This Web Conference consisted of three sessions broadcast via the World Wide Web and telephone September 23, 24, and 25, 2003. It was designed to inform State and community officials about and teach them to use Data Books and tools for monitoring the health care safety net. This initiative consists of a broad range of local area measures related to safety net providers and the populations they serve. The User Liaison Program (ULP) of the Agency for Healthcare Research and Quality (AHRQ) developed and sponsored the program.
Cindy DiBiasi: Good afternoon. Welcome to Data Books for Monitoring the Safety Nets. This is the first event in a series of three Web-assisted audio conferences on monitoring the healthcare safety nets. These events are designed for state and local health officials.
The series is sponsored by the U.S. Department of Health and Human Services Agency for Healthcare Research and Quality, also referred to by the acronym AHRQ or AHRQ and the Health Resources and Services Administration or HRSA. My name is Cindy DiBiasi and I will be your moderator for today' session.
In 2000, the Institute of Medicine released a report describing the healthcare safety net as "in tact but in danger." The safety net is as you know is the nation's system of providing healthcare to low income and other vulnerable populations. In particular, the report emphasizes the precarious financial situation of many institutions that provide care to Medicaid, uninsured and other vulnerable patients. It also examines the changing financial, economic and social environments in which these institutions operate and it looks at the highly localized patchwork structure of the safety net.
One of the five key recommendations in the report focused on the need for data systems and measures to assess the performance of the safety net and health outcomes of vulnerable populations. In response, AHRQ and HRSA are leading a joint safety net monitoring initiative. This initiative involves a three-part strategy focusing on both safety net providers and the populations they serve. As a result, they resolved to create two data books that describe baseline information on a wide variety of local safety nets. They developed a tool kit for state and local policymakers, planners and analysts to assist them in monitoring the status of their local safety nets and identify the data elements that would be needed to effectively monitor the capacity and performance of local safety nets. Information related to the AHRQ and HRSA initiative is available on AHRQ's Website at www.ahrq.gov/data/safetynet.
John Billings: Excuse me. This is John Billings calling remotely. I am in New York and they just had an alarm go off in the building and I have to leave so I apologize.
Cindy DiBiasi: Oh, John, OK.
John Billings: In New York we pay attention to those things.
Cindy DiBiasi: Yes, I bet you do. Go ahead.
John Billings: OK, thank you.
Cindy DiBiasi: Hopefully we will be able to talk to him later in this broadcast.
As we talk about the safety net, it is important to make sure that there is a common understanding regarding what it encompasses. As I said, the opened healthcare safety net consists of a wide variety of providers delivering care to low-income and other vulnerable populations. These include the uninsured and those covered by Medicaid. Many of these providers have either a legal mandate or an explicit policy to provide services regardless of a patient's ability to pay. Major safety net providers include public hospitals and community health centers, as well as teaching and community hospitals, private physicians and other providers who deliver a substantial amount of care to these populations.
In addition to today's event, two more calls will be conducted as part of the series this week. The next call, scheduled for tomorrow, September 24th, will focus on safety net data collection strategies and the topic for the third call, scheduled for this Thursday, September 25th, will focus on How Data can be used to tell the safety net story. I will tell you more about these calls later in the broadcast, but right now let's turn to today's call.
One of the challenges in monitoring the nation's healthcare safety net is that safety net services are provided in a myriad of different configurations, largely at the local level. Today we will be examining the newly-published Data Books, which include information at the county and metropolitan levels. It focuses on 30 states and the District of Columbia. Together these areas cover 75% of the U.S. population. The books use data from a wide variety of sources to describe the status of these safety nets in 90 metropolitan areas and 1,818 counties in these states. The books provide a broad range of measures for monitoring the status of safety nets and the populations they serve.
Let me begin by introducing today's panelists. In the studio with me I have Robin Weinick, senior research scientist and senior advisor on safety nets and low-income populations at AHRQ and Robert Seifert, policy director at the Access Project in Boston. You heard on the phone was John Billings, the director of the Center for Health and Public Services Research at New York University's Wagner School of Public Service. We do hope to be talking to John later in this broadcast. Welcome Robin and Robert.
Before we begin our discussion, I would like to tell the audience a bit about the format of the Web-assisted audio conference. First we will talk with our panelists and then open the lines to take your questions. We will give instructions on how to send your questions to us later in the program. In the meantime, if you experience any Web-related technical difficulties during this event, please click the "Help" function in your window to troubleshoot your Web connection. If it appears that the slides are not advancing, you may need to restart your browser and log on again. If you are on the phone, dial "*0" to be connected to technical assistance. Also, if you have difficulty with the audio stream or if you experience an uncomfortable lag time between the streamed audio and the slide presentation, we encourage you to access the audio via your phone. The number is 1-888-469-5316. This is the same number to call to ask questions when we get to the question and answer portion of the program.
Now I think we are ready to tackle today's topic, The New Data Books for Monitoring the Safety Nets. I would like to talk to Robin Weinick about what did you set out to accomplish with the Safety Net Monitoring Initiative? You were a principle collaborator on these data books and worked to gather this data.
Robin Weinick: Right, Cindy, thanks. What we really found was we started with the IOM reports that you described earlier. They were from the Institute of Medicine and one of their key recommendations was that the data really did need to be gathered and compiled because when they went out looking for information, they couldn't find what they needed. So that was the motivating factor behind their recommendations. As a result, our goal is to begin to provide a national capacity to assess the status of the safety nets throughout the country and the populations they serve and access problems for them. In particular, to really help local communities and state and local health officials by illustrating approaches to how you might measure the status and performance of the safety net and by providing benchmarks that they can compare themselves to.
Cindy DiBiasi: What areas are included in these books?
Robin Weinick: Our goal was to include as many areas as we possibly could and we are limited only by data availability. So we are in 30 states. Within those states, we are in every county. We are also in 90 metropolitan areas, including more than 300 counties and 170 cities in those metropolitan areas, again limited only by the availability of data.
Cindy DiBiasi: To clarify, this data was all data that was existing data.
Robin Weinick: Absolutely. We didn't collect any new data for this. What we did was to go to a wide variety of data sources where we could get information down to the county level and pull those together so people would be able to do additional analyses to understand what is happening with local safety nets.
Cindy DiBiasi: And this is the first time it has all been pulled together in one place?
Robin Weinick: It is the first time.
Cindy DiBiasi: Great. What was the conceptual model you had in mind when you started working on this?
Robin Weinick: Well, we know that safety nets are largely local. When you have seen one safety net, you have seen one safety net. So what you really need it for is our data at the local levels, but the boundaries that matter differ from place to place. In one area, the city might be what matters. In another area it might be a multi-county authority. So we wanted to get our data down to as fine a local level as we could.
We also think that the health of the nation's safety net and its capacity to provide care to low-income populations is somehow the sum of all these different local safety nets. So we wanted to gather data for as many geographic areas as we possibly could. We also thought it was very important to focus on patient outcomes, which is something that hasn't always been done before. Since the factors that influence those outcomes are complex, we knew we needed a broad range of data to tell the story.
Cindy DiBiasi: Tell us how you selected the measures that you included and what types of measures are included in this?
Robin Weinick: Well, Cindy, actually if you don't mind if we talk for just a second about the model that we really used. When we talk about what contributes to optimal health in the safety net, we know that there are a wide variety of things that are components of optimal care, from genetic and the environment all the way down to how people manage their own diseases and conditions. The safety net plays a crucial role here in terms of care availability and provider performance. We know there are also a number of mediating factors such as personal circumstances and community resources that also influence optimal health. So when we picked the types of measures, what we were looking to do was to get as broad a range of these concepts representing the measures as we could.
Cindy DiBiasi: How is the information in these books structured, Robin?
Robin Weinick: Well, the book actually has two major parts to it. The first book has some beginning material, which I am going to show you actually some of the graphs and charts from that where we actually summarized the information that we have learned. We tried to present what the bigger picture is across a wide range of areas. We also have an extensive number of tables where we actually provide detailed information for each geographic area on a wide number of measures. So we actually have very, very large data tables, some for metropolitan areas and the second book is actually specific to counties and states.
Cindy DiBiasi: The books are very well laid out. There is a tremendous amount of data in the books, but as you are going through the books you can see the format is very easy to read and to locate specific information.
Robin Weinick: Right. That was really one of our main goals. We want to make these as usable as we possibly can for our audiences. So whether what you need to find is how two measures might relate to one another across a wide variety of areas or whether you want to know a particular measure for Lackawanna County, Pennsylvania, you can do either of those.
Cindy DiBiasi: I am going to ask you about that because that is my old hometown. I know you don't have time to talk about all the measures, but tell me about what you mean by "demand for safety net services."
Robin Weinick: Sure. We have a wide variety of groups of measures of which demand for safety net for services is one. We also have some information, for example, on financial supports for safety net services, how the safety net is structured, as well as I mentioned on patient outcomes.
In terms of our demand measures, probably the two most important demand measures that anyone I know who works in a safety net thinks about are the poverty population and the uninsured population. So this figure, I am going to actually show you some figures from the book. We have actually taken out the labels of some of the geographic areas just to make them easier to read. Many more of the areas are labeled in the book. So what this shows you is the percent of the population below 200% of the federal poverty line. That is about $35,000 a year for a family of two adults and two children. What proportion of that population is uninsured? What you can see here immediately is a tremendous geographic variation, even within regions of the country. You can see, for example, Springfield, Massachusetts, St. Louis, Tacoma, Washington, at the low end here where Jersey City, Augusta and San Francisco are at the higher end with a larger proportion of their low-income population being uninsured.
I also wanted to take a look at the relationship between poverty and uninsured because people often think if I know one of these two measures for my geographic area, I know what the demand for safety net services is. It was very important for us to point out. This is just a scatter plot that shows you these data relative to one another. Each little dot is actually a geographic area, a metropolitan statistical area. What this shows you is that there is virtually no relationship at the metropolitan level between the poverty population, the proportion of the population that is poor and the portion of the population that is uninsured. You actually need both pieces of information in order to understand what is going on.
Another picture shows you the percent of the population below 200% of poverty that is enrolled in Medicaid. Obviously Medicaid is a federal-state partnership and the states take the main initiative for things like determining who is eligible for outreach programs and that sort of thing. It really varies tremendously from state to state, but you really can't see it here. Again, we have taken some of the labels out to make it more readable. It also varies quite a bit within states.
Cindy DiBiasi: You mentioned Lackawanna County, Pennsylvania. So let's talk about within a state, if we wanted to find out what percentage of the population over 65 in Lackawanna County was living under the poverty line.
Robin Weinick: Right.
Cindy DiBiasi: You would be able to drill down to that?
Robin Weinick: We can do that. I can look up in the books. I can look up in the electronic data books or in the safety net profile school that I will share with you a little later on and tell you that 10.6% of the population of Lackawanna County lives below the poverty line. Compared to Scranton, which is a city right nearby, where it is 15%. So things are considerably worse in the central city of Scranton than they are in the surrounding county of Lackawanna. So we can actually drill down and get down to that fine a level of detail or we can talk at a bigger picture level.
Cindy DiBiasi: I understand you also have some measures of what the overall healthcare system is like in different communities. Can you talk a bit about that?
Robin Weinick: Right. Well let me tell you first about some measures of what we know about community health centers and what is happening. What this picture shows, very interestingly, is that community health centers are in fact located where the poor people are located because if you look in counties where fewer than 6% of the population lives below the federal poverty line, 15.9% of those counties have a community health center. By the time you get up to a county with more than 20% of the population living below poverty, more than 90% of those counties have community health centers. This is a measure of the resources that are available and what aspect of what the healthcare system looks like.
So let me tell you about two cities. I like to call this "A Tale of Two Cities." We are looking at Portland, Oregon and Newark, New Jersey, which are cities that are roughly the same size. You can see how very, very different they are with much higher HMO market penetration in Portland but much greater healthcare utilization as well as numbers of physicians per thousand persons in Newark, New Jersey.
For those people who say that that is not a very fair comparison, they are in two different states. State health policy can have a real impact. This is "A Tale of Two Counties". We are looking at Orange County and San Francisco. Again, you can see how tremendously different they are where in San Francisco there is considerable public hospital presence. In Orange County there are no public hospitals, but there is a considerable investor-owned hospital presence.
Cindy DiBiasi: What other data are included in the books?
Robin Weinick: We also have quite a bit of information on community context where we have pulled data in from the U.S. Census to tell us about population size and growth, age distribution and ethnicity of geographic areas, income, unemployment and education. So we really have tried to not only focus on healthcare, but on all of the miscellaneous things that also happen in a community that may impact on healthcare.
Cindy DiBiasi: Does this tell us how well the safety net is actually performing?
Robin Weinick: Well, it is very, very hard to tell how any aspect of a healthcare system is performing. For the safety net in particular, as you mentioned very early on, a safety net is a very amorphous kind of a thing. It involves lots of different institutions, lots of different providers in lots of different locations. But there are a few really well-accepted measures in general of access to care that can help us start to understand what is going on. In particular, understanding preventable or avoidable hospitalizations. These are hospitalizations where if patients are receiving really good ambulatory care in the community, they really should not theoretically wind up in the hospital. For example, if asthma is very well managed in an outpatient setting, you don't expect to find people hospitalized with it. Obviously there are always a few extreme cases, but in general rates of preventable hospitalizations can tell us something about what is going on because we know that preventable hospitalizations occur disproportionately among low-income populations.
We also know that birth outcomes are a really great way to tell what is happening with care in the community. So where you are really looking at the percent of births that have late or no prenatal care or the percent of low birth weight births. That really just starts to give you a sense of what is going on in a community.
So, what we took a look at were preventable and avoidable hospitalizations. These are per 1,000 children, ages 0-17. You can see again just the tremendous, tremendous variation within geographic regions from one city to another. Then we took a look at what the relationship was to poverty because we know that there is a really tight relationship and what you can see here is the relationship for children. Then if you take a look at the relationships for adults, there is actually a much stronger relationship for adults than it is for children. We don't have any way to actually test this, but we are wondering if what we are seeing is the impact of some of the Medicaid and S-CHIP state children's health insurance program expansion is really starting to break down the relationship between poverty and preventable hospitalizations for children.
Cindy DiBiasi: Can we summarize some of these "big picture" findings from the project?
Robin Weinick: Yes we can. We really went to a lot of trouble to do that. First I think the first big point I would make is that federal and state financing of the safety net really does help. We know that greater Medicaid coverage and higher disproportionate share hospital payments are generally associated with lower preventable hospitalization rates and better birth outcomes. We also know that public facilities matter. A larger public hospital presence is associated with better outcomes.
We also found interestingly, that more providers are not always the answer. We know that more pediatricians are associated with lower children's preventable hospitalizations, but we didn't find the same to be true for adults or for the relationship between have more obstetrician/gynecologists and birth outcomes.
We also found the levels of personal distress are concerns. Higher levels of poverty, unemployment, disability, lower levels of education, all of these kinds of things are actually associated with worse outcomes.
Cindy DiBiasi: We are going to come back. We have lots more questions for you, but before you go any further we will let you take a breather. We are going to go to Robert Seifert from the Access Project in Boston. The Access Project supports local leaders in their initiatives by providing them with assistance designed to enhance efforts by a (unclear) Access. Bob, you work with organizations around the country whose primary focus is to strengthen healthcare access. How are these data books going to be helpful to that purpose?
Robert Seifert: Cindy, I think that the data that are collected and presented in these books provide a great common ground and measurable target for discussion of policy initiatives at the local level. Data like these are all too often absent in discussions of local healthcare system and health policy questions. The data can help shed light on where policy efforts might be focused and to support decisions that are based on concrete data rather than on anecdote, which is frequently the way that policy is made in local discussions. It helps us to get past the arguing about the data stage that often befalls a lot of problem solving attempts in local communities by providing this common ground.
The data also offers the foundations of a roadmap to safety net improvement through the analysis that Robin just talked about, the multi-variant analysis relating it to outcome measures and so forth. But also uses the data to remind us that there are a number of important factors such as local economic conditions and racial disparities that go beyond the realm of the healthcare system into much broader policy areas.
Finally, another great use I think of the data is that it can help combat the isolation that is felt by local providers and policymakers and advocates that are struggling with their own problems in their own communities. Providing this information helps to bind these communities together in a way and helps local actors understand their circumstances relative to other communities. It gives them a voice, a very important voice that needs to be heard in policy debates about the future of the safety net.
Cindy DiBiasi: Let's talk a little bit about why aggregate data are not sufficient. What are some ways that the availability of local data improves on, for example, state or national data?
Robert Seifert: Well, Robin touched on it a little bit in your earlier discussion. Variation across geographic areas, regions and states across the country and so forth, is well documented, well known. What is striking in this data book is that it allows us to look at variation within geographic areas. This is crucial, of course, as we have been talking about that because safety nets are of course local facing uniquely local problems and local circumstances. Good information about these local situations is vital to meeting a community's needs.
I think I can best illustrate this with an example that I have taken from the data book. If we look at the situation of people with limited English proficiency, this slide is looking at the rate of people with limited English proficiency in the Hartford metropolitan area in Connecticut and in the United States. You can see that the percentages are relatively the same. Connecticut is not necessarily known as a state with a high limited English-proficient population, but if we look beyond that, deeper into the Hartford metropolitan area, we can then see that the City of Hartford itself has actually a very, very high proportion of people with limited English proficiency. This large non-English speaking population signals a need for access to language interpretation service as part of the safety net infrastructure. This is information we would not have if we were just comparing the Hartford metropolitan area to other metropolitan areas or Connecticut to other metropolitan states. So the ability to look deeper and to look within local areas and the variation within local areas is a great advantage to people who are trying to make improvements in the safety net.
Just one other point that I want to make on this is that the diversity of the data in the book allows us to look at peer communities. For one community with a particular safety net situation to look at another community with possibly similar demographic information, for example, which maybe has better or worse safety net outcomes for comparison purposes, for benchmarking purposes, to look and see. Here is a community all the way across the country that looks like mine, but their safety net seems to be working better and what can we learn from that? Some of the profiling tools that Robin is going to talk about in a little while I think is a great way to make those comparisons using the data in these books in a very simple way.
Cindy DiBiasi: You talked a bit about how valuable the data are in these books and there are 118 different measures, yet you are saying that local data are elusive. Why would you consider that elusive?
Robert Seifert: Well, there is never enough. I think that people who are listening in and looking in on the Web are already thinking of questions about "is there this in the data book? Is there that in the data book? Why isn't there data on uninsured rates in cities and counties?" I think the point is, as Robin said, these are data that are compiled from existing sources. There is what there is; we'd always like there to be more. In the absence of some of those things, though, the fact that there are so many measures in this one place helps us because we can use substitutes. One of the things that Robin and John talk about in the book are indicators of personal distress, which actually turn out to be very good prophesies for uninsured, the uninsured rate in local areas that we don't have. So, yes, we wish there were more. Maybe someday there will be if the resources are there to collect it. But for the time being, the fact that all of these data are in one place, we can look and have the data say things that we want them to say without necessarily having the exact indicators that we would like to have.
Cindy DiBiasi: As we talk about the purpose of these data books is to contribute to efforts to preserve and strengthen the safety net. Do you think that they are serving that purpose and what else would you like to see?
Robert Seifert: I do. Again, I think that the major benefit of these data is that they are all in one place. They give us a very local perspective. I meant to say the word "local" a lot more times than I have so far because I think that is critical to this and really the important point here.
I think these data are very useful for making comparisons with other communities, for providing warning markers of potential problems in local safety nets and as a way to monitor improvements or deterioration in safety nets over time.
I want to emphasize though, as a card-carrying data nerd, it actually pains me to do this, but the data don't tell the whole story. No set of data can tell the whole story. They are starting points. As comprehensive as they are, they really can't capture all of the important features of a local safety net. Some of the data aren't there, as we have talked about, and some things are just not easily measured. I want to caution users of this data not to fall into the measurement trap of saying if there is not a measure for it, it doesn't exist. Because there are important things that you can learn from the community. Go into the community. Learn from the providers and from the consumers in communities. Some of the texture that these data point to but really can't tell you the whole story about. Use them as a jumping-off point, but then go and learn from the community itself what is behind the data. Why are the measures as they are? I think that is the step beyond the data book that I would suggest people take.
Cindy DiBiasi: Thank you, Bob, and we will come back to you. Before we move on, I would like to ask Robin about the availability of this new data. Now that all the data is available, how can our listeners access the product?
Robin Weinick: Well, first I really want to thank Bob. I am excited about this, obviously having worked on it for a long time. But he makes it sound like the best thing since sliced bread. We hope our users think that as well. So there are a variety of ways to access the data. We do have two print volumes available. The first book focuses on metropolitan areas with the beginning of that book providing some big-picture overview and then getting into all the details and a second book that focuses at the state and county level. This includes rural areas as well so it is every county in 30 states. We have a bunch of electronic files as well. The books themselves are actually available in an electronic version as well as Excel spreadsheets for folks who do statistical calculations we have the data available in a number of statistical formats as well. All of this information is available on the Website, which is www.ahrq.gov/data/safetynet. The information we have there will tell you how to order copies of everything. It will tell you how to get a copy of the print books, which we will send to anyone for free. All they have to do is ask and there is a variety of ways to get that information. If you look at that Website it is going to show you right on the first page of that Website. Just scroll down. It is going to tell you how to order them. There are also ways, for example there, to get technical assistance with the data or provide feedback as well.
What I would really like to tell you about is the Safety Net Profile Tool. This is for people who just want quick, online access. I know it is a little hard to see on the screen. Unfortunately there are some technical difficulties. But what this tool lets you go in and do, you say I am going to start a query and then I am going to go in and I am going to choose what type of geographic area I want to look at. That could be a metropolitan area, a county or a state. Then what it will actually let you do is go ahead and compare different geographic areas. For example, if you happen to know, you mentioned you are interested in Lackawanna County, Pennsylvania, that that is an area that you are interested in and you want to compare to other counties that may be similar but you don't know where else. For example, there could be a similar county out in California. We actually provide you with a tool where what you do is say find me some similar counties. It is going to look through the data set based on population, size and poverty rate which are two very key aspects to what the safety net might look like and it is going to find some suggested comparison areas for you. You can choose any or all of them or you can choose your own comparison areas from any of the states, any of the counties included in the data book. It will then ask you to choose which of the 118 measures you would like. You can get a few of them if you would like or you can have all of them. Then it is going to go ahead and actually produce a table that just has the information that you have asked for. So if you don't want to page through a book, this is a quick and easy way. For example, if someone asks me what percent of all admissions in Polk County, Iowa occur in public hospitals? I can click through and in probably 30 seconds or less have the answer, 10.8%. It really gives you the ability to access just the data you are looking for, just when you need it.