On July 9, 2009, Ernie Moy, Foster Gesten, and Keely Cofrin Allen presented
a Web conference on the AHRQ 2008 State Snapshots Tool. This is the
transcript of the event's presentation.
July 9, 2009
1:00-2:30 PM ET
Margie Shofer: Hello, everyone. I'm Margie Shofer with the
Office of Communications and Knowledge Transfer at the Agency for Healthcare
Research and Quality, also known as AHRQ. Thank you for joining us for this
Web conference on the 2008 State Snapshots tool. The State Snapshots,
which AHRQ has been releasing since 2005, are dashboards that offer visual
presentations of health care quality measures. They provide users with a summary
overview of States' performance on: types of care (preventive, acute,
and chronic); settings of care (hospital, ambulatory, nursing home, and home
health); and care by clinical areas (cancer, diabetes, heart disease, maternal
and child health, and respiratory disease). Users can also drill down into
the individual component measures of the summary to see how their State performance
compares to all other States, best performing States, and selected States in
their region.
This Web conference is part of a project to share AHRQ tools of interest to
State policymakers. In addition to sharing our products we are also offering
technical assistance in using the product so if after learning more about State
Snapshots today you're interested in further assistance in using this
tool, please let us know. I also want to mention we'll be holding another Web
conference on another AHRQ tool, the Asthma Return-on-Investment Calculator,
in the early fall. Please stay tuned for an announcement about that event later
on this summer.
Today we'll spend some time demonstrating how to use the State Snapshots,
primarily examining the latest features of the tool, which include revised
graphics to display summary measures, the ability to calculate the potential
return on investment of the asthma care quality programs for Medicaid, State
employees, and the privately insured, as well as diabetes programs, and a link
to examples of innovative practices to improve quality of care in the area
of diabetes and asthma, drawn from AHRQ's Health Care Innovations Exchange.
We will also discuss the data underlying the tool. We will have some time
for questions after this presentation before moving on to presentations highlighting
two States, New York and Utah, which have used the tool in their quality improvement
efforts. We will end with another question and answer period in which we will
address as many of your questions as possible. We really would appreciate your
active participation today since we want to make sure you're comfortable with
using the Snapshots after this Web conference ends. That said, there
are a large number of individuals on this call so if we don't get to your question
please e-mail it to us. The e-mail address will be posted at the end of the
Web conference along with the link to access the 2008 State Snapshots.
We promise we'll respond to all questions we receive after the Web conference
ends.
Now I'd like to introduce our three speakers. Today's tool demonstration
will be given by Dr. Ernie Moy, who is a Medical Officer in the Center for
Quality Improvement and Patient Safety at AHRQ. His work has included directing
the development of the annual National Healthcare Disparities Report [NHDR]
and the National Healthcare Quality Report [NHQR], and supporting
AHRQ's excellence centers for the elimination of ethnic racial disparities
and AHRQ's Patient Safety Organization program. Ernie's research interests
include disparities and access and quality of care, particularly as they relate
to academic medical centers, patient safety, and technology diffusion.
Our next speaker will be Dr. Foster Gesten, who is a Medical Director for
the Office of Health Insurance Programs in the New York State Department of
Health. Dr. Gesten provides clinical direction and leadership for a team of
professionals engaged in quality oversight, performance measurement, and clinical
improvement with health plans and public insurance programs in New York. Major
initiatives include the development of State-wide public reporting systems
for commercial, Medicaid, and child health managed care programs on quality,
access, and satisfaction. His interests include population health, health services
research, and quality improvement projects directed to prevention services
and chronic care.
Our final speaker is Dr. Keely Cofrin Allen and she is the Director of the Office
of Health Care Statistics at the Utah Department of Health. Her department is
responsible for a variety of health care quality data such as in-patient hospital,
emergency room, and out-patient surgery data, as well as HEDIS [Healthcare Effectiveness
Data and Information Set] and CAHPS [Consumer Assessment of Healthcare Providers
and Systems] data from Utah's HMOs [Health Maintenance Organizations] and PPOs
[Prefered Provider Organizations]. The office publishes hospital comparison reports
on six different health conditions and an annual HMO performance report on commercial
Medicaid and CHIP [Children's Health Insurance Program] health plans and
annual reports on hospital in-patient, emergency department, and ambulatory surgical
visits. So without further ado I'm going to hand this over to Ernie.
Ernie Moy: Thank you, Margie. And I'm going to start the
demo by shifting over to my desktop. Hopefully you all can see the main page
for the National Healthcare Quality and Disparities Reports. I'm supposed
to remind you that if you want the captioning panel, you should click on the
icon all the way on the right—to the right of the question mark—to
open that particular panel.
So, I start here—even though today's talk is about the State
Snapshots—because I wanted to remind you and remind myself that State
Snapshots are part of a panel of tools that we have at AHRQ, the core
of which are these two reports: the National Healthcare Quality and Disparities
Reports. These two annual reports to Congress on the status of quality
and the status of disparities try to provide people with baselines, benchmarks,
as well as trends in quality and disparities over time. The Reports themselves
have a couple of parts so I wanted to highlight several things that
do relate directly to the State Snapshots. They take all of the
measures, roughly 250 measures we track in the two Reports, at the
national level and for those measures that we have State data for—not
all have State data but many of them do—we then focus in on the State
aspects in the State Snapshots. All of the tables that go to support
the State Snapshots are located in the two Reports, specifically
in two places. In the data tables part of the Report Appendices you
can find the series of tables that go along with each measure. A product
that we call the NHQR/DRnet is a place where you can develop customized tools
to look up each measure and pull up the information that you're interested
in related to that particular measure, including all of the State data tables.
I'll now head into the 2008 State Snapshots. This is the main page
where you'll arrive. Again, the purpose of the State Snapshots is
to allow States to see how they are performing, but more importantly to see
how they're performing relative to other States. Here we're looking at an individual
State compared with other States. I'm going to give examples of this tool using
our two guests, Utah and New York, and I'm going to start with Utah. I'm going
to go to the State selection map and pick Utah. When you first come in, you
come to this dashboard, but I'm actually going to skip past it and come back
to it. You'll see on the left hand menu how everything is organized, and this
is consistent with what Margie had mentioned earlier. I'm going to walk you
quickly through the different parts of it. Later on if there's a desire to
go back to something I think we can do that and answer specific questions.
I'll go through it really quickly because there are lots of parts. I'm not
going to go through all parts either; I'll just try to key in on the ones that
are most important and newest.
What everybody typically wants to see first is overall quality. This summarizes
overall quality of care across all of the different measures we track in a
State. Again, the emphasis here is that we track many measures. As a consequence,
we don't drill particularly deeply into anything but we have measures that
cover everything from ambulatory care, hospitals, home health, nursing homes,
and hospice—a very broad range of different kinds of settings. And we
sum it up into this metric.
Here, you see Utah is doing about average; the dotted line indicates the baseline
year, and the solid line is the most recent year. One feature built into this
is, if you click on each of the meters and everything is shown in the same
pattern, you see the underlying summary data table that goes to create the
average you see. What we see for the State of Utah is that they have 112 measures
which we have State information for and of those 112 measures, they're better
than average for 32, average for 53, worse than average for 27, and missing
information for 8. That's just a very overall high level distribution, and
then immediately underneath it we have the measures, all of the measures categorized
by where they're better, about the same, and what I'm mostly interested in,
maybe you too, is where they're worse than the State average. I won't dwell
on it but this where people might look for opportunities for improvement in
their State.
I'll show some of the other features that Margie mentioned. This compares
a State, Utah, with all of the other States, but maybe you don't want to know
about how you compare to all of the other States because your part of the country
is different. You can pick your census division and here this shifts it from
Utah compared to all States to Utah compared to just the Mountain States, those
parts of the census division that are in the mountain region.
The last part that Margie also mentioned is the ability to compare to the
best performing States. If you click on this you see the overall performance
again. Utah's overall performance on the meter was a 52 and you see the best
performing States—this is available for all of the different subscales
that we have as well. In addition to seeing how their State is doing, users
can see how they stack up against the best performing States and then the 25th,
50th and 75th percentile. We particularly put the best performing States up
there because I might want to contact a different State to see how these States
were able to achieve their high performance levels.
Another feature that we've had for a number of years is the strongest
and weakest measures. The strongest measures are measures where the State performs
the best relative to other States. You see how Utah does, and the weakest measures
are those measures where the State does the worst and you can see our assessment
of areas where Utah is weakest relative to other States: Pap tests, fecal blood
testing, quality of dialysis, diabetes, eye exams, and the nursing home measure,
a broad spectrum of measures but at least three of them relate to primary care
activities.
I'm going to walk you very quickly through the different categories that we
have for Utah and for all States. One categorization is by type of care, so
this would include preventive care, acute care, and chronic care. I'll click
on preventive care and you can see Utah is above or about average. And if I
click on the underlying data, we can see how they are doing compared to all
of the other States. You can also see what their summary looks like for preventive
services only, and again, about average looking. I'll scroll all the way down
to the bottom and you see areas popping up. I think these are some areas that
we saw earlier on the worst performing aspects of Utah, but again, you see
some other ones popping up as worse than average as well. Again, a lot of preventive
services: blood testing, cholesterol screening, prenatal care.
Under acute care, Utah is about average. Clicking to see areas of opportunities
for improvement potentially, we see a number of heart attack related measures
and a number of surgical antibiotic measures are down here at the worse than
average range. For chronic care—this is chronic management of a disease—they're
a little bit better. The measures are in the upper range of average, getting
into the strong sector, but opportunities might exist here as well; they typically
do. We see some of the things that we saw before; diabetes related services,
and nursing home related activities where they are performing worse than the
national or State average. That's one way of looking at a State to try to get
some insight about where to target, in this case maybe preventive care as opposed
to chronic care or within a particular type of care, specific types of conditions
or measures that might benefit from some improvement.
Another way to look at it is by setting of care: hospital care, ambulatory
care, nursing home care, and home health care. Hospital care is about average.
Scrolling all the way to the bottom, we're seeing again those heart attack
measures popping up, as well as the surgical antibiotic measures popping up.
And there is a strong relationship, as you can imagine, between acute care
and hospital care. For ambulatory care, we're seeing again the cancer screening
measures and then a couple of diabetes measures, more screening measures, prenatal
care, and suicide deaths. Again, the notion here is to give a State some insight
into the areas where they might be able to see improvement.
Nursing home care is more of a problem. I think we saw that earlier when we
saw a long list of the worse than average nursing home measures. And here we
see them. For many of the measures of nursing home care, it's worse than average.
And home health care is very strong. Every State has something they're really
good at, and home health care measures for the State of Utah are no worse than
average.
Another way that we have for people to examine their State is by clinical
area. These are the main domains that we examine and allow examination: cancer,
diabetes, heart disease, child health, and respiratory disease. I think we
already saw a lot of cancer, diabetes and heart disease pop-up. Maybe I'll
click here for maternal and child health—strong performance. A measure
that's a problem: prenatal care. Everything else is average or above average.
Respiratory disease—also strong performance. Hospitalizations for influenza
are higher than the all-State average.
And now, we'll show you the way that you would probably do it. Instead of
walking through each of these individual different panels, although I like
to do that to just show you exactly what's in each panel, just go to the State
Dashboard which is where you first come in, where you'd see this overall quality
of care meter. Underneath it, you would see all these rows that show the performance
along types of care and settings of care by clinical areas, these match up
exactly with each of the meters I showed you earlier. Now, it's been flattened
out into this spectrum instead and you see where Utah currently performs and
you see where the baseline year is. This is a way to look across all of the
different dimensions to see areas where there might be the most opportunities
for improvement. You see preventive care and acute care are a little bit worse
than the chronic care. You see nursing home care being the worst of the different
settings of care, and you see maternal and child and respiratory disease measures
being better than cancer, diabetes, and heart disease. This is kind of a similar
picture but summarized all into one place. You can still click on these and
get the underlying tables as well.
So that's Utah and I'm going to switch over to our other State, New York,
because they have some interesting data related to some of our focal areas.
Quickly, the Dashboard for New York looks like this: overall, average kind
of quality, and you can see areas that they're doing the best and worst in,
acute care measures, hospital measures typically as well as home health, respiratory
disease, and diabetes seem to be potential problem areas but I wasn't going
to focus on that.
I was going to focus on those things that New York has really interesting
information about. One of these is our focus on diabetes. Since we've had this
now for a number of years, it's been refined a little bit. In the top part
we try to show some information related to diabetes in the State of New York.
And then down here we have a series of resources and links outside of this
particular tool.
To give people a little bit of background about diabetes in New York, here
is our prevalence map. It shows how New York is doing relative to the rest
of the Nation in terms of overall prevalence of diabetes, and shows that diabetes
is a concern in New York State. It's on the higher range in terms of prevalence.
We can look at the quality of care for diabetes—this is the meter that
focuses just on the diabetes measures, now differentiating process of care
from outcomes of the care—so we can see for the process of care New York
just does average on the different process of care that we can track. It is
right in the middle. But in terms of outcomes of care you see New York State
typically has higher rates of hospitalization for diabetes compared to the
Northeast and to the Nation, and as a consequence of that, there are quality
improvement opportunities.
For the diabetes example, we focus on potential cost savings related to State
government employees, something that many States are interested in and so we
produce an estimate of the proportion of expenditures on health care for State
government employees that relate to diabetes care. There are opportunities
for improvement for about 3 percent of those costs or this amount of money
is at risk. Then from the diabetes calculator, we make projections about how
much money could be saved per year with the implementation of a disease management
program that achieves different levels of improvement. We have average results
for an overall average improvement in the level of .5 percent, a reduction
of .5 percent. These are the associated direct costs and productivity gains
that could be achieved through that level of investment. If they could bring
down their hemoglobin A1C on average by 1 percent then these are the corresponding
amounts of money that could be saved each year. A caveat: many of these savings
may not be realized for some time after implementing a disease management program,
and so these are optimistic projections and longer term projections, not necessarily
things that anyone would see right away.
Under the focus on diabetes, we have a section that looks at disparities in
treatment. Instead of looking and sorting States by how well they perform on
measures of diabetes quality, we're looking at the gap. The difference in this
particular case is between residents with high income and low income. No actually,
I take that back. This is actually based on self-reported income, so the difference
is between people who report low income and high income. We see that New York
is about average on this particular metric and you can look down and see what
it actually looks like. For the proportion of diabetics in New York who get
a hemoglobin A1C for the low income versus high income, you see a gap, but
the gap is about comparable to what we see in other States, both in the Middle
Atlantic and all States overall. And then we can see the similar contrast based
upon ethnicity and it's about average for the Black/White difference and about
average for the Hispanic/non-Hispanic White difference. Here you see the metrics,
the actual measures: Hispanics and Blacks are less likely to receive hemoglobin
A1C testing but the gap is about comparable to what's seen in other States
as a whole.
I promised to go through these resources and innovations so I'll mention them.
One is, now you've identified diabetes as a potential problem. What are things
you can do? One thing you can do is look at our Resource Guide for State
Action for Diabetes Quality Improvement. This is a resource guide and
workbook that walks through what a State might want to do to improve quality,
starting from making the case for improvement to actually developing an intervention.
This is the diabetes cost calculator, which was the basis for making the estimates
for the excess costs of diabetes and is a new link for the Diabetes Innovations
Exchange. This links to AHRQ's Healthcare Innovations Exchange tool. This is
a place where people have logged innovative and interesting stories about quality
improvement, less successful attempts at quality improvement, and then tools
that are useful for quality improvement purposes. This pulls up the ones that
relate to diabetes, so if you want to know what to do next this might be one
place to go.
Now, the focus on asthma. Asthma prevalence is another condition specific
focus area that parallels the focus on diabetes. Again, New York is about average,
a little bit on the high side in terms of asthma prevalence, and quality of
care has very high rates of hospitalization for children, adults, and elderly
adults. Here we have a calculator that allows people to calculate return-on-investment
for improvements in asthma care. It's different from the one related to diabetes
and I think it allows one to drill down a little bit more financially. It breaks
the populations into children and adults and then separates them by insurance
groups, into Medicaid, State employees, and private insurance. The State employee
adults is similar to what I think we saw with diabetes but presented in quite
a different fashion. Instead of giving you overall cost savings, this gives
you approximate cost savings per person if the 253 adult State employees with
asthma were able to be entered a disease management program that improves asthma
management. These are the program costs over 3 years and these are the estimated
savings in terms of direct costs and productivity gains. Here it gives you
the return-on-investment, one for direct costs, just savings in terms of direct
costs, and the other one incorporates the concept of improved productivity
and also gives you estimates for the Medicaid population: 4,955 adult Medicaid
individuals as well as privately insured individuals in the State.
Again, just like for the diabetes, we have a link to a State-oriented resource
guide and workbook that can walk you through how to potentially improve the
quality of care in the State as well as the link over to the Innovations Exchange.
This time links pull up interesting stories, those success stories related
to asthma quality improvement, as well as tools that might be useful for asthma
quality improvement.
I'm going to skip over these pieces and go to another new part of the State
Snapshots: the focus on disparities. Just like I showed you earlier
for diabetes, focus on a given measure is on looking at the gap between different
populations as opposed to overall quality of care. This series—this
focus on diabetes series—now looks at a number of potentially avoidable
admissions and looks again at the gap between different populations as opposed
to overall quality. This shows that looking just at the Black/White difference,
New York tends to be a little bit worse. What does that mean? Well we can
click on the meter and you can see that gap is defined as a relative rate,
the Black rate divided by the White rate, and you see that for New York,
Blacks do worse than Whites, that is Blacks have higher hospitalization than
Whites for a number of the measures, only two or less. Moreover, if you compare
the New York rate with the United States rate, the difference, the relative
difference, is larger than in the United States overall. Divide the New York
relative rate by the U.S. relative rate and when it is over one, New York
has a larger relative disparity compared to the United States as a whole.
You see it across many of these measures, it exceeds one overall. Our assessment
is that the Black/White difference is larger in New York State overall than
it is for the U.S.
We also have comparisons for other groups: Asians versus Whites, and here,
New York does very well. Hispanics versus Whites: here we see that New York
State again has some disparities issues. For differences based on income, in
this particular case it's the income of the neighborhood in which patients
live, we see that it's the biggest issue. There are very large differences
between hospitalization rates in high income versus low income communities.
In general, for most of these the rate exceeds the United States difference.
Another aspect I was going to show is this ranking table for 15 measures that
all of the States supply information for. This is the State rank and you see
a huge variation in the ranks—which is typical. What we typically see
is across all measures, including the ones that we rank, is that there is no
State that's best in everything. There is no State that's worst in everything.
Every State has some things they're really good in and some things they're
really bad in and this ranking table brings that to light.
The last part I'll mention about the State Snapshots is the last
link, which has contextual factors. This is meant to show information not about
quality of care but about some of the underlying issues that might drive quality
of care in a given State. Here we see New York has relatively high levels of
poverty, relatively high levels of Medicaid, and a slightly older population.
These might be things that affect the assessments of quality of care. While
we don't adjust for differences across States in their demographics, we try
to provide some information about the importance for interpreting the context
of their quality measurement.
The last thing I'm going to mention related to State resources is "other
information." This is where you can access technical assistance and methods,
guides for interpreting the results, as well as the mega tables. If you want
to download all of the measures and all of the results for all of the States,
go to this mega table and download everything. And with that, I am going to conclude
this portion and return this control over to the boss, who would be Margie. I'm
going to put up the question slide to open up for the first panel of questions.
Margie Shofer: Thanks, Ernie. Thanks for your presentation.
As Ernie said we're going to break for questions so let me quickly remind you
how to ask questions. As Zac mentioned, you can submit a question by typing
your inquiry into the Q & A box located on the right hand tool bar on your
screen beneath the participant list or ask a live question—asking live
questions is great—by clicking on the "Raise Hand" button
located at the bottom of the box containing the participants list. That will
place you in the queue of questions and Zac will introduce you.
Question: One of the purposes of the State Snapshots is
to compare States against each other, identify various strengths and weaknesses,
and communicate with other States of interest. Does the Snapshots tool
contain information on relevant contacts in each State?
Ernie Moy: At this time, we don't have that information included
in the State Snapshots. One of the areas that we're trying to develop
are linkages to other tools both within AHRQ and outside of AHRQ that would
be potentially useful. Right now we're developing a linkage to a report card
compendium, for instance, which is a number of local report cards which would
then include linkages to local organizations that are doing a quality assessment.
You saw in the areas of asthma and diabetes we have the linkage to innovations
where people are doing interesting things related to a particular condition
and those obviously would be resources. But we don't have a list of specific
contacts for each State to go to at this time.
Margie Shofer: I'll just add that that would be a wonderful
thing to do in theory, but technically it's probably pretty hard because the
folks responsible for the different measures in States that actually have programs
related to those measures are probably different people in each State. I'm
sure we could also try to be creative and certainly if anybody on the phone
is interested in being a contact for your State, that's great. We could always
collect a list and as people come in and ask us for further contacts, we could
share that. But we'll do a little bit more thinking on that too here at AHRQ.
Question: How is AHRQ assessing the quality of services and
care received by children, adolescents, and adults identified through newborn
screening as having a disease requiring long term care?
Ernie Moy: Well, the simple answer to that is that we're
not. As I'd mentioned we track a huge number of measures across many types
of settings and care, and that's just not one we've gotten to quite yet. But
if the questioner has a specific measure or way of capturing that population
we would be interested seeing if there is a way to include that population.
I'm just not familiar with a data source or measure that's applicable specifically
to that population.
Question: What software did you use to generate the dashboard?
Rosanna Coffey : All of what you see on the dashboard or
the Web site is custom work that Thompson Reuters has done for AHRQ. We have
worked with one other State interested in using this kind of display information—the
Maine Quality Forum—and we built some different looking Web sites for
them, mainly because they were trying to produce a different type of information.
AHRQ is certainly willing to share any of the software that we developed for
this if you wanted to take it and apply it directly in your case. If you're
working at the State level and then trying to drill down to a hospital level,
you probably need something closer to what we did for the Maine Quality Forum.
Question: There were some measures that did not have a baseline
rate. What does that mean?
Ernie Moy: There are a couple of different reasons potentially
for that. For some of the measures, we add-on new measures all of the time
as they develop and for some measures there simply is only 1 year of data and
so there's no baseline rate as a consequence for that particular measure. In
addition there's some State variability in terms of the availability of data,
so if a particular State did not have information for our baseline year, then
again, that line would not be present.
Question (Libby Kessler): Hi, thank you. We're just wondering
if you have county level data available as well.
Ernie Moy: We are very interested in county level data as
a next step. Unfortunately, when we go from the national data we collect for
the reports to the State level, we lose a bunch of measures, and as we go from
the State level down to the county level, we lose even more measures; in fact
we get attenuated very dramatically. While we do have county level information
for a small sub-set of the measures that you see in the State Snapshots,
it gets to be a pretty small number of measures. We are thinking about doing
something allowing further drill down at that level, but data availability
is a severe problem at this point.
Moderator: And we have
another live Q & A request from Nancy Solis. Nancy, your line is open.
Please proceed.
Question (Nancy Solis): Yes, we have a question about data
sources. While this is the 2008 Snapshots, we're wondering if there
is some data lag depending on the data source. Where would you find information
about what years are being compared?
Ernie Moy: In each of the tables, if you go to underlying
tables it should tell you what year is the most current data that we're reporting
on as well as what year is the baseline data, so that's where you can get that
information. There's a tremendous data lag in terms of what we're showing and
the information for the 2008 State Snapshots is information that we
received roughly a year to a year and a half ago. It just takes that long to
process it all as well as to get all of the clearances that we need to go through.
Before it got to us, of course it had to go to the individual data stewards
who are actually collecting the data, so there's a long trail before it gets
to us. At the State level you probably, in fact I'm sure, have access to more
recent data than we do. The flip side of that is that typically, most of these
quality problems and most of these disparities problems don't disappear overnight.
I think that for a lot of the things that we're seeing—especially those
that apply to multiple measures—the same basic pattern of a problem probably
still persists. Most of these things do not change dramatically from year to
year, especially in relative terms.
Moderator: And we have a live Q & A request from Maria
Rossi. Your line is open. Please proceed.
Question (Maria Rossi): Hi. You have specific information
on asthma and diabetes. I was wondering if in the future you'll have specific
data information for Alzheimer's, obesity, and other diseases.
Ernie Moy: That is one of the developmental pathways that
we're interested in and we haven't chosen a next focus on condition, so if
people have suggestions for that, that would be something to send into us.
Basically, there are two lines of development that we're considering. One is
focusing on an additional condition and the other is trying to allow drilling
down to a finer geographic detail. That's something I would be interested in
obtaining the listeners' feedback on. Would you look for finer geographic
capability or would you look for more conditions to focus on in addition to
the diabetes and asthma? Those are the two major developmental pathways that
we have open to us.
Question: Are there performance measures for or related to
behavioral health issues?
Ernie Moy: The answer is we do have some measures, although
we really have to look hard for them at the national level, and when we drop
down to the State level, unfortunately, we have very limited information available
to us. I think we can track some mortality rates like suicide mortality, but
that's not particularly satisfying for an assessment of behavioral health.
It's primarily a data problem and as people know about good data that can provide
measures of quality of behavioral health, we would love to hear about it at
the State level. But we find the data in that particular area to be very limiting.
Question: Is there a list of the assumptions made or actual
methodologies or formulas used for the Snapshots? I'm thinking this
might be a Rosanna question.
Rosanna Coffey: Yes, there definitely is. It's on the Web
site, and I don't know whether you can go back, Ernie, but it is at the very
bottom of the left hand menu of the "other information" that he
pointed out. There is a section there on methods, very detailed methods, that
explains how the measures were selected for each type of care or clinical area,
and then how the estimates were developed.
It's a very simple methodology. It really takes all of the measures in one
area, looks at each one, determines whether or not a particular State is above,
at, or below average based on that standard statistical test. For each measure,
where a State is above average, you get a point, and for each measure that
you're at average, you get a half a point, and for each measure that you're
below average, you get 0. Those are summed up across the measures in a particular
group and the scores come out to be between 0 and 100 and we translate that
into 180 degrees for that fan-type meter or just the 100 degrees for the horizontal
meter. It's divided into quintiles, then that score and the "very weak" is
the lower 20 percent, the "very strong" is the upper 20 percent
and so on.
Question: Where is baseline information from? Are all States
using the same data or same year?
Ernie Moy: Yes. That should be typically true. There are
maybe some subtle variations in that. For instance, all States don't necessarily
do BRFSS [Behavioral Risk Factor Surveillance System] every year, so I think
there are times when it's not exactly the same year, but typically they would
be comparing the same baseline year.
Question: What is the data source for each State's
information used in the reporting?
Ernie Moy: We use many different data sources, but if you
look at the tabular information as opposed to the actual meters I think you
should be able to identify what the specific data source was. It should be
listed, but we use many different data sources.
Question: Can someone just compare some States only, for
example run comparisons on two States only?
Ernie Moy: That's difficult to do inside the tool itself,
let's put it that way. There's not a built-in capability to compare self-identified
States. It's something that we've thought about doing. Of course the other
option is simply to open two browser windows, which is something that I sometimes
do to compare across two States.
Question: I see information for Medicaid, commercial, and
State employees. Do you have Medicare data or are you planning to get Medicare
data?
Ernie Moy: Rosanna, could you help me with that? That's specific
to the asthma calculator.
Rosanna Coffey: That's correct. It was in the asthma calculator
where we developed those particular populations. All of that information came
out of the Thomson Reuters Market Scan data, where we have information for
privately insured and for Medicaid, and it would be another developmental effort
to put the Medicare population in there. I suspect we would want to look at
the actual studies and how different the results were for just the Medicare
population. We did look at some 52 studies in coming up with that calculator
methodology, and in those 52 studies, we did look at the estimates differently
by Medicaid, privately insured, and commercial—excuse me—and State
employees. It would be an effort to go back and do it. It might be possible.
Ernie Moy: And isn't it right, Rosanna, that when we started
to do the calculators we tried to put the focus on what States—as opposed
to the Federal level—would be most interested in, and thought they would
be most interested in State employees?
Rosanna Coffey: That's exactly right.
Proceed to Next Section
Return to Contents
540 Gaither Road Rockville, MD 20850 