|
Berenson Web Exclusive
M E D I C A R E R E F O R M :
C H R O N I C C A R E
W E B E X C L U S I V E
22 January 2003
Confronting The Barriers To
Chronic Care Management In Medicare
A proposal to make some changes
to Medicare’s existing structure,
while we await the results of a new round of demonstrations.
Robert A. Berenson and Jane
Horvath
ABSTRACT:
This paper examines the ability of the current Medicare program—both traditional
fee-for-service and risk-based contracting—to address the needs of beneficiaries
with chronic conditions, who represent almost 80 percent of program enrollment.
Grounded in indemnity insurance principles, including concerns about “moral
hazard,” the traditional Medicare program faces difficulty evolving to
support of a chronic care model of health care practice. Although capitation
may be the most desirable platform to support provision of care to beneficiaries
with chronic conditions, the current structural limitations and problems faced
in the Medicare+Choice program limit capitation’s use at this time.
Nobel laureate Kenneth Arrow’s 1963 article, “Uncertainty and the
Welfare Economics of Medical Care,” is often cited to make the case that
the special nature of medical care makes markets for it different from more
typical competitive markets.1 Special characteristics
of these markets include the existence of uncertainty in the incidence of disease
and in the efficacy of treatment, as well as the asymmetry of information between
buyer (the patient) and seller (the physician).2
Less well remembered is a section of Arrow’s paper entitled “The Theory
of Ideal Insurance.” Here he emphasizes the core importance in insurance
of the concept of “moral hazard”: “What is desired in the case
of insurance is that the event against which insurance is taken be out of the
control of the individual.”3 When the insured
person wishes the event to occur and can affect its occurrence, there is moral
hazard. Arrow emphasized that the value of insurance increases as the uncertainty
of the risk being insured against rises. This is the reason, he explains, for
putting greater emphasis on insurance against hospitalization and surgery than
against other forms of medical care.
In this context, Arrow specifically commented on the merits of insurance against
chronic illness. “On a lifetime insurance basis, insurance against chronic
illness makes sense, since this is both highly unpredictable and highly significant
in costs. Among people who already have chronic illness, or symptoms which reliably
indicate it, insurance in the strict sense is probably pointless.” In essence,
he was relying on the moral hazard argument that patients with chronic illness
desire to use health services and that physicians and others want to provide
the care to them, thereby undermining the function and actuarial soundness of
insurance.
Medicare was enacted less than two years after Arrow’s article was published,
and adhering to the prevailing principles of insurance was important to the
political debate at the time. Indeed, the Johnson administration’s King-Anderson
bill provided only for inpatient hospital services. Along the legislative path
to enactment, there were proposals to include physician services but to limit
coverage to those services provided by hospital staff—for example, for
inpatient surgery.4 At the last minute, Wilbur Mills,
chairman of the House Ways and Means Committee, added voluntary Part B coverage
for physician services. Medicare provides more complete coverage for more uncertain
risk—hospital care through Part A— than for more known risk—physician
services through Part B. For the most part, outpatient drugs are not included
at all.5
In statute and operations, the traditional Medicare fee-for-service (FFS) program
reflected indemnity insurance coverage and benefit principles. There was no
explicit commitment to preventing disease or maintaining the health of people
with chronic illness. However, the program—then as now—strays from
Arrow’s concepts of insurance in the strict sense: It is a risk pool segmented
from the general population and one where the prevalence of chronic conditions
is quite high.
Despite some loosening of strict insurance principles through small changes
in benefits and attempts to move the program to risk-based capitation, Medicare
is not well positioned to improve service delivery for patients with chronic
conditions precisely because of its roots in indemnity insurance. Yet the Medicare
program is in reality a program serving people with chronic conditions—typically,
multiple chronic conditions—for whom traditional indemnity insurance principles
and coverage are not appropriate and whose health status presents a challenge
for both cost and quality of care. A recent analysis using 1999 Medicare claims
data showed that about 78 percent of Medicare beneficiaries have at least one
chronic disease; almost 32 percent have four or more, and they drive almost
79 percent of program spending.6 Among elderly Medicare
beneficiaries (not the younger disabled), Jennifer Wolff and colleagues found
that the more chronic conditions a patient has, the greater the likelihood of
hospitalization for an ambulatory care–sensitive event.7
Medicare is geared toward paying for these hospitalizations as acute events
but not for activities that might reduce the need for hospitalization, many
of which are correlated to chronic conditions.
In contrast to today’s Medicare program, with its acute care orientation,
a Medicare chronic care model would use payment and coverage tools to deliver
an integrated array of services across settings designed to prevent, or delay,
declines in functional and health status. When an acute episode of illness does
occur, the chronic care model would work to return the patient to the highest
possible level of functioning. But for the most part, Medicare is precluded
from applying the coverage, payment, and delivery system policy tools some private
health plans and provider groups now use to manage care more rationally and
effectively for special populations with ongoing care needs.
Moral Hazard Concerns Persist
To illustrate the difficulties a traditional insurer like Medicare has in transforming
its policies and procedures to support the delivery of care to patients with
chronic conditions, consider what might at first appear to be a straightforward
decision to reimburse physicians and other professionals for telephone and e-mail
communications with patients. In chronic disease care management models, phone
calls have been viewed as essential to high-quality and efficient care.8
Unfortunately, from the viewpoint of an indemnity payer, coverage and payment
for telephone calls and e-mail raise a series of troubling issues. For both
provider and payer, the transaction costs associated with submitting, paying,
and collecting in most cases would be far more than the actual dollar amount
of the reimbursement. In addition, there could be major problems achieving program
integrity—assuring that the payments were being made appropriately for
services actually rendered. The audit activities that would need to be established
to assure proper payments for telephone and e-mail communications would be daunting
and certainly more intrusive even than the much-criticized oversight requirements
for relatively straightforward office visits.9
But most importantly, paying for routine phone calls and e-mail communications
would likely produce an unprecedented moral hazard problem. It is one thing
to cover preventive services that beneficiaries and providers desire and can
control, although even here, statutory provisions typically place evidence-based
limits on the frequency with which these prevention services may be provided.
For discrete services that involve a person’s physical presence, such as
office visits, patients typically experience “time costs,” inconvenience,
discomfort, and other unpleasant effects, which act as a natural brake on excessive
utilization. Not so for phone calls and e-mail messages from the comfort of
home.10
From the practitioner’s point of view, “excessive” patient-generated
phone calls and e-mail messages to the professional team caring for patients’
chronic diseases would probably be desirable, consistent with the importance
of promoting patient “self-management” responsibility and skills.11
When, as now, telephone and e-mail communication is not reimbursed, conscientious
physicians who increase the amount of such contacts will surely suffer financially.
In a very real sense, then, the FFS payment restrictions on reimbursement for
nonvisit contacts freezes innovation in how clinical care is practiced. Yet,
from the indemnity payer’s point of view, the financial exposure could
be disastrous, or, in a context where there are spending limits, such as in
the Medicare payment system for physicians, spending could shift among physicians
in politically unsustainable ways.12
Benefit Limitations For Chronic Care
Beyond issues of moral hazard and the indemnity nature of Medicare, another
limitation the program is that it still lacks coverage for most prescription
drugs. Even though there are certain statutory and regulatory loopholes that
permit limited, albeit important, coverage for certain drugs that are used for
specific chronic diseases because the drugs are not self-administered and are
provided “incident to” a physician service, prescription drugs that
are the staples of chronic disease management are effectively not covered. The
1999 surgeon general’s report on mental illness emphasized the point that
the most important improvement in Medicare for care for beneficiaries with chronic
problems, such as depression, would be a prescription drug benefit.13
Many beneficiaries obtain reasonable drug coverage through supplemental coverage
particularly retiree health benefits and enrollment in Medicare+Choice (M+C)
managed care plans.14 However, retiree health programs
are being sharply scaled back. M+C is experiencing difficulty, and most plans
are reducing the generosity of drug benefits, to reduce costs and to avoid adverse
selection.15
Although much of the debate about Medicare prescription drugs involves concerns
about the cost of such a benefit expansion, prescription drugs are discrete,
clearly identifiable services for which beneficiaries will face substantial
out-of-pocket costs. Prescription drug coverage has become commonplace in commercial
products based in indemnity insurance; it is the kind of benefit that Medicare,
relying on intermediaries—in this case, pharmacy benefit managers (PBMs)—is
accustomed to providing. The challenges raised by expanding Medicare benefits
to include prescription drugs are substantively different from those that accompany
new, improved forms of care delivery to patients with multiple chronic conditions.
The Chronic Care Improvement Model
Edward Wagner has described a chronic care improvement model that, in contrast
to typical medical practice, emphasizes early identification of patients at
risk through specialized assessment tools; greater attention to treatment planning
that provides a schedule of tasks and delineation of roles; evidence-based clinical
management; greater attention to techniques that promote patient self-monitoring;
and sustained, proactive follow-up.16
Implementing this model would require important delivery system changes, including
greater reliance on clinical information systems; patient self-management interventions
that rely on expanded responsibilities for nurses in education and patient support;
delivery system redesign that modifies traditional practice roles and promotes
a team orientation to care; and various decision-support aids.
Most of these services would not be covered or reimbursed under current statutory
authority. The Centers for Medicare and Medicaid Services (CMS) has procedures
for deciding whether a service will be covered and therefore reimbursed.17
The agency first decides if the proposed service fits into a statutorily established
benefit category. For example, Medicare does not cover self-administered outpatient
drugs or preventive tests, regardless of their diagnostic and therapeutic benefit,
because there is no basis for these benefit categories in legislation. If there
is an appropriate benefit category for a new service, then the CMS and its contractors
must determine whether the service is “reasonable and necessary” to
diagnose or treat an illness or injury. Explicit criteria for this decision
do not yet exist, despite the CMS’s attempt to establish criteria through
rule making, first in 1989 and then in 1999.18
Once a proposed service is found reasonable and necessary, it must be given
an appropriate code. The codes serve as the basis for determining payment for
services both in episode-based payment systems, such as diagnosis-related groups
(DRGs) for inpatient care, and based on fee schedules and allowed charges, such
as those for paying hospital outpatient departments and physicians. Finally,
once a service is coded, it is eligible for payment, whether under episode-based
payments (for example, hospitals and home health agencies) or under fee schedules,
which are used for physician services, clinical laboratory services, and durable
medical supplies.
The Medicare statute is very specific about which “providers” and
“suppliers” are eligible to be paid and under what circumstances.19
Non-physician personnel not specifically recognized as eligible to receive Medicare
payments might be recognized if they provided services “incident to”
a physician’s service. But the “incident to” provision is very
narrow: “services and supplies (including drugs and biologicals which cannot,
as determined in accordance with regulations, be self-administered) furnished
as an incident to a physician’s professional service, of kinds which are
commonly furnished in physicians’ offices and are commonly either rendered
without charge or included in the physicians’ bill” [Section 1861
(s)(2)(A)].
The CMS’s elaborate, statute-based review of new procedures would make
the kind of services in Wagner’s chronic care model difficult to support,
without major legislative change. For example, patient education to promote
self-management is not a covered benefit, except in very limited circumstances,
such as diabetes education. More generally, the “incident to” provisions
might restrict the kinds of activities for which nonphysicians could be reimbursed.
Similarly, multidisciplinary team conferences to review and plan would likely
face a concern that the statute only contemplated reimbursement for services
provided to patients, not services about patients. In short, the rules governing
benefits and payments in Medicare, based in statute, limit innovative approaches
to the care of beneficiaries with chronic conditions.
A statutory change to the traditional FFS program to support the chronic care
model, while desirable, should not be undertaken lightly. Even seemingly small
changes in Medicare have great fiscal impact and tend to affect other payers
and medical practice. Payment per service generally requires fairly precise
service standardization—that is, an expectation that the content of the
service provided by thousands of professionals who bill Medicare is comparable.
To this point, the content of multidisciplinary team conferences and patient
education, the application of information technology, and other innovations
are not standardized. Indeed, it may turn out that different health care organizations,
with different cultures, will (and should) adopt different approaches to adapting
the chronic care model to their own circumstances, thereby making FFS reimbursement
for the components inappropriate.
Inability To Influence The Delivery System
The traditional Medicare program has very limited flexibility to influence the
nature of the health care physicians and other health professionals actually
provide. As part of the political deal to achieve passage of Medicare, Section
1801 of the original Medicare statute established that “nothing in this
title shall be construed to authorize any federal officer or employee to exercise
any control over the practice of medicine or the manner in which medical services
are provided.” Section 1802 provided guarantee of beneficiary freedom of
choice—in current parlance, Medicare was to be an “any willing provider”
program.
Although the federal government was authorized to exert some regulatory authority
in some areas (for example, Conditions of Participation for Part A providers,
enforcement of the False Claims Act), the traditional program remains a passive
payer, precluded from using even basic managed care tools to try to induce the
delivery system to improve beneficiaries’ care. Thus, for example, the
agency administering Medicare cannot designate certain “centers of excellence”
for the provision of chronic disease care and provide these particular institutions
additional payment and greater flexibility in how services are provided.20
Rather, as a general proposition, the program rules must be applied uniformly
across the country. Exemplary performance cannot be rewarded, while poor performance
is tolerated. This general program constraint naturally makes policymakers reluctant
to provide a new set of services that constitute state-of-the-art care for chronically
ill people when those services would require a change in clinical practice to
improve clinical coordination and ongoing care management. Although particular,
innovative professionals would provide additional care coordination services
in a highly competent manner that would improve quality efficiently, under current
rules all professionals with the correct license would be eligible for additional
payments, whether or not they actually attempted to implement the principles
of chronic care management.
Another barrier in the way of supporting care for chronically ill beneficiaries
is the program’s predominant orientation toward providers’ interests.
There are a number of manifestations of this provider orientation. Improved,
often prospective, administered pricing systems have been implemented for most
providers, but these provider-specific payment systems typically pay providers
based on their historical costs, regardless of patient benefit. Thus, for example,
three different rates are paid for the same ambulatory surgical procedure, depending
upon whether it is performed in a hospital outpatient department, an ambulatory
surgical center, or a physician’s office.
Reliance on FFS reimbursement for services limits delivery system innovations
that are available outside Medicare. Even the prospective payment systems discussed
above may not improve the delivery of chronic care, in that, as in the straight
FFS payments these new systems replaced, the silo nature of clinical practice
is still reinforced: “These different payment structures…create strong
incentives for providers to focus inward on their own activities and function
in self-serving ways, regardless of the cumulative effect on costs across settings
or on the overall quality of care received by a patient with multiple providers.”21
Capitated Prepayment As An Alternative
The logic of capitation as a platform for launching innovations in chronic care
is compelling. “In a capitated environment, organizations bearing financial
risk have strong financial incentives to identify their high-risk members early
and to provide them with special care designed to optimize their health and
avert health-related crises. They have longer-range incentives to promote continued
good health among older enrollees who are not chronically ill.”22
Advocates of market reform built on managed competition principles had envisioned
delivery systems that integrated the insurance function with the delivery of
care and were recognized as clearly identifiable health care delivery organizations
with distinctive characteristics that differentiated them from other competitors
in the market for enrollees who had free choice of all health plans. Although
not all of these competing organizations would actually engage in mutually exclusive
contracting between health plan and provider groups, as in the Kaiser Permanente
model, there nevertheless was an expectation that health plans and provider
groups of all shapes and sizes would get together in special relationships.
In a system of health plans contracting with one or at most a few provider organizations,
capitation to the provider organization would then be a logical method of payment,
under which an actual or virtual multispecialty physician group would be able
to redesign care to better serve patients with chronic conditions, unconstrained
by the inevitable limitations of FFS reimbursement. As we all know, markets
did not proceed as many had envisioned. Instead of engaging physicians in special
relationships, most health plans wound up essentially contracting with all available
physicians and hospitals, more or less reverting back to the status of claims-paying
insurance companies. The result is that the quality of care provided to patients
with chronic disease by health maintenance organizations (HMOs) is similar to
that provided in the FFS sector.23
Assuming that performance could improve, to pursue private-plan contracting
as a prime vehicle for better chronic care would still require an overhaul of
the current regulatory regime and payment methodologies that govern M+C.24
Given the current instability in the M+C program and the uncertainty of adoption
of health status–based risk-adjusted payment, a private health plan–based,
capitated approach to introducing innovation in the care of chronically ill
Medicare beneficiaries is in doubt.
In addition to risk contracting, a number of care coordination demonstrations
involving capitation payment have taken place in Medicare. These programs have
focused on a particular subgroup of frail elderly beneficiaries who are nursing
home residents or candidates for nursing home placement, and, as such, they
represent programs focused on dually eligible Medicare and Medicaid beneficiaries.
The evaluations of these demonstrations have been mixed, with only the Program
of All-Inclusive Care for the Elderly (PACE) now incorporated as a standard
M+C option with unique payment features. Although these demonstrations have
shown promise, and PACE has moved out of demonstration status, they affect very
few beneficiaries, which raises the issue of how generalizable they are to much
of the Medicare population, even to the frail elderly for whom they were designed.
Further, these programs focus on the needs of beneficiaries who have functional
limitations and are eligible for nursing home care, which raises the issue of
applicability to most of the nearly 80 percent of beneficiaries with chronic
conditions, most of whom reside in the community.
Chronic Care Experimentation In Traditional Medicare
Although capitation approaches would seem to offer the proper platform for new
models of care delivery for patients with chronic disease and the frail elderly,
for the foreseeable future it appears that efforts to improve care must focus
on changes to the traditional FFS program, despite its serious limitations.
Medicare’s roots in indemnity insurance limit the program’s ability
to be an agent for change in the delivery of health care, although attempts
have been made in the traditional program to spur such delivery system change.
The CMS lacks the authority to do many things that might improve reimbursement
for practitioners who target the problems of patients with chronic illness.
For example, by law it must pay all physicians the same amount for the same
service; it can’t pay differentially based on specialty or performance.
By long-established convention and contract, the CMS uses the Current Procedural
Terminology (CPT) coding system of the American Medical Association for virtually
all physician services, despite code definition problems. Finally, grounded
in the statutorily based resource-based relative value scale (RBRVS) methodology,
the CMS cannot modify payment rates to try to achieve specific policy goals,
such as increasing the volume of home visits by physicians.
The National Academy of
Social Insurance, the National Bipartisan Commission on the Future of Medicare,
recent administrations, and relevant House and Senate committees all have discussed
more fundamental change to the traditional FFS program, generally referred to
as “Medicare modernization.” Many of the modernization proposals involve
providing additional benefits, such as prescription drugs; contracting reform;
competitive bidding for certain services; and Medicare governance changes, among
others. Regarding chronic care, there were proposals to grant the CMS new legislative
authority to establish case management or care coordination programs, disease
management programs, and provider-physician collaborations—that is, a form
of bundled hospital/physician payment.
These approaches would represent basic departures for the Medicare program,
mostly because they involve completely new payment approaches. The proposals
come in many varieties. Some envision case management as modeled after primary
care case management (PCCM) programs in Medicaid and some HMOs, in which a designated
primary care physician either is paid a case management fee or receives a higher
reimbursement schedule to be the patient’s care coordinator.
Most of these models do not involve primary care capitation payments for services
now paid for under the physician fee schedule, although the traditional Medicare
program pays end-stage renal disease (ESRD) facilities and renal physicians
a monthly capitation payment, not FFS. Other case manager or care coordination
proposals would designate nonphysicians, usually nurses, as care coordinators.
Some suggest that local agencies, such as Area Agencies on Aging and health
departments, might play a role in organizing the coordination of care, in addition
to provider groups.25 Based on the apparent success
of private-sector targeted disease management programs for specific conditions,
such as congestive heart failure (CHF), diabetes, and asthma, there is interest
in direct contracting with disease management firms to help manage care in Medicare.
Experiences With Chronic Care Demonstrations
In demonstrations, the CMS has had some experience with these new approaches
to chronic care management. In 1990 Congress mandated the provision of case
management services to Medicare beneficiaries with high-cost illness. The demonstrations
were conducted for two years ending in November 1995. The three demonstrations
were a CHF-focused program administered by a large insurer, a CHF and chronic
obstructive pulmonary disease (COPD) program administered by a peer review organization
(PRO), and a program targeting eight diagnostic groups administered by a tertiary
care teaching hospital. All sites included patient assessments, coordination
of care, patient self-management, caregiver education, and psychological supports.
One important finding was beneficiaries’ lack of interest in participating,
in many cases because of resistance by physicians involved, especially in the
program sponsored by the PRO. In summary, these case management demonstrations
found no notable effects on costs, health outcomes, or levels of self-care.26
The Balanced Budget Act (BBA) of 1997 required the CMS (then HCFA) to evaluate
best practices in the private sector for methods of coordinated care, and then,
based on the findings of the study, design demonstrations to evaluate models
of care coordination for beneficiaries with chronic illness. Importantly, Medicare
spending under the demonstrations was not to exceed what would have been spent
without their presence.
Mathematica Policy Research (MPR) conducted the review and evaluation for the
CMS and identified two main types of coordinated care programs—case management
and disease management—and developed a conceptual framework applicable
to these delivery models.27 This MPR review found
that patients amenable to the two interventions differ in important ways. Case
management programs serve a more select group of frail, disabled patients, at
risk for recurrent, costly, adverse medical events.28
Disease management programs target people who tend to have a specific condition,
although the programs must be able to address common comorbidities. Corresponding
to the different populations, case management programs tend to individualize
care, relying heavily on the judgment of the case manager. In contrast, disease
management programs tend to be highly structured and emphasize the use of structured
protocols and clinical guidelines.
Another differentiating factor in the efficacy of case management programs is
whether the programs build on generic, unstructured case management or clinically
sophisticated case management, regardless of whether directed at a specific
disease or to frail people with functional limitations. According to Wagner,
the literature demonstrates that generic case management proved ineffective
at reducing costs or altering the utilization patterns in rigorous trials, whereas
clinically sophisticated case management performed by nurses with specialized
training in geriatrics was effective.29
Based on the MPR report, the CMS issued a Solicitation for Proposals for the
Coordinated Care Demonstration in July 2000 and recently made fifteen awards.
Combined with two awards made last year for a BBA-mandated case management demonstration,
the CMS effectively will oversee seventeen demonstrations of various case management
and disease management demonstrations.
The July 2000 notice solicited formal public comment on the contents of the
notice and demonstration design. These comments are important for consideration
of how Medicare can best promote improved care for chronic illness. Many comments
referred to the difficulties of providing care management services under the
current Medicare FFS payment system. Almost all respondents suggested some sort
of risk-bearing system in which providers would be paid a fixed fee per enrollee
and would share in any savings to Medicare. Some also suggested that reimbursement
be linked to patient outcomes.
Accordingly, the CMS chose to use a monthly all-inclusive rate to pay for the
proposed coordinated care services, which might include coordination with community-based
services, transportation, medications, noncovered home visits, and equipment.
Statutory Medicare services will be reimbursed as usual.
Lastly, the Medicare, Medicaid, and SCHIP Benefits Improvement and Protection
Act (BIPA) of 2000 set up additional demonstration programs testing disease
management programs for beneficiaries with advanced-stage CHF, diabetes, and
coronary heart disease. The CMS recently solicited applications under this program,
which could result in three awards covering up to 30,000 beneficiaries at a
time. Prescription drugs would be covered under these demonstrations, and the
demonstrations would be required to meet strict budget-neutrality requirements.
Opportunities For Incremental Improvements In FFS Medicare
While waiting for results of the major coordinated care demonstrations and considering
a reorientation of M+C to reward plans that manage the care of beneficiaries
with a high burden of chronic illness, there may opportunities to make modest
changes in the current, indemnity-oriented Medicare program.
A fifth of beneficiaries have five or more chronic conditions and account for
nearly two-thirds of Medicare spending. Beneficiaries with multiple chronic
conditions have very high service use; this indicates that clinical care coordination
may be lacking. For example, beneficiaries with five or more conditions fill
an average of forty-nine prescriptions in a year, have an average of thirty-seven
physician visits, see fourteen different or unique physicians in a year, and
log more than 7,000 inpatient days per 1,000 people.30
Although service use is high, their care is not coordinated across providers
and settings, for all of the reasons discussed earlier. We need to think about
incremental ways to improve care coordination for medically complex beneficiaries
in the absence of benefit expansion or other program reforms.
Beneficiaries with five or more serious chronic conditions could be the initial
target group for any incremental policy change. Further analysis could permit
the creation of a subset of chronic conditions, associated with higher costs
and with provision of services by many professionals, that would be used as
“qualifying conditions” to determine eligibility for additional payment
or services. Consistent with planned implementation of health status–based
risk adjustment in M+C, physicians would be expected to identify patient diagnoses
through assessment and documentation, within their scopes of practice.
Increased payment for office
visits. For beneficiaries
who qualify based on the presence of the requisite number of serious conditions,
payments for office- based care would be higher. This increased payment could
be billed by any and all unique physicians who see the patient for each office
visit. The higher payment would compensate physicians more generously for the
greater amount of time they and their staffs need to care for patients with
serious chronic conditions and to coordinate with other professionals caring
for the same patient.31
Clinical care management.
Unlike a broad-based payment available to all physicians, a more targeted and
intensive approach might be a clinical care management model, whereby one treating
physician accepts the added responsibility to coordinate the clinical care provided
by all treating physicians in return for an administrative payment.32
In the managed care environment, this approach has received the pejorative appellation
of a “gatekeeper,” because of the emphasis on requiring the designated
physician to approve all referrals to specialists and for many ancillary tests
and procedures (although it should be noted that twelve European countries require
patients to see a designated physician—a gatekeeper—who functions
as the defined point of entry to secondary care).33
As noted earlier, many Medicaid programs have a similar mechanism, now called
the PCCM model. A Medicaid beneficiary selects or is otherwise assigned to a
primary care physician, who acts as a care coordinator and primary care provider.
Physicians in this role are paid in one of two ways: a monthly per person management
fee, which is separate and apart from billing for specific services rendered,
or a monthly capitation payment to the physician for a range of primary care
services and care coordination activities.
Applying the PCCM model
to Medicare. A
number of design issues would have to be considered in applying a PCCM-type
approach to Medicare. Whereas enrollment in these programs is typically required
in managed care and in Medicaid applications, the strong Medicare tradition
would be to make it voluntary for the beneficiary, perhaps in exchange for reduction
of some cost-sharing obligations or discount off of the Part B premium. Although
the desirability of having a single physician coordinate care might be relevant
for all Medicare beneficiaries and might be promoted in program guidance and
educational materials, specific reductions in cost sharing or premium requirements
might be limited to people with a certain number of chronic conditions, as discussed
above.
For the clinical care manager to have any meaningful ability to reduce unnecessary
services, as well as to reduce the likelihood of errors that result from care
provided by too many noncommunicating professionals, any Medicare PCCM-type
program should require the designated clinical care manager to have prior authorization
authority. However, in contrast to the manner in which many gatekeeper programs
in managed care plans work, a Medicare program could be designed to permit much
more flexibility for care managers’ decision making. For example, a Medicare
clinical care manager might be allowed to selectively designate certain chronic
problems, such as glaucoma, for ongoing care from a specialist (in this case,
an ophthalmologist), without the need for recurring authorizations.
For their part, physicians could participate as clinical care managers to the
extent that they agreed to follow certain administrative procedures to track
and monitor all aspects of a beneficiary’s care, act as a referral agent,
receive and coordinate clinical reports from others involved in the patient’s
care, maintain a robust medical record, be available to provide greater consultation
time surrounding a qualified beneficiary’s care, and have appropriate staff
and administrative capabilities to do so. An outstanding issue is whether specialists
who agree to these requirements should be designated as the care managing physician
for Medicare beneficiaries, given the prevalence of certain chronic conditions
that are commonly cared for by specialists, such as cardiologists. Precluding
specialist participation and inserting yet another physician (a generalist care
manager) to act as the clinical care coordinator into the mix of specialists
already caring for a beneficiary with multiple chronic conditions may not be
warranted if one of the specialists is willing and able to carry out the coordination
functions this model requires.
Supportive services. This model could be expanded to facilitate leveraging
non– Medicare covered supportive services for the benefit of patients needing
such services. To that end, the administrative payment could go to providers
willing to have staff knowledgeable about the availability of other resources
in the community, and how to get access to those resources, make referrals or
coordinate access to those services. These supportive services can be very important
to successful medical treatment and can improve quality of life.
Payment options. A logical payment approach for Medicare would be a monthly
fee for care management services, in addition to standard FFS reimbursements
for discrete physician services covered under the Medicare fee schedule. This
approach does not require changes to the physician fee schedule and does not
inherently induce greater service use to gain access to the clinical care coordination
services. An alternative would be to bundle standard primary care services into
a much larger monthly capitation amount. As discussed earlier, capitation provides
greater flexibility than FFS payment does and may be more conducive to implementing
delivery system innovation, along the lines of the chronic care model. However,
primary care capitation can have untoward incentives to skimp on care and, depending
upon whether capitated physicians are at risk for referrals and hospitalizations,
might actually provide an incentive for inappropriate referrals.
Both payment options could be employed for maximum reach and effect. There could
be a monthly capitation care management fee paid to designated physicians while
maintaining FFS reimbursement for discrete services by physicians practicing
in solo and small-group practice, while encouraging the expanded capitation
option for physicians practicing in large multispecialty group practices that
have the administrative infrastructure and financial wherewithal to manage larger
capitation amounts. Under either payment structure, the model would require
some sort of provider designation, so that participants would have to meet certain
standards for care, quality, and administrative capabilities, a form of conditions
of participation or eligibility criteria that has generally not been applied
to physicians. Depending on how the criteria are structured, the requirements,
coupled with new payments, could promote greater use of Web-based medical communication
and documentation systems.
A new home visit benefit.
Beyond administrative structures, it also may be appropriate to consider benefit
design that can facilitate greater clinical care coordination and management.
One such approach would be a modified home visit benefit. The current home health
benefit is for people in need of extended home nursing and personal care services
and who meet a technical definition of homebound. The current sixty-day
episode-of-care payment reflects the benefit’s extended nature.
There may be need for another type of benefit that is not as extensive or intensive
as the current home health benefit. Although current rules require direct physician
supervision of ancillary personnel seeing Medicare patients, such direct supervision
is not practical in some circumstances. A modified benefit would promote the
chronic care approach if physicians could authorize their office nurses or physician
assistants to periodically conduct home visits to check on patients. This benefit,
then, would be limited in scope to infrequent medical monitoring when a patient
is not able to come to the office because of temporary or otherwise acute health
conditions, but it would allow the physician to have more direct knowledge of
a patient’s health status and functioning than would be possible if the
service were delivered through a separate agency.
The benefit might need some limitations, perhaps by allowing a limited number
of visits per beneficiary per year, by defining the qualifications of practitioners
who might make such home visits, and by restricting services, perhaps to medical
assessment, medical monitoring, and medication management. Further, the visits
might be limited to follow-up associated with acute exacerbations of chronic
conditions or to periods when a patient’s treatments have been altered
because of a change in health status.
This benefit is not intended to replace the current home health benefit but
rather is intended to be a limited tool by which physicians can better coordinate
care. The benefit must be crafted so that it provides a useful tool for greater
clinical care coordination and does not spawn a new cottage industry. In addition,
payment for any such benefit would need to recognize differential costs and
efficiencies between rural and urban areas. Although the coordinated care and
disease management demonstrations correctly are designed to implement broad-based
coordination, it is likely that information gained in the demonstrations would
assist in crafting specifications for this narrow expansion of permitted home
visits.
In this paper we have attempted
to show that there is a mismatch between the chronic care needs of the majority
of Medicare beneficiaries and Medicare’s historical grounding in an indemnity
insurance model. Although the more innovative proposed changes to the program
would involve moving toward organizational accountability for caring for beneficiaries
with chronic conditions, through capitated payments—either for all covered
services or for the services specifically related to care coordination activities—such
changes will depend upon results of demonstrations, some of which have begun
only recently. In the meantime, there may be an opportunity to make incremental,
yet important, changes to the current traditional program that would better
recognize the needs of beneficiaries with multiple chronic conditions.
This paper is based on a paper Robert Berenson wrote for the National Academy
of Social Insurance study panel on Medicare and Chronic Care, and a paper he
and Jane Horvath prepared for the Center for Medicare Advocacy’s March
2002 conference on Medicare coordinated care. The authors thank the many people
who provided helpful comments on the initial papers and two anonymous reviewers.
They especially thank Tom Hoyer, who recently retired from the Centers for Medicare
and Medicaid Services, for his insights on Medicare payment rules.
NOTES
1. K.J. Arrow, “Uncertainty and the Welfare Economics of
Medical Care,” American Economic Review 50, no. 5 (1963): 941–973.
2. The Journal of Health Policy, Politics and Law dedicated
an entire issue (October 2001) to a reconsideration of this seminal article.
3. The term moral hazard is often misconstrued to imply
personal blame or lack of appropriate morality. No such connotation is intended.
4. R.J. Myers, “Why Medicare Part A and Part B, as Well
as Medicaid?” Health Care Financing Review 22, no. 1 (2000): 53–54.
5. R.M. Ball, “What Medicare’s Architects Had in Mind,”
Health Affairs (Winter 1995): 62–72. Reportedly, a main reason prescription
drugs were not included was the relatively high transaction costs associated
with submitting claims for then inexpensive pharmaceuticals. The same issue
pertains to consideration today of direct reimbursement for telephone calls
and e-mail.
6. R.A. Berenson and J. Horvath, “The Clinical Characteristics
of Medicare Beneficiaries and Implications for Medicare Reform” (Paper
presented at Medicare Coordinated Care Conference, sponsored by the Center for
Medicare Advocacy, Washington, D.C., 21–22 March 2002). Among the list
of chronic conditions are those that are real but do not have current clinical
manifestations, such as disorders of lipid metabolism, or are usually easily
managed with low-cost prescription drugs or other interventions, such as hypothyroidism.
Nevertheless, virtually all of these conditions are treated with often costly
medications that need monitoring, and most produce symptoms that benefit from
medical interventions.
7. J. Wolff, B. Starfield, and G. Anderson, “Prevalence,
Expenditures, and Complications of Multiple Chronic Conditions in the Elderly,”
Archives of Internal Medicine (11 November 2002): 2269–2276.
8. E.H. Wagner, B.T. Austin, and M. Von Korff, “Organizing
Care for Patients with Chronic Illness,” Milbank Quarterly 74, no.
4 (1996): 511–544.
9. J.K. Iglehart, “The Centers for Medicare and Medicaid
Services,” New England Journal of Medicine 345, no. 26 (2001): 1920–1924.
A complicating issue would be the collection of the 20 percent coinsurance that
is required under Part B payment rules. If this were required, patients would
have to be billed for cents. If it were waived because the administrative costs
of collection would be too great, any potential deterrent effect of coinsurance
on utilization would disappear.
10. Arthur Garson, dean of academic operations, Baylor College
of Medicine, at the workshop, “Creating a Vision: The Academic Health Center
of the Future,” sponsored by the Commonwealth Fund, 25 October 2001, described
the phenomenon of physicians experiencing “e-mail fatigue” caused
by the increasing onslaught of e-mail from motivated patients.
11. E.H. Wagner, “Care of Older People with Chronic Illness,”
in New Ways to Care for Older People, Building Systems Based on Evidence,
ed. E. Calkins et al. (New York: Springer Publishing Company, 1999).
12. The most straightforward and practical way to compensate
physicians and their staffs for engaging in nonvisit-based communications with
patients would be with a monthly clinical management fee that would be made
when beneficiaries have a high burden of chronic care that needs special coordination.
This approach would, in essence, put the physician at risk for “excessive”
communications.
13. U.S. Department of Health and Human Services, Mental
Health: A Report of the Surgeon General (Rockville, Md.: US Public Health
Service, 1999).
14. Henry J. Kaiser Foundation, “Medicare and Prescription
Drugs: A Chartpack,” 12 June 2002, www.kff.org/content/2002/6048
(2 January 2003).
15. L. Achman and M. Gold, “Medicare+Choice 1999–2001:
An Analysis of Managed Care Plan Withdrawals and Trends in Benefits and Premiums”
(Washington: Mathematica Policy Research, February 2002).
16. Wagner, “Care of Older People.”
17. Lewin Group, “Report 2: The Medicare Payment Process
and Patient Access to Technology,” No. 2 in a Series of Reports Prepared
by the Lewin Group for Advamed (Falls Church, Va.: Advanced Medical Technology
Association, 2000).
18. The proposed rules published for comment in 1989 were withdrawn
and never published in final form. Another set of proposed rules was issued
in 1999 but never finalized. In 2000 the CMS (then HCFA) published a notice
of intent to issue a proposed rule that laid out criteria for covering a proposed
new service. Under this formulation, a proposed product or service would be
covered if it (1) falls within a Medicare benefit category, (2) can demonstrate
medical benefit based on evidence of effectiveness, and (3) provides added value
to the Medicare population. The CMS thus far has not responded to comments on
this notice, which, as in the 1989 notice, introduced notions of cost-effectiveness
in coverage policy.
19. Technically, physicians are considered suppliers to the
Medicare program.
20. Beginning in 1987 the CMS took regulatory action through
a series of national coverage decisions pertaining to transplant procedures
(heart, liver, and lung) that limited medically necessary transplant services
to those performed in facilities meeting certain requirements, a type of center
of excellence. See Centers for Medicare and Medicaid Services, “Medicare
Coverage Policy, NCDs, Transplant Centers: Re-evaluation of Criteria for Medicare
Approval (#CAG-0061N),” 26 July 2002, www.cms.hhs.gov/coverage/8b3-aa4.asp
(2 January 2003).
21. R.J. Bringewatt, “Concept Paper—Modernizing Medicare
for People with Chronic Conditions” (Unpublished paper, National Chronic
Care Consortium, Minneapolis, August 2001).
22. C. Boult et al., “Innovative Healthcare for Chronically
Ill Older Persons: Results of a National Survey,” American Journal of
Managed Care 5, no. 9 (1999): 1163–1172.
23. A few studies, including an oft-cited one by John Ware
and colleagues, have found that HMO patients with chronic illness had somewhat
worse outcomes than those in FFS have had. See J.E. Ware et al., “Differences
in Four-Year Health Outcomes for Elderly and Poor Chronically Ill Patients Treated
in HMO and Fee-for-Service Systems,” Journal of the American Medical
Association 276, no. 13 (1996): 1039–1047. However, a comprehensive
literature review by Robert Miller and Hal Luft comparing HMO and FFS performance,
which was updated in 2002, consistently finds little difference in performance,
even when studies are stratified by categories, such as prevention, chronic
care, heart disease, and so on. See R.H. Miller and H.S. Luft, “HMO Plan
Performance Update: An Analysis of the Literature, 1997–2000,” Health
Affairs (July/Aug 2002): 63–86. Indeed, Stephen Jencks and colleagues
found that interstate variations on twenty-four quality measures in the Medicare
FFS program, which included measures of chronic disease management, were far
greater than the minor variations found in studies comparing HMO and FFS performance.
See S.F. Jencks et al., “Quality of Medical Care Delivered to Medicare
Beneficiaries,” Journal of the American Medical Association 284,
no. 13 (2000): 1670–1676.
24. R.A. Berenson, “Medicare+Choice: Doubling or Disappearing?”
28 November 2001, www.healthaffairs.org/WebExclusives/Berenson_Web_Excl_112801.htm
(2 January 2003).
25. P.D. Fox, “The Medicare Fee-for-Service System: Opportunities
for Applying Managed Care Techniques,” in Medicare: Preparing for the
Challenges of the Twenty-first Century, ed. R.D. Reischauer, S. Butler,
and J.R. Lave (Washington: National Academy of Social Insurance, 1998).
26. J. Schore, B. Brown, and V. Cheh, “Case Management
for High-Cost Medicare Beneficiaries,” Health Care Financing Review
20, no. 4 (1999): 87–101.
27. A. Chen et al., “Best Practices in Coordinated Care,”
Report to the Health Care Financing Administration, Contract no. HCFA 500-95-0048
(04) (Washington: Mathematica Policy Research, 22 March 2002); and “Medicare
Program; Solicitation for Proposals for the Medicare Coordinated Care Demonstration,”
Federal Register 65, no. 146 (2000): 46466–46473.
28. Terminology in this area is very confusing. This form of
case management, often carried out by nurses, needs to be distinguished from
primary care case management, which refers to a physician, usually a primary
care physician, who usually acts as a gatekeeper, responsible for approving
referrals.
29. Edward Wagner, Center for Health Studies, Group Health
Cooperative of Puget Sound, personal communication, 1 November 2001; J.A. Ferguson
and M. Weinberger, “Case Management Programs in Primary Care,” Journal
of General Internal Medicine 13, no. 2 (1998): 123–126; A.J. Gagnon
et al., “Randomized Controlled Trial of Nurse Case Management of Frail
Older People,” Journal of the American Geriatrics Society 47, no.
9 (1999): 1118–1124; S.G. Leveille et al., “Preventing Disability
and Managing Chronic Illness in Frail Older Adults: A Randomized Trial of a
Community-Based Partnership with Primary Care,” Journal of the American
Geriatrics Society 46, no. 10 (1998): 1191–1198; M.W. Rich, “Heart
Failure Disease Management Programs: Efficacy and Limitations,” American
Journal of Medicine 110, no. 5 (1998): 410–412; and A.E. Stuck et al.,
“A Trial of Annual In-Home Comprehensive Geriatric Assessments for Elderly
People Living in the Community,” New England Journal of Medicine
333, no. 18 (1995): 1184–1189.
30. Berenson and Horvath, “The Clinical Characteristics
of Medicare Beneficiaries.”
31. A more straightforward approach would be to permit physicians
to apply to a modifier to evaluation and management codes for patients who required
more time than average because of the burden of decision making and coordination
requirements that patients actually presented with. Theoretically, the modifier
could be used to reflect a range of issues that can produce a more complex visit
or consultation, such as language difficulties. Proceeding with a broadly applicable
modifier, however, would raise serious program integrity concerns that the modifier
would be applied much too broadly and inconsistently by different physicians
and would deflect from the objective of recognizing the unique coordination
and patient education activities associated with patients who have multiple
chronic conditions.
32. A clinical care management fee would avoid several of the
problems that arise in adding a new service or benefit, as discussed earlier.
First, this administrative payment would be outside the benefit structure. It
would apply to clinical management on behalf of a specific group of beneficiaries.
Standardization of service across providers is less of an issue in that this
payment is an average to compensate for a range of fairly well known activities.
The monthly payment is structured to acknowledge that the amount of activity
needed on behalf of individual patients will vary over time. Creating a clinical
care management fee would involve provider designation, which, while politically
thorny, builds on the precedent of organ transplant center designation in Medicare.
33. R.B. Saltman and J. Figueras, “Delivering Services
Efficiently,” in European Health Care Reform: Analysis of Current Strategies
WHO Regional Publications, European Series no. 72 (Copenhagen: World Health
Organization Regional Office for Europe, 1997).
Robert Berenson is senior
adviser at AcademyHealth and an adjunct professor at the University of North
Carolina School of Public Health and Duke University. He directed the Center
for Health Plans and Providers at the Centers for Medicare and Medicaid Services
(formerly HCFA) from April 1998 to October 2000. Jane Horvath is deputy director
of Partnership for Solutions at the Johns Hopkins University.
©2003 Project HOPE–The People-to-People Health Foundation, Inc.
|
