to make of it?
MS. GREENBERG: We did that webinar on the original report, on the community as a learning health system. We were welcomed to do future webinars. I’d have to look at what we actually presented in that webinar to see how different– whether we have enough for a second webinar. I’m assuming, Debbie, that we can do kind of like — it’s just a summary, but we could probably package the summary of the meeting with the appendices in the way that we packaged this, as opposed to an actual bound report or something, with a nice cover. Then we’d send it to the APHA session, and send it anywhere you want us to. That’s one little piece, what you want to do with the summary, because the summary itself, I think, is quite useful and rich.
MS. KLOSS: I thought about whether we’ve got enough new meat– I’ve thought about whether there’s enough product here in this report as I reread it for another go around with the NeHC webinar, and I didn’t think we were ready yet. I don’t think so. I think we could take the original slide deck, update it with some new thinking, and perhaps go to a new audience, and that’s to some extent what you’re doing with APHA. I think there are a lot of audiences that could benefit from it.
MS. GREENBERG: That’s a nice idea, to update the slide set based on this– that would be doable.
DR. MAYS: We watched what OCR did today, so the question is – we could shoot something in LA for less than $38,000. I’m not sure why it was so expensive but you have to do it street-wise, but in all seriousness, I guess what I’m suggesting is maybe it’s time for us to think about whether or not with one of the community partners or a couple of the community partners, that you couldn’t get them to do some little YouTube things about what they found interesting and what they took away. It could at least be on the community side.
On the other is I think it would really be good to package the summary and get some feedback from primary organizations. For example, if you’re going to be in Boston for APHA, PRIM&R is meeting right after APHA. PRIM&R is Public Responsibility in Medicine and Research. They’re the group that oversees all of the training around the IRB, the CFR, and all this other kind of stuff. I sent our survey to them to put online, and by the time I got it through the approval process, it was afterwards. But the number of emails that people sent to me asking questions. I said you could still fill it out, but don’t get concerned about we can still use it.
So for example, that might be a place for you to get more feedback around some of these privacy issues and how they’re dealt with. We should probably come up with a list. The list is we give the summary, but then we present, or we’re there, and we get feedback from a group of people at a meeting or a conference. Then I think that would really help us. I think for the community people, I’m really serious, you can do YouTube. You can do them with your telephone.
DR. COHEN: I just want to insert as folks put their tents up to talk, if we circulate this somehow, what are the key themes and content messages we should be focusing on in terms of our learnings from the Roundtable. Just add that to your list of feedback that I think we want as the Population.
MS. MILAM: Bruce, we have about 22 minutes left, do you think we should move to getting the report-outs on the two conferences, the two meetings, and then move to discussion of next steps? We have until 5:00, and then Privacy goes 5:00-5:30.
DR. COHEN: We’ll take a couple of extra minutes maybe, but sure, I’m happy to do it that way. One of them is– Vickie and Susan, on our agenda is the community health data summit. Vickie you already gave us a really great update on APHA, so I’m just going to ignore that on the agenda.
MS. KANAAN: I am passing out an agenda. This is a very short report on the Healthy Communities Data Summit for the West Coast, which was held at UCSF on May 21st. I went at the committee’s behest and with the committee’s support. Vickie was also there, as well as Leah. I feel a sense of urgency about sharing a little bit, and there really isn’t time to give you an adequate overview, but I want to urge you to look at this agenda carefully when you have a moment.
Both the themes, the supporters, the topics, the presenters, the participants, this was an extraordinary meeting. I decided that this is really environmental scan part two, from my standpoint. Because looking at the same content that we all have been looking at, but really from a rather different perspective or through a different lens and bringing in some players and some perspectives and activities that are somewhat different, at least a different emphasis than what we saw. I don’t think it invalidates anything that we have seen or that we’ve learned.
Some of the bigger things are in terms of the participants– of course the technology sector, the private sector, and another really interesting piece is data-savvy individual community members, co-creators, community activists, but much more atomized.
I think the meeting painted a fairly different picture than the one that we came up with, which was really about community driven change. It was much more I would have to say atomized or individualized. I would like to have an opportunity to think more about what I saw as the same and what was different.
But this was a very high level meeting in terms of the production values. It was excellent use of technology and so on. On the end we saw a lot of the same themes, including the centrality of partnership, and many different sectors were represented there, including community organizations, a lot of government.
San Francisco was the star, and many wonderful examples of the use of data in policy, and to really engage and inform individual citizens, not so much community-based activity I don’t think, which for me was an interesting gap in the meeting content, but a lot of emphasis on actionable data. One of the themes that I thought was interesting that we come back to from time to time is the importance of pairing data with communication. Those are different sets of skills. If I had more time I would continue, but I think I’ll leave it at that. That’s probably about all the time there is. I hope there will be another opportunity to share more with you.
One of the things I do want to tell you– and I brought a little bit of the swag that was passed out at the meeting, if you have a chance you can look at. There were some folks there from UCSF who do community-based participatory research. They have three booklets that I think might be particularly interesting to the Privacy and Confidentiality folks, the same content aimed at different constituencies, one for clinicians, one for community-based organizations, and the third for researchers, sort of cover the same material. I thought it was an interesting template that we might look at. These are very interesting booklets. I’ll just have these if you have a chance to look at them. Vickie, anything you want to add?
DR. MAYS: What Bruce actually asked that I do is to really talk about this stuff relative to next steps. I knew that Susan was going to do this, and I wanted to do it in terms of that. I also asked Leah to come to the table, because Leah was there, and there are some things I want to ask Leah to comment about.
First of all, what I saw that’s very different but I think is a case study for us is really the role of the government. Here what I’m talking about is that San Francisco really has hired within the department of public health the people that we have like on our workgroup. Those were the people that were presenting as part of the culture, because we’re so close to Silicon Valley up there. There’s a case study for us to figure out how do you change government.
I think that’s really important for us to think about. I had lots of conversations when I was there. I was talking about why is it LA County, because LA County is very competitive and we usually try and best everybody out on everything because we’re big and we have lots of resources. What’s interesting is that what I was told is that we’re below the radar trying to do this. There are reasons that our Department of Health and Department of Public Health can’t just suddenly put a lot of money into re-engineering and doing lots of high tech things. It reminded me to some extent of NCHS, what it takes when we want to re-engineer a survey. You have a lot of pushback. I think trying to understand from that meeting how government can move would actually be useful to the work that we’re doing.
The second, whom I think we’re not taking advantage of, and I don’t know Marjorie if there’s a line between them, but that is the foundations. Probably 50-60 percent of what went on there is because of funding that’s come from various foundations. I talked to your friend, Adam? The guy that runs the healthcare foundation? Leah knows all these people and was nice enough to make introductions, so I talked to them at length about different things. The foundations are actually looking for the ability to solve.
It’s interesting because I wrote myself a note when I was thinking about this, which is we’re looking to protect the data or to kind of put the data out there. They’re looking to solve a problem. It’s almost like the reason they get so far is that what they’re really saying is we’ve got problem A, B, and C, and therefore this is the way to use the data. They get people on board a lot easier because everybody’s looking at solving a problem. I’ll give you an example.
I don’t know here, but in California we have– the Department of Health actually rates our restaurants. In LA we have A, B, and Cs. Up there, they have numbers. They were trying to decrease food-borne illnesses. What did they do? They made a partnership with Yelp, so with the restaurant ratings also comes this information. They decreased food-borne illnesses by 30 percent. So now everybody’s on board about this helps businesses, this helps public health. That’s an example of the kind of thing that even the foundations are saying if you’re solving applied problems that really impact the community, that’s what we want to see. I know we’re the data committee, but it may be we need to think about the data for what? For what it is that we can make a difference in?
The next suggestion I’m going to make, and I already Justine I was going to do this one, is I think that having the members of our workgroup present more would actually help us. Even if it’s just Lily’s here more– they’re in another world. We joke about it when they’re here as a workgroup and they’re spitting them out. Leah took me into her world– I’m serious. I went to a code-a-thon, and I didn’t go home until 11:00, but it was a code-a-thon. I went to Yelp and Google and all these things. It really is a different world. People are eager to help with problems. I think we ought to try, I don’t know how hard it’s going to be, to convince them that they should sit with us more and talk to us, and us to understand what they’re saying. I think where they can take us actually can happen faster than we’re going.
MS. GREENBERG: May I respond to that? Without going into detail, but since the last reading Justine and I and Lily and Jim and Susan had a call with Greg Downing, who is really, along with Todd Park when he was there and Bryan Sivak, are the sponsors of the working group. Sometimes I think we get frustrated about what products should be coming out of any of our subcommittees or out of that group or whatever, and he was very positive about the benefits that have already improved. Basically to me it was a real eye opener, because the bottom line I heard was you don’t really want the workgroup to be acting just like another NCVHS subcommittee. I always think of NCVHS as being ahead of the curve or being whatever, but that NCHS is more traditional.
Everything is relative. Compared to a lot of groups NCVHS is not traditional or trying to think ahead or being ahead of the curve, but really the way we operate is more traditional. What is really useful is to have this more disruptive– in a positive sense. I’ve had to learn that disruptive is a positive word as a parent, innovative. Exactly what you’re saying, so challenging all the more traditional ways that the committee approaches issues and approaches products and all of that. Some of them may be embraced by the committee as a whole, some may not be, but it stirs up the pot, and it opens your eyes, as you said.
It sounds like you were having way too much fun for going to a meeting. I think in order for that to work we do need to have to the extent possible more members of the working group sitting in on the full committee part of the meetings. I know people have limited time, et cetera, but do you want to add to that?
DR. COHEN: I really feel that we need to–
DR. TANG: Can I just do one thing so I can pass this prop along? This thing basically is reaffirming that we need to understand problem solved. This has the problem to solve and the data to back it up, that’s all.
MS. MILAM: So we’ll hear from Leah now?
DR. VAUGHAN: There are so many thoughts. I thought Health Communities was a wonderful conference, and it brought together and highlighted some of the things that are happening in the Bay Area overall, and that it’s not an either-or. You can actually in an innovative way bring together technology, community, and traditional structures to solve real problems. I think it was wonderful to have Vickie and to be the cultural anthropologist.
We covered a lot. We went through a lot of small tech companies. We went to Code for America. We got to a code-a-thon, and we had hacking, and then took her to a women’s technology even at Yelp, and that was a whole other group of really wonderful people. The Yelp example is a great one. It takes what most counties publish as their food ratings, which are regular restaurant inspections monthly, quarterly, semi-annually, and they publish it as open data for some years. Rather than just have it sit there to be discovered by an engineer at some point, they approached Yelp, developed a data standard working with the tech community and then just started publishing San Francisco’s restaurant inspections on the Yelp site. If you’re looking for a restaurant, you know if you’re going to get food poisoning there.
Shockingly, compliance with the food inspection recommendations has gone up dramatically. Food poisoning incidents have gone down greatly. I get a lot out of being here, and part of that is because I’m at that intersection of traditional and tech, but I also am looking forward to opportunities to perhaps be the cultural anthropologist for members of the full committee and developing some opportunities to do that. It’s hard to explain unless you actually are there.
MS. KANAAN: I just want to add one thing that I didn’t say before. We probably heard at least very briefly a dozen or more use cases during the meeting. It was stunning. They were all– this was really drinking from the fire hose. The emphasis at the last working group meeting on the need for use cases was really– they’re out there.
DR. FRANCIS: Could I make two quick comments about the Yelp example? First of all, it’s public information already.
DR. MAYS: It is public information, but nobody was using it.
DR. FRANCIS: That’s a very different question than information that isn’t already public. A lot of what we’re worried about is in the latter area. I’m just saying I think there are different issues that are raised by it. I honestly don’t know how good the data are, but if there were questions, as there are, we’ve talked about them with some of the health data about how good the data are; we might be having a really different conversation about that Yelp stuff. I’m not saying one way or the other. I’m just saying I think we need to be really careful about what’s a model for what.
DR. COHEN: We’ve got four minutes to figure out next steps for the population subcommittee.
DR. VAUGHAN: Let me briefly add it’s not so much that it’s a model for all data. It’s a model for the possibilities of cooperation and collaboration and partnership to solve problems.
DR. COHEN: Thanks Leah, that’s a great segue into what I’m hearing. A couple of key things about the space we can be in as the population subcommittee and the national committee, one is making these connections. What’s the role of government, particularly the federal government in helping make these connections? It’s not that the data aren’t there, we just haven’t figured out how to get it to the right people. When we talk about Larry’s charge, which was developing the framework to enable communities to become health-learning systems, these seem to be the sweet spot for our committee. Those are areas that I think we might want to consider exploring. I really want to hear what everybody’s ideas are about where they think the population health subcommittee should be taking our learnings from the roundtable.
MS. KLOSS: I think we came away from the roundtable leading with a recommendation from the group that the first step be an inventory of what kind of useful tools already exist. I still think that is the high priority starting point. We want to disseminate what’s there and perhaps think about innovative ways to fill in the gaps where there isn’t anything.
DR. COHEN: Our initial focus has been to look at what technical assistance the federal government provided, but after this discussion I think that’s too limiting even though that might be what we should do first, but without understanding what technical assistance foundations are providing and other private sector enterprises if we want to figure out where to go we need to really understand what’s out there from these resources as well.
MS. GREENBERG: Beyond what Susan did in the environmental scan?
DR. COHEN: Yes. This is really what technical– the focus here is on useful tools and technical assistance that already exist to help communities use data for change.
DR. STEAD: What if you took one of these examples and then decomposed it into what they did to make it happen, what data they had and how they got at it. Then you could ask what could be done to make those data sources more easily and appropriately accessible. You could also ask, what are the things they couldn’t accomplish that they wanted to accomplish. That might inform something about what else we could do. That way it would be problem-driven, but you would still be trying to– problem-driven, story enabled, and you would actually be trying to learn your framework as we went. That’s maybe one way to parse the problem.
DR. COHEN: Those could also easily be made into YouTube videos. Jim actually had a similar idea he shared earlier with me, the Keystone Beacon Community, and how they tried to incorporate data into their developments. I like the idea of coming up with use cases or examples and play it out, what tools folks have, what technical assistance they were given, what the gaps were, and understand what’s there and what’s not from using real world examples.
DR. CARR: One of the things that– Lily has put a nice summary together of the different approaches that we’ve had, and Bill started us off on thinking about what’s available and what it’s good for. We also thought about here are examples of how people use these data. Actually CMS just came out with a YouTube video, I might add, that shows you how to use the quality data. It is very good for people who are not technically savvy. I think at least– we’re talking about two ends of the spectrum, but at the start here’s some data, here’s who used it, here’s what they also used is another way of getting at this, just at the origins of the data.
DR. COHEN: So we could start with some of the examples from the community and what the problem was, and another stream to follow is starting at the data and how that trickled out and was used.
DR. WALKER: So responding to Vickie’s excellent point about scope, one of the things we might be smart about is what we’re better at and what other people are better at. I’m guessing other people are better at videos. If it was a $38,000 video, my guess is $35,000 was lawyer fees. That’s how I think about stories. We can encourage communities that were at the roundtable and elsewhere to tell their story in a video, and then encourage communities to vote on the videos in terms of their relevance and usefulness.
DR. COHEN: A data video challenge.
DR. VAUGHAN: There was a similar ongoing program at TechSoup Global, which is a non-profit technology organization. It’s actually global, and they have a whole lead in with how to do and with low cost and free resources. Some of the results will make your heart soar.
MS. KLOSS: What was the name of the company?
DR. VAUGHAN: TechSoup, and they have soup recipes, too.
MS. MILAM: We’ve been hearing over and over again that we really need to settle on what the problem is, who the audience is, are we going to help the community or the region or state-level, is it public health? We have this synergy with the stewardship committee, so I’m wondering would it make sense for us to develop two to three use cases together hearing from Beacon communities. Also, I think some of the CTSA folks, because at least the CTSA grant that I’m aware of, they’ve dealt with some really interesting privacy issues in terms of community consent and wrestle with questions of running notice of different research and the papers so at least the community would be aware when people– one of the groups was looking at doing something in the emergency department.
So people are coming in unconscious, so no way to get consent, so how do you deal with that? I know that they’ve wrestled with those issues. I’m wondering if we could have some folks through webinar, maybe CTSA’s Beacon communities inform us, and then we jointly develop these use cases that would be helpful to people in the field, and then go about the work to put the meat on the bones for the stewardship framework and whatever reports we might want to do from a population health perspective. Is it small area estimates? Is it identifying sources for technical assistance? Whatever the gaps are, and make sure it’s really targeted and we’re clear on what we’re trying to fix.
DR. GREEN: Another process comment, I just rescanned the second report. Listening to the conversations, I’m not convinced that all of us have digested our prior work here. I think we have an uneven understanding of what we’ve already learned from the roundtable and the previous workshop. We’ve got two years of work here. When you go back and look at the next step statements, and the stuff that we heard, a lot of the conversation is going on as if that doesn’t exist.
One of the secrets to NCVHS, staying on task and scope, is to build on its prior work. I guess this is close to whining or complaining, but I think we’re at risk for not taking advantage of work we’ve already done. That’s probably not a good idea. I really like this Leslie. So they don’t know how to get consent then. The rules for consent do not work in community-engaged research. We know this. We don’t have to go keep studying this. They ask how to do it properly. They get a run-around answer. We know this. We know that there are contradictory requirements for the use of data in different federal data sets. We know there are different custodians, stewards, and communities that don’t know who the hell to talk to about it.
They don’t know how to reconcile this sort of step. We know they don’t have infrastructures. We know they don’t have data stewards. We know they don’t know what to do after the data move from one location to the other. We don’t know who’s responsible for securing them. We know they have virtually zero knowledge about security. We know the rest of the world is roaring right along as if those problems don’t exist.
I’ll stake out territory to get refuted on. I think one of our sweet spots as NCVHS, as the National Advisory Committee about Health Data, is we need to identify those spots where people in communities, when they become learning health systems, are actually unable to act and are at risk because someone doesn’t. Those are problems to work on, in my view. These are captured to a very large extent in the two reports I’ve already got. My plea to both sides of this discussion is to please build off of what we’ve learned.
DR. FRANCIS: Could I just ask you for a concrete example so I can understand a little more, an example of something you’re talking about that’s captured?
DR. GREEN: Consent. When has a community consented to be part of a study? How do you know? We’ve held five meetings, four focus groups. We asked the county health department, and they all said yes, are we good to go? And when you did that and everyone said yes, but one neighborhood is stigmatized by the result and they say you didn’t ask us, what’s the proper response legally and ethically? I don’t think these questions are answered. These use cases are very useful for revealing them, but my point is a lot of the stuff that got passed by here are little mini-use cases that show these issues. I think we’ve got a sweet spot in there around proper development and use of data. I don’t think the crowd that she lives in is all that worried about that. I think this crowd over here is.
DR. TANG: Just to build on that a little bit, on things that already exist, we talked about the Agricultural Extension Center, the regional extension center, which is normally after that for HIT, already exists. Some are better than others, but the ones that work, the practices that were engaged had twice the rate of qualifying. What these communities need help with is data. We might actually be looking at the mechanism for getting this out right there. They even have the doc partnerships, and just because we want to reinvent another thing again– we know the problems they’ve asked for. We know where solutions may exist. We don’t know how to connect to them. This funding is drying up. We need to continue on a different mission. What do you do with the data you’ve got in your EHR, is sort of what we’re saying. They might have the right people in place, some of them.
MS. KLOSS: We could talk about convergence, because we talked about that yesterday in the context of ICD 10 and some of these other regulatory changes that– small practices just don’t know what’s about to hit them.
DR. COHEN: I’m turning it back to you two.
MS. KLOSS: I just have one higher-level question for the group. Maybe I’m just sensitized because it’s the week of the NSA stories. There’s a vulnerability here, too, that seems to me that needs to be raised up, that we really have stewardship obligations. We know what came out of the roundtable was some really good discussion on the public good. These are themes that should really frame the work we do on this going forward. How do we– whatever use cases, how do they advance the public good, and how do we encourage reasonable stewardship so we don’t have a backlash at some point, when people wake up and say what are you doing with my health data, which could happen? Our standards carpool, we talked about this over the last couple of days. That could be the headline.
DR. GREEN: That could be considered a new NCVHS invention, the standards carpool. That’s part of the process.
MS. KLOSS: There is really some urgency.
MS. MILAM: And on the horizon I was thinking about what Paul said, using data in electronic health records. We did not really hear at any of the hearings we’ve had about people doing that yet. It may not be the physician who has the electronic health record. How much research is the average practicing physician doing? It will be somebody else outside of the treatment relationship doing this research, and somebody possibly outside of HIPAA doing research. Perhaps that’s an area that’s ripe for exploration, that’s high with risk, that people are moving towards quickly, and will be there in a year or two. Maybe that’s something to think about.
DR. FRANCIS: This is just a question for people. I have shied away to some extent, and maybe I’m wrong, in terms of thinking about the research paradigm. I’ve thought more in terms of the public health paradigm. That’s why I haven’t thought of going to thinks like PRIM&R. I see a lot of nods. I won’t correct my thinking.
(Whereupon the breakout session ended)