[Senate Hearing 108-295]
[From the U.S. Government Publishing Office]
S. Hrg. 108-295
SQUARE PEG IN A ROUND HOLE?
DISEASE MANAGEMENT IN TRADITIONAL MEDICARE
=======================================================================
FORUM
before the
SPECIAL COMMITTEE ON AGING
UNITED STATES SENATE
ONE HUNDRED EIGHTH CONGRESS
FIRST SESSION
__________
WASHINGTON, DC
__________
NOVEMBER 4, 2003
__________
Serial No. 108-26
Printed for the use of the Special Committee on Aging
91-382 U.S. GOVERNMENT PRINTING OFFICE
WASHINGTON : 2003
____________________________________________________________________________
For Sale by the Superintendent of Documents, U.S. Government Printing Office
Internet: bookstore.gpo.gov Phone: toll free (866) 512-1800; (202) 512�091800
Fax: (202) 512�092250 Mail: Stop SSOP, Washington, DC 20402�090001
SPECIAL COMMITTEE ON AGING
LARRY CRAIG, Idaho, Chairman
RICHARD SHELBY, Alabama JOHN B. BREAUX, Louisiana, Ranking
SUSAN COLLINS, Maine Member
MIKE ENZI, Wyoming HARRY REID, Nevada
GORDON SMITH, Oregon HERB KOHL, Wisconsin
JAMES M. TALENT, Missouri JAMES M. JEFFORDS, Vermont
PETER G. FITZGERALD, Illinois RUSSELL D. FEINGOLD, Wisconsin
ORRIN G. HATCH, Utah RON WYDEN, Oregon
ELIZABETH DOLE, North Carolina BLANCHE L. LINCOLN, Arkansas
TED STEVENS, Alaska EVAN BAYH, Indiana
RICK SANTORUM, Pennsylvania THOMAS R. CARPER, Delaware
DEBBIE STABENOW, Michigan
Lupe Wissel, Staff Director
Michelle Easton, Ranking Member Staff Director
(ii)
?
C O N T E N T S
----------
Page
Opening Statement of Senator John Breaux......................... 1
Panel of Witnesses
Stuart Guterman, Director, Office of Research, Development, and
Information, Centers for Medicare and Medicaid Services,
Washington, DC................................................. 2
Mark Miller, Executive Director, Medicare Payment Advisory
Commission, Washington, DC..................................... 6
Jeff Lemieux, Centrist Policy Network and Progressive Policy
Institute, Washington, DC...................................... 9
APPENDIX
Statement from the American Academy of Family Physicians......... 27
(iii)
SQUARE PEG IN A ROUND HOLE? DISEASE MANAGEMENT IN TRADITIONAL MEDICARE
---------- --
TUESDAY, NOVEMBER 4, 2003
U.S. Senate,
Special Committee on Aging,
Washington, DC.
The Committee met, pursuant to notice, at 9:58 a.m., in
room SD-628, Dirksen Senate Office Building, Hon. Larry Craig
(chairman of the committee) presiding.
Present: Senator Craig.
OPENING STATEMENT OF SENATOR LARRY CRAIG, CHAIRMAN
The Chairman. Good morning, everyone. Let me thank you all
for coming to the Special Committee on Aging's forum on disease
management.
We have, for a period of time, studied disease management
and now, of course, are increasingly concerned about fitting it
into a variety of health care approaches that we are looking at
here in the Congress of the United States. Disease management
for people with chronic illness is a critically important
emerging innovation in America's health care system. Last fall
this Committee held a hearing on the challenges of moving
disease management into Medicare for those with chronic
conditions. This forum is a follow-up to last year's hearings.
I am Senator Larry Craig and as Chairman of the Aging
Committee, as we studied these things through hearings we have
decided to institute this forum because of the progressive
nature of these interests and developments. A variety of think
groups have worked their way through these developments and
provide them for the Congress. Of course, as you know, we are
not an authorizing committee. But we believe the special
committee can play a critical role in building an informational
base and a record by which the authorizing committees can work.
We have done that in a variety of ways in the past. Certainly,
this is one of them.
Today's panelists will focus on the technical details of
evaluating disease management demonstrations and how good a fit
disease management may or may not be with the Medicare fee-for-
service program. Our panelists today are Stuart Guterman,
Director for the Office of Research, Development and
Information at the Centers for Medicare and Medicaid Services;
Mark Miller, Executive Director of the Medicare Payment
Advisory Commission; and Jeff Lemieux of the Centrist
Organization and the Progressive Policy Institute.
The Senate and the House Medicare bills both include
provisions for disease management and we hope this forum will
aid the conferees in their deliberations. As you know, those
deliberations are hopefully shaping to some finality as we
begin to look at the final product. This forum, I think, is a
great opportunity for all of us to listen and to learn while
building a record on policy considerations.
So with that in mind, let me introduce to you Bruce
Steinwald. Bruce is the Director of Economics and Payment
Issues in the Health Division at the U.S. General Accounting
Office. So, Bruce, we do appreciate your willingness to chair
and moderate this panel. Thank you very much.
Mr. Steinwald. Thank you, Senator. I am very pleased to be
here today for this Special Committee on Aging panel discussion
of the role that disease management might play in the
traditional fee-for-service Medicare program. I would like to
thank Senator Craig for his foresight in setting up this panel
discussion and for inviting GAO to participate.
Senator Craig has already introduced me and the panelists
so I will skip over that. Basically the ground rules will be,
each of the panelists will speak in turn. The representative
from CMS will have 10 minutes roughly and the other panels
about eight. We will not take any questions. We will get
through all the formal remarks first. Then we will have a
session of questions and answers. In my role as moderator I
will ask the first set of questions. But in the meantime, the
Committee staff will circulate three-by-five note cards giving
members of the audience an opportunity to write down their
questions, which will be passed forward. The Committee staff
have asked me then to cull those questions, summarize them, and
then ask them on your behalf, if that is OK.
For most of the session we expect to have a dialog. That is
to say, the panelists will individually speak but I expect that
for most of the remainder of the morning we will have a
discussion among the panelists and the moderator. Our goal is
to generate--through generating this dialog is to raise
awareness of what we know and what we do not know, and what we
need to know about how good the fit is between disease
management and Medicare fee-for-service, and what we need to do
and what we need to learn to make that fit better.
So with no further ado we will turn to our first panelist,
Stuart Guterman.
STATEMENT OF STUART GUTERMAN, DIRECTOR, OFFICE OF RESEARCH,
DEVELOPMENT AND INFORMATION AT THE CENTERS FOR MEDICARE AND
MEDICAID SERVICES, WASHINGTON, DC
Mr. Guterman. Thank you, Bruce. I want to thank Chairman
Craig and Senator Breaux and the other distinguished Committee
members for inviting me here today to discuss Medicare's
efforts to improve the care provided to its beneficiaries
through disease management.
The Medicare population increasingly consists of people
with chronic conditions. Researchers at Johns Hopkins
University determined that 78 percent of beneficiaries have at
least one chronic condition. Chronically ill beneficiaries are
heavily burdened by their illnesses and they often have
multiple conditions and have to deal with multiple providers.
The same Johns Hopkins study found that 20 percent of Medicare
beneficiaries have at least five chronic conditions, and that
those people see an average of 13.8 different physicians in a
given year.
In the traditional Medicare fee-for-service program,
however, physicians are paid for the individual services they
provide and there is no incentive to provide the coordinated
care that chronically ill beneficiaries need. In fact because
such coordination involves effort and resources and there is no
payment corresponding to that effort coordinated care is in
effect discouraged.
The Medicare+Choice program should be an appropriate for
providing coordinated care but the current payment system and
some of the rules under which plans currently operate in fact
penalize them for enrolling beneficiaries who are chronically
ill and, therefore, expected to be much more expensive than
average.
Chronic diseases play a large role in generating both the
growing level of utilization under Medicare and the finances of
the program. The 78 percent of Medicare beneficiaries with at
least one chronic condition account for 99 percent of Medicare
spending each year, and the 20 percent of beneficiaries with at
least five chronic conditions account for almost two-thirds of
all program spending.
We need to find better ways to coordinate care for these
beneficiaries. Toward that end, CMS is looking at disease
management approaches, which have been developed and applied in
the private sector, to combine adherence to evidence-based
medical practice with better coordination of care across
providers. We are developing an array of demonstration projects
to test our ability to apply these approaches in the context of
the Medicare program.
Let me lay out our objectives. There are multiple
objectives in this approach. One is to improve access to the
care that Medicare beneficiaries need. Another is to improve
the coordination of the care so it can be provided more
effectively and efficiently. Another is improving the
performance of physicians by making them more involved and
responsive to the patient needs. Finally, we are trying to
improve the ability of patients themselves to be involved and
participate in their own care.
These demonstrations will need to test and evaluate what
needs to be done in order to get disease management programs up
and running, how best to provide disease management services,
which of these services work and which do not in the Medicare
context, which conditions lend themselves best to disease
management initiatives, and the impact of different approaches.
This involves answering several sets of questions. One is
what should be the focus of disease management programs? How
should these approaches be designed and applied? Another which
is a major question in the current context of Medicare is what
are the data requirements and how can it be achieved?
The data requirements in disease management context are
related to at least three sets of activities. One is to
identify potential enrollees. A second is to monitor their
needs as disease management is applied. A third is to be able
to evaluate the results of disease management approaches. This
is a major challenge for the Medicare program. We have a very
rich database, as most of you know, but that database has been
designed to process and pay bills for fee-for-service providers
mostly. Our challenge is to redesign our ability to use our
data to be able to apply them to these three functions.
Another question we have to address is what organizational
structures work best? How do you establish the appropriate
networks of providers? How do you conduct enrollment? How do
you provide disease management services in the context of
Medicare?
Another question we have to address is which disease
management approaches work best? Who contacts enrollees? What
do they do? How do they make sure there is follow-up, et
cetera?
We also need to know how payment can be designed to be
compatible with these approaches. That is another major
challenge, because again, under the fee-for-service program we
have a payment system that pays for individual services, and
under the Medicare+Choice program, as I will address in a
couple of minutes, there are some drawbacks to being able to
coordinate care the way Medicare+Choice should be able to.
There are several alternative approaches to take. One is a
disease management fee that you can pay specifically for
disease management services. Another is to incorporate more
encouragement within the capitated payment systems to do so.
Another question we have to answer at the end of all this
is how can these issues be appropriately evaluated? How do we
know if we have succeeded? How do we know what that indicates
about how to proceed with these programs?
Let me briefly review where we are today. The first in our
series of disease management activities was the coordinated
care demonstration project. It was mandated in the Balanced
Budget Act of 1997. It has 15 sites, including commercial
disease management vendors, academic medical centers, and other
provider-based programs. It focuses on beneficiaries with
congestive heart failure, heart, liver, and lung diseases,
Alzheimer's and other dementia, cancer and HIV/AIDS. We have
sites in both urban and rural areas. This is in a fee-for-
service context. These organizations are paid a disease
management fee, and currently they have about 13,000 enrollees
that are evenly split between intervention and control groups.
The sites that are included in this coordinated care
demonstration include Carle Foundation Hospital in eastern
Illinois, Medical Care Development in Maine, Health Quality
Partners in eastern Pennsylvania, and Washington University
Status One in St. Louis, Missouri. There is also a separate
demonstration that is not formally part of the coordinated care
demonstration but incorporates approaches that are parallel,
being conducted by Lovelace Health Systems in Albuquerque, NM,
to provide coordinated care services to Medicare beneficiaries
with congestive heart failure or diabetes.
We are testing whether these coordinated care programs can
improve medical treatment plans, reduce avoidable hospital
admissions, and promote other desirable outcomes. As of now,
all 16 of these plans have been in operation for at least one
year. As I said, the total enrollment is about 13,000.
Initial findings indicate that beneficiary recruitment in
the fee-for-service market can be a challenge. While the
majority of plans are at or near their enrollment targets,
several plans have invested more time and resources than they
had initially expected to get there. The most successful plans,
we find, have established close ties to physicians and other
providers and have focused on obtaining physician support
during the planning stages of the project. Although evaluation
results are not yet available, the programs overall appear to
be well-received by both participating physicians and
enrollees. We are planning to submit a report to Congress on
this demonstration in the spring of 2004.
Another project that we have just recently gotten waiver
approval for is the disease management demonstration for
beneficiaries with advanced stage congestive heart failure,
diabetes, or coronary heart disease that was mandated in the
Benefits Improvement and Protection Act of 2000. This
demonstration includes three sites. What is interesting about
this is that these sites receive a disease management fee that
includes drug coverage not just for drugs related to the target
conditions but for all drugs being used by enrollees in the
program.
Participating organizations in this demonstration are at
risk for total Medicare costs. The demonstration is limited to
30,000 enrollees total. The three sites which have just had
their waiver applications approved by the Office of Management
and Budget are CorSolutions of Buffalo Grove, IL; Diabetex/
XLHealth of Baltimore, MD; and HeartPartners of Santa Ana, CA,
which is a joint venture among PacifiCare, QMed, and Alere
Medical. We are hoping to get these projects started in January
2004.
Another project which is in the pipeline is the physician
group practice project that was mandated in the Benefits
Improvement and Protection Act of 2000. We released a
solicitation in which we announced we would make at least six
awards. We have received the applications which are to
coordinate Part A and Part B services. We are planning on
sharing savings if outcomes are improved and if savings are
achieved. The implementation of this demonstration project is
pending waiver approval but we are in the stage where the
waiver package is being developed and considered.
We have got a couple of other projects in process that
address the issue of disease management in a capitated
environment. We have got a capitated disease management project
for which we put a solicitation out on the street. Under this
project, we would be paying a 100 percent risk-adjusted
capitated rate to organizations that take on the responsibility
for patients with specific diseases and coordinate the care for
those patients.
This project is actually designed to overcome two barriers
that currently exist in the Medicare+Choice program. One is
that, as of now, only 10 percent of the payment rate is risk-
adjusted, which means that plans that attract more chronically
ill beneficiaries who are going to be more expensive face a
financial penalty if they do so. So this capitated disease
management will involve 100 percent risk-adjusted payment rate
to be able to adjust the payment to suit the enrollee
population more appropriately.
The other is that the current rules prohibit plans from
specializing in enrollees with certain conditions. Under the
capitated disease management project, we will waive that
prohibition so that we will be seeing plans that are able to
specialize in particular groups of beneficiaries that they feel
they can treat more effectively.
We have also got a project that will do the same thing for
patients with end-stage renal diseases. We are currently in the
process of evaluating the proposals that we have received in
response to our solicitation.
In the future, we plan to piece all of these projects
together in order to develop a new role for Medicare as of more
aggressive purchaser of health care for its beneficiaries. We
want to achieve better coordination of care. We want to be able
to collect information, to be able to measure and disseminate
information on quality, and we want to be able to eventually
pay for quality of care. All of these pieces fit together in
order to try to fit into our efforts to make Medicare a better,
more aggressive purchaser of health care, to be able to provide
better care for our beneficiaries.
Thanks for having me here, and I will take questions at the
appropriate time.
Mr. Steinwald. We will hold onto the questions for a little
bit. Now we will hear from Mark Miller from MedPAC.
STATEMENT OF MARK MILLER, EXECUTIVE DIRECTOR, MEDICARE PAYMENT
ADVISORY COMMISSION, WASHINGTON, DC
Mr. Miller. I would also like to thank Senator Craig,
Senator Breaux, and the members of the Committee for holding
this forum. I am the Executive Director of the Medicare Payment
Advisory Commission. As probably many of you know, we make
recommendations and put analysis in front of the Congress in a
couple of reports during the course of the year.
I am pleased to be here this morning to discuss the role of
disease management and coordinated care. But the forum comes at
sort of an awkward time for us. We have just begun our work on
this issue, looking at disease management and how you integrate
it into the traditional Medicare program. What I can say at
this point is that the Commission believes that these programs
do have promise to improve the quality of care for Medicare
beneficiaries, but there are a number of questions that need to
be addressed. As part of my talk today I am going to discuss
those issues.
I will try and walk through our work plan and the questions
that I think are outstanding. You can think of the questions
that we are going to address in the following way: This is
summarized, of course. Issues of targeting the program to
beneficiaries most likely to benefit from them; the notion of
defining the relationship between the physician, the
beneficiary, and the program; aligning payments to program
objectives; and measuring the effectiveness of the program.
Many of the things that I am going to say coincide with what
Stu has said, so some of this should sound relatively familiar.
Our analysis is going to approach this in a couple of
different ways. We are going to look at Medicare claims data.
We are going to look at Medicare beneficiary survey data. We
are going to review the literature on disease management. There
we are also going to work with stakeholders out in the field to
gather information on what the state-of-the-art is out there.
The reason that I think this issue attracts policymakers
and the Commission's attention is that when you look at fee-
for-service, and some of these statistics are going to sound
familiar, you see things like 5 percent of the beneficiaries
account for about half of the expenditures in Medicare; in the
case of congestive heart failure, 14 percent of the
beneficiaries accounts for 40-some-odd percent of the
expenditures; beneficiaries with three chronic conditions or
more account for more than 80 percent of Medicare expenditures.
I think it is statistics like these that capture policymakers'
attention and make them want to think about how disease
management models can be applied to traditional Medicare.
What the Commission is interested in is how disease
management can improve the delivery and coordination of care,
whether it can maintain or improve the quality of care and
functioning of patients in Medicare, and whether it can reduce
program and beneficiary expenditures.
Disease management comes in a lot of different varieties
and has a lot of different ways that it approaches the issues,
but fundamentally, beneficiaries who are either high-cost are
targeted--or you target beneficiaries who are either high-cost
or expected to become high-cost, and then you engage in things
like coordinating care, patient education, monitoring signs and
symptoms, assuring that the patient follows their treatment
regimen, and then ultimately help the physician practice
evidence-based medicine.
There are a number of populations that MedPAC is going to
focus on in its analysis. We are going to look at the claims
data to identify beneficiaries that have certain conditions.
These are the conditions that disease management programs are
usually designed for, now in the non-elderly population:
congestive heart failure, diabetes, end-stage renal disease,
conditions like that. But we are also going to try to look at
the population and determine whether we can identify
populations that have conditions that lead to more serious
conditions. So, for example, we will look at chronic kidney
disease because that is the precursor to ESRD.
A couple of other sets of beneficiaries that we will look
at are beneficiaries with multiple chronic conditions and
beneficiaries who are high-cost as identified through claims
analysis. The notion here is that if you can identify a
beneficiary who is already expensive or has chronic conditions,
the question becomes, does it still present an opportunity for
disease management to have an impact either on their quality of
care, or the program or beneficiary expenditures.
Two other populations that we are going to look at are dual
eligibles, beneficiaries who are eligible for both Medicare and
Medicaid, and beneficiaries at the end of life. Expenditures in
the last year of life in the Medicare population obviously
account for a lot of expenditures in Medicare in total.
We are looking at these populations, and a broad-brush they
can be identified, they are often growing, they account for
large blocks of expenditures, and either have conditions that
could benefit from evidence-based medical care, or
alternatively from better coordination of their care.
What I would like to do with my remaining time is walk
through a couple of policy issues. This list is not exhaustive
and I think in many ways overlaps with Stuart's list. There are
issues of targeting, which I have already mentioned. I have
tried to identify for you at the macro level the kinds of
beneficiaries that we are going to focus our analysis on. But
we also need to get inside disease management programs because
they also target within those populations. So you may have a
set of diabetics, rank them according to their risk, and have
different interventions for different levels of risk. There are
issues like that that we think need to be explored and put
before the Congress.
Stuart touched on this issue as well. We think that the
physician's role is absolutely critical. If the physician is
involved and accepting of these programs, these programs have a
much greater degree of success. But a physician's involvement
can be thought of as a continuum. You can think of consulting
with the physicians at the front end to create protocols,
making them aware of the program, but have the disease
management program delivered by other professionals such as
nurses. Or you can try and build your disease management model
around the physician, have the physician target the patient,
educate the patient, and so forth. Most often what occurs in
the field is that physicians are consulted and know about the
program but the programs are delivered by nurses. But as I
said, it is important to have the physician's buy-in.
There is a set of questions that involve the beneficiaries.
I will not go through all of them, but one key question will be
whether these programs should be designed as opt-in or opt-out
programs. An opting-in program maximizes the beneficiaries'
choice. They can choose whether they want to go into the
program. An opting-out program is where the beneficiary is
identified as a candidate for the program but chooses to get
out of it. There the notion is that you may have a greater
ability to capture a population that might be affected by
disease management.
Payment issues are huge and I am not going to go through
all of them but a couple of issues are: who should get paid,
what should be the size of the payment, what services are
expected in exchange for the payment, and then finally, whether
the disease managers bear any risk at all. Currently in the
field, if a disease management organization accepts risk it
accepts it only for its administrative fee. It does not accept
insurance risk, at least that I am aware of.
There are also some administrative issues. These include
questions like what entities in traditional fee-for-service
would be delivering the disease management. Will it be
physicians, group practices, disease management organizations,
managed care plans, that type of thing. Next there is the issue
of data. Programs will have to start with Medicare claims data,
which is not unreasonable, but there are questions about how
the data have to be enriched. The timeliness of the data will
be important in order to support these programs.
Here I am mostly talking about delivering data to whatever
entity is administering the disease management. There also
other data issues obviously.
The very last issue I will mention, and again Stuart has
already done this is, the notion of measurement and
accountability. It becomes very complicated--within a
demonstration environment you can do things like identify a
control group. If you have a national program, how you measure
success will become a little bit more difficult if you do not
have a clearly defined control group. It is not impossible but
it becomes more difficult.
Then finally, some of these programs have to have a time
period in order to demonstrate results and there is a question
about how much time should elapse before you should expect to
see results in terms of quality or savings, if they
materialize.
With that, I look forward to your questions.
Mr. Steinwald. Thank you, Mark. Now let us hear from a non-
fed, Jeff Lemieux.
STATEMENT OF JEFF LEMIEUX, CENTRIST POLICY, NETWORK AND
PROGRESSIVE POLICY INSTITUTE, WASHINGTON, DC
Mr. Lemieux. Thank you. Ex-fed, I guess. My name is Jeff
Lemieux and I work for a small startup think tank called
Centrists.org, and I also work for the well-established
Progressive Policy Institute. I would like to echo Stu and
Mark's thanks for the Committee members inviting us here to
talk about disease management and Medicare.
Let me start with a few theoretical comments and then move
to the disease management and chronic care initiatives in the
Medicare bills that are being negotiated in conference
committee, and to some thoughts about how those might come out
and what impacts they might have.
The Progressive Policy Institute definitely agrees that
finding better ways to treat people with chronic illnesses is
the next great challenge in health care delivery, and that any
new benefits or other changes in Medicare should work toward
better chronic care and be consistent with chronic care
improvements.
There are two main approaches to chronic care. One is a
disease management approach where an organization targets
people with a particular ailment and reaches out to them and
their families and their health care providers with
information, with self-care initiatives, home monitoring, and
things like that. Then there are more intense case management
services for people with multiple chronic illnesses where they
literally need to have a health provider, a physician or a
nurse or a doctor's office taking charge of their care,
coordinating their care with community services and with
various health providers to try and avoid duplication and avoid
health providers working at cross-purposes.
In economic theory, comprehensive health plans like HMOs
generally have a good incentive to provide excellent chronic
care services. They can evaluate investments in chronic care
improvements and make tradeoffs. They may be able to spend more
money on improving a patient's compliance with medication
regimens on one hand, so they can save money on
hospitalizations on the other. They can evaluate payoffs from
improving chronic care services.
Now in general, traditional fee-for-service in Medicare
does not have this easy way of making tradeoffs. Medicare's
benefit contractors are separated into different silos. We have
Part A contractors that pay for hospitalization, Part B pay for
outpatient services, and if we pass this Medicare bill we will
have a Part D to pay for drug benefits for at least most
beneficiaries. It is awkward for each of those contractors to
say, if I invest more in chronic care services within my area,
someone else will get the payoff. So that is a problem in fee-
for-service.
Besides encouraging HMOs and PPOs to join or rejoin
Medicare, the House and Senate prescription drug bills address
chronic care in the fee-for-service program with two basic
approaches. The first approach is highlighted in the House bill
where instead of having demonstrations they would move to a
complete, full-fledged program for disease management that
would encourage disease management providers, organizational
entities, to come forward and contract with CMS. They would
also break the country up into separate chronic care
improvement areas where the administration of this program
could be devolved down more to the local level. I will get back
to that in just a second.
The second approach is something that was highlighted in
the Senate's complex care demonstration program which would
establish--the program as written in the law is vague but I
think as it is coming out of the conference committee it would
establish--essentially pay for performance procedures like both
Mark and Stuart talked about, which would attempt to reimburse
physicians in a way that I will talk about in just a minute.
Before I get to the details of the conference agreement,
let me go back and just mention this idea of experimenting in
lots of different ways. No health policy analyst can really say
we know how to improve chronic care. Even the clinicians, who
know much more about it than certainly I do, do not have a
magic answer to the chronic care problem. So I think the theme
of everything we do in Medicare should evolve around
experimentation, especially localized experimentation close to
the health providers and the problems of local communities, and
evaluation where organizations like GAO and CBO and MedPAC and
other groups spend a great deal of time trying to figure out
what works so that local areas can do what works and avoid what
does not.
Back to the conference agreement. There is some progress
moving toward trying to go beyond just the standard disease
management organization approach to trying to find ways to
compensate individual health care providers, especially in
geriatrics, taking care of complex cases. The problem with this
has always been a budgetary problem. If we provide additional
reimbursements, how do we know we are going to get more and
better health care as a result? Can we be sure as a public
program that everybody is not just signing up for these extra
reimbursements?
What I think they are moving toward, which I think is a
very good idea, is a situation where providers could enroll
with local quality improvement organizations and say, ``Look, I
will volunteer to make investments in chronic care, in
information technology, and in working with my patients, and in
return I will open the books for CMS and the local quality
improvement organization to evaluate how this is going, so that
CMS can really get a good idea of whether or not these programs
are working.'' This is the sort of thing that I think can work
in fee-for-service where you are not saying to one provider,
you are going to get this reimbursement and denying it to
another. It would be a partnership between the local
organization and the provider.
Whether or the conferees will stick with this localized
structure that was in the House bill is unclear. There was also
a big pool of funds in the Senate bill, and whether or not that
will still be available at the end for chronic care
improvements is unclear.
Finally, I think the conferees will have to make a tough
choice on whether or not they are going to emphasize
straightforward disease management with the organizational
approach or to invest a lot of time in this effort to try to
find a way to reimburse physicians directly for chronic care
services and then evaluate those very carefully for cost
effectiveness.
Thanks.
Mr. Steinwald. Thank you, Jeff.
Once again, I would like to remind you that we are
accepting questions from the audience on cards and I encourage
you to fill out a card if you have a question. If you would
like the panelist to address a particular issue related to
disease management, please avail yourself of this opportunity.
OK, guys--we are a bit gender deprived up here today. Here
is what I plan to do, if it is OK with you. I want to start out
with directing the first set of questions to an individual but
give every panelist the opportunity to weigh in. So I will
start with Stuart, ask him a question and then by weighing in
then we will get our dialog going.
We are short on data about outcomes. We seem to be long on
data about who the potential targets of disease management
programs might be. We know what we spend and for what. We think
that there is a population that could benefit, but the question
is how to get to them and how to affect programmatic changes
that yield the kinds of outcomes that we would like, both for
the beneficiaries and for the program. Stuart, in your
testimony you noted that some of the disease management
demonstration programs to date have had some difficulty
recruiting beneficiaries. I wonder what implications this might
have for a broader disease management program under Medicare.
Mr. Guterman. Actually I heard two questions there, one
embedded in your preamble, so let me address both of them. One
is in terms of enrolling beneficiaries, we found that they tend
to be reluctant to sign up for programs where they are not
familiar with the organization that is running the program.
That is one of the reasons why the disease management
organizations that have worked through the physicians have been
much more successful in enrolling beneficiaries. But I think we
are going to face that issue. Disease management is something
that is different. Medicare beneficiaries particularly tend to
be fairly reluctant to try new things unless they have a very
well-established reason from someone they trust that it is
going to work.
We think as these things build momentum it will be easier
to enroll beneficiaries. We are also trying to focus--one of
the major things so far that we have learned from the chronic
care demonstration is what people have done right and what they
have done wrong in terms of how to enroll beneficiaries. We
hope to take advantage of what they have done right and apply
it in future efforts.
You did mention outcomes, however, and--I guess there are a
couple ways to describe our attitude going into this. One is,
that is why we do demonstrations, to be able to get evidence on
whether things we want to try out work and how they work. One
of the things we will be looking at, clearly, is whether
beneficiaries are better off, whether they think they are
better off, whether they have better health outcomes under
disease management.
But the thing really preceding that is we are very excited
about this because, as a couple of us mentioned, there is a lot
of money on the table here, and it is hard to believe that we
cannot do a better job than is currently being done. If
beneficiaries see 13.8 physicians a year, different physicians
in a given year, it is hard to believe that a system that
emphasizes those physicians coordinating their care cannot do
better than a system that currently does not provide any
incentive for them to interact.
Mr. Steinwald. Mark or Jeff?
Mr. Miller. I think I would emphasize the notion that you
have to have the physician at some level involved, because I
think that beneficiaries will listen to their physicians. Also,
it will help if the beneficiary views it as an additional
service and something that they can benefit from. I do not
necessarily have the answer, as to how that is communicated.
Also I think that the beneficiary has to feel that in making
this choice it does not somehow affect their benefits more
broadly. I think they need to understand that before they will
move towards it.
Mr. Steinwald. Stuart, I am not picking on you, I am just
going in order. The BIPA demonstrations are required to
guarantee savings to the Medicare, as I recall, and that was in
your testimony. One open issue is, what is a reasonable
timeframe to expect savings from disease management programs to
be realized? How is this all set up in the demonstration
program so that you will know whether and when savings are in
fact realized?
Mr. Guterman. Under the BIPA demonstration, actually the
terms were very explicit in the legislation so that we--and
since it is a 3-year demo the assumption, I guess, of Congress
was that we could see savings before the end of that timeframe.
The thing is that the organizations that are participating at
the three sites actually are quite enthusiastic about their
ability to save money right away under this demo, partly
because what they are focusing on is people with more advanced
stages of the diseases that they are focusing on, and partly
because the demonstration project gives them the ability to use
drugs to help manage diseases. Pharmaceuticals have a major
role in trying to manage some of these chronic conditions.
So we expect to see savings right away. Now clearly there
are some other circumstances and other conditions where the
timeline would be different.
Mr. Steinwald. Jeff.
Mr. Lemieux. One of the things I just wanted to point out
is that in the bills being conferenced, they all talk about
budget neutrality, but they do not really define what the
period of budget neutrality is. Some of them have pots of money
available as if they assume that this will require some upfront
investment to get a longer-term payoff.
What I wanted to point out is that for these programs to
work there is going to have to be a fair degree of stability
that could have to last a long time. For example, I mentioned
that HMOs, Medicare+Choice plans, have an incentive built in to
provide good chronic care services only if they believe that
those patients are going to still be enrolled two or three or
four or five or six years down the road. If the program seems
less stable or if they do not know what their enrollment will
be or cannot guess it with some probability, then they will
have less of an incentive to make investments in chronic care.
The same thing would apply to disease management organizations.
The stability and the length of the program, and knowing that
Congress and CMS are behind these programs for a long period of
time would be important in the investments they would want to
try and make.
Mr. Steinwald. Thank you. Mark, you mentioned targeting
several times. I have got a two-part question for you. One is,
if you rely on program data, claims and other program data,
what is the lag between the time that a beneficiary develops
the conditions which might be suitable for participation in a
disease management program and the time in fact the program can
identify them?
A second related question is, is there any reason why we
could not identify these people upon enrollment to the program?
Do we know what proportion of those with chronic conditions
have them upon enrollment, or are they developed after
enrollment?
Mr. Miller. In terms of targeting, I think the claims data
gives you the starting point and you can look at diagnoses and
services used. If that is given to a disease management entity,
they can often start with that population and drill down and
gather additional information to sort out who is at risk. That
is some of what they do. So that is one point.
I think that a person does not necessarily have to have an
acute episode in order to be identified, if a person is going
to a physician and their comorbidities are put onto physician
claims. I know there have been some issues with the diagnosis
codes on physician claims but they do seem to be getting
better. Even before they hit a hospitalization for a diabetic
event you might be able to pull that person out of the data.
But the administrative data is fairly clumsy. It is not
impossible to do that, but it is a clumsy way.
But I do believe that it offers information to begin to
identify populations, and I think it is probably not all that
different than what disease management organizations in the
non-elderly sector start with, except that those databases
often have drug data in addition to this and that can be an
indicator.
As to enrollment, I think if you listen to disease
management organizations--and I am not saying that this is my
view--if they will tell you that you can discern a lot of
information in a conversation with a beneficiary just about
their home situation, what kinds of drugs they are taking, how
they are feeling. I mean, literally whether they are feeling
depressed and those types of things. One would argue that
perhaps when somebody is enrolling in Medicare you could do
something like that. But I think there is a real issue of
resources if you are going to do that for every person who is
enrolling into Medicare, and there may be issues of privacy.
Mr. Guterman. On the use of data, I think that is one of
our major challenges. I think that is clearly something we are
going to have to deal with. As we develop future projects we
are trying to figure out better ways to get a hold of our own
data faster so that we can use them more effectively to
identify beneficiaries who would be more amenable to this kind
of approach.
But if you look at the data, there are still--these are
chronic illnesses. They are not acute illnesses. So even under
current conditions you can still capture a lot of manageable
time with a condition with the data we have. So I think there
is always an incentive to get better and to capture these
beneficiaries before their first spike, the first serious
hospital admission. But even if you wait until the clinical
data are in through the claims system as it stands now, you can
still capture these people for a long enough time that you can
use to manage their conditions better than they are being
managed.
Mr. Lemieux. I just wanted to say, Mark's comment about
drug data I think is extremely important. It is a little bit of
water under the bridge, but the idea of having a drug benefit
under which virtually every Medicare beneficiary has a card
with electronic transfer of information would be a pretty good
early predictor of people with certain conditions that CMS
could turn around very quickly.
I just also wanted to mention the importance of physicians
in the community knowing about these programs and being able to
refer their patients at that point.
Then finally, consider for patients with multiple chronic
conditions. These are the most expensive patients, the patients
most in need of having care coordination. Even the lagged
claims data that CMS currently has is going to be able to point
out these people. Maybe not in time to do a lot of prevention,
but certainly in time to do some management and care
coordination.
Mr. Steinwald. Let me follow up on the matter of the drug
benefit. Mark, in your view, and others as well, if the
Congress passes a prescription drug benefit, does that heighten
the need for or the potential benefit from disease management?
Mr. Miller. I do not know that it heightens the need. I
think it gives two additional tools to a disease management
effort. He provides data that you can use to identify patients,
just as Jeff and Stuart just said, and it is one of the tools
that disease management organizations use to manage patients.
Mr. Steinwald. Jeff, in your testimony you observed that
disease management is ideal for an integrated care environment.
But you also observed that so far Medicare+Choice plans have
had mixed success in implementing disease management. Would you
elaborate on that? Why is that? Should we be more or less
optimistic about disease management in an expanded managed care
proposal as envisioned in the Medicare reform bills?
Mr. Lemieux. Let me take off that very last clause, as
envisioned in the Medicare reform bills and just speak a little
bit more theoretically. I think that the HMOs and
Medicare+Choice, and even to some extent looser managed
integrated health plans like PPOs, if they have the full range
of benefits available, if they can afford to provide the full
range of benefits under the Medicare system, then they can
initiate programs that would in fact involve tradeoffs.
Investments in home monitoring, investments in compliance with
drug regimens, and so on, can pay off in other areas. I think
that is a powerful investment.
I think that over the last five to 10 years the
Medicare+Choice program has been volatile enough that only some
of the plans are making these investments. Some feel as though
they have a stable enough situation in Medicare that it pays to
make investments in a big way. Others rushed into the program
in the mid-1990's to take advantage of what were perceived to
be favorable reimbursement rates. When the reimbursement rates
turned unfavorable, they left just as quickly. So it has not
been a stable place where an executive of these plans can say,
look, we can make investments that will pay off five and 10
years down the road because we are in this market for good.
Now in another semi-theoretical point, if we were to go to
a working competitive system that encouraged private plans to
enter Medicare, we would probably also benefit by giving the
Government-run fee-for-service plan more flexibility. If it
turned out, under a competitive, say, premium support system,
that these private plans were more efficient than the
Government-run plan, then Congress would be sure to give
Medicare the sorts of flexibility that it needs to compete with
them. This could be a win-win situation.
Whereas, if we had a stable place for the private plans,
they could do disease management better. If we had a more
flexible situation for the public plans, the public plan
managers could really go to work on this in a bigger way.
Mr. Steinwald. Stuart.
Mr. Guterman. As I mentioned in my testimony, there are two
non-theoretical barriers that Medicare+Choice plans face now.
One is that, currently plans, the rates that plans are paid are
only risk-adjusted for 10 percent of their payment, which means
that--and the rest of it is determined by a rate that is more
or less geared to the cost of the average Medicare beneficiary.
At least they do not get paid more for enrolling costlier than
average Medicare beneficiaries. So we think that as we
transition to a fully risk-adjusted payment rate that will
provide the financial cover for plans that engage in programs
that make them more attractive to chronically ill beneficiaries
and remove the current financial disincentive for them doing
so.
The other is that in the capitated disease management
project that we have got on the street, the other feature is
that we are trying to encourage plans to specialize in groups
of beneficiaries that they can deal with most effectively. We
think that giving them that kind of leeway will allow us to see
how the demonstration works out, and it might also open up the
Medicare+Choice environment- or whatever it involves into--to
being more open to disease management.
Mr. Steinwald. Jeff, in your testimony you suggested that a
national disease management program should have a degree of
decentralization. Why do you believe that Medicare's
centralized structure hampers coordinated care? What do you
think are the tradeoffs between a centralized and decentralized
program?
Mr. Lemieux. I do not have any evidence. It is a
speculation. But I think it is a speculation that is based on a
decent logic. Different areas of the country have different
issues that might be the most important issues. In certain
parts of the country, diabetes might be extremely important. In
other parts of the country, the most important disease
management might be remote monitoring and communications in
rural areas.
So it just seems to make sense to me to do as much as we
can in Medicare to get local Medicare officials, administrators
and clinicians, medical directors, on the ground in as many
different localities as possible to see what is available, to
see what the local hospitals and clinics and community
organizations can do, and to try and coordinate those efforts
locally rather than through headquarters.
That is not a knock against headquarters. Headquarters has
enough to do. I am advocating that we expand Medicare's
administrative capability to include a local administrative
component that would be specifically charged with working the
chronic care and disease management angle so that it would just
be better able to find out what is available and try as many
alternatives as possible and then evaluate them to see what
works.
My feeling is that if we put an administrator in northern
Louisiana and another one in southern Arkansas, and the
northern Louisiana folks could show dramatic improvements in a
variety of areas, and the people in southern Arkansas were not
doing very much, then we could take a couple people from
northern Louisiana and put them in Arkansas. In other words,
just see what works in different areas and let the areas almost
compete against each other to try and make progress in these
areas.
Mr. Steinwald. Mark or Stuart, any reaction to that?
Mr. Miller. I think this goes back to Jeff's earlier
comment, along the lines of creating a stable environment for
plans and giving greater flexibility to the traditional program
in order to administer these types of programs. I think that is
an issue that is key if they are going to have any success.
Mr. Guterman. On the issue of doing things at the local
level, all of the disease management organizations that are
participating in our demonstration projects have specified
service areas. We recognize that it is necessary to work pretty
close to the ground to be able to perform disease management.
We anticipate that even if we ended up getting to the point
where we could apply this program-wide it would still be
locally administered in terms of actually doing the disease
management, and maybe centrally administered in terms of
coordinating the different areas.
Mr. Steinwald. Thanks. Those are my initial questions. Now
I would like to turn to questions from the audience, if that is
OK. We still have an opportunity, if anyone would like to ask a
question, write it on a card and send it forward. We can still
accept them. Keep those cards and letters coming, please.
To what degree is the goal of disease management--by the
way, I think every one of these questions is directed to the
panel, so anybody who wants to, jump in. To what degree is the
goal of disease management cost reduction? Is there a cost-
quality tradeoff in disease management?
Mr. Guterman. I can take that. I have frequently been
asked, is the purpose of this to provide better care and
perhaps save money, or to save money and perhaps provide better
care? My answer is, to provide better care and perhaps save
money, although I think it that we can do both. We have got
possibilities, an area under the possibilities curve where we
are where I think we can improve both in terms of coordinating
care better and providing better quality care for beneficiaries
and in saving money. In fact, they are not necessarily even
tradeoffs because--let me give you an example.
If a person with congestive heart failure goes to his
cardiologist and gets a blood-thinning medication and then
feels light-headed and goes to his internist to get a blood-
thickening medication, that is neither good for the patient nor
is it good for the program. If the person ends up in the
hospital it ends up being a major expense for the program. If
we could have avoided that conflict and avoided that
unnecessary utilization, then we would make both the program
better off in a purely fiscal sense, but also, more
importantly, make the patient better off because they would not
have had to go through the illness in the first place.
Mr. Lemieux. Just a quick comment. Your question points out
what was so interesting about the $6 billion pool of money that
was put in the Senate bill. It started in 2009, and if at that
point we had done enough experimentation to know that we had
lots of ways to do disease management and chronic care that did
save money or that were budget neutral, we could continue
those. But if we found some that were not budget neutral, or
that we could not prove were budget neutral but they still
seemed like very worthy initiatives that were really good for
beneficiaries and good for the program, then that pool of money
would allow those to continue, which is an interesting idea.
Mr. Miller. I think this is the kind of question that
ultimately has to be answered, in my instance, by the
Commission. It is not an answer that I can give. But I think a
way to think about the question for the Commission and the
Congress ultimately is, if all of these things improve quality
but did not raise costs any more, they would probably still be
interested in moving forward with them. That would be my sense.
Mr. Steinwald. This is a longer one so pay attention. The
private sector (insurers and corporations) for the working age
population seems to be moving away from the disease management
approach with the exception of congestive heart failure. They
seem to be finding that disease management does not save money.
First, do you agree with this characterization of the private
sector experience? Second, what makes us think that Medicare's
experience will be any different?
Mr. Guterman. This issue has come up, and I am no expert as
to what is happening in the private sector but I will cite some
testimony that was actually made before MedPAC, I think at
their last meeting when Glen Mays of the Center for Studying
Health Systems Change pointed out that there was a lot of
excitement in the private sector about disease management but
that there was relatively little evidence that it saved them a
whole lot of money at this point. You probably would want to
look at, for those who have become disenchanted, with disease
management, what they did and why they became disenchanted.
But I will address the issue of what makes Medicare
different because I think that is a very important distinction.
I think Medicare has two big advantages. One is that the
chronically ill population is a very large chunk of our
beneficiary population, and a growing one. So I think we have
the mass there to make it more worthwhile to devote some effort
to figuring out how to deal with those chronic illnesses
better.
The other is that unlike private plans which have enrollees
for maybe a year or two at a time, we have them for life. So we
can make the kind of investment that it takes to really get
these programs to work because we have a longer timeframe where
we know that we will be the recipient of the benefits, to the
extent that there are financial benefits, and to the extent
that there are benefits in terms of improved treatment of our
beneficiaries.
Mr. Miller. I was going to say that I think the experience
in the private sector has been mixed. I agree that there is
no--I think I say that both in the testimony and I think this
came out at our public meeting--there is no clear evidence that
it saves money. I think there has been a mixed experience. I
think some of our work in going out and talking to stakeholders
is to try and get a feel for some of that, why some of them
went into it, why some of them walked away from it, if we can
identify those people. The other part of my answer was going to
be what Stuart said, you do end up with the beneficiary for a
longer period of time in Medicare.
Mr. Steinwald. Could you discuss the role that e-health
currently plays in disease management, and how it could be
better utilized in the future?
Mr. Guterman. E-health is a very interesting subset of
this. I believe that disease management organizations can use
e-health to improve their ability to manage conditions for an
enrolled population. We have in fact got a demonstration
project underway to look at it the impact of better
communication and computerized communication techniques to
manage populations.
But it has been asserted that we should pay for those
services, because after all how much does an e-mail cost and so
many physicians have found e-mail to be a very effective way of
communicating to the populations. The answer is simply that we
have found that if you pay for it, it will happen. It will not
be restricted to e-mails that are tremendously productive.
So our approach right now is to encourage disease
management organizations to use whatever methods they can
without specifying what methods they should use, so that they
have the freedom to put together a package that is successful
for them, and to just put them at risk for, in most cases, for
some part of their fee or their whole fee and let them deal
with the consequences.
Mr. Lemieux. I was just going to add, in addition to
communications through e-mail, it seems like it would be
important to try and use these programs to leverage the
creation of electronic medical records. It could be a
requirement on the part of patients wishes to enroll in these
programs and getting additional benefits and services, then you
have to have had developed for you an electronic medical record
that you keep. Or it could be part of the physician's
responsibility in one of these programs. But using these
initiatives to spur the creation of electronic medical records
to prevent duplication or working at cross-purposes would be a
nice outcome if it could be arranged.
Mr. Guterman. Actually, in response to Jeff's comment,
there is a tremendous amount of interest, not only in CMS but
throughout the Department of Health and Human Services, in
improving the state of information technology in the health
care sector. We are actually trying to develop some projects
where we can combine the advances in information technology and
the ability to manage people with chronic conditions. The
physician group practice demonstration is one example of that,
where all of the sites in that demonstration will be large,
multispecialty group practices with pretty sophisticated
information technology systems. We think that that is something
that eventually is going to really be a big help in trying to
manage people with chronic conditions as well.
Mr. Steinwald. Here is a question about nutrition, although
I think it maybe could be enlarged a little bit to include the
treatment of potentially beneficial services that are not
currently benefits. The specific question is, nutrition is not
typically incorporated into disease management protocols even
though it has proven to be cost-effective. How will nutrition
be incorporated into disease management programs? As I say, we
could enlarge that to include other kinds of services that are
not typically covered.
Mr. Lemieux. This brings up the same point that Stuart just
made on e-mail. If you provide payment for it, lots of it will
happen. This is why it is so important to make sure that these
initiatives are tightly controlled and evaluated, and that they
involve some investment on the part of the providers or the
patients or their families, so that it is not just a new fee
that Medicare is going to pay that does not really turn out to
change much clinically on the ground.
Mr. Steinwald. If you build it and pay for it, they will
come.
Risk adjustment. How effective are risk adjustment methods
for Medicare beneficiaries, and have they ever been implemented
with any success in disease management models?
Mr. Guterman. Let me address the first part of that, as I
understand it. We are unveiling a new, improved risk-adjustment
methodology beginning January 1 in the Medicare+Choice program.
It is going to be applied to 30 percent of the payment rate for
Medicare+Choice plans. We think that it does pretty well in
capturing the higher cost of riskier beneficiaries. We are
going to continue trying to improve that approach, because
there are always questions of how precise you have to be, and
there are always limits on how precise you can be in terms of
predicting each individual's cost.
But we think it overall makes the system much more amenable
to having payment reflect the populations that are actually
enrolled in various Medicare+Choice plans. We think it is
important, particularly in Medicare+Choice, in encouraging
plans to do more disease management because, as I said, right
now there is a fairly substantial penalty potentially for
attracting beneficiaries who are chronically ill.
In terms of the fee-for-service side, I do not know
particularly of any risk-adjustment applications. I have heard
of disease management organizations stratifying enrollees with
particular conditions. So they will take a congestive heart
failure patient and they will group them into stages and charge
a higher fee for managing patients who have more advanced
stages of their condition. That is a sort of risk adjustment.
But we are not planning to formally apply risk adjustment
beyond anything like that to our fee-for-service disease
management projects.
Mr. Miller. Actually I do not think I have much to add to
that. That is what I was going to say. Within the capitated
plans you have the issue of risk adjustment, and I am not aware
on the fee-for-service side in the private sector specifically
beyond the tiering of risk among patients with certain
conditions. So I am not sure I follow the question.
Mr. Lemieux. Maybe the question is, can risk adjustment
really ever work? Because there is a group of people who are
skeptical that it is really going to work well, and a group of
people who are very optimistic that it will continue to get
better and better and less burdensome for the health plans and
more predictive.
Certainly drug data, if we had it in Medicare, might be
able to help risk adjusters a little bit, and maybe
significantly. I think that any time we are looking forward to
additional use of private health plans, or if our disease
management organizations turn toward extensive case management
and near capitation or partial capitation, then we will have to
have good risk adjusters.
In the past, the health plans have viewed risk adjusters
as--in mid-1990's they felt as though a risk adjuster might not
be so good for them because they felt as though maybe they had
healthier than average patients. Then the risk-adjusters got a
reputation in the community for being very burdensome and also
being coupled with other formula changes that would reduce
private health plan payments, so there became a fear factor.
No one knows but it seems probable, just looking at the
marketplace, that the enrollees of private health plans have
gradually moved toward something that is very close to a
neutral risk with respect to comparisons with the fee-for-
service program. I think if health plans will get over their
fear factor a little bit, begin to really embrace risk
adjustment in the coming five to 10 years as they realize that
it could be to their advantage.
Mr. Steinwald. Just to follow up, what about instead of
risk adjustment, risk assumption? A number of the disease
management models involve the assumption of risk by some entity
who presumably performs services for a fee or a capitation. Is
that essential in your view, to make disease management work in
Medicare, or is that just one of several models?
Mr. Guterman. I think, another way of putting risk
assumption is that you are paying these organizations a price
for managing the beneficiaries that they work with, and that if
they are going to do that effectively I think they have to have
some incentive to be able--some responsibility for the outcome
of what they do. In return for that, I think we would rather
not be very prescriptive about what they do to manage the
conditions as long as it leads to better care.
So I think it is a less intrusive way to give the
organizations the freedom to do what they do, and of course,
protect the program, but also to put the responsibility with
the organizations that are doing the disease management.
Mr. Miller. The hard answer of yes or no, the way you put
the question, would be one that I would have to have the
Commission opine on, but I certainly think that it would be one
of the models that they would want considered in fee-for-
service.
Mr. Steinwald. We have a couple of questions here that
relate to the Medicare bill and the Medicare conference. Jeff,
they are both directed to you although I am sure Mark and Stu
will want to weigh in. What are the best and worst disease
management proposals in the house and Senate bills?
Mr. Lemieux. They are all good. They are trying really hard
to do the right thing. It is not easy. Fee-for-service is not
naturally amenable to chronic care. But the House bill, as I
said, has a permanent program with a regional evaluation and
organization structure that seems very promising. It is more
dedicated toward organization disease management than
individual physician disease management. The Senate bill is
moving toward, or has at least one component that would create
a large demonstration that would involve individual physician
or clinic-based payment systems. As I mentioned, I think they
are working with CMS and others to try and figure out how best
to do that so that physicians can actually enroll and be
monitored so that we will be sure that these programs were
achieving a good cost-effectiveness.
But I think they are doing exactly what needs to be done,
which is experimenting in as many ways as they think might be
workable and then trying to evaluate them very carefully and
see what works, because we cannot sit down and write a law that
will be the right law, but we can try lots of different things
and see what happens.
Mr. Steinwald. Given that the conference will need to
reconcile the House and Senate approaches, where would you like
to see that come out, within the realm of political
feasibility?
Mr. Lemieux. Personally, I think that the more investment
we try to make in this area is probably a good idea. So I would
try and take the House's proposal for a permanent program
instead of the Senate's demonstration for organizationally
based disease management. I think with the right safeguards,
the Senate's physician-based disease management program could
also be included.
As I mentioned before, the extra pot of money that they put
in at the last minute in the Senate after 2009 is also very
intriguing. It would give the flexibility to continue some
programs that did not necessarily save money, but that could be
very effective from a clinical point of view, and that might be
worth considering.
Now the flip side of all this is that the bills already
probably cost a lot more than $400 billion, which was the
ostensible limit, so not all these things are going to be kept
in the bill. I would just hope that disease management is one
of the things that they do try to emphasize.
Mr. Steinwald. A follow-on to that though is, given the
fact that the demonstration programs currently underway have
not yielded findings yet, is it not somewhat premature to be
enacting such a large-scale continuation of demonstrations or
even programmatic changes? That is for all of you.
Mr. Miller. The only thing I would say about that is that
our analysis is headed toward the June 2004 report. If they
have acted prior to that point then I think our research agenda
will be how to make these programs work. For our organization,
we would just turn to that type of analysis as opposed to the
analysis that we are doing now which is exploring the
landscape.
Mr. Steinwald. Of course, the conference might still be in
session in June 2004.
Mr. Miller. I just want to be clear that you said that.
[Laughter.]
Mr. Lemieux. You have to consider the alternative. If the
alternative is to not try to make changes in fee-for-service
Medicare and to allow fee-for-service Medicare to continue to
be a payer that is essentially blind to how things are working
out, and that allows uncoordinated care to flourish, and that
does not compensate coordinated care very well and the extra
efforts that are needed, that is not really a tenable situation
in the long run. So the investment that we make now to try to
teach fee-for-service how to better manage its care and be more
like a private health plan, in a sense, that is a good
investment.
Mr. Steinwald. Let us take a different tack. Here is
someone that is thinking outside the box. We admire that. Is
there any evidence that disease management in Medicare fee-for-
service could reduce variations in practice patterns across the
country? Should this be a goal?
Mr. Miller. I am not aware of any evidence in the
literature that would drive it to reduce the geographic
variations. Arguably, to the extent that there is evidence-
based medicine and that becomes much more widely understood and
practiced, in theory it might have that outcome. But I am not
aware of any evidence that I could cite that it actually does
that.
Mr. Lemieux. Let me just throw in that the goal of having a
decentralized evaluation structure is precisely to drive that
change, to force people to look at different regions of the
country and see what they are doing right and wrong, and try
and mimic those successes and avoid the failures, in a more
systematic way.
Mr. Steinwald. We have not talked a lot about models from
other programs. What about Medicaid in its primary care case
management demonstrations, and even its less well known cash
and counseling demonstrations? Are there any lessons there for
Medicare? Are these models applicable to Medicare?
Mr. Guterman. I am far from an expert on the Medicaid side
of it, but my understanding is that at least some of the
programs are fairly excited about what they are experiencing in
terms of improved results, and savings as well. Certainly, we
need to consider what the Medicaid program is doing in order to
apply to Medicare. In fact there is some overlap because some
of the projects that we have got in development are
specifically designed to look dually eligible Medicare-Medicaid
beneficiaries.
Mr. Steinwald. Anything that constitutes low-hanging fruit?
Let us say from the experience of private sector programs, if
we could figure out a way to fit it into Medicare that would
almost certainly yield benefits to beneficiaries and/or the
program, in particular certain disease areas? I suppose the
ones that were selected for demonstrations were selected for a
reason.
Mr. Miller. I think that the private sector has gone after
specific conditions like congestive heart failure, diabetes,
end-stage renal disease because one of the ways that they can
actually have a benefit relatively soon is to forestall
hospitalizations that are associated with those conditions.
Hospitalization are expensive and they see that as a way to get
savings and improve the quality of care for the beneficiary.
I think your point stands. It is those programs that are
most common because I think they believe that they are the most
approachable on the admissions and most clear where there is
evidence-based medicine to guide the decision.
Mr. Guterman. Actually, I would point out a private sector
program that has hit the street already called Bridges to
Excellence where General Electric and Verizon and Federal
Express and several other large employers have collaborated.
They are currently operating in, I believe, Cincinnati, in the
Cincinnati-Louisville area. Their broader plan is to have three
components, and this brings in the IT component as well. They
are going to have an office system called physician office link
where they are going to have physician practices sign up and if
they meet certain conditions they will be willing to put money
on the table for each patient that is enrolled that is a member
of one of those employers plans. Then they also have separate
components that they are developing for cardiac and for
diabetes patients.
We are interested in that actually and we have been talking
to them about how we could adapt that kind of approach to
Medicare. But it is clear that if these private employers are
willing to put money on a table for better management of these
conditions that they view them--the argument is convincing that
they might reap some benefits from that.
Mr. Steinwald. At this time we have exhausted the questions
from the audience and most of mine. I would like to give the
panel an opportunity, each individually, to make one closing
statement. Maybe the orientation of this is what you think the
audience should take home from today's panel. Stuart, why don't
you begin?
Mr. Guterman. Our approach is to try and work to improve
the way the Medicare program operates, and we are developing
initiatives in several areas that we are excited about. One of
them is disease management. We have an array of disease
management projects. We have got a couple more in the pipeline
that are going to come out soon. There is overlap between
disease management and developments in information technology
that I think ought to be recognized because you need--the
better the information systems available, the more able you are
to manage people with chronic conditions, where you need to
communicate across providers and you need to keep information
over time for these beneficiaries.
Also, I would mention, the initiative we have to collect
information on quality, and to disseminate that information,
and then even in one of our projects to pay for performance
based on that information. These things all link together
because I think the availability of information technology
makes it easier to coordinate care. In fact one might argue
that it is necessary to coordinate care. It also makes it
easier to have information available on quality of care by
allowing providers to report that information and the Medicare
program to disseminate that information and act on that
information eventually by modifying the payment system to
reflect quality of care.
So all those things fit together and we envision that in
the future we will be able to work toward a better Medicare
program. One that is more effective both in terms of delivering
higher quality care and also in terms of being more efficient,
and a more efficient.
Mr. Miller. What I would say is just to remind listeners
that we are the front end of our work and we have just started.
I think the Commission thinks that this is an important area to
look at, both for private plans in Medicare but also for fee-
for-service Medicare, and that it can be useful in
strengthening that program and making it a more rational
purchaser and a more rational provider of care.
While disease management may have promise, there are a lot
of questions that are unanswered about whether the models work
in and of themselves. Even if they do work, how do you
integrate them into Medicare fee-for-service's current
structure? That is some of the issues that I tried to lay out.
I also think, and this is sort of along the lines of what
Stuart was saying, we also have other recommendations that we
believe play into this. We have made recommendations in terms
of linking payment and quality outcomes as well as other areas
where we are trying to make fee-for-service a more prudent
purchaser. This could, if the evidence supports it, become part
of that overall picture.
In the work that we are doing for the June report, if there
is no action on the part of Congress then we will help lay out
the issues for the Congress. If the Congress does take action,
then we have done the groundwork in terms of building data and
looking at the literature to get in behind that and begin to
figure out how to make those programs work.
Mr. Lemieux. I would just like to try and remind the
audience of what we originally meant by Medicare reform. In the
original vision of Medicare reform, which started percolating
the mid-1990's, there were two basic ideas. One was to try and
slow down the growth of Medicare costs, and the other was to
try and accelerate the modernization of Medicare benefits. The
idea was that we should have a large component, or at least a
substantial component of private sector involvement in Medicare
precisely to spur innovation on both of those fronts, to see
what sorts of benefits people like, and see what sorts of cost-
saving possibilities can be invented in the private sector.
Then the other half of it was to try and give the
Government-run program, which most beneficiaries are going to
remain in for decades and decades, a more flexible and
business-like approach and be less involved with having to
write regulations for tightly prescribed Federal laws, and more
involved with running the program as a business. Now the flip
side of giving the Government-run program that sort of
flexibility is accountability, either directly through public
evaluation of results and so on, and then accountability
through competition to some extent.
But the original goals were to do precisely these things
increase the speed of benefit modernization and slow the growth
of costs, and that is why we are having this discussion and
that is why it is so important that we think about disease
management as a way to do both of those things.
Mr. Steinwald. Thank you, panelists. I guess where I come
out, relating back to the title of the panel session, square
peg in a round hole; well, maybe, but sometimes you can force
that peg in and make it stick. We seem to be somewhere between
a feeling of cautious optimism and open-minded skepticism about
the workability of disease management in fee-for-service
Medicare. I guess we do not have the answers to those questions
or that question, but at least we have the prospect of
obtaining more information in the not-too--distant future, and
we certainly have a Congress that is interested in the concept
and pursuing it.
So please join me in thanking our panelists for their
presentations. [Applause.]
Thanks to the Senate Special Committee on Aging for
sponsoring this session. Thank you. We are adjourned.
[Whereupon, at 11:29 a.m., the Committee was adjourned.]
A P P E N D I X
----------
Statement from the American Academy of Family Physicians
The 94,300 members of the American Academy of Family
Physicians offer this statement for the record on ``Disease
Management in Traditional Medicare.''
The Academy applauds your Committee for holding a hearing
that recognizes federal funds are increasingly directed toward
beneficiaries with chronic illness within the Medicare program.
The Robert Wood Johnson Foundation Partnership for Solutions
initiative estimates that about two-thirds of Medicare dollars
go to participants with 5 or more longstanding conditions. This
is a startling figure for a program that not only costs
taxpayers billions of dollars, but also is not geared toward
chronic care management. Examining ``what works'' for chronic
care is crucial as Medicare costs spiral upward and budget
pressures to hold down spending increase.
The Academy has a continuing and expanding interest in
improving chronic care. Specifically, we are in the second
generation of a project entitled ``Quality Enhancement
Program'' (QEP), which includes a focus on improving chronic
illness care within family physicians' offices. The impetus for
this initiative has been the Chronic Care Model (CCM), which
was developed by Edward Wager, MD, who is National Program
Director for the Partnership for Solutions program. The model
is premised on the fact that care for most people with chronic
illness takes place in primary care settings.
In November 2003, the AAFP convened an advisory committee
to discuss specific initiatives to help family physicians
design, document and be recognized for quality health care.
Although quality improvement was the reason for the meeting,
chronic care management was a particular focus. As a result of
this 2-day session, the Academy has decided to improve family
physician training through web-based information, one-on-one
interventions, new and innovative residency programs and the
media. The new program will begin as a demonstration project
targeting 2500 members. While a significant activity for a
private organization, our focus on improving chronic care
management for our members is only a fraction of what federal
support could do to support disease care in the US health car
system.
IMPORTANCE OF PRIMACY CARE TO HEALTH CARE IN THE US
According to the Graham Center for Policy Studies in Family
Practice and Primary Care, 82 percent of Americans have a usual
source of medical care and a majority of them, 62 percent, name
a family physician as that source of care. And, regardless of
self-reported health status, people benefit from having a usual
source of care even if they are uninsured. Specifically,
utilization of health services by individuals with a usual
source of care (doctor's office visits, admission to a hospital
and purchase of prescription medicine) was higher than for
those without a usual source of care. Moreover, difficulties
obtaining health care and doing without needed services was
higher in people without a usual source of care.
Furthermore, according to the Center, of people 65 years
and older who report a usual source of care, 60 percent
identify a family physician. The Medicare population not only
relies heavily on family doctors but 72 percent of respondents
identified an individual. Family physicians were also more
likely to be identified as the source of care for rural and
Hispanic seniors, and those with less than a high school
education.
Moreover, another Graham Center study shows that Americans
depend on family physicians more than any other specialty. Of
the 3142 counties in the United States, 1184 (38 percent) are
designated full or partial county HPSAs, which translates to
more than 41 million Americans. In a hypothetical exercise, the
study removed all family physicians from the US counties. When
this was done, that figure nearly doubled--the large majority
of US counties became full or partial county HPSAs.
These data lead to our view that care for patients with
chronic diseases must be integrated by the primary care
physician--rather than disease management companies or other
entities.
IMPORTANCE OF CARE INTEGRATION
Additional information from Partnership for Solutions
reveals that 66 percent of Americans over the age of 65
currently have a least one chronic condition, and the majority
go on to be afflicted with a number of illnesses. Data from the
Medicare Standard Analytic File (1999) shows that the number of
physicians actually increases with each condition.
Specifically, Medicare beneficiaries without chronic conditions
saw an average of 1.3 physicians in 1999. Beneficiaries with a
single chronic illness saw an average of 3.5 physicians while
those with two saw an average of 4.5 physicians. Seniors with
six chronic conditions saw an average of 9.2 physicians in
1999. These figures argue for a single primary care physician
who can provide cost-effective and integrated care for those in
Medicare.
In addition, the most common chronic conditions, i.e.
hypertension, heart disease, diabetes and arthritis, can be
cared for and controlled in ambulatory settings. When care for
these conditions is poorly coordinated, increased
hospitalization and higher costs are the result. For example,
an article published in the Annals of Internal Medicine (Wolff
et al.) showed that 7 out of 1,000 Medicare beneficiaries with
a single chronic condition were hospitalized for a condition
that was unnecessary. Eighteen out of 1,000 beneficiaries with
two conditions were needlessly admitted, while 36 out of 1,000
seniors with three conditions were sent to the hospital. The
figure for Medicare beneficiaries with six conditions was 161
out of 1,000.
Appropriate care integration could ensure patients get the
care they need, keep them out of the hospital and save them and
the Medicare program substantial amounts of money.
ACTIONS THE FEDERAL GOVERNMENT CAN TAKE
The AAFP has made it a priority to support the efforts of
family physicians in the US who every day coordinate the care
of their chronically ill patients. After all, it is the
nation's family doctors who provide the most care to the
chronically ill. The federal government should support these
efforts, as well. Specifically, this support should come in
additional Medicare demonstration projects that provide
financial, consultative and best practice models to encourage
primary care practices to move from their current acute care
focus to one that is based on managing chronic care. While
individual practices develop novel and innovative ways to
manage care, these efforts are not supported, systematized or
made available nationally.
CURRENT LEGISLATION IN THE US CONGRESS
The House and Senate Medicare bills both contain provisions
on chronic care management. Some provisions focus more on
disease management companies, while others rely more on a
``payment for performance'' structure.
Our concern is that legislation focusing primarily on
disease management companies, absent the integral role of an
integrating physician, is counterproductive. We believe that
federal support of disease management entities will take
chronic care in the wrong direction: many insurers will work
with many organizations, which will manage many diseases--in
many, many different ways. The health care system will become
further fragmented, more costly and without any perceptible
health benefit for the elderly, chronically ill patient, who,
in fact, should be our chief concern.
The Academy is concerned particularly about a focus on
disease management organizations and single, targeted
conditions. Not only do these entities lack experience in
managing multiple conditions, but Medicare beneficiaries
typically have more than one disease.
We are also concerned about the lack of physician
involvement in some of the disease management programs; systems
vary. For example, under the current House Medicare bill
provisions, the Center on Medicare and Medicaid Services (CMS)
would contact the beneficiaries, inform them of these voluntary
disease management programs, and then ask disease management
organizations to follow up absent any requirement for primary
care physician involvement. Not only do patients enroll at
higher rates in these programs when a physician they know is
involved, according to current CMS demonstration projects, but,
more importantly, a patients' primary care physician must work
in collaboration with the patient to coordinate and oversee the
care.
While we understand that there is an impetus to incorporate
the Institute of Medicine's call for patient-centered care,
enrolling patients in several contracts and agreements to
manage their diseases is directly contrary to the notion of
patient-centeredness. What patients tell us is that they want a
physician who will work with them, as an individual, to manage
and coordinate their care.
Consequently, the Academy is more supportive, in general,
of projects that focus on physicians serving eligible
beneficiaries. Specifically, ``pay-for performance'' programs,
such as those in S1, the Senate Medicare bill, that allow
physicians to contract through quality improvement entities
seem better geared toward primary care. Physicians would serve
as the primary contact; maintain health information; meet
outcome measures; promote self-care and report quality and
outcomes measures electronically.
The Academy is at the vanguard of a movement to incorporate
electronic medical records in physicians' offices. This project
is part of our goal to provide tools to physicians so that they
can redesign their offices for our current health system, and
provide better quality care. Specifically, on November 12,
2003, we announced an initiative called ``Partners for
Patients,'' which involves a number of strategic business
alliances to provide electronic health record technology to
medical practices. We believe that in small- or medium-sized
medical practices, an electronic health record system is the
``central nervous system'' for clinical patient management,
including chronic care. The electronic health records systems
that our partners and we are developing will help to ensure
patients receive the most timely, appropriate and efficient
medical care available.
CONCLUSION
The AAFP understands that the demonstration projects in the
Medicare bills are different means to seek additional
information on what constitutes effective chronic care. We also
realize that CMS currently has several demonstrations in
progress that seek answers to the same question, and that the
Medicare Payment Advisory Commission (MedPAC) is also seeking
data on ways to provide quality chronic care in a cost-
effective manner.
As an emerging issue of crucial significance, care for
chronically ill patients should be on everyone's policy agenda.
Our belief, however, is that any new system must use primary
care practice as the foundation on which to build this care.
Americans are currently receiving most of their health care
from primary care physicians and they are satisfied with it.
The federal government's role should be to ensure that primary
care is at the core of any chronic care program.