Multilevel Interventions in Health Care Conference: Presentation by Richard Warnecke, PhD


Uploaded by NIHOD on 05.05.2011

Transcript:
>>>DR. STEPHEN TAPLIN: So our next presenter is
Dick Warnecke. So we're going to give you the fire
hose approach today, this morning.
And then we'll have a break and have a chance for discussion.
I think what you just heard was a nice overview of the papers.
Hopefully you'll get the papers, you'll get a feel for
what's going on. I think there are some important, obviously
the important change in the context that Kelly represents,
the experience that we have in VA's are both critical to our
understanding of what we can do in multi-level and what areas
we need to inform in the future. So Marty did a great
job of bringing those points out. Now we're going to hear
from Richard Warnecke who's a professor emeritus of
sociology, epidemiology and public administration at the
University of Illinois at Chicago. He has conducted
cancer control research with populations of color and low
socioeconomic status since 1968. Currently he is the
co-principal investigator at the University of Illinois at
Chicago Center for Population Health and Health Disparity.
It's one of ten centers, five funded by NCI and five by the
National Heart, Lung and Blood Institute. Dr. Warneke's
continuing long range vision is to test and establish
interventions that address determinants of population
health disparities by approaching them with
multi-level and multi-disciplinary population
health strategies. And I think it was pointed out yesterday
that the first morning we managed to get through without
mentioning disparities or ethnic differences and their
influences, and I think now we're getting around to
something we all realize is critically important in our
understanding of our health care delivery,
and we're very happy to have Dick and his experience
represented her today. Welcome, Dick.
>>>[APPLAUSE]
>>>DR. DICK WARNECKE: Thank you and good morning.
I want to begin by thanking the co-authors on this paper,
particularly Steve and Sara Gillard who really made a big
difference on what came out in the paper. I also want to thank
the quality care consortium at the Metropolitan Breast Cancer
Task Force and the (inaud.) Health Institute who provided
some of the data that I'm going to talk about this morning.
The purpose of this paper is to examine an example really of
how multi-level analysis and local data can influence health
policy intervention. And as I was listening to the two
previous papers, I kept asking myself, well that's a wonderful
model for overall health, but how is that going to interact
with the, particularly the agencies that are dealing with
the under served. And so this may be an example of that,
understanding how local context effects access and quality of
service, and how mobilization based on local data can change
policy. This is a picture of what's going on in Chicago in
breast cancer, and it continues to today, although this goes
through 2005. And I put that arrow there because when we
were doing the first C comp evaluation and so forth, one of
the things that was happening, was that they were publishing
the first trials of early cancer treatment and minimal
surgery, and some of the other things that changed breast
cancer care. And there were also changes going on in the
way in which cancer was detected. And in some senses it
represented a paradigm shift because in 1974 during the
Betty Ford, Happy Rockefeller meetings, there was a big blow
up about whether or not you could treat breast cancer
systemically. Now I'm old enough to remember all of that,
and actually to have attended those things. But it's still a
very interesting thing because I think we're headed for
another paradigm shift as we introduce genetics and risk
measurements into both treatment and care. And so
again, we need to ask the question, it may effect
population health, in fact it may improve everybody's health.
But will it effect the disparities, the differences,
or will, like mammography, the disparities just persists
because all the boat shot rise at the same level. So that's
really what I'm concerned about. So when we started doing
our research in this area we began by looking at the
patterns of late stage diagnosis and the predictors
using the census, and the Illinois State Cancer Registry.
And what we found, not surprisingly, was that African
Americans were at risk for late stage diagnosis compared to
non-Hispanic white women. And this occurred at any age
overall, and that there was a similar pattern for Hispanic
women. We have enough Hispanic women so we can get some
estimates of that. But when you introduce poverty, most of the
effects of race disappear, except that there's an
interaction between poverty and late stage diagnosis among
young women of color where they're more likely to get
aggressive breast cancer and more likely to present with
late stage diagnosis. There's a poster by the way out there
showing how we extended that to six other cities and replicated
the results. So this is the model we've been using and this
has caused my friends on the review committee quite a bit of
consternation, and I'd like to clarify some more about it.
In that top box there really should be two things. On the
one side there are policies, and on the other side there are
social conditions that some people have called fundamental
causes. That no matter what you do, these things still crop
up and cause disparities. Social context is the place
where those policies and programs and the disparities or
the causes of disparities meet at the point of delivery.
And this is usually in a social context like a community or
something like that. And then finally you see outcomes at the
biological or genetic, some level at that point, patient
level, maybe in risk behavior. So then we look at some
indictors of access. This was data again from the registry
and the census tract. And we were looking for things that
would cause disruption in neighborhoods, because the big
literature that talks about disorderliness and problems in
neighborhoods effect health and also effect cancer. And what we
found was that women living in gentrified neighborhoods, and
women living in areas of high immigration were both more
likely, those that lived through these processes, were
more likely to be a late stage diagnosis than women who lived
in stable neighborhoods, even at the same socioeconomic
level. And after we considered a bunch of things, we started
with the hypothesis that this was stress related and that was
the result, and we didn't find any indicators of stress.
But we did find that when these populations change, the new
populations often do not have eligibility for Medicaid and
other things that pay for these services, so the services moved
or reduced their services quite a bit so that there was loss of
access. We think that is the problem. We're studying now the
safety net institutions to find out more about that. But that's
what we think. I'm going to talk a lot about what we think
and less about what we know. Surveys of mammography services
in the area that was done by the Metropolitan Chicago Breast
Cancer Task Force found that there were some deficiencies.
Non-Hispanic, African American women were more likely than
non-Hispanic whites to be screened at non-academic
facilities. And these facilities were less likely to
have digital mammography which is important for detecting
cancers in young women with dense breasts. And women
screened at these non-academic facilities were less likely to
have their mammograms read by trade specialists. We happen to
have, we interviewed about 900 newly diagnosed cancer patients
using this rapid case ascertainment and the state
cancer registry to do this. And we geocoded these people
into areas that were medically unserved and designated.
Other areas which we happen to have in Chicago that would be
eligible for under service but were never designated, and then
areas that were general population, the state coded
this for us. And when we did that, we compared residents of
MUA's with the general population and they were more
likely to present with late stage diagnosis. But when we
control for utilization of services and socio-demographic
things, the residents of the MUA's looked, their probability
of late stage diagnosis was the same as in the population,
general population. Whereas the residents of the undesignated
areas, we could not erase the disparity. They continued to
have a higher risk for late stage diagnosis. Which suggests
something about quality of the services that are available.
So potential explanations are one, that with the
neighborhoods that access was limited before gentrification
occurred. That's not the case because the areas that we
looked at were areas that were designated as medically under
served, so they had facilities. One of the things we noticed
was that the areas that were designated had somewhere around
180 safety net facilities in those areas, whereas the areas
that were undesignated had under 20. So we think that
there's something about the ecology of safety net
institutions in these undesignated areas that needs
further exploration, and that's what we're doing. Access was
adequate, but adequate was poor is the second explanation.
The initial assessment of quality indicated it was poor,
but access to quality was not jointly assessed. That is, the
access and quality were not jointly assessed. And finally
in the paper we explore some biological explanations which I
won't go into here. So, the community effected rates of
late stage diagnosis. They formed a task force that was
led by very influential leaders in the health care community
but was supported by a huge range of different groups in
the community. They produced a report that produced
legislation that removed some barriers, and then with help
from Avon and Coleman, they established a quality
consortium where they collected data and used four points that
are listed here from each of the safety net screening net
institutions in Chicago, including the Department of
Public Health. The thing is that they didn't get, although
about 70% agreed to provide this data, the provision was
slow. The Illinois Department of Family Services, however,
became aware of what the consortium was doing and so
they changed their policy and said that they weren't going to
reimburse providers who didn't provide this data to the
quality consortium. So what that does is, I think to me it
points out the importance of having community stakeholders
involved in these things. And the community stakeholders are
not just the lady next door or the person down the street.
They range from the very powerful people who have lots
of influence, and have lots of influence that we as
investigators can't use. So finally, my focus for
discussion is, is it necessary for multi-level analysis to
lead to multi-level interventions. The quality
forum is clearly an intervention. It's effect can
be measured at the health care organizational level and at the
individual level. Whether or not it produces change, what
they're doing is collecting the data and taking each reporting
institution and sending them a report comparing their rates on
those four items to the general outcome. So they're at least
trying to get everybody on the same page. My question I guess
for the people that were talking about the high level
system is, how are these ACO's and other programs that are
included in this legislation, what is the way in which
they're going to be linking to these community health centers
that are going to be established and the other
programs that are directed toward the people who don't
have insurance, who don't go to these large scale providers.
Thank you very much.
>>>[APPLAUSE]
>>>DR. STEPHEN TAPLIN: Excellent, Dick. Thank you for
brining up the issues and beginning to look at how you
look from the bottom up at multi-level effects.