Pay-for-Performance / Quality Incentives: Discussion Paper For The MMA 623e Advisory Board

Pay-for-Performance / Quality Incentives
May 24, 2005
Pay-for-Performance /
Quality Incentives
Discussion Paper for the

MMA 623e Advisory Board
Prepared by CMS/ORDI/MDPG
May 24, 2005
Pay-for-performance (P4P) is a major priority for the current Administrator of the Centers for
Medicare & Medicaid Services (CMS) who believes Medicare should seek opportunities to
encourage improvements in the quality of care provided to Medicare beneficiaries.
The ESRD program has a long history of concern for quality of care. Medicare, the National
Institutes of Health, the National Kidney Foundation, the American Society of Nephrology,
the Renal Physicians Association and others have actively participated in efforts to develop
data systems that support the measurement and improvement of quality. These efforts have
identified opportunities for quality improvement and have achieved notable improvements in
patient care and outcomes. These organizations also candidly acknowledge that
opportunities for improvement remain.
The history of quality improvement efforts, the availability of data systems and quality
standards, and consensus on opportunities for quality improvement combine to make
ESRD a good candidate for possible P4P initiatives. ESRD is, in fact, the focus of a CMS
Break-through Initiative in the area of quality. The FistulaFirst initiative is the first of several
break-through initiatives that CMS is undertaking to improve quality and implement the
strategies outlined in the recent quality chasm reports from the Institute of Medicine.
MMA 623e: ESRD Bundled Payment Demonstration
Page 1
May 24, 2005
Quality / P4P Incentives
Purpose of P4P / quality incentives
Traditionally payment reflects process

P4P allows outcome to affect payment
The right care for every patient every time
Safe
Effective
Patient-centered
Timely
Efficient
Equitable
May 24, 2005

Slide 2
P4P represents a fundamental break with traditional thinking about provider payment.
Traditionally, Medicare and other third parties have based payment on the process of care.
The specific services and procedures provided to patients determine provider payment.
Questions about medical necessity and effectiveness enter into coverage decisions, but
payment is traditionally determined by what was done to a patient. The goal of P4P is to
allow outcomewhat treatment does for a patientto influence payment. Under P4P,
providers who do more for their patients in the sense of providing better outcomes would
be paid more.
CMS has defined the goal of its P4P strategy as promoting the right care for every patient
every time. Following the lead of the Institute of Medicine, it defines the right care as care
that is safe, effective, centered on the patients needs, timely, efficient, and equitable (i.e.,
independent of personal characteristics such as gender, ethnicity, geographic location, and
socioeconomic status).
A P4P component of the bundled payment demonstration would seek to address
opportunities for improvement in any or all of these six areas. For example, P4P incentives
might be used to promote safer and more effective care such as reliance on fistulas for
venous access unless fistula is not a viable option for a patient. A significant question,
however, is whether (or the conditions under which) P4P incentives should address aspects
of performance that lie somewhat outside the scope of a bundled payment.
Page 2
May 24, 2005
P4P and Bundled Payment
Relationship of P4P and bundled payment
A matter of perspective
The goal / purpose of bundled payment

The goals / purposes of P4P
P4P as an add-on to bundled payment
Bundled payment as a P4P initiative
How can P4P support bundled payment?
Issues bundled payment cannot address

Issues bundled payment may create
May 24, 2005

Slide 3
The question of what dimensions of performance to address in P4P is large and complex. A
basic question for the bundled payment demonstration concerns the relationship between
the purposes and goals of bundled payment and those of an associated P4P initiative. Are
the goals of P4P independent of those bundled payment? Is P4P something that is merely
grafted on to the bundled payment system? Or is P4P more properly viewed as a
potentially complementaryor even integralcomponent of a bundled payment system?
P4P principles can be seen as supporting the purposes of bundled payment. But it is also
possible to view bundled payment as itself one element of a broader P4P initiative. For
example, the goals of bundled payment include the promotion of:
More effective management of significant complications and co-morbidities of dialysis

(e.g., anemia, infections, etc.).
More efficient use of resources (e.g., appropriate use of EPO, elimination of

duplicative laboratory testing).
Promotion of medical practices that involve less risk for the patient (e.g., conservation
of venous access).
Given the close potential relationship between the goals of bundled payment and P4P, it is
possible that P4P could support the bundled payment demonstration in two ways.
It could offer a means of addressing issues that bundled payment cannot readily
address such as the alignment of incentives among physicians and dialysis facilities.
It could offer a means of mitigating unintended consequences of bundled payment

such as potential adverse effects on quality or access.
Page 3
May 24, 2005
P4P Technical / Policy Questions
What are P4P incentives for?

How should P4P incentives be funded?
How large should P4P incentives be?
How should P4P be earned?
What performance measures will be used?
Whose performance will be measured?
May 24, 2005

Slide 4
Whether and how specific and explicit P4P principles can be used to promote the broader
goals of the bundled payment demonstration is likely to depend on the answers to a number
of narrower technical questions related to the design of any P4P system.
What are the P4P incentives for?
How should the P4P incentives be funded?
How large should the P4P incentives be?
How should the P4P incentive be earned?
Whose performance will be measured?
What aspects of performance will be measured?
Page 4
May 24, 2005
What are P4P incentives for?
Using P4P to support bundled payment
PPS creates incentives to improve efficiency

P4P creates explicit incentive to improve quality
Adequacy of renal replacement therapy

Effective management of co-morbidity
Using P4P to reach beyond the bundle
Management of vascular access

Coordination of care for underlying diseases
Encouragement of appropriate modalities
Appropriate use of inpatient hospital care
May 24, 2005

Slide 5
It has already been noted that the purposes of the bundled payment demonstration can,
broadly speaking, be seen as part of an effort to improve quality by promoting safer, more
effective and more efficient care. The quality objectives that guide CMSs P4P initiatives
also influence the design of the bundled payment system. Ideally, a well-designed
prospective payment system would promote safe, effective, patient-centered, timely,
efficient, and equitable care. However, additional and more explicit incentives to improve
quality could be incorporated into the bundled payment system.
Prospective payment creates strong incentives to improve efficiency. Improvements in

efficiency may be seen, however, as being achieved at some risk to quality or access.
P4P might be used to mitigate these risks.
P4P incentives might be designed to explicitly recognize superior performance on the

adequacy of RRT, the management of anemia, management of infection, or other
measures of quality and access.
There are limits to the improvements in quality that can be achieved within the feasible
scope of a bundled payment. Indeed, some improvements in safety, efficiency,
effectiveness, etc., involve providers and resources that are outside the scope of the
bundled payment. The question is whether and how to use P4P to achieve improvements in
quality that involve resources outside the bundled payment. For example:
Management of vascular access and underlying diseases involves providers that are
only indirectly influenced by dialysis facilities and associated practitioners.
It may be difficult to encourage adoption of appropriate modalities relying solely on the

resources and incentives inherent in the bundled payment system.
Encouraging appropriate use of inpatient hospital care may be more effective if

savings from eliminating inappropriate use were made available to dialysis facilities
and their associated practitioners.
Page 5
May 24, 2005
How should P4P be funded?
Implicit efficiency incentives
Efficiency of renal replacement therapy

Effectiveness of renal replacement therapy
Management of complications / co-morbidity
Other dimensions: safe, timely, patient-centered
Explicit quality improvement incentives
Additional funding
External savings
Withhold / set-aside for quality performance
May 24, 2005

Slide 6
Perhaps no issue precipitates as much animated discussion as the question of how to fund
P4P incentives. The options are fairly simple to list. In descending order of preference (from
at least a provider perspective) these options include: (1) additional funding, i.e., new
money; (2) savings from improved quality and reduced use of expensive but potentially
avoidable services; and (3) a withhold, set-aside, or reduction in existing payment levels.
Before discussing these three options, it may be worth drawing attention to the distinction
between the implicit funding of incentives to improve efficiency and related aspects of care
that is inherent in a prospective payment system and the explicit funding of quality
improvement incentives. As noted earlier, prospective payment inherently rewards
providers who can more efficiently deliver effective renal replacement therapy, successfully
reduce the frequency and manage the cost of complications and co-morbidity, and promote
safer, more timely, and more patient-centered care. This opportunity represents a kind of
implicit funding of implicit (if intended) P4P incentives.
A P4P program would explicitly fund incentives to pursue specific improvements in
performance. These incentives might be supported by new or additional funding. In
establishing the bundled payment demonstration, Congress explicitly authorized an
increase in payment rates although it did not direct CMS to use these funds to support P4Plike incentives.
A second source of funding might be savings that are achieved from reduced use of
services other than those provided by the dialysis facility. A good example of such savings
would be reduced use of inpatient hospital services arising from better management of
vascular access, complications (e.g., infections), or co-morbidity (e.g., anemia and
diabetes).
A third source of funding might be a withhold or set-aside from the bundled payment
amount. This funding method would probably make sense only if the performance
objectives and incentive payments involved activities that were themselves covered by the
bundled payment amount.
Page 6
May 24, 2005
How large should incentives be?
Purpose of incentive payment
Considerations
Source of operating revenues

Bonus for investment in improved service
Compensation for cost of improving performance
Incentive needed to motivate change in behavior
Allocation to specific objectives / measures

Allocation among providers
May 24, 2005

Slide 7
Pay-for-performance arrangements differ widely in terms of the size of the incentives that
are offered. Some P4P initiatives have started by putting a relatively small amount of money
on the table, but have adopted a goal of expanding the percentage of revenue that
providers will be expected to derive from performance incentives.
A fundamental question concerns the purpose of the incentive payment. Some proponents
of P4P believe that the incentive payment should be an integral component of a providers
total revenue. Others see the incentive payment as akin to a bonusrevenue that the
provider can use to improve service and quality but that is not essential to support a basic
level of day-to-day operations. A somewhat larger incentive payment might be justified if it
is viewed as supplying part of a providers operating revenue, but only if the provider can
confidently expect to achieve performance targets.
A further set of considerations run somewhat parallel to this fundamental question. One
view would peg the size of the incentive to the investment that is needed (or the cost
associated) with improvements in quality. According to this pragmatic view, the size of the
incentive should be sufficient to cover costs that are incurred to earn the incentives. A
slightly different perspective would peg the incentive to an amount that is deemed
necessary to motivate a change in provider behavior. Of course, some economicallyminded analysts might see little difference between these two perspectives.
Two further considerations are related to the question of how much money should be set
aside to support P4P incentive payments. The first is the extent to which the incentive
payment is allocated to specific objectives or performance measures. The larger the
number of measures and the more tightly incentive payments are tied to specific indicators,
the larger the incentive pool might need to be. In general, as the number of measures rises
a providers confidence in its ability to earn the full amount of the incentive may decline.
Similarly, to the extent that incentive payments must be allocated among multiple providers
the minimum size of the incentive pool may tend to rise.
Page 7
May 24, 2005
How are P4P incentives earned?
Absolute / consensus-based standard

Competitive standard
Improvement standard
Hybrid approaches
May 24, 2005

Slide 8
A fundamental question for any P4P system concerns the nature of the standard that is
used to earn or distribute incentive payments.
One alternative is to focus on the provision of care that meets a consensus-based standard
of quality. Those providers who meet the standard will be rewarded. Providers who do not
meet the standard will not. A common objection to this kind of standard is that the providers
who do not meet the standard could potentially make the best use of, additional resources
to improve quality. A response to this objection is that performance incentives represent a
return on a providers investment in quality improvement, not an grant to support those
investments.
As an alternative to professional or industry consensus, the actual performance of providers
could be used to set an empirical benchmark. Competition among providers would both
determine the standard and determine the distribution of incentive payments. A provider
would earn an incentives by performance that is superior to its competitors. A critic of this
approach would object that it denies additional resources to those providers whose
unsuccessful performance may point to a need for additional resources. However, a
proponent might respond that the successful competitor will expand access to higher quality
care as its market share increases.
A third alternative is to focus on improvement in quality. A provider whose performance falls
short of an absolute or competitive standard could still earn an incentive payment if it
improves its performance. Proponents of this kind of standard note that it provides poor
performers with an the means and opportunity to improve performance but gives even
those with superior performance an incentive to strive for further improvement. An
objection to this approach is that it implicitly rewards performance that is below either an
absolute standard or that is inferior to the attainable performance of other providers.
In practice, the various approaches can be combined. Providers might earn an incentive
payment both for meeting a consensus-based target and by improving performance over a
prior period.
Page 8
May 24, 2005
P4P Quality Measures
Number of quality measures

Possible clinical / performance indicators
Adequacy of renal replacement therapy

Incidence of complications / co-morbidity
Vascular access
Anemia management
Modality choice
Criteria for selection of measures
May 24, 2005

Slide 9
The selection of measures to use in a P4P arrangement is an enormously difficult but

critically important task. Even basic questions concerning the number of performance
measures to use can be controversial. An approach that relies on many measures
increases the likelihood that a provider can achieve at least some improvements in quality,
but it also diffuses quality improvement efforts and may reduce the ability of a provider to
implement strategies that are based on a confident expectation of earning incentive
payments.
The specific measures of performance to use in a P4P arrangement will, of course, depend
on its goals. Possible measures might focus on the adequacy of renal replacement therapy,
the incidence of complications and co-morbidity, promoting safe and effective methods of
vascular access, successful management of anemia, and choice of modality. No doubt,
many other measures can be identified. For example, the ESRD Disease Management
demonstration proposed a quality incentive that used five measures: Kt/V, hemoglobin,
serum albumin, AV fistula, and bone disease.
A final difficult issue concerns the criteria that should be used to guide the selection of
measures. This is a subject that has received considerable attention from professional
groups representing ESRD physicians and providers. It has also been the focus of attention
by MedPAC. The central issue is whether it is possible to set performance targets that
accurately reflect the performance of the providers whose payment is at risk under the P4P
provisions.
Page 9
May 24, 2005
MedPAC Measurement Principles
Use of accepted, evidence-based measures

Collection without undue burden
Acceptance of risk adjustment methods
Possibility of improvement
Apply to a broad range of care

Performance should be under providers control
Focus on areas needing improvement
May 24, 2005

Slide 10
The MedPAC in its March 2005 report (pp 186-187), proposed four criteria to guide the
selection of quality measures.
First, quality measures should be based on evidence and should be accepted by
independent quality experts, purchasers, providers and consumers.
Second, the data needed to implement these quality measures should be collected without
undue burden on either providers or on CMS. To the extent possible, currently available
data should be used, but it may be necessary to collect additional data.
Third, when risk adjustment is needed methods should be adopted that are accepted as
sufficient to avoid creating barriers to access. Generally, MedPAC believes that outcome
measures are more likely to require risk adjustment than measures of quality that focus on
the process of care.
Fourth, most providers should be able to demonstrate improved performance. MedPAC
believes that this means that the measures should be applicable to a broad range of care
and a substantial percentage of all patients. A providers performance on a measure should
be under the providers control, not the influence of factors that a provider can do nothing
about. Finally, measures should focus on areas needing improvement, not areas in which
performance is already high.
Page 10

Pay-for-Performance / Quality Incentives: Discussion Paper For The MMA 623e Advisory Board

Uploaded by

Copyright:

Available Formats

Pay-for-Performance / Quality Incentives: Discussion Paper For The MMA 623e Advisory Board

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Pay-for-Performance / Quality Incentives: Discussion Paper For The MMA 623e Advisory Board

Uploaded by

Copyright:

Available Formats

Pay-for-Performance / Quality Incentives

May 24, 2005

Discussion Paper for the

MMA 623e: ESRD Bundled Payment Demonstration