[This Transcript is Unedited]
DEPARTMENT OF HEALTH AND HUMAN SERVICES
NATIONAL COMMITTEE ON VITAL AND HEALTH STATISTICS
SUBCOMMITTEE ON QUALITY HEARING
OCTOBER 19, 2010
National Center for Health Statistics
3311 Toledo Road, Auditorium A
Hyattsville, MD 20782
CASET Associates, Ltd.
Fairfax, Virginia 22030
(703) 266 8402
- Panel IV:
- Dr. Barry Bershow, Minnesota Community Measurement
- David A. Stumpf, United Health
- Dr. Arnold Milstein, Mercer
- Dr. Joachim Roski, Brookings Institute
- Roundtable – Deliberating the Issues
- Introduction of the DHHS Office of Healthcare Quality – RADM Steven
- Environmental Scan – Ms. Kristine Martin-Anderson
- Introduction of the DHHS Office of Healthcare Quality – RADM Steven
- NCVHS Discussion
P R O C E E D I N G S
DR. MIDDLETON: Good morning. My name is Blackford Middleton. This is the
National Committee on Vital and Health Statistics Subcommittee on Quality
Hearing, second day, October 19, and we have Panel IV speakers ready and
willing to go but before we start off please let’s just go around the table and
around the phone and learn who is with us. Please announce yourself and
describe your role and whether or not you have any conflicts.
I am Blackford Middleton. I am Co-Chair of the Quality Committee and a
member of the full Committee and I have no conflicts.
DR. TANG: Paul Tang, Palo Alto Medical Foundation, member of the full
Committee and Co-Chair of the Quality Subcommittee and no conflicts.
(Introductions around room)
DR. MIDDLETON: And do we have any attendees on the phone this morning?
DR. MILSTEIN: Yes, Arne Milstein, Medical Director, Pacific Business Group
on Health and Professor of Medicine, Stanford University. I am one of the
DR. MIDDLETON: Good morning, Arne. Welcome.
DR. ROSKI: This is Joachim Roski from the Brookings Institution, no
DR. MIDDLETON: Good morning. Welcome. If there are no others, well actually
one more just in time for introduction.
DR. GREEN: Larry Green, member of the Committee, member of the Subcommittee,
DR. MIDDLETON: Why don’t we go ahead and jump in. Yesterday we had three
terrific panel presentations addressing various dimensions of the roadmap for
quality measures over the next three to five year timeframe looking at it from
the consumer perspective, the provider perspective, professional organizations
and accrediting organizations addressing the question of what information and
measures do those various stakeholders need to measure quality and value in the
years to come.
This morning we will be asking payers and group purchasers and other thought
leaders to address the question what information do payers and group purchasers
need to measure the value of health care and use it for decisions about
coverage. A terrific panel has been lined up with David Stumpf, Barry Bershow,
Arne Milstein and Dr. Roski and Dr. Bershow, you are leading off. Thank you.
DR. BERSHOW: Very good, thank you for having me here. When I mentioned I
have no conflicts that actually makes my wife sad that I do not have all of
these extra sources of revenue. I am as you will see on the slide, Chair of the
Board Minnesota Community Measurement but that is a not for profit and that
position is unpaid.
Lake Wobegon, we are going to have a little bit of apparently MAC/PC
dysfunction. There was a picture of Lake Wobegon on my submitted slides. But in
Minnesota those who are MPR devotees know it is the town in central Minnesota
where all of the women are strong; all of the men are good looking; and all of
the children are above average. In health care we have always had the same sort
of thinking where all of our female patients are sicker than average; all our
male patients are noncompliant and despite that myself and all of my colleagues
deliver above average care. It was not until Minnesota Community Measurement
came along that we were able to pretty much bust that myth and show people how
things were really happening.
This is screen shot from our current website with the breakdown of just some
of the things you can look up: depression; diabetes; vascular disease. We have
quite a few others but across the top notice that you can also find out about
medical groups as a whole. For example, our large integrated system at Fairview
where we own 40 clinics but you can also look and drill down and get
information on each and every one of our individual clinics which is of course
much more useful to the patients and a little bit more on that later.
We are a cooperative, collaborative body that includes employers, state
government, provider groups, hospitals, health plans, consumers, QI
organizations, all working very collaboratively. I put this partial list on
just to give people a flavor of the diversity on the board and also on our
reporting committee. The name you cannot see up there probably on your slides
is from 3M. We have representatives from the Mayo Clinic, Public Health
Consultants, individual health plan representatives, union representatives;
Terry Cahill is a doc in a small rural practice in Minnesota, the VP of HR at
the U of M, a large university, and the head of our Medicare QIO from Stratis
Health. We have the Director of Insurance from the state handling both Medicaid
and state employees and then representatives from the MMA, the Minnesota
Hospital Association and the Minnesota Council of Health Plans. Again, that is
just a partial list. You probably wonder how we get anything done and we
recently had a retreat and the facilitator wondered how we were going to get
anything done either with this size board but we impressed her quite a bit.
Because of the diversity and the collaborative nature of our organization
that same data can be used for quality improvement, for peer group rankings,
for pay for performance and this is now both external and internal so in our
own organization we are already lining the four sites with a major shift in our
compensation redesign where we are no longer paying people for volume but
instead for quality. This year 40 percent of their pay will be based on quality
rankings that tie back to what Minnesota Community Measurement is reporting and
what we are doing internally.
DR. TANG: When you say they, who is the they? Who are you paying that is
going to be?
DR. BERSHOW: I am sorry; Fairview Health Services is paying their employee
physicians, which is in the four to five hundred range. We seem to be growing
every time I look back at them. So it has reduced data collection burden for
the groups doing the reporting and I will talk about that in a second whether
it is the health plans coming in and doing the traditional HEDIS audit types
for some of the measures or the direct data submission coming out of the group.
We also avoid a lot of confusion for consumers. Before Community Measurement
came along, there was one year when Fairview was ranked highest in the state by
one health plan in vascular disease and lowest in the state in vascular disease
by another health plan. It drives us crazy and certainly was not helpful to
anyone trying to make a decision about where to go for medical care.
The direct data submission is the process where we are mining our own data.
Most of us in Minnesota do have electronic medical records but we have a number
of clinics who are on paper and who audit a sample of their charts and report
and actually do quite well. So it does not have to be just an EMR. But the
data, because of that self reporting, is granular down to the clinic level
instead of having to be grouped just by the medical system or the tax
identification number and it is clinical rather than administrative which I
think everybody here agrees is far superior. I will I think an interesting
anecdote on that in just a couple of moments as well.
So what does it look like when we do this? Well it looks pretty ugly
actually despite the fact that we are pretty sure Minnesota is among the best
in the country in quality. When you look at clinic to clinic variation, it is
really quite extreme. Here you will see the x-axis, the second column in. That
is just the diabetes blood pressure test. That was just whether that doctor
took a blood pressure on the diabetic when they came into the office on their
last visit. So this is not being able to keep them in control. It is not an
issue being able to forward your medications or compliance in taking them. It
is just did you even do what was necessary absolutely to take care of a
diabetic. So you will see way over on the far right on the graph, the blood
pressure under 130 over 80. The worst clinic in the state is in single digits.
The best clinic in the state is somewhere around 60 percent but if you just
look at how often they took the blood pressure test and then the next x-axis
over did they even get an LDL test forgetting about whether it was in control.
It was interesting to me that twice as many people got an LDL test rather than
took a blood pressure when we get a much bigger bang for the back for
controlling blood pressure in diabetics than controlling their lipid level.
I stole this slide from one of the health plans that was presenting at a
conference I was at. I stole it because I noticed coincidentally my name was at
the top as one of these reports. So what happens if we do not report ourselves
by grabbing a hold of the reigns and working in a collaborative nature? Other
people report on us. These are just some of the reports that are available. Now
if everyone will be honest and not look at their handouts for a second and just
stare at the screen, this I gave a talk in Cincinnati and the night before the
talk I plugged in some of the docs I knew were going to be at the conference to
see if I could get some ratings and I found this new pharma-stat site. I had
never heard of that before so I put my own name in to see what was happening
and you will see the rating categories of knowledge, helpfulness, punctuality
and staff. Does anyone who has not read ahead know what I have covered up as
the fifth rating category? Cost? No. It is hot or attractiveness. This is the
honest to God website and this is a screen shot right from it. I did not make
this up. I am sad to report I have no stars across any of those but I like to
think that no one had rated me yet rather than no one thought I was hot.
DR. MIDDLETON: Just send a link to your friends.
DR. BERSHOW: So the question posed to our panel is what information do
payers and group purchasers need to measure the value of health care and use it
for decisions about coverage. Well people always argue about what value is. A
lot of people think it is quality plus patient experience over cost. Whatever
it is, it certainly has to include all three elements of the triple aim. So we
have to have info on the quality of care, the cost of care and the patient
experience in receiving that care. One thing we do not do well in Minnesota but
what I encourage people to look into is getting data across the entire spectrum
so that patients and employers can look at health plans and their service and
their quality because that is actually the first big thing that people do often
upstream of when they need to access and so it would be useful. It is too late
once you are in the middle of the year and you discover you signed up for the
wrong health plan and it cannot be about structure or process. It really needs
to be about outcomes.
Long term outcomes? We are going to need more time to see are the things we
are doing really saving the heart attacks and strokes and other chronic
illnesses but at least we can use proxy outcomes now. Blood pressure is
according to protocols and LDL is according to protocols. Admin data is
In Minnesota we have something called the diamond project. It is a
depression treatment plan. The health plans pay the clinics a PMPM to take care
of the depressed patients with a psychiatrist on hire, available for consult
keeping the patient right in the primary care office and having a care
coordinator to reach out to the patients frequently. There is some aggressive
medication titration protocols. I would guess 90 percent of the patients
enrolled in this program are on an anti-depressant. At ICSI, the Institute for
Clinical System Improvement, we were auditing what was happening with this
project and so I have screen shots from what three health plans sent us. This
is health plan number one. The important number is in the lower right hand
corner. The percentage of patients with anti-depressant scrip’s: 89 percent at
one of the three biggest health plans in the state; 70 percent at another, so
probably statistically significant or at least within the believable range; and
the third system, 19 percent just really not credible. If you give this kind of
number to physicians and ask them to act on it let alone peer group them or
base their pay or their quality rankings on numbers like this, it is no wonder
they have not bought in and are fighting back.
This slide actually made up because the question was what information do
payers and group purchasers need and I am a physician and I was doing a bike
ride with somebody from Bridges to Excellence Minnesota who represents large
employers and she said, why did they as you to tell what payers and group
purchasers need. Why didn’t they as us? So then she thought about it for a
second and she said well at least they know you are reading from the same hymn
book that we are. What she said and so I included it is that for large
employers with coverage needed in multiple states, or for patients interested
in domestic medical tourism national measures would be best but we do not have
to have national measures if we have to settle for the lowest common
denominator in some regions. So one of the health plans in Minnesota recently
came out with a rating system where the end could be as low as five on your
quality and they said this what the national employers are asking for and what
they are getting in other states so that is what we have to do. The lowest
common denominator to me is really painful and I think it is work avoidance to
say that it is too hard to collect the data or unfair for those with harder
patients is the other thing that you will often see lowest common denominator
being set at for reporting.
So in Minnesota, this was in 2006 but the variation continues, this is a
composite diabetes score; highest clinic in the state this year was Oxborough
at about 36 percent. They did not make the x-axis too long on this slide. There
were a lot of zeros out at the end. Not one patient in their entire clinic hit
on all five diabetes measures on the composite. So knowing how hard that is,
this slide is for those who say you cannot get the poor or ethnic minorities or
others to do well. We have got to set lower thresholds for those clinics that
take care of them. If you are on the slide, it is if you are on the handout you
cannot see the pink but is the one that in 2006 was at 605. So if you look over
in 2005, they were at 0.54 percent so one of 200 of their patients hit on all
five diabetes measures. Sadly they were not even the worst in the state. As I
just showed you there were some zeros. But if you follow them along just two
years later, they were over eighteen percent. They were actually the second
best system in the state and this is a network of community based clinics where
half of their patients are on Medicaid, half are uninsured, most of them do not
have English as a first language. So it is really possible once you set your
sights and really get down to the work and avoid work avoidance to do a good
job with these minorities.
The other reason you will see lowest common denominator is associations
represent all of their members and sometimes want to protect the lower
performing members so they will continue to pay dues. So we have a saying here
in Minnesota, aim high and you will wind up in high places. That is true both
for the actual performance on the measures of the clinics so when we set our
internal pay for performance at Fairview, we set some pretty aggressive
thresholds but also here I am talking about just making sure the measures that
we are publishing are of high quality because otherwise if you aim low you are
going to be disappointed even if you get what you set out to accomplish.
I would like for people to stop saying we can’t do that because we have been
doing it in Minnesota since 2004 and our task on this Committee as I understand
is what do we need to do over the medium term: three to five years. That is
actually just about perfect for replicating what we did in Minnesota. That is
about how long it took us to get it right and to get buy in from providers. You
have to set up the infrastructure. You have to do a year of reporting that is
blinded so everyone can kind of see what their numbers look like and set about
trying to improve them. Then you have to tweak them and do a lot of cultural
selling. But within five years no one sneers at community measurement anymore
and they would much rather be reported there than on Angie’s List.
So what information do we need? It needs to be at the clinic level for
primary care because we believe chronic disease is a team based sport and you
need to keep your N’s high but probably doctor specific for the procedural
specialties. If the N is not high enough for them, I think that in itself, is
Waste and appropriateness are important as well so not just reporting on
what we are doing and how well we are doing it but whether we are even doing it
when we shouldn’t be. No use touting a surgeon’s success in stenting if the
procedure is not called for in the first place. I think most people in the room
know evidence pretty shoddy to none existent that stenting will prevent heart
attacks that loses its advantages over medical therapy quite quickly for
angina. So in Minnesota we are getting ready to put out our first report on the
frequency of high tech diagnostic imaging studies which are appropriate based
on ACR and ECC guidelines. And then it has got to be timely so useful to both
consumers for having the latest information. It is no use having two year old
information which is a lot of what is out there and also for providers to
understand what is possible to achieve so that they know what they are peers
are doing currently which brings me to Barry’s dolphin rule.
I am always flabbergasted when I speak to groups and find no one in the
audience knows how you teach a dolphin how to jump through a flaming hoop. Very
quickly you throw the hoop on the bottom on the pool. When he swims by it
accidentally you throw him a herring. Then you make the hoop vertical. Then he
accidentally has to swim through it. You throw him a herring; half out of the
water, same thing. All of the way out of the water he learns to jump through to
get his herring. You light it on fire and you are there. Well how effective do
you think it would if you threw him a herring twelve months after he first swam
through the hoop and really of course he would never, never do it. When we are
working on old data and when we are not getting rapid feedback it is very
difficult to improve.
So that last point on the bottom of this slide with all of the others is it
has got to be credible. Thinking back to that admin data story I told, it has
got to be transparent. The providers have to have faith in how it is collected.
It has to be evidence based and it has to be nimble in its ability to keep up
with advancing science. So we want to not have the state legislature or
Congress pass rules that saying it has got to be this because then when the
guidelines change it is too hard to move. So in Minnesota we have been able to
adjust as the ACORD study and later the advance study came out and some others
so it has been evolving.
So thank you very much. I am not sure if we have time for questions now or
DR. MIDDLETON: What we would like to do is take a little time now for
questions and there probably will be time at the end as well. Paul?
DR. TANG: Thank you, Barry. Appreciate it very much. I think we share a lot
of the same concerns about administrative data and the benefits, almost the
quote free benefits of credible clinical quality measures because everybody
wants to improve when they believe in the measure.
One of the things you also mentioned was the, you talked about the
associations having certain biases in terms of some of the measures they bring
forward to get approved or endorsed. The dichotomy, we ask this question, or I
asked this question yesterday. I think you were here and I want to give you an
opportunity to answer it: the dichotomy between measures that we use for
internal quality improvement and those that are separately designed for public
reporting. Sometimes you just in what you said, in the publicly reported often
times it is designed to be a race to the bottom. In other words you want to
protect everybody and the consumer, even probably yesterday let off saying you
know consumers do not have any value and everybody is 90 percent. It just
doesn’t help anybody. On the other hand in the quality improvement activities
normally people just rise to the top. Is there a difference in the measure?
Well, there are differences in the way measures are constructed and defined.
What is your opinion in terms of should there be differences and what measures
are you used for one purpose or another?
DR. BERSHOW: Well it is more I think the speed that you can get the
information back to yourself and then act on it. We use in my organization the
same measures that are reported publicly so that we have some peer group
references. When we first put our pay for performance in place at Fairview and
at that time it was much tinier. It was ten percent potential bonus on top of
an RVU based system which has now gone away. We only said get to ten percent
because we were in the low single digits and we thought ten percent seemed like
a reasonable hurdle given where the state was. We would have thought we were
asking for their first born child. We can’t do it and we can’t make our
patients come in. We can’t make them take the medications but in the end when
they stopped whining and we didn’t back down, they actually finished that year
at fourteen percent. So it probably would not have been reassuring to patients
to see that number but we did not care in a way. We went out and published our
bad results on the Fairview internet page: not internally, even before
Minnesota Community Measurement was publishing because we thought from an
ethical standpoint as health care providers we always want to know who is the
best doc, who do I go for an orthopedic procedure. Who is the best
cardiologist? Why wouldn’t we want our patients to also have that privilege?
Why should we have a special privilege that they don’t have? So that was our
The next year we, after they finished at 14 and a half, we said okay, we
will get to fifteen. More whining, more whining. We can’t do it. We can’t do
it. They finished that year at twenty-two. So when you actually get out and
publicly report what you are doing I think it helps boost that quality
improvement because otherwise people are more slow to take it up and we are
going to take the fairly brave step we think this January of posting quality
cost of care and patient satisfaction scores in all our reception rooms so that
patients can look at them while they are waiting for the doc. The docs are
starting to come up with some talking points on why the numbers are not higher,
but we think it is the right thing to do.
DR. TANG: So you commented on the use of quality measure reports in changing
the culture but what about the definition of the measure itself. So for
example, the quote a publicly reported diabetes measure is the percent of
diabetics whose A1C is greater than nine percent. That is not very helpful to
someone trying to improve score. It does not compare to the guidelines.
DR. BERSHOW: Right.
DR. TANG: So how do you use what I thought I heard you say is you use the
publicly reported measures inside and then made it public. Do you find the
public reported which are largely based on administrative?
DR. BERSHOW: Maybe I was not clear enough on what Community Measurement
does, and I hope this doesn’t sound too un-humble but we kind of scoffed at
some of the national measures and so we never chose the HEDIS limits to report
on. So out of the gate we used an A1C of less than seven as passing, not on
less than nine and with recent changes in the literature we modified that to
less than eight.
DR. MIDDLETON: Loved the presentation and very thankful for your coming and
sharing it with us and your leadership obviously in this area.
One of the challenges I think I would be interested in hearing a little more
color commentary on is the direct data submission process. Whether it is from
an health IT source or from a pay for base source, could you talk a little more
about that and then the return trip. I gather these are publicly reported but
what other advisories come to individual docs who are trying to make quality
improvement efforts and then the last part of this is really how do you get the
doctor assignment or attribution right? Is this my patient or not is always a
common refrain immediately after my patients are sicker than everyone else’s.
So basically how does the data flow? How do the reports come inbound and how do
you make sure you have the right patients for the right doc?
DR. BERSHOW: I will try to remember all of them. Feel free to prompt me
again. With the direct data submission, if it is coming out of the EMR and our
system is Epic so I know more about that, we can create a report in the
application called Clarity and it is exported and there is a portal site at
Minnesota Community Measurement that the data is directly exported to. Then
Community Measurement does a spot audit. The first year they audited everyone
but then once they were confident the code was right they would do just like
eight charts and if all eight were exactly what they expected by from an
eyeball chart audit then they assume the rest of it, the data was good.
I know in Cincinnati that Minnesota Community Measurement is serving as an
aid in getting their system up and running. They have some groups having their
paper charts scanned and then natural language processing extracting some of
the data and then getting it into the portal. I cannot speak to how well that
it is working. They have been up one year on that system but even if you do not
have an EMR that is an option. For those clinics in Minnesota that do not have
an EMR, we let them do thirty charts hand audit randomly selected and again
those charts are re-examined.
I am sorry the second piece was?
DR. MIDDLETON: The return trip, the quality report to an individual doc.
DR. BERSHOW: Because it is our data going in, we know what it is even before
it is published and we also have a lot of confidence in it and we can give them
in terms of the attribution. We got a lot of whining at first, well this is not
my patient or the data isn’t right but when we put out the reports internally
we give it by all of Fairview, by the geographic area, by the clinic, by the
doctor and then the list of the patients that go into that doctor underneath
it. So they can go in and they can look at each and every patient. After they
saw the reports were pretty good they stopped complaining, but the attribution
problem has been taken care of from us. We have some denominator definitions
that Community Measurement puts out so for example some conditions they have to
be seen twice in the last eighteen months and once in the calendar year being
measured. They vary depending on what the initiative is.
Internally at Fairview we have taken a little tougher stand. If you see them
once, they are yours. Go get them. If their numbers are not good it is your
responsibility for tracking them down, getting the records from another doctor,
outreaching to them. So that is the stand that we have taken.
DR. MIDDLESTON: One quick follow on, so a lot of quality reports HEDIS as
well you know have data which is not really or include data elements for the
reports. They have to be from the clinician or from the provider; diabetic foot
exam. That is not a laboratory data that can easily pick from a database and
send wherever you want to send it. So those data elements that are really from
the clinician, how are they getting into the reports? Is there a special
structure data entry in the record obviously or elsewhere? Is it the physician
or is it some other allied health professional, etcetera?
DR. BERSHOW: It can be anyone on the team. Again we are moving towards team
based care so everyone, the room medical assistant or nurse, the physician, the
diabetic educator, the pharmacist, whoever is touching that patient is
responsible. If the data is not in there when it comes to them, get it in
there. So blood pressures are all collected in discrete data fields not
dictated into the note. If we did a clunky work around for foot exams and eye
exams where the docs actually place a non-chargeable order for that which then
can be tracked, Community Measurement just for clarity does not include that as
one of pieces of its composite. We felt the data was not 100 percent reliable
as to the value of the doing that and documenting it. But anytime there is not
an easily track-able data field already built into the EMR we just build a fake
order for it and the team puts it in. Then if they fail because again they are
getting the reports every month, it doesn’t take too long before they see I am
not doing that well and I am looking bad compared to my colleagues and because
we pay them money to do well they pay a lot of attention to it.
DR. MIDDLETON: Barry that was beautiful. I have two questions: one about
appropriateness and one about one of your slides. What is your opinion about
the current state of the art in the measures for appropriateness of care and
particularly do you know what do you think about appropriateness of care when
the problem is not a evaluating a disease but a complaint or a symptom?
DR. BERSHOW: I think that is an area that we are immature on certainly and
it is one of the reasons it is our last report out of the box at Community
Measurement. It was not until there was a state-wide project going on to use a
common vendor for content and to be able to look at our high tech diagnostic
imaging appropriateness as rated by American College of Radiology, American
College of Cardiology that we had something we thought we could really point to
and say if not everyone agrees at least most people agree if you get a scan for
this indication and it has almost no utility at all, it is not likely to give
you some information and worse yet you are probably going to need to do another
scan that you just should of ordered in the first place. We felt comfortable
that that could be something we concentrate on first.
We are also looking at trying to delve into reporting unnecessary frequency
of colonoscopies. The gastroenterologist and colorectal surgeons in the state
have said we probably have the capacity to do a colonoscopy on everyone who
needs one if we weren’t actually redoing people more often than the guidelines
suggest. So we are looking at people who didn’t need one the next year or three
to five years later when they could have had one seven to ten years later. That
may be our next appropriateness measure. But to get to the core of your
question I think the science here is still more confusing than in other areas
on what is waste in the system.
DR. GREEN: Thank you, a question about one of your slides. I was really
struck by that improvement slide what you referred to the community health
centers. I am wondering what your opinion is about what explained the
improvement in particularly what type of pay for performance system was the
community health centers in and did it have anything to do with it?
DR. BERSHOW: Well as I understand they did not use any pay for performance
element but they felt a very firm commitment to have their patients get state
of the art care and so they set about an aggressive pattern and I think their
main technique was the use of care coordinator. They made sure that they had
people who could speak the patient’s language, who would reach out into the
community with phone calls if they had a phone, door knocking if they did not
trying to get them in, working with them, helping them navigate with a medical
care system that can be very confusing and that is how they did their major
DR. GREEN: So slip this back to one of the first things you told us about
the need for the same data to be used for QI pay for performance peer group
rankings. Are you sure we need to have the same data to do pay for performance?
DR. BERSHOW: Well I believe it was you yesterday that was making the case
for measure once, use three times. So I am assuming that you are just prodding
me a little bit with this smile on your face. But I think from a waste
standpoint in the medical system it cost money in human resources and cash to
build our high diagnostic imaging reporting system back into Community
Measurement was a 30 to 50 thousand dollar project for us. We are big system
that is around in here but it is too large for smaller groups. We do not want
to be collecting or reporting data unnecessarily that cannot be used for
multiple purposes unless we feel the one purpose it is used for is so
important, so worthwhile it can justify the cost of it. One man is on the
DR. TANG: I just have one quick question. You talked about going in for
primary care, the granularity going down to the clinic because you thought it
was team based and then other times you talked about going down to the
provider. So how many providers are in a clinic and how do you reconcile the
DR. BERSHOW: Inside our system we have clinics as small as two physicians
and we have clinics that are probably as large as twenty, twenty-two so it can
vary. While it is true that clinics specific is not going to help you if you
are an individual patient. You are looking for an individual doctor and if I
had another half hour I could tell the sad tale of variation within our system.
Jack Windberg’s geography as destiny comment, is even true to which hallway you
are roomed at one of our clinics, to an extent. But because we report
internally un-blinded, we report externally by clinic. No one wants to be in
the clinic that is not doing well so there has been a lot of provider peer
pressure wandering down the hall and kicking their partner in the chin they are
not doing well.
I can tell the story after 2005; our first year of pay for performance,
there was one clinic where every single doc in the clinic missed a $17,000
bonus because one doctor blew off the initiative. He offered to resign. They
turned him down but said we do not expect this ever to happen again, met next
year. They were a high performing clinic. So we think that over time that
internal peer pressure has really narrowed the variation from where we started
out to where it is now. Still not perfect but it is getting there.
DR. TANG: So did you say that you then do measure by provider internally?
DR. BERSHOW: Oh internally yes, yes I am sorry. We do our clinic rewards in
our pay for performance by clinic but we report internally to them on the
provider level so that they know how they are doing and they can see the
patients attributed to them and they can work that list so that they can see
who is not doing well and try and reach out to them if they are not coming back
DR. MIDDLETON: Mike, could you introduce yourself?
DR. FITZMAURICE: Michael Fitzmaurice, the Agency for Health Care Research
and Quality. I think I heard you say that if there is an non-chargeable eye
exam for example that it is not counted in the quality measure. So I am
wondering how do you communicate to the doctors exactly here is how we want the
data, here is how it is defined, here is how we want it represented and what
you report to us?
DR. BERSHOW: Are you talking about at the Minnesota Community Measurement
level or at the internal Fairview level?
DR. FITZMAURICE: Um, you have got me because I am not sure it was what you
were referring to.
DR. BERSHOW: Okay, that measure because it is not something that is reported
to Community Measurement there are no guidelines or standards for how that
comes in. We wanted to track it internally and so we put out some education so
we have got videos. We have got emails. We have got paper handouts. We have got
clinic administrators going around and showing them again you know you are not
doing well in this area.
We do not think it is because you are not ordering them as often as your
colleagues. You are just not getting the data here. This is where you have to
put it. So it has been a continuum. When I talk about three to five years, all
of this did not happen in the first six months after we went live so a lot of
provider education. Minnesota Community Measurement puts out a set of standards
at the beginning of each measurement year. This is what we are going to be
looking at. Here is the denominator definition. Here is the numerator
definition. Then they have consultants available to come out and help you if
you don’t understand but it is usually pretty clear.
DR. FITZMAURICE: And you get into the ICD-9 codes and the CPT codes?
DR. BERSHOW: Correct, correct.
DR. FITZMAURICE: Thank you.
DR. OVERHAGE: Blackford, this is Marc. I would like to raise my hand.
DR. MIDDLETON: Sure. Please who is on the phone?
DR. OVERHAGE: Okay, I just didn’t know if I was in the queue, I think. Marc
Overhage. A few things for the presentation I will say I learn a lot and I
think you have alluded to this. Could you give us a sense of the timeline for
change? You started down this road and talked about the yearlong sort of run in
and those kinds of things. I wonder if you could just go through from
inception, report delivery, clinician engagement, to the kind of changes you
were able to achieve, summarize the timeframe for those.
DR. BERSHOW: Okay, thanks. I am sorry I actually have slides on my computer
but none available for this presentation showing that. So Community Measurement
started their first report blinded. Each group only got to see their results
and the blinded high and low and average in the state. So that was year one
just so that they could learn how to do it. Year two, there was public
reporting at the system level and the state average at that time was I believe
six percent on the diabetes composite. That was the first report out of the
gate. This last year the state average was twenty percent I believe give or
take one percent and the higher performing groups, the highest group in the
state last year was at 60 percent. There were quite a few groups over 40
percent. I bet twenty to twenty-five clinic sites in the state over 40 percent
but it took that long and it was a gradual, continual improvement. There was
not the one year jump up. It went from like six percent to ten percent. Then I
think it even stalled in the low teens for a bit then into the high teens and
to right around twenty percent. So that is kind of the timeline for
The timeline for how this is done internally, a lot of education is needed
to the provider community early on. You have to get the buy in; lots of hand
holding and individual selling. People like myself and some other colleagues in
the state who felt really passionate about this going out to other systems,
talking to them, explaining as I did about the Angie’s List what is going to
happen if we don’t do this ourselves. If we do not have measures we think are
important, if you are seeing this does not represent quality, whose fault is
that? It is not Community Measurement or Angie’s List or the health plans or
HEDIS. It is our fault. We really have to grab hold of these reigns; define
what we think is quality; define how we think we should measure it. So I kept
making that pitch over and over and over again. I do not want to take full
credit. I was not the only one out there and I was actually at the knee of some
other people who were teaching me along the way. Just gradually we got everyone
to swing from I hate this. This is all wrong. It is terrible to this is really
pretty good. What does Community Measurement say about this, is the common
phrase I hear now in the state. It has really been remarkable.
DR. MIDDLETON: So why don’t we shift gears. Thank you very much, Barry,
really a terrific presentation and rich discussion. Thank you all and while we
are switching gears for David Stumpf, Marc would you mind introducing yourself
on the phone?
DR. OVERHAGE: Thanks Blackford. This is Marc Overhage for the Regenstrief,
Indiana Health Information Exchange, member of the full Committee and the
Subcommittee and I have no conflicts. Thank you.
DR. CARR: Blackford, this is Justine Carr, Caritas Christi Health Care,
Chair of the Committee, member of the Subcommittee, no conflicts.
DR. MIDDLETON: Good morning to you both. Anyone else on the line who has not
yet been introduced? Okey Dokey. Well, while David is just teeing up I thought
we might plan for the 10:45 session. Dr. Steven Solomon will be joining and
discussing perspectives from the DHSS Office of Health Care Quality and if time
allows in that block, I would ask if the Booz Allen folks might present the
environmental scan perhaps the second half of this session. That would be
terrific and David, if you are ready please take away.
MR. STUMPF: Thank you Blackford, and thank you to the Committee for this
invitation to speak. I might actually to get a little continuity pick up on the
last point that was made by Barry that the way that this quality measuring
needs to be done is by the providers in those environments at the points of
care and so on. We as payers of course have been in this field because of
pressure we have received from the purchasers, the employers. But we have said
all along that we would like the providers to do this themselves. So what I am
going to focus on is sort of the ideal future state if you will and some of the
steps we are taking to get there.
I am here on behalf of the United Health Group which is a Fortune 25 health
and wellbeing company and it has two major arms. One of those is the payer
segment which is actually merging. Now these two buckets, the individual
employers and our federal and senior programs are actually moving together for
some of the reasons that you mentioned. They are trying to create
accountabilities across all of these because doctors do not make this division.
They take care of all of these groups. Then we have what we call our services
group, the enterprise services group which is where I spend most of my time in
eugenics and Optum Health, and together we now serve actually over 80 million
Americans with the services that we provide through these channels.
My role in the company is strategy. I do not have direct operational
responsibilities. I spend a lot of time outside of the company with quality
organizations as well as inside the organization connecting dots because we are
a very big company with lots of silos and a lot of value from my participation
comes from connecting those dots. I try to assess the environment, identify the
gaps and advise our executives so they can create their own strategies. I do
not tell them what to do but I do think I can influence some of that. Of course
one of the big issues is change and how one manages change and creates more
diversity of how we actually do business is one of the biggest challenges I
think we face.
This presentation was designed around the charge that I was given: to look
three to five years out and try to identify the paths that we need to go down
and what are some of the gaps that we need to know and address and what kind of
information should be useful. So I have distilled that down into how do we
presently link value to benefits. What is that ideal future state and where are
the gaps, challenges and opportunities?
Our current state, I think most of you are aware is fee for service. It is
fees negotiated with providers and sold to employers. This system has many,
many problems. We did not invent it by the way. We often get blamed for the
fact that it doesn’t work but we did not invent it. We implemented what
providers told us they wanted to do.
We have a lot of projects in the mill right now to try to help redesign the
system. We have over a dozen medical home pilots going in various markets that
are based on NCQA requirements and each of them is a little bit different,
testing various hypothesis and models. They continue to have some fee for
service. They have a flat rate which I emphasize flat, per member per month
payment. They incentivize shared savings in some programs and they also have
incentives for population outcomes, not just individual ones.
One of the companies we just set up as a start up is Liprint down in
Arizona. This is a laboratory if you will. They are going to be setting up a
chronic disease clinic and creating accountabilities for the providers in those
clinics and it really becomes a laboratory where we can investigate what works
in accountable care organizations. Some of the models I will get to in a minute
will be a good test bed for those. They obviously will be looking at again
these various payment mechanisms but also some business processes that will be
different than what we now have.
We like the Fairview model that you just heard about. We have some large
provider organizations where they actually set up their own quality reporting
and incentive programs for their doctors and we create for them a quality pool
so we don’t just pay all of our money for fee for service to those providers
but we give them a lump sum at the end of the year and they distribute it
according to their internal quality metric thing. So this is a step in that
direction I talked about from us stepping away from the measurement and reward
system and pushing that out to the point of care.
For those who do not have that capability we still have some systems that
will push our quality analytics out to the market place; gap reports that we
can identify in our data which is a good bit administrative but we also have
laboratory data. We also have prescription data and we, by the way, know when
we do not have the data so we do not report on gaps when we do not have the
data. We try to do this in a way that connects to some of the meaningful use
targets. I want to talk for a minute too about the targets of opportunity that
we are addressing because we are trying to align those with national drivers
which I will get to a little bit later.
So the ideal future state I think for all of us and us included is a work in
progress. I am actually leading a workgroup in eugenics which is looking at the
ideal future state. Our timelines are meaningful use phase two and phase three;
where do we want to be at those points in time. But I can give you a few
illustrative examples about how this work is in progress. We do collaborate
with health information exchanges and we started off like most people I think
just jumping in the water and found out that there were some problems and so
that led us to lay out a few principles about getting to the value, getting to
sustainability and those have been clearly annunciated I think by our company
and we provide data to HIEs as well as some of our analytics under certain
circumstances and some of our companies help HIEs actually move data around.
We have an emerging business group which is actually quite new in the
company. It is only a few months old and they are still actually defining
themselves but this is a group that is going to have access to capital to
invest in innovation. They are still refining how they are going to do that but
it basically gets into we build it ourselves. We buy it or we collaborate and I
put the emphasis on the “collaborate” because I think you are going
to see a lot more necessity of collaborating in order to get where we want to.
This group will have some funds to set that up.
In eugenics we are really looking at point of care analytics. How can we
push the analytics out to the point of care? How can those analytics be done
when we do not own the data but we provide the tools that the point of care is
using for that. We also are looking at work flow enhancements which we have
already heard something about that. How can one merge work flows? How can one
repurpose data for accomplishing multiple purposes? Then we have a lot of
capital and we are out buying companies right now.
I think you are going to see a consolidation in this industry around HIT and
outcomes. So for instance we acquired Quality Metric which is a company that
does a lot of work with functional outcome measures that the SS surveys are all
there. We are incorporating those now into some of our analytics and some of
our business strategies. We acquired Axolotl which does a lot of work in the
HIE space and is a platform on which a lot of things can be delivered as well
as just moving data around. We have acquired two companies that are at the
point of care right now. One of those is Pisces which is a company that is in
emergency rooms, operating rooms, ICUs getting streaming data coming out of
those units and applying analytics for decision support.
The other interesting company we acquired is Executive Health Resources and
EHR is in emergency rooms applying analytics to raw data in emergency rooms:
chief complaints; symptoms; laboratory data and rendering decisions about where
that patient should end up. Do they go home? Do they go into observation? Do
they go into an inpatient or an ICU bed? This company was set up and designed
to maximize the efficiency of the administrative process for a hospital so that
if they place somebody in this unit, they are not going to get hassles about
their claim. It is going to run through smoothly. It also maximized their
reimbursement but when you look at that model, it actually has a lot of other
applications for how one can make those decisions based on quote raw data, you
know unrefined data because as we all know especially in those acute settings
the data is not refined and is probabilistic. Diagnosis are uncertain at that
We also have an HIT department if you will really a large component of our
company that is working on some innovative things like the use of multiple
devices, super computers that we are applying to some of our analytics and
working on our service oriented architecture strategies. Then we have within
Optum Health a bank and that bank is looking at some interesting financial
services which we think can be transformative in this industry as well.
I want to now kind of shift into what are these drivers of changes. We
looked at the national agenda for that. This is not something we create again,
and I kind of put them into three but major buckets here. What are the unmet
needs? We participated in and believe in the national priorities partnership
and the gaps that they identified there. They are part of our key strategy. I
am a particular fan of the National Quality Forums Coordination of Care model
for some reasons I will get into in just a second. This model has the
capability of addressing most of what ails health care in my opinion including
the data requirements that we have going forward. I would just refer you to
their website for the document.
Then one of the RT drivers in looking at the ideal future state is to
implement impactful technologies. So we were active in the HITSP process and
what we have decided as we develop things because we have tended to be product
oriented. We want to have this product and that product and that product. What
we are no focusing on are capabilities. This gets into really some service
oriented architecture strategy. You want to have the fundamental capabilities
and then link them together to create products and services. That is a
fundamental shift for us that I think will over time give us more agility and
flexibility in responding.
So let me run you through one of the things that I am working on right now
and I think our company is. The only revision I would make in this slide I
think right now is rather than patient centered coordination plan is to make
that person centered coordination plan because most of people would prefer not
to be patients but rather healthy individuals who have a limited need for the
health care system and in fact most health care occurs outside of what we would
traditionally think about is a health care system. But one of the big
challenges, how do we get structured data? How do orchestrate the process? And
this model actually I think provides a mechanism, a mechanism, not the only
one, for how this can be done.
This is structured document presumably in the way it is being envisioned and
developed, an HL7 Version 3, CDA adherent document which has sort of an
unlimited amount of nesting that can occur. It has two kind of key components.
One of those is the patient characteristics and that includes the typical
things you would find in a health summary but it can also contain other things
like functional status, beliefs, desires and intentions of that individual
which will drive their behavior in the marketplace and so on.
The second component are the tasks. These are obviously can be linked. This
diagnosis or problem or issue will be linked to this task and the patient
characteristics can be utilized on an individual patient level to create a
dataset, really to extract the dataset of similar individuals or a cohort on
which you can then look into your crystal ball which in the modern era we would
call a probabilistic outcomes and choices so that if we have this 53 year old
diabetic, hypertensive person we can look at our data already and find hundreds
of thousands of those people. We can then look at the characteristics of them
and where they are headed in the health care system and we were trying to now
work on how can we use that data to look at the paths, the various paths that
those individuals take in the system. Of those 10,000 diabetics, you know 1,000
of them have a stroke and 9,000 do not. What are the things that distinguish
them and what are the modifiable decision points that you are going to get at
in that population?
The other thing we can do with this is calculate the risk adjusted PMPMs
which provide an incentive for providers to generate those characteristics and
also provide the data which is necessary to create the task. The task reports
then have three key elements: the task itself; the accountable entity, who is
going to do this and that frequently will be the patient; and what is the
outcome metric, and those outcome metrics can take on all kinds of
characteristics. They can be EBM based, consensus based. They can be a business
process or they can be a functional outcome to name a few. All of this then can
be linked to the outcome that you anticipate which can also be rewarded in a
whole variety of ways. One of the fundamental things that is emerging from some
of the groups trying to implement this is the need for a business process
management and the ability to orchestrate this whole scenario.
Here are what I see as some of the high level gaps. The workforce, I think
what we are bumping up against is how do you translate all of this clinical
data into the kind of HIT analytics that you need and clinician informaticists
are absolutely essential for this. You cannot do this with people who are used
to administrative data and there are too few of them. They are too dispersed
and getting critical mass in this field I think is one of the real challenges.
The sociology is another very big issue and in fact I think the biggest
issue. I quoted the Intel Chairman you know that culture eats strategy and I
certainly find this inside my organization. Changing a culture is the hardest
thing. The incentives need to be robust, pervasive and very directional and I
think they need to apply to everybody.
We also focused on business process management and some of the issues we
talked about before and then the last point here is the capabilities need be
turned into applications that have several requirements. They have to have a
sustainable production environment where professionals develop these and I
could comment more about that but I do not have time. There is a real need to
stimulate this area and we also I believe need to get the plug and play so that
the capabilities that we develop will enable the implementation of what we
could call a platform as a service so that these services can be delivered
seamlessly to the points where they are needed.
So with that I will stop.
DR. MIDDLETON: Thank you, David and I want to apologize for having to
shorten your presentation a little bit. I miscalculated myself not recognizing
there were four speakers for our block this morning and I was timing for three.
So thank you for your acceleration on that.
Is there one burning question for David? Yes, Mike.
DR. FITZMAURICE: Do you use quality measures in the measuring value and
where do you get the data?
MR. STUMPF: We do have quality measures. We have some 700 EBM rules that we
run, can run. We use about 150 of those in our quality reporting for
physicians. The data comes from several sources. It comes from the
administrative claims. It comes from pharmacy data and it comes from laboratory
data. The emphasis right now is on getting more laboratory data. We get about
50 percent of our claims for labs are associated with results and those results
are increasingly important in the analytics. We also get it from some patient
input through our consumer portal. We also want to increase that and part of
the incentive for these care plans, coordination plans from our perspective is
the ability to integrate some of the clinical information in designing the
incentives. Again, the idea here is to push that out to the point of care so
the doctors and other providers are telling us what is the task I am trying to
accomplish here? What is the outcome I am expecting so that we can incentivize
DR. FITZMAURICE: Thank you.
DR. MIDDLETON: So David, thank you very much for a terrific presentation;
very rich and we may have questions after the fact based upon your testimony
and the slides. I know we can reach you via email. Thank you again.
Let’s now turn to two presenters from the phone: first Dr. Arne Milstein and
then Dr. Joachim Roski and we will endeavor to wrap up at 10:45, sharp. So
Arne, please take it away.
DR. MILSTEIN: Thank you Blackford, and out of respect for constraining time
I think I will skip over some of the introduction, other than to say that
purchasers, large employers, large labor unions, they are self insured. Large
state government is self insured for their employee benefit plans. I think most
of them would agree with the long term vision for what the U.S. and health
information systems needs to look like that was articulated in the NCVHS May 26
concept paper. We thought it was terrific. So what I have done this
presentation to kind of hone in on a shorter list of U.S. health information
system enhancement that would be especially useful in efforts by large
employers, large unions and their employees to improve the value of their
health care spending and broad brush the enhancements I am going to recommend
fall into three broad buckets.
The first is scope of data collection. That is what is getting collected.
The second is things needed to make the data more useful. And last but not
least is the issue of due speed because I think we have all been a part of that
for I think it probably was some of the visionaries among us not including me,
probably long ago could have written something very close to the May 26 concept
paper, that even in which you approached it has proved to be for the pi side of
the market meaning that large employers and the union workers who are competing
against more efficient health systems in other countries has been not what we
would hope it would have been. Let me just hone in on these suggested near term
priorities and then I would be glad to take your questions.
The first category is the aim to enhance the scope of what gets collected.
We think there are opportunities to improve that in four ways in the near term.
Our first recommended target would be the routine collection of standardized
electronic information on patient reported health status and experience of
care. These two components of the triple aim represent consumers’ essential
purpose in seeking health care. That is they want to get better and that they
want to be treated reasonably in the process. We do not think there is any
inherent reason why this information cannot be routinely and economically
collected at the inception close of most patient contacts of the health system
and at intervals thereafter. That is someone who is given healthy intervals
might be for collecting status, health stats for patient status would be
obviously, it would be more infrequent than for someone who was here for
We think electronic versions of the Dartmouth pull up(?) instruments
illustrate one feasible element of such a data collection system that could
apply to multiple but not all health care service types but it is really
illustrative. We are not recommending that as the actual standard.
The second opportunity we think near term opportunity that should be pursued
to improve the scope of data collection was either routine collection of
standardized information on whether any high risk or high cost, non-emergency
clinical service meets current appropriateness criteria and was subject to
robust share decision making process. While we are waiting for universal
adoption of sufficiently robust electronic health records to enable such
routine information collection, we think we should in the interim move ahead
with the establishment of clinical registries to serve this purpose. We think
registries should also be initiated to track outcomes and care problems for
health conditions generating substantial disability and health spending that
may be unrelated to specific high cost or high risk discrete health services.
A third recommendation on scope of data collection is the routine collection
of standardized information on the payable charge for every service or in a
bundled reimbursement arrangement of a service bundle. Right now, particularly
in California we are facing some behaviors by typically market dominate
providers that we think will go done in the annals of infamy, in the history of
American health care. Some dominant geographically dominate health systems that
have placed a gag rule on the sayers including self insurance lawyers
disclosing the amounts that they are able to collect due to their advantaged
market position for whatever unit of service they are being paid at whether it
is a capitated rate for professional services or all inclusive cap rate or fee
The fourth opportunity we think near term that we think should be
prioritized in terms of data collection is the routine collection of
standardized information on each patient health care relevant quality of live
and end of life preferences I see as such collection is authorized by a
patient. The reason for doing this we think is self evident given the
substantial evidence of non-concurrence or maybe it is a better way of putting
it is the inability of providers to intuit what patient health relevant values
or preferences are. We think the only way to do that is to routinely collect
such information and preferences on a regular basis particularly over the
course of a life evolution.
Again we finished up four of the eight specific recommendations for near
term priorities for health information system improvement. Let me move to three
recommendations that fall under the category of data usefulness. The first of
these would be the conversion of electronic health information into structured
data in standardized format. Many of the clinicians managing information
systems on an inpatient and ambulatory basis point to this is a tremendous
problem that is a lot of the useful and usable information are now being
collected electronically is in a non-structured format and therefore a real
challenge to make good use of.
The second recommendation in terms of short term priorities under the data
usefulness category is the linkage of all personal relevant health and health
care information on a real time basis so that it is available to inform the
health and health care decisions of patients as well as of their clinicians and
the in-home care givers when authorized by a patient. This is obviously been a
problem. Fortunately there has been some initial progress. For example, the
blue button is now available to veterans in the U.S. This gives them a version
1.0 axis to priority care information and now application developers are free
once that information becomes available to begin to create a more useful
personal involvement in health and health care information about each person.
Last in the data usefulness category is the following priority. We believe
that for patients and clinicians that request it, we should begin to aim for
the continuous generation of active decision props whenever there is a new
opportunity to take an action likely to improve health or health care
affordability for the patient. Our vision is that most people go about their
daily lives and do not want to occupy their mind with the question of gee I
wonder if there might be a way that given updates in my current health status
that I might improve my health or improve the affordability of health care to
me. Accordingly we think that really quite parallels with the evidence that it
is really active decision prompts for clinicians that make the biggest
difference in improving performance. We think the same system ought to be
pursued so that a patient and clinicians on a real time basis can have new
information about them accumulate that they are alerted anytime there is an
opportunity to take an action either by the clinician or the patient to improve
patient health status and or improve the affordability of health care for a
The last and eighth recommendation in terms of near term priorities that we
would make a big difference in the ability of private purchasers, private
sector purchasers and patients to play a more active role in encouraging
doctors and hospitals to excel on the value of care they deliver would fall
into to the category of a reduced fee and I think what we would like to see is
use of the Secretary’s authority, referring to the Secretary of HHS, to adjust
Medicare and Medicaid and SCHIP payment as needed to encourage most U.S.
providers to meet the above described seven priorities within a reasonably
prompt period of time. We think that most of the things on that list if they
were nationally prioritized could be in place within a period of four or five
So that is, I met your time limitations but the there are recommendations
for near term priorities fall into this category of scope of data collection
and data usefulness and in due speed and we make those recommendations while
simultaneously endorsing the much broader and more inclusive long term vision
articulated by the NCVHS in its May 26 concept paper.
DR. MIDDLETON: Thank you, Arne. That is really terrific: a big vision and a
really thoughtful across many dimensions of the problem. I am very thankful for
your testimony. I wondered, did you actually are you producing a written
testimony as well or even a power point?
DR. MILSTEIN: Yes, I forwarded that to staff last night.
DR. MIDDLETON: Okay, so we have it but I don’t have it. Maybe it is in my
stack. Okay, I was copying furiously, Arne as you were talking.
DR. MILSTEIN: I sent it to Matthew Quinn and Debbie Jackson.
DR. MIDDLETON: Yes, they have it. I probably have it too. So let’s just take
a couple of minutes then for burning questions for Dr. Milstein. Mike?
DR. FITZMAURICE: Arne, Mike Fitzmaurice here. On number three, routine
collection of standardized information on the payable charge for every service
or service bundle, by payable charge do you mean the amount paid for every
service or service bundle? I mean forget the charge –
DR. MILSTEIN: Yes, sometimes when there are prolonged, the reason I phrased
it as amounts payable is sometimes there is a prolongation of the collection
process and most providers know based on the nature of their contractual
relationship with any given payer what the amount payable is. It is analogous
to the field on the new UBO core which is labeled amount due. The amount
payable, amount due, it is the amount that under whatever contractual
relationship exists between the provider and the payer that is payable for the
service by virtue of the contract. The only place where this is a little bit
challenging to implement is where the patient does not have insurance and
presumably that is going to be a much lower order of magnitude problem going
DR. FITZMAURICE: Many states are starting to set up all payer databases to
try to get at this payable charge so I think that that is one of your many good
DR. MIDDLETON: Matt?
MR. QUINN: Thank you so much for your testimony doctor. This is Matt Quinn
from AHRQ and I had a question under data usefulness number five, conversion of
all EHR information into structured data in standardized formats. Our last
hearing that the quality subcommittee had was on meaningful measures and one of
the things that came out of it is that for the ability to use electronic health
records to their potential to support quality measurement we needed to have
standardized data formats underneath.
One of the things that we struggled within this is you can standardize
everything but then where do you start in this in building it out to balance
the needs of the measurement community and the needs of quality improvement but
also with balancing the needs to maintain not too much work for the individual
DR. MILSTEIN: Well I guess it is in a sense that obviously a good point and
I think probably in the interim we will probably need to rely more heavily on
software that can make reasonable guesses based on an inferential logic. That
is obviously not the ultimate solution but I think in your question you
correctly articulated that this is a problem that is eventually solvable and
the only question is what, when we expect it to happen and how tolerance of
error we want to be with respect to interim fixes such as software that could
make a reasonable guess based on the electronic text as to what the answer is.
I acknowledge this is something that needs to be finely calibrated. If you move
too fast on accelerating the workload implications on clinicians associated
with getting everything into the records in standardized electronic format, you
could induce an unfavorable reaction.
So I think it is the kind of thing that Office of the National Coordinator
have laid out with essentially signatures and imagine this could be woven into
the definition of the escalator level so that the greatest reward is for
providers that achieve a higher rate at this whether they do it through being
more rapidly insisting that providers use and adhere to this discipline of
standardized format or whether they in the interim fall back on software that
can take a pass at the electronic text and draw out reasonable inferences.
DR. MIDDLETON: Last quick question, Larry Green.
DR. GREEN: Arne, thank you for being so clear again. I have a question about
the registry idea. Could you go back to that one – the registry of at risk or
high cost patients amenable to intervention. What is your thinking about how
that gets formed? Does it exist in real life? Is there someplace we could see
one of these and touch it?
DR. MILSTEIN: Sure, I think in some ways we do have some great national
models for how this could happen and work. I think for example the Vermont
Oxford Network of NICUs. We have some very nice expansions on that to other
facets of perinatal estimates in California being led by Jeff Gould and
colleagues at Stanford. The American Colleges of Cardiology at registry on
interventional cardiology are comparable and preceding registry that was
established by the SDS. These all advise you and show the way and show that it
There is a regional, there is escape for it and that providers can be
engaged. I should have thrown in by the way another I think really admirable
step forward that has been launched over the last several years and that is the
American College of Surgeons’ adoption and now promotion of NSPT(?) These all
illustrate that since these things are doable in the U.S. and they also show
that once they are in place they can be infinitely expandable into things that
patients care about. For example the evolution of measurement of risk adjusted
outcomes that move beyond mortality against the measure complications both in
during hospitalization and in some cases during a reasonable period of time
DR. MIDDLETON: Thank you again Arne, and now quickly switching gears, Dr.
DR. ROSKI: Yes, hello. I am happy to be speaking to you today and I am
really sorry I could not be there in person. We had some last minute scheduling
issues that prevented me from being there.
What I thought I might be able to contribute today is offering my views on
the trajectory of performance measurement and how we can capture value. In
order to be efficient in my presentation what I would like to do first is
introduce a few thoughts about where I think we should be in three to five
years and then offer some perspective on that vision and what might be a
pathway to get there.
What I will try to do also is keeping in mind with my comments where if you
will the majority of the country will be in terms of capacity to reach that
vision rather than focus on what a few advance systems might be able to do. I
will reference in my comments the work that we have involved in here at
Brookings in working with payers and at hiring accountable care organizations
which to some extent may represent some laboratory with some specific
incentives to produce useful information. And I will also talk a little bit
about the very early impressions about what is happening with the Beacon
Community Program along with Booz Allen and the Institute for Health Care
Improvement. Brookings will be offering some technical assistance to these
communities over the next few years.
Let me start with where we need to be in a few years. I think one of the
things that we need to figure out is how to develop a more patient centric view
on quality rather than a physician centric view of quality. We have heard a
little bit about this earlier today about rather than focusing on individual
physicians we may want to be focusing on clinics or care teams as a unit of
measurement. What I am suggesting is that it might be even more useful to talk
about a patient centric view that tracks people or patients along the episode
of care be that a cardiac episode and so forth so that we are going to be able
to tell more specifically keeping in mind of how a particular form is or is not
successful if patients along there continuum of care receive all of the care
that they needed and were able to achieve the outcomes that were hoped for.
What we clearly will need to do for that is over time move away from the
often process focus measure that we have to move towards outcomes be they
intermediate or quote unquote real outcomes that we can track across care
settings and for groups of patients look for particular conditions’ needs and
secondarily figure out ways not only to capture information about the
biological stage of these patients but also gather patient feedback about
preferences and values that these patients have in receiving care and
particularly as it pertains to preference sensitive care so that we can more
effectively initiate decision making and so forth and I am specifically also
addressing not only the need to employ quote unquote after the fact patient
survey but develop more real time opportunities to collect patient information
that can be stored alongside clinical information to track quality of care.
Clearly what we will need to do in order to get to a value based assessment
of care is link these outcomes with the resources tendency that was associated
producing. That has got to be a clear focus of where we need to be including
developing more accurate ways to measure cost of care at the various units that
might be useful as we have heard from Arne and others about potential ways to
get at that. Then in my book what we need to lay out is given if that was the
vision, how we can build a trajectory to get there assuming that it is not
going to be an on and off switch that we will have available to us to make this
become reality but figuring out ways of what measures could become available,
by when, given some intensive focus.
Given that vision, let me offer some perspective on what we have learned in
working with organizations around the country. We are currently working with
about 80 aspiring A-fields around the country that are trying to become
accountable care organizations, or at least have a great interest in learning
more about what this may be. As part of that we are working with five A-field
pilots and their care partners in a very intensive format to help them actually
implement contracting provisions, building a network, figuring out what the
metrics are going to be and so forth.
What is very clear is that given the, that there is a great deal of interest
in this particular format of care that may get away from kind of a procedure by
procedure point of view to a more comprehensive view of care, that there is a
great deal of interest in getting to, having some guidance, getting some
consistency, the ability to receive feedback about performance not in
relationship only to where that same organization was at the previous time
point but also in relationship to others.
What is clear also as a result of that is the great hunger for some
consistency that will allow for the such comparisons to take place which then
may lead you to think about what could be some core set of measures for example
that would be consistent across many of these organizations and what might the
other measures that could be augmented based on local needs and requirements
and so forth.
The site that we were working with I would describe as representing an
advanced group of organizations but not a vanguard group of organizations. The
reason why I am stressing that is that that was a very deliberate decision on
our part because we think that if these are going to be successful, they have
to work for a very large part of the country who are otherwise it is going to
be more of a nation of programs that is not going to transform health care in
the way that I think we are all hoping it can be transformed.
One thing that we learned very early on is that these organizations who are
very committed to transforming themselves are at very different points in their
trajectory of being able to measure themselves ranging from organizations who
have recently formed and are trying to either install or reconcile different
standards that are in place between different health IT systems that have been
put in place in these organizations to entities who actually have an EMR system
or EHR system that spans much of their organization. It is very clear that most
of these organizations are very early on in their process.
What that has led us to pursue relative to the measurement then is to lay
out a trajectory of where we need to be and where in the first round of
measures that are going to be used we are going to be relying on measures that
could be computed to administrative records where the payers actually will
compute these results on behalf of the ACO covering demands such as
effectiveness, preventive care, pharma care safety, overuse and so forth and
feed that information back to these ACOs in a consistent way.
In parallel to that we are setting up an additional mechanism to get to
additional measures that will require more time for these organizations to
compute it and that is intermediate outcome measures that will require
information from about lab results and various other components that would not
be contained in administrative data.
Interestingly enough there is no one pathway to get this information from
these sites. As it turns out, one of the more effective ways may be in having
the payer identify quote unquote the denominator for the particular population
and the provider side or the ACO providing quote unquote the numerator or any
exclusion criteria that may be relevant that could not be captured in the
When I say that there is no one pathway what we have learned is that many of
these organizations have already developed some internal mechanism to produce
information for feedback purposes that differs between these organizations
ranging from figuring out how to more directly access electronic lab fees so
that information becomes more available to working with web portals where
physicians might be sending specific pieces of information that have been
prioritized to some data warehouse to use some internal and external registries
and last but not least, several of the organizations still rely on manual
extraction if they cannot get at the information in any other way.
Hopefully the meaningful use programs and other incentives will accelerate
that but one thing that I have certainly observed over my career in health care
is that the hope for more information becoming available tomorrow once some new
technology has been implemented has frequently been thwarted meaning it has
been much slower than one would have thought. I think Arne’s comments sort of
spoke to that as well.
One of the issues that we are going to be working with or that we are going
to be considering of course if once we have intermediate outcomes available for
example, those might replace process measures that otherwise are now being
produced through administrative data. For all of these organizations we have
been working with it is probably true that this is a heavy lift even for the
measures many of us might consider very boring and run of the mill. For many of
these organizations that we are working with it turns that simply producing
them is a significant effort for them, particularly for organizations that are
just now forming as opposed to let’s say more mature medical groups that may
have been working together in different parts of the country such as Minnesota
or California where some of this may already have been underway for quite some
What I would also observe is that what we are seeing here with these ACOs I
think is actually fairly representative of what we are going to be seeing all
around the country. In other words if this is where organizations are at that
are trying to by ACOs who integrate here and come together in ways that are
unprecedented if you will, it will certainly apply, or let me put it this way,
then organizations who are less evolved, who are more fragmented or represents
small clinics, these issues will be even harder to implement.
I have had an opportunity, and I would encourage Marc Overhage who is
leading one of these Beacon sites to comment on this as well. I had an
opportunity to attend one of your early meetings of the Beacon communities and
one thing that became clear to me is that there is a great deal of variability
in what these communities can achieve, where they are at at the moment in terms
of their IT infrastructure and what they can produce and it will be interesting
to see how quickly they can realize the aims that they have set out in their
proposals. The Beacon Community is of course by definition if you will were
meant to represent the most advance communities we have in terms of producing
information that currently is not being able to be produced and so I think we
need to keep a very close eye on what is possible in those communities thinking
about them as a laboratory if you will for what might become more possible in a
few years time in many more regions of the country.
DR. MIDDLETON: That was really terrific. Thank you for your thoughtful
comments and insightful, as well. If you have a couple of minutes, can you
entertain a couple questions?
DR. ROSKI: Absolutely.
DR. MIDDLETON: Okay, super. Are there any questions for Dr. Roski? One
question comes to my mind, I guess thinking about the standardization problem,
taking it down a level which is inevitable sometimes for this group to get into
the weeds, do you have thoughts about merging those clinical and administrative
data to make more insightful measures and intermediate outcome assessments?
Which way do you see the winds pointing if you have thoughts there.
DR. ROSKI: How to create more meaningful measures you mean?
DR. MIDDLETON: Well and standards relevant to those more meaningful
DR. ROSKI: Yes, I am sort of two minds if you will. One thing that I was
very struck by in working with the various organizations is that some of these
organizations have developed this work around if you will that are trying to
get around the issue of not having fully operational EMR systems in place. To
some extent while we are working on standards for how EMR systems may want to
capture information we may want to also think about a transition period where
information might be coming from several different sources and figuring out how
for example in one instance information that is directly set by a lab to a
claims database could be considered a standard and how that should be
implemented figuring out work around using web portals to transmit some pieces
of information rather than making a lot of pieces of information available and
so forth. Hopefully that will only be needed for let’s say three to five years
until EMR systems are more widespread and have implemented the type of
standards that are all ready being worked on now.
DR. MIDDLETON: Larry Green?
DR. GREEN: Thank you very much. I would like to ask you to link your
thinking if you could and share it with us from one of your first comments
about patient centric care, moving from provider centric to patient centric
where you, if I heard you correctly mentioned the need to know about episodes
DR. ROSKI: Yes.
DR. GREEN: And then jump to the needs of the aspiring ACOs.
DR. ROSKI: Yes that is actually a very good point and very important. It
would seem to me that as a payer or as a country we would want to know that
people with heart attack or people with particular procedures six months after
the fact are as healthy as they could be or as functional as they could be.
That information would in my view, be very difficult to obtain if we simply
focused on whatever is in one cardiologist’s records or in one surgeon’s
records or in one hospital’s records. Instead it would be information that
would need to be compiled from the ambulatory physician, let’s say a primary
care physician, the hospital, the rehab unit, the ambulatory site, or
cardiologist, who takes on the ongoing care of a patient with cardio conditions
and so forth.
And examples that we have worked on for example with the ACC is figuring out
the American College of Cardiology, are there ways for example that information
could be more effectively linked across the continuum of care. So for example,
if we were to link information from registries which in this case procedure
registries which for the most part contained information that is very detailed
and which about a small part of the episode to information that have some
record on what happens to the patient after the fact. Is that patient on the
right medication? Is that patient alive and so forth? What we worked on for
example is to link this registry information with information that is contained
in for example administrative records, linking it to on the one hand string
together an episode of care and on the other hand use the clinical information
to enrich and adjust quote unquote information on the administrative side and
enrich the information sitting in the registry with long term follow up
information at a somewhat thin level but at least at some level that is
currently not available in most registries. Certainly that would by true for
the ACC registry.
In my book if we could for example in the context of an ACO achieve and
promote those types of linkages so that at the end of the day we could say
there is a patient with a heart attack six months after the fact lead as
healthy a life as she could then that would be the information that I think
would tell us more about health care value than we currently have available.
DR. GREEN: Would you go so far as to say that the meeting the aspirations of
the ACO depends in part on the ability to discern, define and operationalize
episodes of care?
DR. ROSKI: I would say if an ACO has the heart. I would say that an ACO who
does not have the right information available along the continuum of care, they
may not in the end produce the value and cost savings that I think we are all
hoping for. So to some extent we are going to be entering into an
experimentation stage here over the next two years where we will try to see
maybe facilitated through the center for innovation that is being built at CMS
what seems to be working, what type of programs that are being implemented
including those relying on data linkage will get us to the right outcomes.
DR. MIDDLETON: Paul Tang.
DR. TANG: Hi, Joachim.
DR. ROSKI: Hi, Paul.
DR. TANG: Very good presentation. I liked your thinking on that; also
enjoyed the request to pull in patient both the preferences and real time
collection of their information in terms of outcomes. May I also make a plea in
terms of you made reference to saying well maybe as far as capturing some of
the measures the payer would provide the denominator and the clinician provide
the numerator exclusions. My concern about the payer providing denominators is
when you define it using those administrative you know the two encounters
within a measurement period, etcetera, you miss a lot of folks the very people
that we probably need to reach out to. The corollary to that is you end up
artificially inflating the score because you are only going to include in the
denominator the people you are already working on.
DR. ROSKI: Yes, in the case of ACOs it works a little bit differently than
at the, if you will, and the individual physician level. Here you actually
assign whole populations independent of their condition. So what you would do
is you would assign, or the way that we have doing it is to define the overall
population that is quote unquote enrolled in that ACO and then within that
group look for particular series of events.
I would go back to what an earlier speaker said is that that is not the way
we would ideally want to do it. Ideally what you would want to do is have the
ACO themselves generate this information. However what we are finding is that
they cannot do it. They cannot reliably at this point find all of the people
with diabetes or all of the people with particular other conditions you name it
and that is kind of an interim strategy in order to get to some denominator
that the payer believes the ACO should be accountable for.
DR. TANG: In other words, so if you are doing it. I did not know which
numerator you were referring to. So if you are doing it by condition let’s say
diabetes, could use other data like the lab tests or even the medications in
diabetes for example and that is more accessible to the provider than it is to
the payer. So I can see identifying the quote assigned group from the payer
point of view but most everything else where you want to measure whether it is
the health or the disease or the chronic condition state, you might be better
able to help them find out then do it on your own from administrative data.
That is all.
DR. TANG: Thank you.
DR. MIDDLETON: Again, thanks again. Blackford Middleton here. Terrific
presentation and you were very generous with your time. We are now ten minutes
over. I am going to suggest we conclude this session and propose a five minute
bio break. I do not know if anyone else feels the same way but let us try to
reconvene in five minutes and we will take it from there. Thank you.
DR. ROSKI: Bye.
DR. MIDDLETON: Thanks everybody for coming back quickly after a short break,
but we want to continue on our agenda which is just running a little bit late
with Dr. Steve Solomon, welcome to you, from the Department of Health and Human
Services, the Office of Health Care Quality. I think part of the goal with this
discussion is to learn more about your thoughts on all of the questions and
your office’s activities and help you understand more and I apologize we have
not met but about NCVHS activities and its current efforts. So please take it
Agenda Item: Roundtable – Deliberating the
Issues – Introduction of the DHHS Office of Healthcare Quality
DR. SOLOMON: Thank you very much and I am sorry I cannot be with you in
person. Normally we are in Washington but we are here in Atlanta at a
conference on health care follow the elimination of health care associated
So I would just say a little bit. In this time I am going to give you about
a ten minute or so overview of the office and some of our activities and then
perhaps we can have some dialogue and you can ask some questions and we can get
really at some issues that you may be most interested in.
The office got its start in 2008 and 2009 as a result as you know very, very
well of rapidly increasing interest in the issue of health care quality and the
decade long growth in the tremendous interest in patient safety and the patient
safety movement. The office was established specifically to improve
coordination across the HHS agencies on the subject of health care quality to
serve as a focal point for coordinating and leveraging the agencies’ activities
and to serve as a point of contact for external groups looking to interact with
the Department and its agencies around the issue of health care quality and to
provide leadership and innovation as well as to try to increase the efficiency
and effectiveness of the work that HHS does in health care quality.
Our initial and real focus has been on health care associated infections
which was an area of extraordinary interest and continues to be obviously both
in Congress and in the executive branch. We have also worked, I have staff
working on public health quality issues in recognition of the fact that these
are really health systems issues and that a lot of the improvements in health
care and health care quality are going to result as a result of better
community services and improvements in the public health system.
The structure that was set up to manage activities across the department was
a steering committee, executive level steering committee broadly representative
of HHS’s agencies and offices and other components of the Department with high
level representation from those various components. That steering committee has
a number of subcommittees and working groups that contribute to its overall
efforts and the first major accomplishment was the development of an HHS Action
Plan for the prevention and hopefully eventual elimination of health care
associated infections. In the beginning that Action Plan addressed only work in
acute care hospitals and focused on five areas of specific types of infections.
Those were three device associated infections: blood stream infections,
catheter associated urinary tract infections and surgical site infections and
two organism specific infections: Methicillin-resistant Staph aureus and class
treating of difficile infections, set some very ambitious goals including a 50
percent reduction over five years in the incidence of catheter related blood
That got off the ground in 2008. The baseline for a number of these measures
is 2008 data collection. The primary data collection source is the National
Health Care Safety Network run by CDC although some of the measures are drawn
from the HCUP Project managed by AHRQ, the SKIP Project managed by CMS and the
Emerging Infections Project also managed by CDC.
So the work in acute care hospitals was deemed the tier one effort. In 2009,
at the end of 2009 and during this year, 2010, we have expanded the scope of
the action plan to include the so called tier two activities of which extends
us into the outpatient area and those are specifically infection prevention in
ambulatory surgical centers and in end stage renal disease dialysis centers and
we can talk if you wish about why those were chosen, as well as extending into
the area of health care personnel health and the area of influenza vaccination
among health care personnel.
Specific chapters have been developed for the Action Plan to expand the
Action Plan. Those draft modules have recently been published in the Federal
Register. I think they were released actually just about two weeks ago in the
Federal Register. We are now reviewing the comments on those drafts. The Action
Plan itself is undergoing an extensive revision and we hope to have a new
version of the Action Plan which includes those new modules and the extension
into the outpatient area published sometime in the first quarter of calendar
In terms of the measurement and the goals that were set in 2008, 2009 and as
I know you have been taking about and we are taking here at the meeting in
Atlanta very, very important to have good measures to be able to track progress
and we recently completed a meeting, a stakeholders’ meeting that we held in
Arlington, in Washington area just the end of last month which and I was able
to hear some of your earlier discussions this morning about a more consumer
centric view. In fact we had a very good representation from consumer groups
and patient advocates at that meeting and worked very hard to engage them. At
that meeting we presented our progress over the last two years since the
targets were set.
I mentioned the 50 percent reduction and we are in fact on track. The
progress that has been made of two years has seen substantial improvements in
the rate of catheter related bloodstream infections as measured by the National
Healthcare Safety Network and if we can maintain that pace of change, we will
hit that target. And we are also on track to meet the target for reduction in
surgical site infections. That is a more modest target. That is a twenty-five
percent reduction over five years of what we are on track for that one as well.
So we continue to monitor that progress.
The new modules for aim at surgical centers and end stage renal disease
dialysis centers as well as for flu will set very specific targets and goals
for achieving improvements in infection rates in those settings and for those
infections and we continue to work on that as sort of the core of the work we
do from the standpoint of promoting innovation.
We give seed money primarily to the agencies and to other groups at HHS, to
fund projects that would not necessarily be a part of their regular mission but
certainly that encourage cross agency collaboration. Those are again pilot
projects or seed money for things that are done both in the area of research
and in demonstration projects. We are now supporting a similar type projects in
some of the HHS regions and we actually very, very proud of those because they
are also stimulating work. For example to look at infection prevention in small
and rural hospitals in the West, to look at communications strategies for
reaching out to consumers, to enlist consumers in helping in the goal of
improving health care quality and preventing associated infections. So that is
the bulk of our work within the area of health care quality.
I will just mention very briefly, the public health quality work. We have a
very, very dedicated if small staff working on that. They have also done a
terrific job of reaching out throughout the public health system particularly
the governmental public health organizations, state and local health
departments as well as other experts in public health. They have developed a
working paper on priority areas for improvement in the public health system and
that document is going to be released and presented. It is scheduled to be
presented at the meeting of the American Public Health Association which is
coming up in about two weeks in Denver and that will, if all goes according to
plan, that will be presented by Dr. Howard Coe who as you know is the Assistant
Secretary for Health. So that is the other major area that we are working on.
That is a very, very brief overview but I know you as always are always
challenged for time. Let me stop there and see if there are any questions and
then maybe we have some time for me to elaborate on some points that are of
particular interest to you.
DR. MIDDLETON: Thank you very much Dr. Solomon. There probably will be
several questions. Any burning questions on the table? Larry.
DR. GREEN: Can you say more about this public health quality measurement
group and particularly help us understand where you seeing the intersection
with the redesign of the delivery system that NCVHS is very interested,
particularly its Population Subcommittee but also this Quality Subcommittee in
finding the new epidemiology. Have we developed new numerators and new
denominators, collect data once use and use it often reuse it, reuse it, reuse
it, reuse it, reuse it. That is sort of a mantra almost. Can you tell us where
you are headed and what the reach is of this report and to whom it is intended?
DR. SOLOMON: Yes, absolutely. As of this point, it is not measurements per
se. There is a section on public health measurement but it does not get into
the area of choosing measures. It says that those are very, very important and
recognizes a tremendous amount of work is going on. That particular work which
is going to be presented at APHA really focuses more on systems type thinking
and the idea that we need to take a more systems based approach toward
improving public health quality. What needs to be done within the existing
public health system to change some of the models of care and as is often
discussed, Larry and I will let me just digress for thirty seconds by saying
that yes the funding through the American Reinvestment Recovery Act really
sparked tremendous growth within the public health system.
The meeting that we are here is a meeting of state health care association
infection coordinators and what is being discussed is adopting the paradigm of
elimination. Some of you are familiar with other campaigns. For example, the
committee to eliminate tuberculosis in the United States and the long standing
approach within the public health system to have. We have the state TB
coordinators. We have state chronic disease coordinators. We have state
environmental health coordinators. Well as a result of the ARA funding, we now
have state HAI coordinators. That is the group that is meeting here in Atlanta
from all 50 states and territories to discuss what public health departments do
to help eliminate health care associated infections. That is tied in with the
growth in the public reporting of health care associated infection rates from
hospitals and other health care institutions.
Just to complete that digression, obviously the Affordable Care Act contains
a number of provisions that reflect back on population health. We are in fact
the targets that we have are national targets. Those are population based
targets for the NHS and a measurement for example for central line associated
bloodstream infections. I will say that the HHS Action Plan which I mentioned
is actually cited in the Affordable Care Act legislation as the milestone to be
used for measuring progress for health care quality improvement.
So from the standpoint of population health, that is the work that is going
to be presented at APHA does not get into the specifics of what types of
measurement systems but really talks about how to restructure the health care,
some ideas for improving systems work within the public health system. So I
hope that clarifies that.
I think the other part of the question, Larry, was what is the intersection?
Clearly that is as I note you have been discussing is part of the Affordable
Care Act the idea is to keep people out of the hospital. To keep people
healthy. The so called downstream work or upstream depending on your point of
view, the primary prevention that public health departments and communities
need to be engaging in that helps keep people in the community, helps keep
people well and keeps them out of the system. I think that is the broad brush
stroke on that.
DR. MIDDLETON: Steve, Blackford Middleton here. I was wondering if you could
give a little more color commentary on the Action Plan details that will be the
interventions reducing hospital acquired infections or you know the
cath-related blood stream infections, etcetera. What is the intervention that
you are proposing and how is that being implemented across the country? Is an
IHI type thing or what?
DR. SOLOMON: Yes, these individual subgroups that work as part of the
steering committee work across a range of issues. There is an information
systems and technology subgroup, an incentives and oversight subgroup that
talks about things like pay for performance. There is an outreach and messaging
subgroup that is very involved in both consumer engagement and the engagement
of health care providers and there is a research subgroup that talks about what
are the knowledge gaps that need to be filled.
There is no single, there is no magic bullet. However what is being
discussed quite a bit at this conference and as you may know at many other
conferences is the so called CUSP type approach, the Comprehensive Unit Based
Prevention which really talks about having three components. One is what is the
best evidence. What are the evidence based recommendations and what are the
high priority recommendations?
CDC’s Hospital Infection Control Practices Advisory Committee has been
developing guidelines, evidence based guidelines and producing those for many,
many years and in fact there are several hundred specific recommendations and
one of the calls was that there needs to be a prioritization of what are the
most important guidelines. What are the most important things that hospitals
and other health care agencies can do. So we are not only having the evidence
based guidelines. We are prioritizing those.
The second component, is the data is feeding back both national data for
benchmarking and local data which obviously as you know, is the most important
thing. What is happening in your institution? What is happening in your
hospital? What is happening on your ward? What are the rates associated with
your surgeon? That type of local data is extremely important.
And the third component is really the so called culture of safety which
again gets at this issue of either some folks view it as the social science
issues. Some folks view it as a complex systems or a systems thinking issue.
But it is really to address the model of the way we think about health care and
health care associated infections at the local level. So that is a very popular
approach that has been used, developed by Peter Pronovost among others at
Hopkins. It has been used with tremendous success in Michigan in a
collaborative with the Michigan Hospital Association and that is spreading to
some other states. Tennessee has had some good results with that.
So I would not say it is a thousand flowers blooming, but there is a
recognition that within each state, within each community, and within each
facility you have to take these best practices and learn from them and adapt
them. But I would say as very, very well the three cornerstones of that are
having the data, having the evidence based practices and then in fact changing
the culture to be more prevention oriented and variations on that theme does
seem to be working.
DR. MIDDLETON: Matt.
MR. QUINN: Thanks Dr. Solomon. This is Matt Quinn from AHRQ. Just a quick
question. Is there a way that we could have a roster of the folks who are part
of the group and then some of those subgroups around IT and technology and some
of the others?
DR. SOLOMON: Absolutely, absolutely and we can get that for you. The folks
in your shop most involved have been Del Munet(?), Kathy Crosson and Jim Quinn
and Jim Battles who is here at this meeting. Those four folks have really been
in the lead on working with us on the steering committee and on the subgroups.
So I am happy to connect you with them or I am happy to get the information.
DR. MIDDLETON: I was going to see if you could send the whole thing to
Debbie Jackson. That would be great and she can send the list to the group.
DR. SOLOMON: Yes, sure.
DR. MIDDLETON: Thank you. I am going to suggest we move on if there are no
more burning questions, and perhaps the desert for our meeting is the next part
here. Booz Allen has conducted an environmental scan which Paul and I have had
a chance to preview. I think in a way we can begin to think now about to tie
together of all of the four different panels and the various themes we have
heard from all of the expert testimony and now relate this to the environmental
scan and begin to frame a quality letter to go forward to the Secretary for
this vision of a measure roadmap for the next three to five years. So, Kristine
please take it away.
Agenda Item: Environmental Scan
MS. MARTIN-ANDERSON: Great and I was not here in the very beginning so I am
Kristine Martin-Anderson from Booz Allen Hamilton.
I am going to take you through an environmental scan that we started in
about September as the hearing was being planned and explained how we did it
and what we had found. Starting with the executive summary, I think that most
of the summary actually resonates well with me from the panels and what we have
heard. What we wanted to address was where are the main gaps and there are
still many, many gaps around information that can track progress toward an
efficient health care system and then a gap that we talked about a lot
yesterday came up again today. Information that actually can engage and
activate consumer assessment. It is not really health reform. It is
participating in their own health.
We noted a pretty obvious difference between what consumers desire to manage
their health and health care and what is important to regulators and
purchasers. Consumers do not really equate information with measures and they
do not use the current measures but this is actually something really important
that we will come back to which is that a lot of our scan was focused on how do
you get measures that matter? How do you get measures that actually advance
health and health care improvement? So we will come back to that a little bit.
Provider organizations have embraced the need to measure and improve and
demonstrate efficiency in health care systems. It is really not a matter of
lack of desire. It is really a matter of problems still around methods and
measured gaps and this concern about burden of measurement and the potential
that it actually could divert the resources from quality improvement where they
really need to be spending time to advance. So it is again that measure then
reuse, reuse, reuse is something that we hear again and again. Providers have a
particular need for information about how to engage consumers and how to
coordinate their care and how to assess their progress. So we are going to go
through some of the details. But those were sort of the high level findings.
So why do we do this scan? It was to support this Quality Subcommittee in
their development of a letter to the Secretary with recommendations. We really
wanted to gain perspective as the Committee really wanted to gain perspective
around several different stake holder groups. So those who use or consume, does
it provide care, accredit and regulate and then who make decisions about
coverage. So you saw that the agenda was actually structured that way. I am not
going review the questions because are the same questions that are in the
agenda and that our presenters actually presented so I am skipping now ahead to
This is where we really talk about how we approach this scan. The guidance
we got from the Co-Chairs from the very beginning was this need to be forward
looking. This is not supposed to be an environmental scan that looks back upon
all of the things that have already been tried and says you know what needs to
continue. And so in order to do that, the guidance that we got was look for the
measures that matter. Look at places where performance improvement has occurred
and what measures help drive that. Look at places where performance improvement
is not yet getting us where we want to be and look to see what people are
articulating the gaps to be.
So the way we did that really was not a classic literature review. We did
look at the literature but was also did searches that were more oriented toward
improvement had occurred and then trying to look back at those projects and say
what was it that, is there anything in these descriptions of efforts that tells
us what they attribute their improvement toward. How do they get there and what
role did measure play in there.
We had to look at a lot of different information and we also wanted to look
at the roadmaps and frameworks that were already identified because the idea of
doing a measures roadmap is brand new. There have been quite a few done and
there have also been a number of frameworks around how to think about
measurement and so we also then looked at the existing frameworks and wanted to
make sure that we could sink up with what those organizations said were the
gaps then compare that to what we saw the gaps were in practice for those who
actually were advancing quality improvement.
When I organize these results by the four questions and the first question
was what information and measures would help consumers better understand and
become more active participants in managing their own health. The findings here
really very much reflect what we heard in the panel yesterday. They were
looking for condition specific information and also patient specific
information. They wanted personalization, so actually if I flip to the next
slide it makes it easier to summarize it.
For the consumer measures, the message back was they are not so interested
in these fragmented quality measures. They really want holistic measures and
also customized to their own condition ideally linked to some kind of
coordinated guidelines that are not about each individual specialty that they
might come in contact with. They want answers to the questions like how am I
doing relative to how well I should be doing post surgery or what is my likely
outcome given my circumstances. We heard a lot of that yesterday. So that was
not too new.
Really consumers do not use quality measures to select providers. There have
been a number of surveys that prove that. The most recent one we found was a
Deloitte survey that was just done in 2010 where and it was around selecting
hospitals. Quality ranked sixth among factors that consumers use to choose a
hospital. So this was consistent across most of the literature that we looked
at. That is not what consumers wanted. They wanted much more information about
how to coordinate their care and they wanted to know how well their provider
coordinates care. They also wanted to know if the whole person was taken into
account. So we talked about yesterday about functional status and quality of
life indicators. They also want to know something about what are my options
given my insurance choices, etcetera. Very personalized information.
Within the consumer groups, we also saw a bit of consistency in the writing
around linking what we want to give to consumers to what is happening in
community health, so connecting health and care. So, looking for metrics to
help connect clinicians and the communities, and population health metrics, and
potentially composites to measure the impact of community interventions and
medical care improvements. There were a few communities where we found that
already occurring in Minnesota and San Francisco and a few other places but it
was not the common response.
In this journey in talking about how quality improvement, quality measures
and improvement are linked we came across a lot of references to technologies
and when we found those in the what worked thing part of the scan, when we
found those I think the story was that there were lots of improvement stories
that told how they had integrated technologies into moving from measures to
improvement. So these are just a few of them talking about capturing consumer
specific information that was put into PHRs and the accessibility of that.
Functional assessment information, that was also collected through in-home
monitoring techniques. Patient experience information that they were able to
create their own feedback links and also talk to others with similar conditions
through social networks. Then also different ways of doing patient navigation
so one of the main interests of consumers was how do I navigate this health
care system. There are lots of different formats by which they were able to do
that. I think this is presented more for context. If you are thinking about the
role measurement plays in improvement, you also think a little bit about what
else plays a role in improvement and how do we make the whole system work
Then on providers, I am going to switch to just the measures as a way to
talk about it. Okay. We did see quite a bit about customization of quality
measures. I think Barry talked about that a little bit this morning, what you
use internally within your clinic or even within the Minnesota Community
Measurement. It might be a little bit different than what is out there for
public reporting. So they might not always be the same. They sit together
right? These are measures that are often times overlapping with something else
that is a priority at the national level, and that led us to ask the question
in sort of a ruminating way, is how do you provide the least burden on
providers. So if there are for instance process measures that are more likely
to be customized, how do you think about getting you know nesting quality
improvement and public reporting or should they really be the same? This is
really something for the Committee to talk about some more and decide which if
they feel a particular way for their recommendations.
But there were sort of two sides to that. There were some who said look for
QI I really need to be able customize some measures because what matters in my
community is what gets my clinicians to engage. It does not have to be same as
what is measured nationally and others that said look we really need to reduce
the burden and that is the number one concern. So we are going to use the
national measures and we are just going to repeat use of them for quality
improvement. But providers also expressed a need for better understanding of
non-clinical outcomes. So how do they, for their particular population so they
can support patient centered care. They are also aware of the fact that their
view is often fractured of the whole patient.
There was a need expressed for providers to understand how their peer
clinicians are treating the same patient. So the lack of information that flows
back and forth between a primary care physician and a specialist is something
that they saw as a gap and they needed better metrics to think about as they
refer to specialists. How do they know who is the best or who is the right one
for that particular patient.
Providers saw gaps in areas that were mostly do to methodology or
information constraints so there are not good measures for care or coordination
or transitions of care and there have been some focus lately on hospital as
opposed to acute rehab, home health, etcetera. Episodes of care not being
satisfied as of yet with the way that episodes are constructed and that
measurement comes out of episodes; that the methodological issues have not been
Efficiency including patient flow, how do you get a sense of not just
efficiency within an individual site but what is an efficient episode.
Specialty care? There are some that are in proprietary registries and are
not accessible to everyone and there are other specialties where there really
is not a lot of data that is collected. Then trying to segment their population
and disparities and understand how health disparities is impacting outcomes
that a lot of the data that does not exist in order to be able to parse their
own data to understand performance.
We also again as we talked about provider, looked at a provider information,
saw these references to new technologies that are enabling improvement. Most of
the references in this case were around clinical decision support technologies
and helping point of care but also talking about there are some examples of
informed referral systems, electronic referral systems that are more advanced
that met the need. I think the point really between this slide and the last one
is there are examples of technologies that are advancing this. They are just
not necessarily universally used.
Then this consumer provider intersection was another area that we pulled
out. The lack of understanding and measures around self care strategies, and
which ones work. How can providers or health care facilities include patient
input? Measures of patient engagement and shared decision making are not
broadly adopted. Measurements of population specific care coordination and care
transition and then expansion of this patient experience of care survey so that
there is a feedback loop across the management of a medical condition and not
just experience I think that should be with an instance of care. So having a
better holistic view of what the patient is experiencing beyond what they are
experiencing in their office or in their facility.
Now switching gears a little bit we also looked then at professional
organizations, accreditation organizations and regulators we get to a different
place. I am going to go to slide twenty one this one.
The regulators and accreditation organizations are largely who is driving
the national measurement agenda and they desire more measures covering a
broader population. So the mantra that you will hear more measures, more
measures largely comes from this group and the purchasers which we will talk
about as well. But they are not necessarily creating the measures but really
choosing which measures get broadly adopted and implemented. They express a
need for new ways to measure systems of care as care is bundled and
organizations become jointly accountable through ACOs.
The next two are more reflections so I want to make it clear that these are
not things that were said in any one particular report but more sort of fodder
for the Subcommittee to think about. There seemed to be distinction between
procedural and knowledge based care to identify differences about how do you
think about how you measure across those two groups? So for instance what do
you do when you know that volume is a big driver of performance for a
procedure? How does that get incorporated into measurement and would that be
different from other areas where there is not such a high correlation between
volume and outcome?
Reflected on the lack of the examination of the decision making process and
this really has to do with the fact that not all decisions are date driven and
if they are they do not all use the same data. So looking more for some
systematic link between measure and behavior change so if you want to think
about these organizations that desire to move the system in a certain
direction, what measures best moved system and needing to understand what types
of measures matter to different groups.
The last group that we looked at were payers and group purchasers, and here
we saw more of a focus on measures of value and efficiency, the need for
consensus definitions around how you measure value and efficiency. Desire for
more resource utilization measures during an episode of care so how do you get
to overuse and under use?
Integrated quality and administrative data? I think it also should be
clinical and administrative data for quality.
Outcome measures to evaluate the impact of a coverage decision. So there has
already been some research that has shown that what you cover impacts what
utilization of which services will be used but more need to understand how that
lever of changing coverage impacts quality.
Then information on patient satisfaction experience and quality, in relation
to overall care efficiency. But in this particular area, there was lots of
writings about the deficits and the ability to define value because of the lack
of agreement on measuring costs and how cost could be attributed to certain
treatments, products, people, etcetera.
Overall the payer and group purchasers tracked really closely with the
regulators and accrediting bodies on measure means broadly although there was
more focus on value and the need for resource utilization that brings measures.
Measures of value are really challenged by changing perspectives cost to whom,
value to whom, and also you know this discussion we had yesterday around is it
out of pocket costs or is it total costs that really move a consumer. If it is
only out of pocket costs that move the consumer but it is total cost that need
to move to satisfy the payers and group purchasers how do you reconcile that
and have it all pull in the same direction?
Then again this reflection on how do we access value when coverage decisions
influence consumer and provider decision making in really non-random ways?
Take some questions?
DR. MIDDLETON: Thank you Kristine. That was terrific even on a repeat
hearing with all of the nuances that you added in we can consider now after
these two days of testimony. So I guess now what I would like to do is open it
up for discussion across all issues from any member of the Subcommittee or any
testifiers still here. Any parting shots on synthesis or integration of themes
When I think about what we have heard about over the past two days there are
many, many different themes that come to mind for a measurement roadmap. You
know first perhaps the measurement science issues around what will be measured
and what are the constructs of measures both in the temporal sense over time
but over space as well perhaps in a geographic sense. How do we assess the
transitions of care and measure quality across transitions? How do we assess
quality across providers in a care team? How do we assess quality in relation
to a patient context? And where do the social, environmental, clinical and
financial impacts on an assessment of quality? How are patients themselves
contributors to quality through patient activation and how do we measure that
and perhaps even think about measuring provider activation or a system
activation thought that all three might relate to a quality assessment? What
are the sort of the physician attributes whether it is certification,
licensure, or other fundamental assessments of a knowledge base as another
element of a system characteristic physician as a component of the system. How
do we assess that and its influence on quality.
I am reminded of Justine Carr’s quality circle that the NCVHS created a
while back. I think we should surface that and think about it again as it
relates to a quality roadmap. It might be fodder for an extended view of
quality development over the upcoming years. Then how do we even visualize
quality. The compass ideas and the triple aims and the four axis of a compass.
What is meaningful to both the provider seeing a patient and thinking about a
population or a cohort or to a patient who is trying to assess a health care
system or a provider for a particular condition of interest. So those are just
some very quick thoughts. Lots and lots of ideas and extraordinarily fertile
ground I think to actually do something useful and interesting that will help
which is always nice.
But let me open it up for questions, discussions, synthesis from anybody
DR. TANG: So my true lumping style I can only think, I almost bring it we
talk about measures that matter and when we talk about costs and quality I
guess I am in the camp that says if you make appropriate decisions then the
cost will fall out of those. So similarly I think if we pay attention, if we
looked at a systematic sort of look at the various stakeholders in quality
measures, that the people who matter are one, the person we are focused on the
patient in the care situation and two, the folks who most influence that
through their orders, the clinicians and particular physicians. So if I hone in
on those two groups, start with the folks that we are trying to influence in
what they order, let’s say, and in particular docs and time and time again both
hear the story that Barry told and experience it my own organization where if
you give docs credible measures almost for free they want to improve. There is
nothing better than free in the sense if it does not consume any more societal
resources. So, credible data is powerfully motivating. So that is the give
people, give the people you are trying to influence, whose actions you are
trying to influence, the data and credible measures and things will change.
On the patient side, I think really we have done almost nothing for them up
to this point. The plea that is very simply stated was a couple of times which
is just give me a way to know what is going to happen to people similar to me
which is what Eva said and was repeated by Kristine. If I had measures like
that as a patient, not the average patient where everything is basically
normalize and the meaning exact away from it in terms of how can I as a patient
who wants to make an informed choice use this information. So in a sense,
stratifying to some major groups where I can try to match like your MPG, your
miles per gallon in a car, I either drive in the city or I mostly drive on the
highway. Let me know what is going to happen to me. So I either have lots of
comorbidity or I am basically healthy or I am this age group, etcetera. Let me
know how I could I sort of have an idea of what would happen to me if I were
under your care or your system’s care and oversight. I want to put it, I do not
want to only possess care. If I developed a relationship or partnership with
this health system, how would my overall health improve? I would like to know
that as a consumer, as an individual.
So I would almost focus on those two kinds of things and I think actually
the payers would get what they need out of that too. One, what you want I am
looking at Dave Stumpf who represents the payer when you just want high quality
cost effective care. If these other two make that happen, it really doesn’t
make. You can stop running your machines I think. I am somewhat being
categorical but also for the people on the phone, David Stodding.
Now regulators have it a little bit different. I mean we heard some very
interesting comprehensive way to figure out well are you in their case, are you
a good doc. Do you know the stuff? Do you know how to do the stuff? Do you do
this stuff? That is really good information and that is another complimentary
piece of information for a patient to know. I would just like to focus on
getting measures to people who are trying to influence on the professional
side, the docs and clinicians, and the people who are receiving of these
services on the consumer patient side. I think those measures are largely
unavailable in the public sector now meaning publicly reported so what Barry
said is unabashedly said you know what stuff that is handed over and that we
are forced to report on right now just really do not measure up to what I need
to influence the docs. So we use this zip(?) set of measures happen to be the
credible clinical measures. I think that is the message for me. So, sorry that
that was what I took away from Dave’s.
DR. GREEN: Before we lose them, is there any of our testifiers who are still
on the phone and for Barry and David, I wish you would say more about anything
that is going through your head right now.
DR. BERSHOW: Partial question for Kristine and then the rest of the group or
a comment first and then the question. I agree with what Paul just said and
what Kristine had presented that in general consumers have not been paying
attention in our own website hits and if we track where the come from they come
from doc’s health systems, insurance plans, big employers, not from consumers.
In a way you can think well it doesn’t matter if we can really get the quality
up then it doesn’t matter what they see but we know in fact that variation is
always going to exist and even if we get it to be stupendous at a 90 percent
versus 100 percent, there is going to be some who want to know who is the 90
percent and who is the 100 percent.
So my question is, is it that consumers are not really interested or that
they are operating under the false assumption that quality is like water in the
United States no matter what tap you pull it out of it it is excellent. It is
safe. It tastes good. I think if we could get the information in front of them
that on the average looking back at Minnesota five, six years ago that only ten
percent or less of the diabetics were getting the care they need. The
immunization rates are terrible. The cancer screening rates are terrible and
that there are huge variations spot to spot. If they had that information would
they be more interested and then we would not have to drive it to them.
MS. MARTIN-ANDERSON: I do not know the real answer to that but I can tell
you from looking at the literature that we looked at and at the programs that
we looked at that the message was different. It was that at the end of the day
they really care about how they can personalize the information to themselves
and that while they might interested from a societal view of what is going on
in this sort of health are ecosystem that when it came to the point of
decision, they really just wanted to know how do I use any of this information
because I want to know what are my chances of survival. What is the difference
between how I might operate? How functional I might be a month, two months,
three months, after surgery? You know what is the likelihood that I will
experience that complication? So I think it was stated best by Eva, there are
very, they have a lot of other things going on. So spending time to study the
health care system is not on their list.
MR. STUMPF: Dave Stumpf again. So I like Paul’s comments as well, but maybe
just to step back. I mean what we now do as payers is sort of viewed as managed
care and I think that at a high level we are moving to a different model. It is
not, it is like all politics are local. All care is local. You are moving I
think more to a collaborative model where the various players in this
enterprise are collaborating with each other and there is not this management
from on high and part of this is cultural too from the whole industry. I mean
physicians certainly speak very highly and being what I do too of the
relationship between the physician and the patient but at the same time they
really do not know a lot of the details about the patient that are absolutely
critical to their successful care.
One of the models that intrigued me in thinking from a big corporation
standpoint was the FedEx model: not FedEx, the Fidelity. Fidelity collects all
of this money for our 401K plans from our employers just like we might collect
all of this money from employers for health care. And yet they view you as the
individual, as their customer. So you can go to the Fidelity website and see
your account and move your money from one account to another and do things like
that. We have not had that focus on the individual as our customer and on the
payer side. I think that is something that could change.
I think that there is a couple of very high level things to think about that
are low hanging fruit and where there is experiences in other industries that
have changed things. One of those is eliminating middle men in the system.
There is a lot of waste there. Somebody mentioned catalogs of services and
things like that. It is part of the model we are looking at now too. Eighty
percent of the reimbursement in health care occurs with the very limited set of
codes. Those could be put into catalogues and instant payments retail models
and eliminate most of the clearing house functions.
If you listen to the AMA they report that about fourteen percent of a
doctor’s revenue is going into collecting fees and that translates into
billions of dollars that is going into just that administrative process. It
could be simplified.
The idea of insurance exchange is to me the attractive thing about that is
the ability to eliminate brokers. Brokers account for about five to seven
percent of health care costs and they are just merely in the middle of
connecting an employer to a insurer. The travel industry was able to move
brokers out of that role in about two years. Now most of those travel agents
are still working but they are doing it in a different context that has a lot
more value than what they did before and so I am not suggesting the brokers get
totally eliminated but that their role is refocused into and not in a role that
it is currently at.
So I think again I am not speaking on behalf of United Health Care here
because we use brokers a lot. We depend on them and this actually gets into a
third point which is how do you get from here to there. One of stories I like
to tell is I was on the hospital board at Northwestern at a time when we
rebuilding the campus and we had these old facilities and we had this new
facility that we were building. In the context of the capital and operating
budgets which I was on those committees, the question always came up is when do
we stop repairing the leak in the roof on the old building because it was
leaking. We were still admitting patients there but the roof leaked and at some
point we said no more. The same thing is true with the equipment. What do you
move? What do you scrap? What do purchase brand new? We have a lot of those
dilemmas in health care right now. We have products around quality metrics that
make a lot of money for us. They work. The industry wants them. They use them
but they are not suitable for the next generation.
So how do you, when do you stop supporting those and start migrating the
next level of things that push it to the point of care and things like that.
These are very real issues and when I start thinking about shared savings, you
know if a group can push the price down on something, you have to distribute
that to where it is actually going to be impactful. So for instance you know
the experience up in Virginia Mason when they suddenly cut out all of the
imaging for back pain. They were left with a dilemma. They had all of this
capital investment that they had to depreciate and part of the shared savings
it needs to go back into depreciating those assets so that people are motivated
to actually make the move and we haven’t thought I do not think enough about
the cost of the change process because that is a very real cost.
The emphasis I put on the content in my talk is that we do not have enough
content so that we have structured data documents for all of the things that go
on in health. I am a neurologist. I am need a template for a new migraine
patient, a new epilepsy patient, a new this, a new that. My professional
society ought to be creating those templates for me and yet the societies do
not have the resources to do that today and there is not a incentive built in
for the organizations and I am advocating now for more money going into some of
the organizations to create this kind of content because until you get them
funded and empowered to do this, you are going to have I think a real shortfall
in your ability to collect standardized data. That ability to collect
standardized data is critical for quality measures, for maintenance, for
certification processes, creating registries, doing all sorts of things. It is
a fundamental infrastructure thing we need.
DR. MIDDLETON: Go ahead, Justin.
DR. CARR: Okay, thanks. I have appreciated these fantastic presentations.
They are just tremendous. I had a couple of observations. I think that we are
living in this world of clinician generated data and we are working with the
standardized measures that we have, and that is what we are pushing out on the
websites. So we are clearly hearing that the patient wants to understand what
does it mean for me. In the spirit of grid you know I think there is almost
like a two by two kind of thing where outcomes are what physicians tell you
about on the y-axis and on the x-axis is really patient burden. So for me to
live five years with stage four whatever, I can do that but what is the burden
on me in terms of the complexity of the care, the frequency of hospitalization,
the cost of care, the likelihood of complications, and I just think that there
might an interesting model in beginning to move toward representing. The
physician is always going to be thinking in terms of outcome and mortality and
so on but overlaying with that the patient’s burden whatever that might be cost
or complications or hospitalization.
So that gets me to another observation which I think we talked about in that
May 26 summary which is just the need for education at large. We have to
educate our patients and our providers but we need to develop a new workforce
of data developers and part of that is what was just mentioned about template
creation within the specialties societies but then also looking at how do we
depict data and what do we use and how do we risk adjust and how do we make
these dashboards and I think that this is a new discipline that needs to be
part of our universities, part of NCHS and others to really develop that.
Then just a third observation as I am looking at some of our own care and
cost data at the moment it is that FTS and UHC and various aggregators have
very rich risk adjusted broad database data but it is held up with contractual
relationships and even, not necessarily able to be shared so in many ways it
has more granular data but it is held captive because of these understandable
proprietary relationships and so we default back to things that are maybe not
risk adjusted by publicly reportable. So just my three observations.
DR. TANG: I wonder if I could go back to Dave Stumpf because I thought he
said a lot of insightful things and now I am going to challenge a little for a
different relationship because he also mentioned collaboration. So one I will
remind you that payer is a middle man, too.
SPEAKER: Payers are what?
DR. TANG: Are middle people, too. So speaking somewhat categorically, in a
for profit payer you are using this data and your analytic techniques in order
to find profitable arrangements. Tell me if this is true in the health
insurance reform, in theory that should be less important because you will have
to take more risks across the board and that is where you can tell me well
actually that is not true anymore, it is not really true. But if that were true
then it seems that for you to make your margin on your piece of the
transaction, it goes back to our saying if the providers you contract with just
make more appropriate decisions, deliver better care in a cost effective way,
you still make your business and so you would be less inclined to have your
proprietary way of analyzing data and more vested in working with the
providers, collaborating with the providers and helping them get better data to
do their side of the game which is do better care in a more cost effective way.
Is that holding water at all? One, is it true that under insurance reform in
say 2018 will it be less incentive for the payers to try to differentiate
themselves among say other payers by sort of doing their proprietary data bank
and would that, could that actually translate into more collaboration on the
data analysis side with the provider?
MR. STUMPF: Yes, now I understand. It is a great question. What you are
addressing is the fact we really do not have health insurance in America. We
have prepaid health care that then gets thoroughly managed and I think what you
are describing is moving more back towards an insurance model where you have a
more predictable risk and part of the problem right now is the data is limiting
on what you can predict with risk. I mean even with the best analytics that we
can do with episode treatment groupers and quality functional outcomes with the
SS36 all of this kind of stuff you can get to a level of predicting risk. This
may be 40 percent and you miss a whole lot of detail with what you can do now.
So that is why there is a lot of emphasis on managing various aspects of the
system. The mission I think is, so what I think you are going to see as a
movement this is why I think our company happens to actually be well positioned
because we have this big payer component. We also have this big service
component and it is possible to translate a lot of the things that are going on
in the payer side which are costing money for everybody into services that are
actually revenue generating for the people who are able to provide the
So if you can take what now might be a report card to try to hang up and say
this guy can do better and instead you turn that into services where you are
delivering analytics and data and things to the point of care where they are
actually in the beginning at the point of care making the right decision, you
do not then have to manage it on the other side. So I think in addition to kind
of merge workloads and get this data repurposed you want to try to convert what
are now cost centers into revenue centers on the other side and to get people
delivering more services to the point of care instead of trying to manage it
from on high.
DR. TANG: And will reform move in that direction?
MR. STUMPF: Well yes I think it will. I think there are some things that
could be done to stimulate that though myself. For instance, right now the
payers do tend to get viewed as the bad guy here, the heavy hand. You have to
ask yourself what are their incentives.
Now there is always the capitalistic ones but there are other things that
can be put in there so for instance you could think about a tax incentive for a
payer like we heard here who is paying 40 percent of their money is being paid
for quality. That could be incentivized to that payer so you know you are in a
tax free state but much of the United States is not and a lot of health care,
well their health systems are nonprofits in Minnesota.
That is kind of unique market but in the for profit areas where that takes
care of a lot of costly care, there could be incentives actually that are
positive incentives to move payers to a different model. If you were paying,
you know set a target, 20 percent of all your things based on outcome metrics
there is an incentive for you financially to do that.
MR. QUINN: To build on some of your discussion as well and an appreciation.
It is so much that collaborative care is important and we would certainly hear
about medical homes and ACOs in the model, there is sort of decision or maybe
it is not a tradeoff but there is certainly x amount of resources and effort
that can be spent in the next two to five years on improving measurement and
one set of activities could be focused on the health care enterprise and
improving measurement capabilities and measurement validity for health systems,
individual providers and then there is another set that is not you know there
is definitely overlap here but one that I would argue will take concerted
effort and heavy lifting, cultural, political, technological, etcetera, that is
actually connecting communities so that we have good measures of coordination
of care, connecting so that we can look at attribution of care and frankly, an
ACO that includes more than one hospital system or will require that. How
should we balance those and is putting too much effort in the latter going to
do the ability to accomplish the former?
MR. STUMPF: In my opinion, yes. I mean we can invent quality metrics until
the cows come home but is how you implement them that is going to be the most
important thing which is why I put the emphasis on sort of this business
process management and the coordination of care model which gives you
accountabilities and traceable information that allows you to measure the
quality. I mean I think if you create that ambiance or that environment people
will become creative with the quality metrics that they are doing and there
will be a definite stimulation oh this is what we are trying to improve.
So yes, I mean the current pipeline for quality measures is very cumbersome
and it takes years to get them all down through the system. We need something
that is a production environment that is much quicker so that you have this
iPod I have got and you just download the app. This is app to improve the
quality of diabetic foot exams. People will take Blackford’s example. People
will get very creative on how they begin to create and gather that data. That
can then be plugged in as the one of the outcome applications for diabetics. So
I think the infrastructure and how you are going to orchestrate that is very
important to quality improvement. I know people take issue with well it is not
about process improvement but if the process improvement is necessary even to
measure the quality because right now the whole pipeline is very cumbersome.
DR. MIDDLETON: You know David one thing following up your thought I guess
what we did not talk about in this panel or this committee, this set of
hearings was really kind of we talk about roadmap and kind of development
agenda and perhaps the target areas, etcetera. We did not talk about
implementation at all. I think that is okay but many interesting and thorny
issues will become apparent.
When we think about implementation of quality measures writ large whether it
is local, regional, national, the production life cycle, implementation,
methodology obviously with all of the details and then the measurement and
assessment process for the whatever the implementation is, I have a particular
interest in this because I think decision support is obviously the front end of
quality in one way and we are conducting as some of you know some intriguing
experiments about distributing decision support services via the web to anyone
who is interested to publicly subscribe.
Could the same thing happen with quality? Could quality measures be
distributed? Could assessments be done in a distributed way that might help
establish at least the first order of business case for quality is how to get
it done? The second order of business case for quality though I think remains
challenging because one person’s savings is another person’s loss in the
classical dilemma and the patient perhaps is not as sensitive to that as much
yet under varying degrees of out of pocket expenditure, etcetera, but we do see
a trend in it sometimes on sharing of data for quality assessment to view that
patient then, actually a transparent view across providers. So some have
overcome this in very interesting and creative ways but that is the second
thing we did not talk about in this agenda was sort of the explicit business
MR. STUMPF: Quality measures I view as decision support tools. They are as
CDS said, one mechanism of clinical decision support. In this set of
circumstances this is the right decision to make. I like what you are saying
about it is one subcomponent of clinical decision support. Most clinical
decision support does not have evidence behind it. It has got consensus and
business rules and experience, et cetera.
The more we can encode that the better, but even more and it is better
because having encoded it we now can collect data systematically about how that
decision was done and eventually that becomes evidence. That is really I think
the basis for comparative effectiveness for search Q. If you get these various
decisions encoded and one doc is going in one direction and another one is
going another one, how does that node influence the outcome and then you begin
to generate evidence from the standard workflow of a doctor and you create a
learning environment which is exactly where I think we want to be.
MS. MARTIN-ANDERSON: This fits with one of my sort of takeaways of what I
have learned through this past month process and one of the main issues was
where can measures serve to provide the link between assessment of performance
and behavior change whether it be at the clinician or a patient level. It
strikes me, if you think about measures from the perspective – certain types of
measures make obvious links.
So for instance if you think about outcomes, the risk adjustments that you
do to get assessments if you turn it on your its ear it is risk assessment that
can be used at the clinician patient together or consumer alone level to try to
face what about me. If you plug in the same parameters and it can tell you what
is your risk for some are hard, some are easier. Then if you also think about
process composites and if thought about process composites to include not just
what the clinician does but what the patient does.
You have got a broader process composite and you understand which we do not
yet, the link between that process composite and outcome, and the contribution
of each element. If I stop smoking or if you check my eye exam, right? What is
the likelihood that we are going to get improvement in outcomes and you could
also create a calculator around that and say well if I stop smoking what would
happen or if I actually went and got my lipid for my hemoglobin A1C went down
what would that be the outcome.
Then you can also do that on treatment options where you begin to connect
the two. That was one of the aha’s that I had, is I do not think we think
enough about taking measurement, which measures are actually effective in
behavior change and which measures just are not effective because in and of
themselves discharge instructions. You have to take that further than just that
piece of paper because you are going to anything out of that.
The other aha that I had was the meaning of health IT sensitive has become
much bigger for me. I used to think of health IT sensitive in terms of measures
of could you reduce the burden of creating it. So can we get –
SPEAKER: That is enabled.
MS. MARTIN-ANDERSON: Okay so I might not get the right words. I am pointing,
when I think about health IT enable or I do not know what health IT sensitive
means so I am probably thinking on the enable side.
DR. TANG: So let me give you the common law definition of those two terms
and see which one fits. So HIT enabled means that it makes it easier to get or
easier to capture, to retrieve or to capture. So that is enabling. HIT
sensitive means that there is literature that says the presence and effective
use of this HIT solution, CDS reminds me has an impact on the clinical things
you are trying to measure. So I think what you are talking about is HIT
MS. MARTIN-ANDERSON: Right so yes, I think now what is happening for me is
getting the full breadth of HIT enabled and sensitive begins to make more sense
to me in the sense that it is not just, that we have to really think about the
consumer level and the provider level levers that you get through that
technology beyond just the measure itself. I think that is the question is that
both sensitive and enabled in order for it to cross the threshold to really
push. And I was also thinking about the fact that right now we talk a lot about
burden reduction through health IT. But a think we need to talk a whole lot
more about burden reduction around all of the patient education and everything
else that needs to go around the interventions that could happen because of the
health IT which is bigger than an EMR.
Then the last one, the last sort of ah-hah and Paul agree with what you said
in the beginning about providers and consumers. The area that I would like to
see payers and purchasers think about more is the coverage decision element of
how does that really impact. That is one of the ways that they actually pulling
levers that makes a difference in the outcome and how do we articulate what
that might be.
DR. TANG: Of course I will add on to our idea. So if we could make a
composite of a different kind. It is composed of things that the health care
professional have a lot of influence over. If I do not want it to test, can’t
go get it and there are things that user patient have a primary responsibility.
If we could allocate responsibilities and we design the benefits and some of
this is already being done to influencing both of those components to a quote
quality measure or an outcome measure more or less, then we can get everybody
pushing in the right direction and have it aligned with what a patient might
guess is oh, I have a better quality of life and I have a better health status.
That is what I always wanted in the end anyway, but none of these other things
told me that.
MR. STUMPF: The example you gave is something I think that is doable. One of
the clinical decisions support things which is consensus based is
appropriateness criteria and appropriateness criteria can be bumped up against
medical policy and payers and automate that process so this a good case where
the doctor is using clinical decision support something they want and at the
same time is feeding an administrative task that they do not like. So that is
something that we are going to be over at the American College of Cardiology
tomorrow talking about that very thing. How can we mesh what you do with
appropriateness criteria with medical policy to begin to automate the process.
DR. MIDDLETON: Larry.
DR. GREEN: Borrowing from Kristine about her ah-hah moment my ah-hah moment
of the hearing, I want to see if I can give voice to and particularly again
invite our testifiers to react to it. It would go something like this. The
existing health policy of the land, fundamentally is the market will decide
what is best. I am hearing in this hearing in the testifiers that we are
questioning that fundamental highly American principle that it is heretical to
question. When I listen to various presentations earlier this morning laced
throughout it in its ability over a four or five year period to make a
difference was a great deal of cooperation and it was in a specific community.
The New England Journal article two weeks ago about Grand Junction, the
fundamental lessons wrung out of that is what happened here was when the payer,
the IPA, the hospital, the specialist, primary care folks, mental health folks,
the social workers decided that maybe they should work together, they improve
results and reduce their costs. This suggests that in the last twenty minutes
which I basically just loved listening to you guys, you pointed out how
important collaboration is going to be for us to get to where we want to go and
we carefully extended the notion of collaboration to be all the way to the
person who dares become a patient as part of that collaborative model.
So my aha goes like this. We may be stuck in developing quality measures in
our quality roadmap because we have not yet agreed who we wish to advantage by
the creation of the quality measures. It seems to me that this is very
important. The purpose to which we want a quality roadmap probably should be
stated and made explicit and I do not have this quite right I am sure but
listening to David two or actually three times now made comments, it may be
that we have to decide that we want a quality roadmap that advantages patients
instead of where we have been in my opinion where we have been looking for
quality measures that would advantage me whether I am a broker or whether I am
a hospital director or I am a cardiologist or I am a pediatrician. What will
the quality measure do to my piece of the industry and the enterprise and I am
for it if it makes me look good and advantages me in a payment system where I
get paid more because I did that and I am against it if it disadvantages me.
This is I think quite fundamental.
Yesterday afternoon when Kevin was here and was talking about what was going
on at the American Board of Medical Specialties this really bothered me last
night when I started thinking about what we had heard. His presentation and
also the lovely lady from ABIM, they were showing us and we saw it again in
your slides, Barry about the community health center that there is a thing
grounded and taking care of sick people, or people who fear being sick and are
promoting health that is fundamentally based on professionalism in which we do
things that benefit others even if they do not benefit us. We saw glimpses of
the medical specialty boards flirting with actually staking out some
professional territory. We saw in this hearing some examples of where absent
pay for performance just because it was their mission and what they were
supposed to do in their town, community health centers went from the bottom all
the way to the top without ever getting pay for performance dollar out of their
work, if I heard you right. I think this is Committee’s work around this and I
would like to frame it as a question Blackford and Paul. We should be explicit
about for what advantage will quality measures be used.
MR. STUMPF: Well you have raised another important issue which is who are
you measuring and the emphasis has been on providers whether they are
hospitals, physicians or whatever but a lot of health care is delivered today
by non-physicians and in fact I think a lot of what is in fact evidence based
medicine you know use Porter’s term, precision medicine is better done by
non-physicians yet we are not measuring them.
The data that we have now coming in even administratively is pretty rich in
non-physician data but we have not used to get to say you know these nurse
practitioners and nurse anesthetists but we have a nurse anesthetist actually.
It was really controversial. We got ourselves in trouble with the American
Association of Anesthesiologists over what it looked like. What you are trying
to measure is precision medicine, the evidence based kind of stuff and most of
that is not done by physicians and that is an important group to get a handle
DR. MIDDLETON: I am going to give Paul the closing comment because we could
go on for a long time. This is very rich but we do have to move on.
DR. TANG: Well I think I am going to move something that Larry and Kristine
said. I mean basically I would move that this quality roadmap be designed for
the dyad of the patient and the professional health team, and that picking up
on a different thing that Kristine said that we attempt to portion some gross
very qualitative assignment of responsibility and a lot of it is going to be
joint so that we can help educate all of us about our respective
responsibilities on this journey for an individual’s health.
Then in the future it could be used by benefit design folks to help even
strengthen that responsibility and the rewards and incentives for taking action
there. So just like that pyramid of the MOC and of what you know and getting
into the action, a patient has the same thing.
Now we, the people who are paid to do this stuff should be helping to
educate the patient on their triangle just like the professional associations
are trying to measure this. So I think there is just a new theory of quality
measures and the subsequent use of it you know whether it is in rewards or
measurement, whatever it is but the notion is we have got this new dyad that
happens by the way to be empowered because of HIT.
None of us have any information which is why we have gotten down this
roadmap on ridiculous measures is because we had no access to the data. None of
us so HIT has been enabling but what we should do to take the ball and run with
it is to create a whole new purpose for this stuff and who has benefited from
it. Who is the beneficiary which means who do you pay attention to in designing
the new stuff then let the other stuff follow in. Talk about ahas. That is the
thing, the aha got to me in the entire two days. It is a yellow brick road. It
really is opens eyes in terms of this could be a whole new place.
DR. MIDDLETON: So on that high note with that passion, I agree with Paul. It
is a change and I want to thank everybody for taking the time to come and share
your wisdom and expertise and experiences and insights. Members of the
Committee and Subcommittee and support, everybody, thanks so much and we are
(Whereupon, a luncheon recess was taken.)
A F T E R N O O N S E S S I O N
DR. MIDDLETON: Well, good afternoon. Welcome back from the lunch break. This
is the reconvening of the Subcommittee on Quality of the National Committee for
Vital Health Statistics. The next agenda item is to work with the Booz Allen
contractors to begin to draft the final product of our letter to the Secretary
and I think that work has begun and why don’t we just it from the top with
Kristine, whenever you are ready.
Is anyone dialing in from the Subcommittee on the phone? Is anyone dialing
in on the phone?
DR. CARR: Blackford can you hear me? It is Justine.
DR. MIDDLETON: Hey, Justine welcome back.
SPEAKER: Are we going out on the internet?
DR. MIDDLETON: No.
MS. JACKSON: Yes, we are. This is an open meeting.
DR. MIDDLETON: Oh, okay then we still are open to the public as well?
MS. JACKSON: That is right.
DR. MIDDLETON: Welcome any observers. So as soon as we are ready, Kristine.
Or we could talk further about feelings, okay.
Well I guess maybe the first thing to do is just to begin to think about the
structure of the letter and what are our essential components and reviewing
past letters. I haven’t had a chance to review past quality letters but if
there were any in it might be fun to share with the Quality Committee or just
give us those pointers and share with the contractors as well.
MS. MARTIN-ANDERSON: Actually we did. We looked at the past two letters and
then also used the structure of the last letter. I think it is the May letter
to set up the outline for how to write this letter.
DR. MIDDLETON: May is the May information one.
MS. MARTIN-ANDERSON: Just for this outline, right. Isn’t it the last one
that you did right, May?
DR. MIDDLETON: Was that the meaningful measure one?
DR. CARR: This is Justine. I think what we generally did was figure out what
our recommendations were and then develop them and then there is some kind of
customary opening paragraph about NCVHS or this or that.
MS. JACKSON: I can add to that. Thanks, Justine. This is Debbie and that is
kind of the and just when you look at the end of the last after the morning
session when there was so much on the table and we were going in dimensions of
time and space and trying to identify, codify what you want to tackle and what
message you want to make, what story and what journey. There is a lot going out
there so that is how we have seen the best way to handle that just crystallize
as things were kind of coming to their own kind of natural crystallization at
the end of the morning session and to build from there and then you kind of
fill in the pieces.
DR. CARR: I think the last time when we wrote last year’s meaningful
measurement letter, we wrote about what we didn’t hear; not what we did hear.
That was the most powerful thing.
DR. TANG: Okay we are not going to compose the thing in front of us, right?
We are just going to do some outlines?
MS. MARTIN-ANDERSON: Yes, we are going to show the outline up there and then
we can talk about each section. That might be easiest and she is going to real
time take downs so we can write.
DR. MIDDLETON: Okay so maybe just by way of getting a little rhythm going, I
think Justine just pointed out something which I think is very important to do.
Somewhere near the beginning of the letter is to frame what the letter is about
and what it is not about and we had some discussion about that at the end of
the morning. It is not about a business case for quality. It is not about sort
of technical implementation details for quality. What it is about is a quality
measures roadmap from these four different perspectives. Does that sound like
kind a set up in general.
Maybe before we get too much into the weeds we might want to just step back
to a thousand feet and say and Paul started to do this and I have a few
thoughts, what are the overarching recommendations that we think are the key
themes coming out of this. Maybe we should just start there. I think Justine
was suggesting that as well, or Debbie.
DR. CARR: I think that helps a lot and then I will make sure that we all
agree on what we are going to say, what we heard or what we want to put
DR. MIDDLETON: So the sea change, the last moment from the morning session
was the epiphany reached when we recognized that quality is being turned upside
down and we are going to take a patient perspective and quality will be patient
centered from here henceforth, evermore.
DR. TANG: I wouldn’t say it that way.
DR. MIDDLETON: We are just brainstorming, Paul. We are not wordsmithing.
DR. FITZMAURICE: Is this the time for an Amen?
DR. MIDDLETON: Other big themes?
DR. GREEN: Blackford, I think one of the framing big theme issues for the
letter we heard in our conversation and also in testimony was the claim that
there is a crossroads that is occurring because of the intersection of
meaningful use, ACOs, PCMH, the quality movement. There is a lot of fluidity
and a lot of things going on all at once now that answers the why now question
about this letter. We need to do this letter now because of some reason. I
think we heard testimony that supports the notion that it is because new
opportunities are here that are very, very important and are achievable now.
That would set up a conclusion but we do not seem to have a clear vision of
where we are headed. We need a roadmap for that.
DR. MIDDLETON: We can sort of play off the why now, who cares, why me. Who
cares? Stakeholders. We have extraordinary pressure on cost and value of health
care and that is not going to abate. It is going to get worse. We need to be
able to demonstrate value and produce quality, etcetera.
DR. CARR: Right and I think actually following on that we have invested
gigantic amounts of money in this transformation and it has got to come out as
quality access and costs and how does that happen. So we have the building
blocks but just sort of taking and tying exactly the operationalizing it.
Actually you said that at the end, the operationalizing of this is a big thing.
DR. MIDDLETON: Another parenthetical comment perhaps not yet for typing is I
worry a little bit about the fact that we have not benched the quality curve or
the cost curve heretofore. We really have seen only marginal improvements in
quality with the exceptions of course none of you but in general despite a big
investment in quality itself it has been difficult to demonstrate improvements
in quality so I am not sure if that means there is something about how we are
measuring quality or what is being measured that needs to be considered here.
This is still in the brainstorm mode.
DR. CARR: And I also think the measures that we do have of patient
experience of care you know did I understand what I was told. Did I understand,
did I have everything I needed when I left? Did I understand the nurses and the
doctors that you know that the scores are not high. Unlike the core measures
where 100 percent of people get an aspirin on admission for acute MI, the
measures around the patient experience of care are nowhere near that high even
in the best places.
DR. GREEN: I want to build off of what I heard Paul say just before our
lunch when you used the word three or four times about the dyad and supporting
the dyad. That in my mind Paul linked to David Stumpf’s comment about that many
of the currently used products of the health care industry are not suitable for
the coming system. That is because so many of the products are provider centric
and designed to meet their needs and so when we get to your dyad it seems to me
the letter is positioned to call for some transformative approaches to the
development and implementation of quality. Help me out here. I am trying to
build off of what you were saying.
DR. TANG: No, I think that is right and if I look at the work we have so far
before the dyad, it is the same old, same old. That is what doesn’t seem very
transformative. So I think going back to whose behavior are we trying to
influence and there is two. One of them we always pick on which is the
physician and we are expanding that to the health care team and the other we
paid so little attention to and that is the consumer patient. So that is a new
theme. So we expanded it to health care team and we are expanding it to the
other part of the dyad which is the consumer patient. I think that is a big
difference and it goes back to what you are saying of we have always written
everything and targeted everything towards the provider and that is not getting
us anywhere, because it is more than being quote patient centric. One, it is
just so overused and two, you could describe all of our measures as having to
do something with the patient but that is not what we mean. It is really focus
on the behaviors of the two parties of that dyad. And all of a sudden you want
to measure different things. You want different contribution like consumer
generated information you are going to look at all in all because there is a
guide that the health care team wants to know that. So it just sort of forms a
whole new paradigm for how to use this stuff.
SPEAKER: And a shared accountability.
DR. TANG: And the accountability piece. Any easier word than accountability
would be responsibility at this point because it just; it is a more humanistic
and holistic way of looking at it mainly because accountable is so laden with
blame and penalties. We are trying to introduce the patient and the rest of the
health care team who we ignored up to this point and we all ran into
accountability. I am playing with the words here.
DR. MIDDLETON: In a way that I do not want to lose the attachment to what is
in the current vernacular about accountability but I accept a similar point of
DR. TANG: Maybe we could do it the other way around so shared
responsibility, open paren and accountability. Do you see what I am saying. It
is just a different frame.
DR. GREEN: It is a accountable and responsible for what?
DR. TANG: For behaviors associated with improving the individual’s health.
So as the health care professional I have some responsibility and
accountability to act in certain ways. It can be an educator. It could be
someone who orders tests. It could be someone who performs procedures. As the
patient I have some responsibility to learn about things, to modify some of my
behaviors where with the assistance of other parties you know in the whole
DR. GREEN: For me what helps with the accountability responsibility thing is
the prior question about the purpose of the quality road mapping is to work
with the dyad to advantage the patient and share responsibility for advantaging
the patient. Advantage does not seem like such a loaded word to me. The
accountability thing and the patient centered phraseology and responsibility
and personal responsibility we just flaunt those words to death and couldn’t
DR. CARR: So in this letter, are we setting the background for this as a sea
change of where we are today as the measurement needs to now expand to
encompass whatever. Is this a letter about measurement ultimately?
SPEAKER: Yes, yes, yes.
DR. MIDDLETON: Yes, and I think even if we do not use the words put sea
change somewhere at the top because we want to underscore this and make it
almost as paradigmatic a change as we can possibly make it because if we do not
write it that way, it is not going to be.
MS. JACKSON: Sea, S-E-A.
DR. MIDDLETON: Thank you.
SPEAKER: Actually both words could probably work.
MR. QUINN: That was on of the questions that I had is that one side of this,
one of the balances here is that there are paths of less resistance and then
there are paths of transformational change and what we saw from the meaning
measures as well as what we heard is that there are pockets of excellence and
the question how do you develop a plan that is implementable not just in one of
three places but is something that can be adopted very widely. We heard some
examples of that today and so people are thinking about that. The further and
this is a little bit of a different way of thinking about it. Those who are
probably most excellent in the current paradigm of measurement might not be
those who will be capable and excellent in measurement with A different
SPEAKERS: That’s right, right, yes.
DR. MIDDLETON: It is a good point in itself, Matt.
MR. QUINN: There is a lot of investment in the current but if the current
changes, well you got an A all of the way through middle school but now in high
school. You know after the sea change.
DR. MIDDLETON: All of a sudden it is going to be harder grading. What is
that really going to mean? Part of me thinks too, again brainstorming more for
a second, we kind of maybe at some level want to address the fundamental values
for which quality is going to be assessed in this country. It is challenging
given that responsibility is shared between provider and patient but you know
the values with which any benefit assessment are always intrinsic to whomever
stakeholder perspective you are taking, right? So if you do not know the
stakeholder perspective you are taking and you do not know the value structure
with which the assessment us being made from a very traditional sense. So I
wonder if you just want to talk about that, or not but whether we need to
address kind of a values for quality or the values for measurement.
MR. QUINN: Maybe the title for this is the patient centered quality roadmap.
DR. MIDDLETON: Before we get to title, let’s flush it out more. I like dyad
because it is doctor patient. It has a lot of resonance in lots of different
ways as provider or patient but I also think we heard stuff about a broader
perspective quality and patient provider system and really community or we did
not really hear public health but community. We did hear some about all of
those other dimensions too and particularly other dimensions at the provider
side of it you know the provider assessment and knowledge base actions and,
etcetera. On the systems side sort of a traditional structural measures of
system components that influence quality. This is where I think we are going to
see a new world here partly too based on testimony and my own opinion. For
example the NCQA structural measures for the patients at a mental home will be
used to kind of assess patient centered mental homes but that is just the
beginning. There will be other types of structural HIT and non-HIT measures I
think, too. What do you think?
MS. MARTIN-ANDERSON: I was just thinking about one of the concrete ways that
influences around how many just selected for use. What I heard over the last
two days is perhaps we have the wrong criteria around importance to measure and
relevance to measure because that criteria is largely interpreted by the
purchaser. And if you were to interpret it by the consumer or by the dyad,
importance and relevance would have to truly have to do with improvability and
behavior change and it would no longer be as important to measure something
that is really narrow that lets you stratify the provider as it would be
measure something that you could work with the patient to change.
DR. MIDDLETON: I like it.
DR. TANG: That has huge ramifications on the whole endorsement process that
currently exits. That is just a comment.
DR. MIDDLETON: We are going to get there.
DR. GREEN: Let me build off of Kristine’s comment and ask Justine. Justine
are you there?
DR. CARR: I am here.
DR. GREEN: Would really like your insight and advice about whether or not
this letter could take a position akin to the NCVHS position around the NHIN.
Can this letter be a framing letter for it is a vision of the way quality
measurement ought to be in the new world as opposed to the way it was and be
DR. CARR: You know I am new at this but I would say that we when we do that
we are appreciated. I have gotten feedback on a number of things that even the
letter from last year when we kind of put out there you know what is the next
thing. I think that is what we need to do so I do not see a prohibition to
doing in this. I think that was the point of it that we would try to envision.
The one thing I would say is for us to keep a little bit in mind the discussion
that we had, the idea that the Population Committee brought forward. Now,
Blackford did not hear that because came up in the executive session but as we
speak about this, we ought to be keeping in the back of our minds a broader
perspective that goes back to that earlier NCVHS diagram of the role of
community, and that addresses population health and the influences on health
that are outside the health care system. I do not know if that is too much to
think about with this but to answer your question, yes I think it is good for
us to vision and I think we should keep in mind our population health
perspective in some way.
DR. TANG: Another one topic that we talked about. I do not know that we have
discussed this for a Subcommittee opinion is the whole notion of risk
adjustment versus risk stratification. I think risk adjustment turns all
numbers into bland ones and wipes out the information. I think the risk
stratification is one way that first thinking about the patient consumer can
start to find themselves. That is one good point and another one is for the
health care team to figure out where they can spend more of their effort and
time. So that is really another kind of a sea change.
DR. CARR: You know it is tying on a little bit of what we heard yesterday
that risk stratification may be your condition of it may be your environment.
If you are very low income and you have a very high deductible, you are also in
a risk. You may have constraints on accessing care even though you have
DR. MIDDLETON: This one I think is the patient’s like me one. To get
measurement framed from the patient’s perspective so it is not the patient
belonging to a cohort. It is the evidence synthesized from the patient’s
perspective so patients like me with this therapy can experience having this.
DR. TANG: I think it was stated patients similar to me.
DR. MIDDLETON: I know. Apologies to Jamie. We are not going to use patients
MS. MARTIN-ANDERSON: Instead of a hospital comparing nursing homes now it is
patient compare which how do you find yourself in it.
DR. MIDDLETON: Relevance, your word. The other one, I do want to come to
Arne’s comments because he was just so provocative in the sense of describing a
cascade of actions that transformed the way we do quality. We can decide I
guess to which level of detail relevant, appropriate for a letter to the
Secretary but I was just enamored or it was very appealing to me. I have to
digest it more thoroughly but I think we should return to that and think about
what is the boundary up to which we will go sort of recommending specific
actions or methods or techniques as he was outlining.
DR. TANG: So I might ask us whether we want to, I would almost think about
this as a higher level and this is our first volley. So we are cheat sheet
almost everything including the actors of the future, all of the actors in the
future world to try to get specific seems like it is more than we actually did
due diligence on for one and it may be too much to have in the letter.
MR. QUINN: This is a reframing. We are talking about a reframing of the
quality measurement framework. In order for this to have in impact we have got
to discuss some of the consequences if we follow the status quo and keep on the
same course so that would be something else that we.
DR. TANG: There is a good point and maybe that has developed, it is a
rationale that maybe leads up to, well let us see.
DR. CARR: Could you repeat that, Paul? Somehow it just went dead.
DR. TANG: Well Matt was talking about somewhere along the way we have to
compare to what is going to happen if we do not change our quality measurement
paradigm. So I am trying to figure out whether that leads to the summary.
DR. MIDDLETON: I think it is motivation and rationale. I mean in a way what
we know is that the current quality measurement process is insensitive to a lot
of what really matters is important or relevant. So that is what we are trying
to fix. And the secondly is this new perspective sea change. Paul, to your
point about sort of the first volley, if it is one volley of two, three, four,
whomever maybe that is what we need to describe to is a roadmap for the
roadmap. This could be the first sort of reframing we expect in a year. Not too
long, but in a year to talk about implementation details. That would be pure
roadmap and then a year later talk about assessing the implementation of policy
changes needed. The thing that I might differ a little bit though, I think Arne
he really took the message. He had a four year horizon for these changes.
DR. TANG: But they were specific before knowing necessarily. I mean he was
not aware of what we were talking about right now because he was shooting
towards standard this and standard that. You have to figure out what the that
is. Listen to that we are describing different this’s and that’s.
DR. MIDDLETON: He didn’t name the this and that, though.
DR. TANG: He said there should be so there still is the naming the actual
DR. MIDDLETON: It is not an implementation. It is a recommendation.
DR. GREEN: I just re-read Arne’s stuff. I wrote it down. I think Arne’s
contributions align with what Blackford is saying. I do not think it takes in a
different direction but he just starts getting more concrete. It is very good
actually I think we need to give the letters some clout. I mean it is not just
your outcomes starts moving towards the sorts of thing. It talked about the
importance of pay for vital services connects to what you have got to do. You
have to be able to find those data from all of those different sources and
bundle them up. You have to find the patients that are at risk or those that
are expensive in the registry. Those are the sorts of things.
DR. TANG: To me it actually paints it in a payer framework perspective when
we are trying to actually focus. I agree you can but the way it is written is
it perspective? Oh yes, in the brain. And one of the possible directions we are
going is we are going to focus on the first two: the dyad. The other things
will fall from that.
DR. MIDDLETON: Okay so I think we had testimony from four stakeholders and
to leave two off and only focus on provider and consumer I think we are doing
them a disservice. We invited them here. We heard them.
DR. TANG: No, no, no. I mean I did put the motion put forward was okay so
their personality is we can either worry about costs and separately we are
about quality or you can take the perspective you know what if you make the
appropriate decisions you will get both cost and quality and cost
effectiveness. Okay? So that was the analogy. So in this world we can either
design in the extreme we could design measures for each one of the stakeholder
groups however many there are or we can figure out what are the ones that
matter. That was the people that matter. People that matter are the consumer
patient and the professional sides they are trying to influence to provide
services and advice or whatever it is and then the others, which I think the
purchasers and the creditors should be able to get their information as a
by-product. So that is just a theory so I am saying that to not ignore but that
the others are a consequence of the first two.
MR. QUINN: There is an evolving role of purchasers and payers in that dyad
DR. MIDDLETON: Thank you Matt.
MR. QUINN: I was going to say that a lot of the testimony that we heard from
the payer and purchaser group and the accreditation groups was around their
influence or how to shape that dyad interaction and so we heard United talking
about how payer data could help inform both sides of that and we heard about
how in Minnesota they are bringing data from different organizations and
bringing it together to do that and then how the SGIM and other groups are
trying to influence the position of quality measurement activities. So those
organizations have their own data needs but I think they are focused on how to
impact the patient clinician dyad.
DR. TANG: Actually Dave Stumpf gave a concrete example saying in United they
have two lines of business. They have more than two but primarily it was the
payer side and that is where they made their money. In this new world we focus
primarily on the dyad, they could move over to the service business and they
had a payer function. In a sense that is what the agreement really was saying.
The payer perspective is subsumed or subservient to the patient consumer
DR. MIDDLETON: I am still struggling a little bit because I think there are
already payer source of data if you will that we really talked about trying to
look at this in a sort of in a technology neutral way and perhaps data reuse,
reuse, reuse. If we think about getting data from where it is to where it needs
to be, payers have data that is relevant to that provider intervention or the
consumer intervention. So maybe we are limited by this stakeholder perspective
and we should really be looking at it from a quality perspective sort of first
Yes, framing it as just an dyad is an acceptable one alternative approach. I
just worry thought that focusing only there will not get at the value question.
The value question comes from the payer side. I mean creditors, I think we
heard some extraordinarily interesting testimony about quality assessment of
providers themselves. You know how do we measure providers and what data from
the environment under the relevant to accreditation process.
DR. GREEN: What would you think about, I mentioned moving to reconciliation
mode here. As I listen to you guys just try this out. This may totally misfire
and be useless. I think maybe we should conclude that quality is about finding
value and that may mean that quality is everyone’s business. All four groups we
heard from have got to get serious about quality. If you pick out a dyad, pick
out two of them, it is not going to work. Quality over cost. All four
testifiers, the way the hearing was organized. Maybe we should just find that
none of them get a buy on this. It is going to take all four groups if we are
going to envision this new world of quality measurement and advantaging
patients through it. Okay say that was not helpful and we will go back.
DR. MIDDLETON: It was helpful. You got me. I guess I am not the lumber. I
think, now I know my role.
DR. TANG: Oh, man just team me a counter replacement.
DR. CARR: I heard that.
DR. TANG: In its corporate and your perspective so just because someone
performs a payer function has data doesn’t mean it is not important to the
perspective of the consumer. Does the consumer care about costs? Yes. So there
is nothing to exclude them. It is a refocusing of the perspective and the
consequence to the framework that talks about what it means. What does cost
mean to the consumer but not to the payer.
DR. MIDDLETON: But in fact patients are mostly insensitive to real costs and
number two, payers’ assessments of value goes way down into if the patient is
actually interested in. They are interested in a value conversation with the
provider. The patient is almost secondary. A payer provider conversation is
about I am going to give you x amount dollars for a CHF care in your hospital
and with the readmits, etcetera. You are going to have to produce care at this
value level and you need to measure that.
DR. TANG: The consumer is eventually going to feel that and within the next
years the consumer will make these decisions. The other party that is going to
feel that is the other side of the dyad. So the fact that there is not enough
money to keep escalating the amount that providers are paid makes the provider
sector interested as well. There is nobody that does not get interested when
society’s resources are limited. So I do not see anybody that of these two
dyads if we are going to focus on them that would not be interested in societal
DR. MIDDLETON: I do not understand. Am I ever going to pay out of pocket for
my entire CHF hospitalization? You said in a few years I would. I don’t see
DR. TANG: In a few years you will be sensitive to what it costs to take care
of your CHF.
DR. MIDDLETON: Americans already pay more of pocket than any other
industrialized country in the world. We are not going to increase out of pocket
payments in America. It is just not going to happen.
DR. GREEN: Okay I want to try this again. I do not understand why we need.
We have good testimony. I mean it was a rich hearing. We learned signing. We
learned something about all four of these but why would we want to frame the
letter in a way that sort of to find out of our letter of what we learned from
any one of the groups. It seems to me we should frame the letter to be position
to be inclusive. We are going to have a findings section too and we have time.
From my notes I feel like we have good interesting findings from all four of
these groups so I do not see why we would want to frame the letter just around
two of them.
DR. TANG: Because you are not a leopard(?).
DR. MIDDLETON: So we may have to make a directional call here and make a
vote of the Subcommittee which would be totally fair or a straw poll. And
further more, we can do some drafting Paul and see how it looks like. I think
your focus area is right on target because the action is going to be in the
dyad but if we fail, so you heard it already, I think we should include the
secondary dimensions both from the creditor and from the payer and we can make
DR. TANG: That is all I am saying.
DR. GREEN: Wait, wait record it. We heard it.
DR. TANG: Throw away the other pages, you just put it in from your
perspective of how it impacts.
DR. MIDDLETON: Yes, the only thing we still have to see how it unfolds is I
think there will be independent dimensions of the perspective from payer and
creditor that do not pertain to the value specifically. I mean so we will just
have to unfold it.
So now that we got over this temporary roadblock, what are the other key big
chunks of recommendations? Did we capture something Matt about you know the
physician assessment as a measure of quality? I cannot remember.
MR. QUINN: Kevin Weiss.
DR. MIDDLETON: Kevin Weiss and Lipner. That I thought was interesting, new
MR. QUINN: I would almost frame the Kevin Weiss and SGIM testimony is that
there are levers that are pushing that aspect of it. Less says that that is
more of the part of the why rather than exactly the what and between the two
groups that have hard from and the joint commission they have a little bit
different philosophies on low bar, high bar, exactly what to measure. They just
want to see activity being done and I think that if quality measurement was
reframed that they would follow suit. But the important thing is the mechanism,
the lever that this is part of being an accredited physician.
DR. GREEN: I agree with Matt’s last comment. The sea change from that part
of our hearing is that certifying boards for this nations physicians, have all
gone into the quality business. That is the statement. They used to give tests
to protect turf. They are all moving into the quality business. That is part of
this disruption, this liquidity stuff. They are actually thinking that maybe
they have something they each participate in a meaningful way. They are
connected up with meaningful use already. They are already connected up with
CMS around PQRI. They are changing their modus operandi. We detect that as a
significant development for Paul’s dyad. The last statement about the lever, we
should not lose sight of the lever. When these boards defrock a physician there
is hell to pay for that position. It does not usually go well.
It usually has serious disadvantaging properties that come to bear on that
physician so they have their attention and they have a new model called
maintenance of certification in which they have at least four strategies that
they are trying to implement all designed to change practice so that it is
better. Now that is a pretty big deal in my view. Both sides of that equation;
they change their business and they have got big levers. We should notice that.
DR. MIDDLETON: And the new data types part here? You know if you are
examining the record to see what a patient provider actually did with a variety
of patients by type to attest his performance in reality that is a new data
type for accreditation. I am not suggesting we talk about that specifically but
there are new measures that can be developed based new data types such as that
will be relevant to physician attestation or physician certification. The flip
side of this one, which we have heard loud and clear too was patient
activation. I do not know where it fits in exactly but the patient activation
measure was how cool was that. Think about it in fact as a vital sign. I should
think about a Pan score so I think about what diabetes therapy or CHF’s therapy
is going to work for this patient as opposed to treating everybody the same. It
is more about care for me, about me, because of me. Whatever.
DR. TANG: So does anybody know the answer to the does it cost money to use?
SPEAKER: (inaudible off mic)
DR. TANG: Well, that stinks.
DR. MIDDLETON: Well we would have to know that if we were going to recommend
something like that.
DR. GREEN: Well that could be one of our findings. That is one of the issues
to address is the availability of the data that is proprietary. Who owns it and
just because it exists does not mean you have access to it if you needed it.
DR. FITZMAURICE: That is correct we do not want to endorse a certified
product, but it is also an impediment.
MR. QUINN: If you go back to the accrediting group I think that that is more
likely. They are not going to come up with new schemes of measurement as much
as they are going to adopt the ones that seem to be most relevant and the ones
that are going to I think the challenge for them is going to be changing the
paradigm and explaining to their member base why the paradigm is changing. I do
not think that there is for a while now we have been trying to push or they and
everybody else have been to push measures of a certain flavor and there has not
necessarily been acceptance of it so that is the difference.
If you see the dichotomy of the measures that we track internally versus the
measures that we report and it is kind of a big deal if you give them both
versus sort of going through the motions on both. I think the accreditation
boards and joint commission and others are going to have a role in helping sell
the new paradigm especially if it makes sense to their members and their
members’ patients. I do not think that they are going to be the ones out front
on it but they are going to be the ones that are, it is going to seep down and
be in their new programs, hopefully.
DR. MIDDLETON: What was the joint commissions’ lady’s name again?
MR. QUINN: Margaret VanAmringe
DR. MIDDLETON: The other dimension I think is patient provider, the dyad and
its appendage to a payer. Whatever you want to call it. But then the system, so
she had some interesting things to say about what are they thinking about in
their accountability framework for you know clinical integrity and their
processes for a system if you will quality assessment and measurement. Somehow
we have to fold that in.
MR. QUINN: I think that she talked about, that she showed this entire wheel
of different kinds of programs that they have got. Each one, I think Larry
asked, how many measures are encompassed in all of this and I do not know
probably some number of times ten to the third and we asked, I think your next
question was around how many of these are system measures or measures that go
across these and it was like zero.
DR. GREEN: She emphasized the unmet need of the accrediting organizations to
measure quality they need to be able to measure care transitions. That was what
I thought was her fundamental testimony. You must measure care transitions and
then she said, and we have zero of them.
DR. TANG: She said that transition communication was the number one safety
issue. Now the other thing she has talked about was and I did not get a chance
to ask her the definition she called of the accountability measures versus the
non-accountability measures. So accountability measures are things that are
effective like the core measures. Things that are not, she called put them in
the bucket, like smoking cessation counseling and heart failure discharge.
Another way to look at it, these are things that are being gained without any
apparent benefit. So it would be interesting to know what their criteria for
accountable levers versus non-accountable levers.
MS. MARTIN-ANDERSON: Maybe one way to frame that all together is if you
think about where Blackford started about we have not bent the quality or the
cost curve, right. And because of that, a large part because of that the
creditors and the purchasers have sort of put the screws on for measures to try
to get the attention of the providers. That was really important if you want to
say what it looked like five years ago. But now the attention is here and if
the attention is here and the dyad is actually working to get sea results that
are actually going to go in the right direction, then you do not need the same
level of attention on the creditors and payers to make it happen. They still
made a long cost. I am not clear on the value side but I think if you try to
say it is more about taking accountability for the direction of the future. If
you look at how we got here I would say you would look differently at what they
DR. TANG: The corollary to what you just said is I do not know that the
current measure developers at the ones that are going to supply the measures of
the future. That is another one that has been there disrupt a bit. That is
important. See there are so many things that are ramifications — so if we
adopt a framework or a philosophy or a paradigm it is almost worthwhile
starting in this letter to talk about the conceptual and the implications. That
is just going to take so much time to develop that. I do know if getting down
into the weeds, is something we can do in a future letter and I think we should
do more due diligence around it but I just think it has all of the implications
and I do not know whether speculating.
MR. QUINN: Lay out the three that this one is the framing document and say
let’s get our heads on straight on what quality measurement is about. The
second one is prioritization of operations.
DR. MIDDLETON: Implementations, details and implications.
DR. TANG: And what is the third one?
DR. MIDDLETON: And the third one is kind of you know a status check,
evaluation. I like that Paul. The only thing I might do in this one if we did
that part I think that there is this whole issue that you are putting your
finger on of kind of the production life cycle for measures and it is a mess.
We have heard it. Somebody said it is a mess.
DR. GREEN: And they say it has to be accelerated.
DR. TANG: Actually we said it was a mess and we said that there was no
organized way of getting it better so that is nice to build. We can remind them
at our preamble of where we came from. It says you know we tried to figure out
one of the meaningful things we have now is that well, not too many out there
and it doesn’t look like there is a optimism in that.
DR. MIDDLETON: I mean it is just like standards acceleration.
DR. CARR: But you know we also in that email I thought were very EHR centric
in measures in that figure so we may want to think about referencing going
beyond that as well.
DR. MIDDLETON: Thank you, Justine I agree. I think the other thought here is
somewhere in here we should say make illusions to there is a future world of
measures where information will be coming to inform measures from different
places. We heard about it coming from different places. Not only HIE based upon
EMR, but hospital but potentially PHR and potentially other consumer source
MR. QUINN: I think one of the questions for the next one on the operation,
the implementation and implications, we heard a lot about registries.
Registries are different from HIE an those are two different paths and
registries are different than ACOs.
DR. TANG: So the writers, where are we in terms of how much do you have and
what are the gaps you need from today’s meeting?
MS. LESIKAR: I think that what we talked about now is sort of the framing in
the beginning and I guess what I am not clear on is in terms of those specific
recommendations because I think our next step would be to take those
recommendations and then go back, something more concrete or specific, and then
go back through the testimony and find the tie together the evidence that
supports those. So at lease that is the structure of the previous letters where
it had three, four different recommendations and then we can tie the evidence
MR. QUINN: This letter we did an observation and recommendation paradigm. Is
that what you are thinking for this?
MS. LESIKAR: Yes, that is what I was thinking. I thought that flowed nicely
where you had an overview statement sort of a general themes. Here is the three
themes you know or here is the theme of this letter and then three specific
recommendations and the observations is the evidence that led to that
DR. MIDDLETON: We could take the perspective where we made recommendations
around a consumer set of issues, a provider set of issues, and other set of
issues. I am keeping it simple. You know a creditor and payer and that might
give us a little more specificity our you can use that same three and bucketize
some of these recommendations. Or we have to do some iteration by email.
MS. MARTIN-ANDERSON: Can I just add the first couple of things I heard you
all talk about that sounded like recommendations. I just want to check. One was
certainly about reemphasizing the for what. Right? The for whom that quality
measurement exists and saying in that that it was really about the purpose of
the roadmap is to work with the dyads to advantage the patients. In that case
we could talk more about how measure is tied to action and actually improve the
health of the patient actually on behalf of the provider and the patient and
that is sort of sounds like it. It is the focus of the roadmap should be in
Also talked about the lack of involvement of patients. The party observation
was the lack of involvement of patients and consumers and they need to take
something like patient activation and see it as a vital sign and use that, find
ways to connect measurement to how do I engage those patients even if their
stratification of the measures to be how you might change the providers’
behavior or in better understanding why some patients are not actually changing
behavior and how to get them there from an education point of view and you can
take it deeper than that.
The last one that I wrote down that I am not sure how deep you want to go on
is that changing the frame around risk adjustment, risk stratification, the
patient compare idea. How do you really make measurement relevant to patients.
I think it is different than inducing change. It is related but different. So
all three of them have the theme around them of this engagement but they are
Now the last one I did not really get it a recommendation I can still see
disagreement was around the role of the third parties who have their role to
oversee the health system either because they pay for it and want to see their
money well or because they accredit and therefore they are part of a body that
says it is functioning well. We talked a little bit about value and whether or
not that would fit in there or not but I still think that there needs to be
something about what do you see their role in the future to be.
DR. TANG: Let me try to re-edit your lumps. So the first was the focus on
the dyad to advantage the patient. I think that is sort of well stated and we
have all of the consumer, well a lot of the stakeholders, well particularly the
consumer and the payer actually as David mentioned that but anyway, the focus
on the dyad to advantage the patient which is a change from the physician
centric measures of the past using administrative data.
So going down the next one is talking about the dyad so the second
recommendation focuses on producing credible measures that assist physicians in
measuring and improving their practicing, the care team. Exactly, no that is a
good one. So we can go into the implications of that for what kind of measures
are those or they come out of the electronic systems like EHRs in this case as
a by-product. That is another thing that does not exist today. So on and so
forth so you develop that stream.
The third one we can introduce the notion of shared responsibility and that
is where we are talking about the consumer patient. So it is a shared
responsibility with the consumer patient. How do we help them? One, we give
them data that is more meaningful to them which is the risk stratified not
adjusted quality measures and the importance of activation however we choose to
measure that. So that is in this recommendation are three.
Then your forth one and maybe you came closest to finding a way between the
dyad which was we have the primary people in the dyad and then you have the
secondary users of that information. The importance is that they do not create
another silo. They make other uses of the same measures and the same data which
creates the efficiency. They just have a different perspective that is not
divorced from the patient’s or the consumer’s perspective but it has special
attributes in the payer side and the creditors’ side. How does that fit, feel?
DR. GREEN: Can I keep going over that? Your second point about producing
metrics I would like to enrich that just slightly. They have to be national
metrics. We heard that loud and clear. Cannot be one offs. Can’t have a
different one in Nebraska from what we have in Louisiana. There cannot be
variations. There has got to be some national metrics to move the dial and the
other that is the place I think to make Justine’s point from moments ago about
it is not just out of the HR but it is the HR and additional sources for these.
That belongs in your number two I think.
DR. MIDDLETON: I had a similar orientation and I will just give you some
color commentary to help flush it out. I think the first bucket is something
about enhanced to patient measurement. Whatever the right word is, patient
centered, patient something and it is about all of the issues that have been
discussed patients similar to me, stratification, personalization, prediction,
my own care experience, and what I may reasonably expect, roadmap for me, you
know, blah, blah, blah and then the patient contextual stuff measures of the
social context, environment, the usual clinical measures and then the financial
issues that pertain to that patient perspective on health.
I think then we can talk about the dyad and focus on the provider side of
it. Measures of team-ness, measures of coordination of care, measures of
communication, measures of knowledge that is where we can fold in the
accreditation stuff even if it is in other licensure stuff, measures of action
and performance. Do you know what to do and did you do what you know to do or
are you making errors and then outcomes, traditional.
I would like to see something about a system level of measurement you know
does the system support quality measurement with the right data, the right
standards, the right information, the right technologies whether the right
technologies are in place, whether it is paper or EMR and that data is being
managed and that value data is being produced.
I could leave it there or I could throw in one more. I think it was sort of
matched to what Paul was saying. The only one that is outside a little bit is
kind of this whole community and secondary use of data. We want to be able to
drive and reuse, reuse, reuse these data pharmaco vigilance, public health, all
the kind of other meta quality stuff that I would be remiss not to say
something about on this roadmap.
So now you have some buckets perhaps to work with.
MS. MARTIN-ANDERSON: We talked earlier about, before lunch about a new
concept of composites where you recognize the responsibility of the patients
and of the care teams and potentially even of the insurer because they have to
have proper coverage for something to happen. And I wonder whether or not some
of this context of community fits in well there because a lot of the public
health initiatives fall into behaviors that exist in one of those areas.
DR. MIDDLETON: Yes, we had a little bit of this conversation on the phone
too remember about sort of these population effects which can influence
individuals whatever all of that stuff. It may be in this bucket two or
somewhere in here. We should talk a little bit about measurement science
because we heard people complain about the ability to measure an episode. That
is what I meant about temporal dimensions of measurement. You know how do you
tell when an episode starts and stops.
DR. GREEN: And longitudinal doesn’t make sense.
DR. MIDDLETON: And geography, that is, it is hard to tell who are the team
of people responsible for this patient. So measurement science stuff I think
just throw some illusions into and get a lot of research dollars from.
DR. GREEN: From my notes we still are supposedly talking about our
recommendations right? In so many words I heard this out of our various
testifiers that there is a need to fund a two or three of them used the word
fund a movement of measurement of development and mapping by a new workforce
with new depictions of the data. I grabbed that out of about three different
testimonies of this but the idea was is that where were the comments about
needed clinical informaticians and new depictions of the data so that they
could be meaningful to both sides of the dyad and individualization of that and
I forget who it was, I am sorry, but someone sort of crystallized that from all
this and said you know this is going to be a piece of work.
SPEAKER: It was David’s.
DR. GREEN: That is an important tenor for a recommendation in my view.
DR. MIDDLETON: So we can say something about measurement science and
research agenda as part of the roadmap too, just at high level.
DR. GREEN And then going off the ranch here, we did not hear, I did not pick
up much guidance from the CMS perspective in the innovation center that was
there but in my because I know it has been created and has funding, my
imagination goes awry here maybe a little bit but I am just wondering if we
don’t want to call attention that this might be an extremely powerful
application of that funding to measure quality in this sort of way. There might
be a good match up that the Secretary could use.
DR. MIDDLETON: Absolutely. I know we are out of time.
MS. JACKSON: Quick process question before we do the handover. Would that be
looking at the importance of where you are going with some of these issues with
something with a task to you as members of writing down maybe three bullets of
where you think the highlights of where of the points or do you feel that it is
already captured. Where are we in getting it down?
DR. TANG: I think a combination of what Matt put up there but Sandy and
Haley are doing.
MS. JACKSON: You think it is kind of already down in text?
MS. MARTIN-ANDERSON: Why don’t we send back the first version that is just I
am editing that one. Do we all agree we are going in the right direction and
get your feedback there. Then flush the next one because I think we captured
the verbal bullet points. Then we will iterate the next one, we will iterate
and annotate that one until we agree and then we will flush the letter through.
Does that work?
DR. MIDDLETON: Perfect. That is great. Big help.
MS. JACKSON: Because considering where we were with the scan and everything
I think that we have got a good handle on the big concepts and kind of where
you are going from there.
DR. GREEN: I want to ask for help from both of you guys and also you,
Justine. I am going to need help explaining this hearing to the Population
Subcommittee. So help, please. I want to put a space holder out on the table
right now. I think that we learned out of this hearing what the Population
Subcommittee needs to do with the hearing about the reuse of data for the new
numerators and the new denominators needed for public health and an idea that I
got out of this hearing is I have been thinking of it more as we have got to
get to public health world. We have to get the public health world.
I am beginning to think that we should contemplate designing that hearing to
have a balance set of presentations from the provider world and the public
health world about the data that they are collecting that could be dragged down
into new types of measures that would measure population health but would also
be the reused data sources.
DR. TANG: You said the providers and the who?
DR. GREEN: The usual suspects from public health.
DR. TANG: The only question is that would seem to me to reinforce the status
DR. MIDDLETON: Here is a thought. What I heard was in a way sort of
broadening it from just the public health traditional population management
stuff with population data because now management population does a clinical
exercise but this third thing we heard that numerators may come from clinicians
and denominators may come from the payers so I actually would have a –
DR. GREEN: Could you write that down? That is what I am asking for help with
right there. I think we could frame the Population Subcommittee hearings in a
way that was very complimentary to what we are hearing and keep building off of
it in terms of the nation’s statistics. So Justine will you help too please?
DR. CARR: Of course, of course. No I think it is very important that we keep
thematic and there are themes that are emerging that are complimentary and I
DR. MIDDLETON: So Paul is almost out the door.
DR. TANG: Thanks to Matt for putting together this hearing. Thank you.
DR. CARR: As usual, fantastic job. Thank you.
DR. MIDDLETON: Thank you, Matt and thank you Debbie and thank you everybody
for coming. Mike, good to see you and all of who are on, folks, thank you.
Carol, guests and others. So I guess now may I say we are officially adjourned.
(Whereupon, the meeting adjourned at 2:30 p.m.)