NYSHealth Lewin Group

%HQGLQJWKH
+HDOWK&DUH&RVW&XUYH
LQ1HZ<RUN6WDWH
2SWLRQVIRU6DYLQJ0RQH\ JULY 2010
DQG,PSURYLQJ&DUH Prepared by The Lewin Group
$FNQRZOHGJHPHQWV
7
he following report, outlining the estimated savings potential of 10 health care cost
containment options, is based on research, modeling, and reporting conducted by
The Lewin Group. The New York State Health Foundation gratefully acknowledges the
contributions of many individuals involved in the production of this study. From The
Lewin Group, Kathy Kuhmerker and Jim Teisl led overall management of the report,
and John Sheils, Randy Haught, Joel Menges, and Lisa Alecxih led the
modeling and policy analysis. While The Lewin Group modeled the policy
options and provided the estimates contained herein, members of a
technical advisory panel (TAP) provided valuable input throughout
the process. These individuals include:
JOHN BILLINGS
New York University
(Robert F. Wagner School of Public Service)
FRANCOIS DE BRANTES
Bridges to Excellence
MARK CALLAN
The Healthcare Association of New York State
SEAN CAVANAUGH
United Hospital Fund
DAN HEIM
New York Association of Homes and
Services for the Aging
DENNIS NORTON
New York State Division of the Budget
CATHY SCHOEN
The Commonwealth Fund
ELISABETH WYNN
Greater New York Hospital Association
We appreciate the thoughtful review and comment by Diane Meier of the Center to Advance Palliative
Care, Sean Morrison of the National Palliative Care Research Center, John Murtha and Eric Wallace
of the Greater New York Hospital Association, and Qiang Xu, James Sherman, Michael Rynasko,
and Christopher McManus of the New York State Division of the Budget.
Cover photo: Nelson Syozi, sxc.hu
Finally we note that although Medicaid claims data were provided by the New York State Department
of Health (NYSDOH), conclusions in this publication are not necessarily those of NYSDOH.
Support for this work was provided by the New York State Health Foundation (NYSHealth). The mission
of NYSHealth is to expand health insurance coverage, increase access to high-quality health care
services, and improve public and community health. The views presented here are those of the authors
and not necessarily those of the New York State Health Foundation or its directors, officers, and staff.
%HQGLQJWKH+HDOWK&DUH&RVW&XUYHLQ1HZ<RUN6WDWH2SWLRQVIRU6DYLQJ0RQH\DQG,PSURYLQJ&DUH
&RQWHQWV
Executive Summary 1
Introduction 10
Methods 15
Promoting Accountable Care Organizations (ACOs) 18
Modernizing Primary Care 22
Expanding Palliative Care 29
Implementing Mandatory Managed Care For Medicaid Dual Eligible Population 34
Adopting Bundled Payment Methods 39
Imposing A Tax On Sugar-Sweetened Beverages 45
Expanding Hospital Pay For Performance 50
Realizing Administrative Simplification Through Health Information Technology 55
Rebalancing Long-Term Care 60
Using Alternative Delivery Systems 64
Conclusion 68
([HFXWLYH6XPPDU\
&
ontrolling health care costs and their growth has been both a national and a State-specific
issue for many years, but few steps have been taken that have significantly impacted the
overall growth rate. Growing cost pressures raise the need for a serious, focused effort to
fundamentally restructure the delivery of care and associated spending. Massive budget
deficits are expected for the next several years. For 2010–2011, New York State faces a budget deficit
exceeding $9 billion. New York State’s most recent Executive Budget anticipates a $14 billion dollar
deficit for 2011–2012, increasing to $18 billion for 2012–2013, and $21 billion for 2013-2014. While there
are many reasons for these deficits besides health care spending, New York health care expenditures
are significant. Total annual health care spending in New York State exceeded $126 billion in 2004,
with the State exhibiting the fourth highest per capita health spending ($6,535) in the nation.1
Overall health care spending in New York continues to grow at nearly 6% each year. New York’s
high health care costs result in expensive health insurance premiums. New Yorkers, on
average, paid nearly $13,000 in 2008 for family coverage, which places New York among the
most costly states for health insurance premiums. 2 Premiums continue to rise rapidly—from
2006 to 2008 they increased by 11%. 3 Cost containment is more important than ever with the
passage of Federal health care reform, and bringing more individuals into the health care
system requires that care be provided more efficiently.
Despite the high cost of health care, New York State is not performing well on many indicators
of health system performance and quality. The 2009 Commonwealth Fund State Scorecard ranks
New York State 50th of 50 states in potentially avoidable use of hospitals and costs of care, 22nd in
providing prevention and treatment, 18th in health care access, and 17th on healthy lives.
New York State, like many other states, has undertaken numerous initiatives intended to control health care
costs over the last decades. Many of these efforts focused on the State’s Medicaid program—the nation’s
most expensive4—while others, such as New York’s Prospective Hospital Reimbursement Methodology
(NYPHRM), its Health Care Reform Act (HCRA), and its Commission on Health Care Facilities in the 21st
Century (commonly called the Berger Commission), were broader in their purview.
Significant responsibility for identifying and implementing cost-saving opportunities is expected

to continue to fall to states, which have long served as laboratories of health reform. States have
considerable authority to impact the rate of growth in health care spending through their roles as
regulators, payers, and providers of health care services, and their efforts should complement
those undertaken nationally. The Patient Protection and Affordable Care Act provides some new
opportunities to control costs, but their success largely depends on their implementation by states
and providers.
1 State Health Facts, The Henry J. Kaiser Family Foundation, www.statehealthfacts.org. Retrieved February 2010.
2 C. Schoen et al., “Paying the Price: How Health Insurance Premiums Are Eating Up Middle-Class Incomes—State Health Insurance Premium Trends and the
Potential of National Reform,” The Commonwealth Fund, August 2009.
3 The Big Picture Updated, United Hospital Fund, http://www.uhfnyc.org/publications/880650. Retrieved May 2010.
4 New York State Division of the Budget. 2010-11 Executive Budget.
—1—
([HFXWLYH6XPPDU\FRQWLQXHG
This report, commissioned by the New York State Health Foundation, is designed to inform
a State-level discussion of health care savings opportunities in New York. This paper
outlines the estimated impact of 10 scenarios that could help to contain escalating health
care costs in New York State over the next decade while also improving health care quality.
This report shows that billions of dollars in savings are possible. New York State’s health
care cost curve can be bent through policy options that better integrate care and yield better
health care outcomes. While government would realize much of the savings, in many cases,
the savings would also extend to private employers and households.
POLICY SCENARIOS DESIGNED TO HELP CONTAIN COSTS AND

IMPROVE QUALITY OF CARE
The report is modeled after the 2007 Commonwealth Fund report, “Bending the Curve:
Options for Achieving Savings and Improving Value in U.S. Health Spending,” which examined
15 options with the potential to lower health care spending nationally. Identification of the
10 scenarios began with the 15 options included in the Commonwealth Fund report. Several
options were excluded because national policies had already been adopted in the area (e.g.,
promoting health information technology), New York had already made great strides in the
area (e.g., reducing tobacco use through increased taxes), or because they were focused
on policy set at the Federal level (e.g., Medicare and comparative effectiveness). Other
scenarios (e.g., promoting Accountable Care Organizations) had matured in the intervening
years and were considered for this analysis.
Options were focused on identifying approaches that would improve the quality of care
provided, as well as catalyze sustainable reductions while not creating significant
disruptions in the health care marketplace for any one participant (e.g., costs would not be
extracted from nor borne exclusively by one group or another). The 10 options selected are
not meant to be an exhaustive list of approaches. Rather, they represent a range of options to
address various factors that contribute to increasing health care costs and inefficiencies
in existing health care delivery and financing systems.
The final 10 scenarios were selected with the advice of a Technical Advisory Panel (TAP)
convened by the New York State Health Foundation and The Lewin Group. TAP members came
to this project from a broad spectrum of the New York State health care market, including
individuals with knowledge of primary, acute, and long-term care services and providers, and
with experience in government, private industry, foundations, and research. TAP members also
provided valuable perspectives and feedback during the course of the modeling process.
To provide a baseline against which to measure the 10 scenarios, we projected the growth
in health spending in New York State through 2020 based upon Centers for Medicare and
Medicaid Services (CMS) historical and projected spending growth data. CMS projects that
national health spending will grow at more than 6% per year over the next decade. CMS also
provides historical data on the rate of growth in health spending by state for 1991 through
—2—
2004, which shows that New York State health care spending has increased at a slower
rate than nationally: 5.9% per year. Using these data, we estimated that health spending in
New York will grow from $189.0 billion in 2011 to $318.8 billion in 2020 (Figure ES-1),
resulting in a 10-year baseline of almost $2.5 trillion. 5
Figure ES-1. Projected Total Health Spending in New York for 2011 through 2020 (billions)
$350
$318.8
$300.2
300
$282.1
$265.2
$248.8
250 $234.1
$220.4
$198.2 $208.4
200 $189.0
150
100
50
0
2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
Source: The Lewin Group estimates using CMS spending growth estimates.
This analysis recognized that the State’s ability to influence health care policy is generally
bounded by its role as a direct purchaser of health care and as a regulator of health care
providers and insurance companies. As a result, the State has limited ability to influence
the actions of Medicare, private insurance plans, and self-insured organizations. Therefore,
when appropriate, we modeled scenarios two ways: projecting savings if the scenario were
universally adopted in New York (“potential” savings) and if it were adopted only by Medicaid,
Child Health Plus, State and local employee benefit plans, and as the result of provider or
payer regulation by State government (“actionable” savings).
THE 10 SCENARIOS INCLUDE:

• Promoting Accountable Care Organizations (ACOs). Implement the theoretical ACO model,
which creates incentives for providers to emphasize primary care, prevention, and adherence
to evidence-based guidelines. This can take the form of a medical home model, disease
5 At the time that these estimates were derived and modeling was conducted the Patient Protection and Affordable Care Act of 2010 had not yet
been enacted. As a result, baseline figures do not account for coverage changes that will result from this legislation.
—3—
management program, and coordination of care for patients with multiple health conditions.
ACOs have been proposed as an opportunity for health care providers to band together to
coordinate care and share in the savings that can be realized through improvements in the
quality and efficiency of care. The model examines savings of both mandatory and voluntary
ACOs, with potential savings over 10 years for all payers ranging from $10.7 billion (voluntary)
to $49.8 billion (mandatory), and actionable savings ranging from $3.1 billion (voluntary) to
$14.6 billion (mandatory).
• Modernizing Primary Care. Enhance the use of primary care, with particular emphasis on
services provided to the chronically ill. The model estimates savings for four separate, but
overlapping, approaches, including initiatives requiring people to have a primary care “medical
home,” paying the primary care provider to coordinate care for patients with complex health
needs, and administering evidence-based disease management for people with chronic
conditions. We also examined a pay-for-performance program that rewards providers who
show positive health outcomes. Potential savings for all payers over 10 years range from
$1.3 to $33.7 billion, with actionable savings ranging from $0.5 billion to $11 billion.
• Expanding Palliative Care. Require hospitals to establish a palliative care program to

promote better coordinated, higher value care where appropriate. Palliative care programs
have been shown to improve physical and psychological symptom management, caregiver
well-being, and family satisfaction. Studies have also shown that, when given the choice,
patients nearing end of life will often decline costly and invasive treatments that hospitals
may be inclined to provide, resulting in lower health spending. Savings are estimated at
$11.9 billion over 10 years, all of which could be realized through State action.
• Implementing Mandatory Managed Care for the Medicaid Dual Eligible Population. Enroll
New York’s Medicaid/Medicare dual eligibles into a fully integrated coordinated care setting.
Under this model, dual eligibles would be mandatorily enrolled into capitated managed care
organizations (MCOs). These MCOs would be at full financial risk for the entire Medicaid
and Medicare benefits package for their enrolled dual eligibles. This model would require a
partnership between CMS, which manages these programs at the Federal level, and New York
State. We assume that under this partnership Medicare and Medicaid funds for dual eligibles
would be pooled, with the overall savings achieved split 50/50 between the Federal government
and New York State. Over 10 years, this policy is estimated to save $10.8 billion, assuming that
100% of dual eligibles are enrolled in such plans.
• Adopting Bundled Payment Methods. Make prospective payments for entire episodes of care,
potentially encompassing inpatient care, physician services while hospitalized, and post-
acute care services including short-term skilled nursing facility (SNF) and home health care.
By offering global fees, otherwise referred to as “bundled payments,” this reform would
provide an opportunity for hospitals, physicians, and other health care providers to benefit
from reducing complications and hospital readmissions and allow for more flexibility in
—4—
allocating resources. Based on adopting bundled payments initially for a selected number
of conditions, this option has the potential of saving $6.3 billion across all payers during the
10-year period, approximately $1.6 billion of which is actionable.
• Imposing a Tax on Sugar-Sweetened Beverages. Designed to reduce obesity and related health
costs, this option assumes that a sugared soft drink excise tax is imposed in New York State
in addition to the current state and county sales taxes in New York. This proposal is similar to
an item in Governor Paterson’s proposed 2010–2011 budget, which included an excise tax on
non-diet sodas and fruit drinks that contain less than 70% natural fruit juice. Over 10 years, this
option could potentially save $5.6 billion in spending on chronic illnesses related to obesity and
overweight, all of which is actionable by the State.
• Expanding Hospital Pay for Performance. Provide hospitals with bonus payments for
demonstrated improvements in patient care. Under such a system, data would be compiled
by providers for their patients that receive selected health services. The results would be
adjusted for the severity of illness and compared with benchmark measures of outcomes
for these services. Providers with favorable results would be rewarded with higher payment.
Net potential savings, after bonus payments are made, are estimated to save $3.8 billion
for all payers over the 10-year period, with net actionable savings of $1.3 billion.
• Realizing Administrative Simplification through Health Information Technology. Implement

a set of approaches to reduce the administrative burden on health care providers and
insurers alike. We begin by summarizing our estimates of billing and insurance-related (BIR)
costs for New York. We then present estimates of two options for reducing BIR costs through
standardization and improved use of health information technology. Potential savings
over 10 years, for all payers, are estimated at $1.6 billion, all of which are actionable savings.
• Rebalancing Long-Term Care. Restructure New York State’s Medicaid programs for long-
term care, examining both residential and community-based settings for a large population
of beneficiaries with extensive functional and cognitive impairments, and behaviorally and
medically complicated needs. This option would enhance Aging and Disability Resource Centers
throughout the State; create a standardized assessment tool; establish data capability for
planning; develop rate and fee systems for institutional and community-based long-term care
services; and institute a diversion and transition program. This approach would position New
York State to make more far-reaching changes to long-term care in the future. Savings, which
are projected to only affect the Medicaid program, are estimated at $1 billion over 10 years.
• Using Alternative Delivery Systems. Promote the use of alternative, less costly systems
for the delivery of care for low acuity conditions. Many of the services provided in physician
offices, urgent care centers, or hospital emergency rooms can be provided at a lower cost
in retail clinics or workplace clinics. The intent of this policy option is to promote the growth
of these types of clinics throughout the State and to encourage patients to use them for
—5—
low-acuity conditions instead of traditional settings. Shifting more care to these settings
could help reduce health care costs across the State. Potential savings for all payers are
estimated at $0.35 billion from 2011–2020, all of which is actionable by the State.
The following table presents a summary of the savings estimates developed for the 10 policy
scenarios (Figure ES-2).
FIGURE ES-2. Summary of Projected Savings by Policy Scenario, 2011–2020 (billions)

CUMULATIVE CUMULATIVE
BASELINE POTENTIAL SAVINGS ACTIONABLE SAVINGS
SPENDING
$ % $ %
Promoting Accountable Care Organizations
Mandatory $1,096.56 $49.80 4.5% $14.59 1.3%
Voluntary $1,096.56 $10.71 1.0% $3.11 0.3%
Modernizing Primary Care
Mandatory Medical Home $1,620.97 $33.66 2.1% $10.99 0.7%
Voluntary Medical Home $1,620.97 $9.11 0.6% $2.25 0.1%
Advanced Disease Management $1,620.97 $10.74 0.7% $3.30 0.2%
Pay for Performance $1,620.97 $1.33 0.1% $0.46 0.0%
Expanding Palliative Care $235.38 $11.93 5.1% $11.93 5.1%
Implementing Mandatory Mgd. Care for
$496.85 $10.76 2.2% — —
Dual Eligibles*
Adopting Bundled Payment Methods $133.30 $6.30 4.7% $1.56 1.2%
Imposing a Tax on Sugar-Sweetened Beverages $136.30 $5.63 4.1% $5.63 4.1%
Expanding Hospital Pay for Performance $162.03 $3.85 2.4% $1.31 0.8%
Realizing Administrative Simplification through HIT
Standardized Quality Requests $1.57 $0.65 41.5% $0.65 41.5%
Standardized Credentialing/Verification $6.18 $0.92 14.9% $0.92 14.9%
Rebalancing Long-Term Care** $147.28 $1.02 0.7% $1.02 0.7%
Using Alternative Delivery Systems $24.3 $0.35 1.4% $0.35 1.4%
Source: The Lewin Group estimates.
* Savings estimate includes combined Medicaid and Medicare savings.
** Savings estimate includes Medicaid savings only.
METHODS IN BRIEF
Our approach to these analyses was to begin with an extensive review of the literature on
the effect of these policy options where they were implemented. However, some of the ideas
we studied, such as ACOs, are so new that they have never been implemented or even tested
through demonstrations. Here we extrapolated from similar experiences in other settings
that employ similar incentives. The analysis also estimates how savings would be distributed
among various payers, namely the Federal government, the New York State government,
private employers, and households.
—6—
Each of these scenarios required very different data. For example, modeling the impact
of bundled payments required us to perform a detailed analysis of claims for New York
residents, including Medicaid data, the New York cohort of Medicare claims data, and data
from United Healthcare on service use for 1.5 million people with commercial insurance
in the State. For other options, such as improving palliative care, we used State hospital
discharge data. Summary New York State 2008 long-term care claims data were used for
analyses of both the managed care for Medicare and Medicaid dually eligible individuals and
modernizing long-term care scenarios. The dual eligibles scenario also relied on data from
CMS Medicaid Statistical Information System (MSIS).
Finally, we used data from The Lewin Group Health Benefits Simulation Model (HBSM) to
provide detailed data on spending for people in New York by source of payment, type of
service, and demographic group. These data are based upon information on health spending
for New York from the Office of the Actuary of CMS, State health benefits programs, State-
level population demographic data, and the Medical Expenditures Panel Survey data for
the region. For several scenarios, these data enabled us to estimate spending for the State
populations that would be subject to State control or regulation. We developed long-term
spending projections for New York based upon the historical spending growth information
provided by CMS. Using these data, we estimate that health spending in New York will grow
from $189 billion in 2011 to $318.8 billion in 2020.
The actual effects of these options over the next decade could differ substantially from our
projections. Unforeseeable changes in new technology and disease prevalence could dramatically
alter spending growth trends. Our estimates are also based upon the results of demonstrations
and similar initiatives implemented elsewhere, and often in an ideal environment for a given
option, such as size of physician group or physician affiliation with hospitals. Thus, the effect of
an option on spending once broadly implemented may differ from our estimates.
ADDITIONAL CONSIDERATIONS
There are a number of additional considerations to keep in mind regarding these scenarios
and savings estimates.
1. Results are not additive. It is important to note that a number of these scenarios have
overlapping impacts and, thus, savings estimates are not additive. For example, if mandatory
managed care for dual eligibles (which includes both acute and long-term care services) and
modernizing long-term care were both implemented, savings would be less than the cumulative
savings projected for both scenarios because a portion of the cost savings in one would also
be included in the second. This interaction is also true for ACOs and Modernizing Primary Care,
which both involve care coordination, and Expanding Palliative Care and Rebalancing Long-
Term Care which both impact end-of-life health care costs. While we did not attempt to quantify
all of the potential overlaps, there are clearly a number of areas where overlaps likely exist and
that need to be considered when determining which options to implement.
2. Results include potential and actionable savings. These estimates reflect the total
estimated “potential” savings that will result from each scenario and “actionable” savings—
—7—
those that the State can independently affect. In the case of the scenario calling for
mandatory enrollment of dual eligibles in managed care, there are no actionable savings
without the Federal government’s participation. For three policy options (Expanding
Palliative Care, Realizing Administrative Simplification through HIT, and Rebalancing Long-
Term Care), there is no distinction between potential and actionable savings because New
York State can fully advance these options without Federal action.
3. Federal health care reform includes new opportunities. With the recent passage of the
Patient Protection and Affordable Care Act, the Federal government has taken steps that
have the potential to slow the rate of growth in health care spending at a national level.
As noted earlier, the 10 scenarios presented here were determined, and the analyses
conducted, prior to the point at which the Federal reform law had taken shape. This analysis
does not quantify the potential impact of the new law. However, we note that several
initiatives contained within this legislation will likely complement and, in some cases,
facilitate implementation of the scenarios that we have modeled. These include:
• CMS Federal Coordinated Health Care Office: Authorized to more effectively integrate
Medicare and Medicaid benefits and improve coordination between the Federal government
and states to improve access to and quality of care and services for dual eligibles.
• Innovation Center at CMS: Authorized to test, evaluate, and expand—in Medicare, Medicaid,
and the Children’s Health Insurance Program (CHIP)—different payment structures and
methodologies to reduce program expenditures while maintaining or improving quality
of care; includes payment reform models that will attempt to improve quality and reduce
the rate of cost growth.
• Authority for ACOs, Medical Homes, and bundled payments: Incorporated to improve
accountability for the overall care of Medicare and Medicaid beneficiaries, define
processes to promote evidence-based medicine, and coordinate care.
• HIT Requirements for Administrative Simplification: Incorporated to address the need for
an automated system that will simplify the process of communicating transactions among
providers and health plans.
4. There are intrastate differences. We also note that while the scenarios were modeled
for statewide impact, New York City clearly differs from the rest of the State. Policymakers
may want to consider such differences during implementation, perhaps by targeting certain
aspects of the initiatives regionally.
5. Effective design and phased implementation is assumed. For the purposes of this paper,
savings are presented on a consistent 10-year timeframe (2011–2020) to permit comparison;
however, realistic timeframes to implement these options vary. In addition, the efficiency
and effectiveness of the implementation process will impact the rate at which savings can be
realized. For the purposes of this analysis, we have assumed that programs are designed and
implemented effectively and, where appropriate, the savings estimates reflect a phase-in period.
6. Realizing savings assumes that costs are not shifted elsewhere. Savings to providers or
insurers are not savings to consumers unless the savings are passed back in the form of
—8—
lower prices and/or premiums. In this analysis, we assume that the savings would come as
a reduction in the rate of growth in health spending. For example, savings from simplification
of provider credentialing would effectively reduce the increase in provider prices over time,
resulting in reduced premium growth. However, to the extent that savings are retained by
providers as income, or made up for by increasing the utilization of other services, these
savings would not represent a net reduction in health spending and premiums.
CONCLUSIONS
Health care costs in New York State continue to escalate at an unsustainable rate and squeeze
government, employers, and households. In the face of sizable State budget deficits, rising
health insurance premiums and expanding public insurance enrollment, it is urgent to start
addressing health care costs now. The enactment of Federal health care reform will further
increase the number of New Yorkers accessing the health care system, increasing cost
pressures and the need to find more efficient, higher value approaches to health care delivery.
This report shows that New York State has the ability to trim health care cost growth by
billions of dollars. A wide range of policy options are available to the State, and a cohesive
combination of them will be needed to address the negative incentives of the existing health
care system. Implementing these policy options would slow the growth of health care
spending in New York State, but do so in a way that does not destabilize the foundation of the
delivery system. Rather, each of these scenarios would promote a substantial improvement
in the manner in which care is delivered, resulting in a more efficient, integrated, patient-
centered, and quality-oriented system. Obtaining these savings requires active participation
by government, providers, and payers working together and not shifting costs. Estimated
savings from these policy options would largely benefit Federal and State governments,
but would also ease growing cost pressures among private employers and households.
Federal health care reform raises the urgency that New York State and providers tackle
health care costs while implementing expanded coverage provisions. Fortunately, it also
offers new opportunities, including Federal partnerships, to advance many of the policy
options presented in this paper. The time is right for New York State to move beyond the
status quo and toward a more efficient, effective, and higher value health system.
—9—
,QWURGXFWLRQ
3
erennial increases in health care costs present budgetary and other challenges for
numerous stakeholders including governments, employers, consumers, insurers,
and providers. Further, escalating costs restrict access to necessary health
services, which may potentially have a negative impact on the public’s health. Rising
health care costs correlate with significant drops in health insurance coverage, and national
surveys show that the primary reason people are uninsured is due to the high and escalating
cost of health coverage.6
The facts largely speak for themselves. From 1965 to 2007, health spending as a percentage
of gross domestic product (GDP) steadily increased from 5.9% to 16.2%. Recent projections
see the average annual growth in national health spending to be 6.2% through 2018—2.1
percentage points higher than the average annual growth in GDP. At this rate, national health
care expenditures are expected to reach $4.4 trillion by 2018—more than double the amount of
spending in 2007—and 20.3% of GDP.7
New York State, in particular, is faced with staggering health care expenses. Total annual health
care spending in New York State exceeded $126 billion in 2004, with the State exhibiting the
fourth highest per capita health spending ($6,535) in the nation.8
The Centers for Medicare and Medicaid Services (CMS) estimates that total personal health
care spending for New York residents was $126.1 billion in 2004. Of this total, $22.8 billion
was paid by Medicare, $40.1 billion by Medicaid, and $63.2 billion was covered by other
sources, including private insurance, other public programs, and out-of-pocket spending
for families. Hospital care accounted for 36% of spending with physician care accounting for
20.1% of expenditures.
We estimate that spending in New York would grow to $189 billion in 2011 under current law
(Figure 1). This figure includes total spending by all payer groups for New York residents
including payments to health care providers and the cost of administration for insurance and
public programs (excludes public health research and construction).9 Of this total, $66.5 billion
is estimated spending for hospital care and $36.7 billion for physician care.
Long-term care spending would be $35.4 billion in 2011 including nursing home care, home
health, and “personal services.” New Yorkers will spend approximately $35.4 billion on
outpatient prescription drugs, which is equal to approximately 12.9% of total spending. The cost
of administration for private insurance and public programs would be $11.6 billion, which is
6.1% of total spending.
6 The Uninsured: A Primer, Key Facts about Americans without Health Insurance, Henry J. Kaiser Family Foundation, April 2009.
7 A. Sisko et al., “Health Spending Projections Through 2018: Recession effects Add Uncertainty to the Outlook,” Health Affairs, 28, no.2 (2009): w346-w357
8 State Health Facts, The Henry J. Kaiser Family Foundation, www.statehealthfacts.org. Retrieved February 2010.
9 This definition of spending is termed “health services and supplies.
—10—
,QWURGXFWLRQFRQWLQXHG
FIGURE 1. Projected Spending in New York by Type of Service...
Home Health ($17.2) Administration ($11.6)

9.1% 6.1%
Nursing Home ($18.2) Hospital ($66.5)

9.6% 35.2%
Medical Equipment ($2.0)
1.0%
Physician & Clinic ($36.7)
Prescription Drugs ($24.3)
19.4%
12.9%
Other Professional ($5.2)
2.8%
Physician & Clinic ($7.3)
3.8%
...and by Source of Coverage in 2011
Worker’s Comp ($2.4) MediGap ($1.3)

1.3% 0.7%
Out-of-Pocket ($22.0)
Other Public ($2.8)
1.5% 11.7%
CHAMPUS/Vet ($4.7)
2.5% Employer Workers ($52.7)
27.9%
Medicaid ($52.8)
27.9%
Employer Non-Workers ($7.2)
Other Public ($39.2) 3.8%
20.8%
Total Spending = Non-Group ($3.7)

$189.0 billion 2.0%
Source: The Lewin Group estimates using data provided by the Office of the Actuary of the Centers for Medicare and Medicaid Services (CMS).
Numbers may not add to totals due to rounding.
—11—
New York’s Medicaid program is the most expensive in the nation. New York spends more per
capita ($2,360) than any other state in the country and more than twice the national average
($1,077).10 In the absence of any changes, total Federal, State and county Medicaid spending is
projected to reach $53.2 billion in 2010–11, an increase of more than 5%. Medicaid enrollment
has grown by nearly 13% since September 2008 and is expected to continue growing in part due
to expansion enacted by Federal health care reform. Increased Medicaid spending contributes
to the State’s budget deficits, which are now projected through 2014.
New York’s high health care costs result in expensive health insurance premiums. New
Yorkers, on average, paid nearly $13,000 in 2008 for family coverage, which places New York
among the most costly states for health insurance premiums.11 While the reasons for New
York’s higher levels of health care spending are not completely understood, according to
the Dartmouth Atlas of Health Care Hospital Care Intensity Index, New York outranks every
state except New Jersey in the amount of time patients spend in hospitals and the intensity
of physician services delivered during hospital stays.12 In the New York City metropolitan
area, the Dartmouth Atlas finds that Medicare payments to hospitals are high due to Federal
payments for graduate medical education and to payments for caring for disproportionate
shares of low-income patients.13
Despite the high cost of health care, New York State is not performing well on many indicators
of health system performance and quality. The 2009 Commonwealth Fund State Scorecard
ranks New York State 50th of 50 states in potentially avoidable use of hospitals and costs of
care.14 For example, in comparison to other states, New York has the second highest rate
of home health patients with a hospital admission. It ranks 31st of 50 states in the percentage
of adult asthmatics with an emergency room or urgent care visit in the past year, suggesting
this chronic condition is not as well managed as it should be in primary care settings. Also,
the Scorecard ranks New York State 22nd of 50 states in providing preventive care and other
treatments, 18th in health care access, and 17th on healthy lives. These rankings suggest
opportunities to provide more efficient and effective care.
Ten Options to Bend the Cost Curve

This paper outlines the estimated impact of 10 health care cost containment scenarios that
could help to contain the escalating health care costs in New York State over the next decade,
while simultaneously leading to improved care coordination and quality. While each of the
10 scenarios is summarized here for policymakers, the paper is accompanied by a detailed
technical appendix that contains additional information regarding the estimates and the
assumptions upon which they are based.
10 New York State Division of the Budget. 2010-11 Executive Budget.
11 C. Schoen et al., “Paying the Price: How Health Insurance Premiums Are Eating Up Middle-Class Incomes--State Health Insurance Premium Trends and the
Potential of National Reform,” The Commonwealth Fund, August 2009.
12 Dartmouth Institute for Health Policy and Clinical Practice, “Hospital Care Intensity Index,” The Dartmouth Atlas of Health Care, 2008.
13 Daniel J. Gottlieb, et al., “Prices Don’t Drive Regional Medicare Spending Variations,” Health Affairs March 2010 29:3, pages 537-543.
14 Commonwealth Fund State Scorecard on Health System Performance, 2009.
—12—
This analysis is based on a similar national study published by the Commonwealth Fund in
2007, which examined 15 options with the potential to lower health care spending nationally.15
Identification of the 10 scenarios began with the 15 options included in the Commonwealth Fund
report. Several options were excluded because national policies had already been adopted in
the area (e.g., promoting health information technology), New York had already made great
strides in the area (e.g., reducing tobacco use through increased taxes), or because they were
focused on policy set at the Federal level (e.g., Medicare and comparative effectiveness). Other
scenarios (e.g., promoting Accountable Care Organizations) had matured in the intervening
years and were then considered for this analysis.
Options were focused on identifying approaches that would improve the quality of care
provided, as well as catalyze sustainable reductions while not creating significant disruptions
in the health care marketplace for any one participant (e.g., costs would not be extracted from,
nor borne exclusively by, one group or another). The 10 options selected are not meant to be
an exhaustive list of approaches. Rather, they represent a range of options to address various
factors that contribute to increasing health care costs and inefficiencies in existing health care
delivery and financing systems.
The final 10 scenarios were selected with the advice of a Technical Advisory Panel (TAP)
convened by the New York State Health Foundation and The Lewin Group. TAP members
came to this project from a broad spectrum of the New York State health care market,
including individuals with knowledge of primary, acute and long-term care services and
providers, and with experience in government, private industry, foundations, and research.
TAP members also provided valuable perspectives and feedback during the course of the
modeling process.
New York State has sought to control health care costs for decades, particularly within its
Medicaid program, and has achieved some success. The scenarios described in this paper,
however, are intended to broaden the scope of potential savings beyond those that the
State can impact as a direct payer or purchaser of health care services. We note that there
is variation among the scenarios in the ability for the State of New York to directly impact
savings. For example, by implementing a tax on sugar sweetened beverages, State action
could lead to savings throughout the health care system. In the case of bundled payments,
however, savings may be limited to payers that the State can impact directly (i.e., Medicaid,
and State and local government employees). We also note that while the scenarios were
modeled for statewide impact, New York City clearly differs from the rest of the State,
and policymakers would likely need to account for such differences during implementation,
and may wish to consider targeting scenarios to particular regions where they may be
most effective.
15 C. Schoen et al., “Bending the Curve: Options for Achieving Savings and Improving Value in U.S. Health Spending,” The Commonwealth Fund,
December 2007.
—13—
Finally, we note that the amount of work to ramp-up to implementation of each scenario
varies considerably. For example, the strategies under “Modernizing Primary Care” assume
that a sufficient professional workforce is in place. In reality, however, workforce development
resources may be required to recruit and retain sufficient providers to make the scenario
successful. For the purposes of this paper, savings are presented on a consistent 10-year
timeframe (2011–2020) to permit comparison; however, realistic times to implement vary
and will be considered in the next phase of this project, which is to develop high-level
implementation plans for the most promising scenarios.
Now is the right time for New York State to become a leader in controlling future health care
cost increases. The following 10 scenarios should act as a catalyst for conversation and
discussion among leaders both within and outside of the State’s health care industry. These
scenarios are not presented as recommendations, but rather analyses of proposed cost-
containment policy options that may benefit New York State. Only through sustained control
over health care cost increases will the State be able to continue its myriad public health
insurance programs and improvements in the public’s health.
—14—
0HWKRGV
,
n the following sections, we present our estimates of the financial impact of the 10 cost-
containment scenarios. Estimates are provided for a 10-year period (2011 to 2020). For
each option, we include a brief discussion of the scenario, how savings were estimated
(including assumptions required to model impacts), and a brief discussion of limitations
and potential implementation considerations, such as actionable steps that the State could take
to achieve these savings. Further detail regarding assumptions, data sources, and methods can
be found in the detailed technical appendix.
It is important to note that a number of these scenarios have overlapping impacts and, thus,
savings estimates are not additive. For example, if mandatory managed care for dual eligibles
(which includes both acute and long-term care services) and modernizing long-term care
were both implemented, savings would be less than the cumulative savings projected for both
scenarios because a portion of the cost savings in one would also be included in the second.
This interaction is also true for ACOs and Modernizing Primary Care, which both involve care
coordination, and Expanding Palliative Care and Rebalancing Long-Term Care, which both
impact end-of-life health care costs. While we did not attempt to quantify all of the potential
overlaps, there are clearly a number of areas where overlaps likely exist and that need to be
considered when determining which options to implement.
We do not assume that these proposals would necessarily save money, and we recognize
that some could actually increase costs if the upfront investment does not result in greater
efficiencies or reduced utilization of services. We have dealt with these issues by designing
variations on these cost savings proposals that focus on the patient groups where the potential
for net savings is strongest, such as people with chronic health conditions. Within each
scenario, we identify our data sources for purposes of the analysis. In general, we rely upon
estimates of the impact of similar proposals described in peer reviewed literature.
Our approach to these analyses has been to begin with an extensive review of the literature
on the effect of these policy options where implemented. For example, elements of pay-for-
performance and disease management (DM) have actually been tested in demonstrations
conducted by Medicare and some commercial insurers. Here we can base our estimates
on the net savings that were documented in these demonstrations. In addition to
demonstrations, there are independent studies in academic and professional journals that
document the impact of alternative approaches to patient care, such as palliative care.
However, some of the ideas we studied, such as Accountable Care Organizations (ACOs),
are so new that they have never been implemented or even tested through some form of a
demonstration. Here we extrapolate from similar experiences in other settings that employ
similar incentives. For example, the bonus systems envisioned for ACOs are similar to the
bonus systems used now in many HMOs, which have been studied and found to reduce health
services utilization. Studies of these systems provide a basis for estimating the impact of
policies that seek to change provider incentives, although careful adjustments are required to
reflect the unique features of each policy option.
—15—
0HWKRGVFRQWLQXHG
Some of our estimates were developed using a model that incorporates results from several
studies. For example, we simulated the effect of a tax on sugar-sweetened soft drinks in
several steps including: 1) a study of the effects of changes in the price of soft drinks; 2) studies
of how calorie intake affects the prevalence of overweight and obesity; 3) studies of the effect of
changes in overweight and obesity on the number of people with related health conditions; and
4) data and studies of how reductions in these conditions affect costs.
Each of these scenarios required very different data. For example, modeling the impact of
bundled payments required us to perform a detailed analysis of claims for New York residents
including Medicaid data and data from United Healthcare on service use for 1.5 million people
with commercial insurance in the State. For other options, such as improving palliative care,
we used State hospital discharge data. Summary New York State 2008 long-term care claims
data were used for analyses of both the managed care for Medicare and Medicaid dually
eligible individuals and rebalancing long-term care scenarios. The dual eligibles scenario also
relied on data from the Centers for Medicare and Medicaid Services (CMS) Medicaid Statistical
Information System (MSIS).
Finally, we used data from the The Lewin Group Health Benefits Simulation Model (HBSM)
to provide detailed data on spending for people in New York by source of payment, type of
service, and demographic group. These data are based upon information on health spending
for New York from the Office of the Actuary of CMS, State health benefits programs, State-
level population demographic data, and the Medical Expenditures Panel Survey data for
the region. For several scenarios, these data enabled us to estimate spending for the State
populations that would be subject to State control or regulation. These data also provided
information on Medicaid enrollment and spending in New York, which was used to estimate
the impact on Medicaid spending for several of the scenarios that did not require claims
level analysis.
We projected the growth in health spending through 2020 based upon CMS historical and
projected spending growth data. CMS projects that national health spending will grow at
roughly 6.3% per year over the next decade. Separate spending projections for New York are
not available, but we developed long-term spending projections for New York based upon
the historical spending growth information provided by CMS. These data describe the rate of
growth in health spending by state for 1991 through 2004, demonstrating that the rate of growth
in health spending for New York has been on average approximately 0.4 percentage points lower
than the national average. Based upon this historical trend, we projected health spending in
New York would be equal to the national growth rate of 6.3% less 0.4% for an assumed growth
rate of 5.9%. Using these data, we estimate that health spending in New York will grow from
$189.0 billion in 2011 to $318.8 billion in 2020.
When necessary, we prepared two sets of estimates for each policy option. We first estimated
the impact of implementing these policies throughout the State for all New York residents
who are not already covered under similar programs, regardless of the source of payment.
These estimated savings are described as potential savings, or those reflective of the full
universe of savings if the policy options were implemented for all payers (Federal, State, and
—16—
0HWKRGVFRQWLQXHG
commercial). The second set of estimates corresponds to the cost impact of applying these
policies to the groups that are subject to State control or regulation. These estimated savings
indicate those that could be realized though specific steps that the State could take (termed
“actionable” savings throughout this paper). Estimates were distributed across payers based
on the baseline distribution of health spending for types of service affected by the option.
For example, hospital P4P savings were distributed based on the distribution of baseline
hospital spending by payer. Unfortunately, none of the data sources used permit us to develop
estimates at the sub-state level.
The actual effects of these options over the next decade could differ substantially from
our projections. Unforeseeable changes in new technology and disease prevalence could
dramatically alter spending growth trends. Our estimates are also based upon the results
of demonstrations and similar initiatives implemented elsewhere, and often in an ideal
environment for a given option, such as size of physician group or physician affiliation with
hospitals. Thus, the effect of an option on spending once broadly implemented may differ
from our estimates.
Also, it is possible that the loss of income under these policy options could cause spending
shifts and increased utilization for other services. For example, some physicians may increase
the number of services prescribed in other ways, such as adding new imaging equipment
to their office, to replace the lost income. Hospitals experiencing reduced admissions may
seek to fill beds through elective surgeries. To bend the health care cost curve, a focused policy
effort is needed to ensure reduced spending is not inappropriately offset by new spending.
The growing emphasis on reporting and health outcomes will make it difficult for physicians
and hospitals to provide services that do not enhance outcomes or health status.
In addition, savings to providers or insurers are not savings to consumers unless the savings
are passed back in the form of lower prices and/or premiums. In this analysis, we assume that
the savings would come as a reduction in the rate of growth in health spending. For example,
savings from simplification of provider credentialing would effectively reduce the increase in
provider prices over time, resulting in reduced premium growth. However, to the extent that
savings are retained by providers as income, these savings would not represent a reduction in
health spending and premiums.
—17—
Promoting Accountable Care Organizations (ACOs)
BASELINE CUMULATIVE POTENTIAL SAVINGS CUMULATIVE ACTIONABLE SAVINGS
SPENDING (2011–2020) (2011–2020)
SUBCATEGORY (billions) (billions)
(2011–2020)
(billions)
$ % $ %
Mandatory $1,096.56 $49.80 4.5% $14.59 1.3%
Voluntary $1,096.56 $10.71 1.0% $3.11 0.3%
Baseline spending includes FFS primary and acute care spending for all State residents.
BACKGROUND
$
ccountable Care Organizations (ACOs) have been proposed as an opportunity for health
care providers to band together to coordinate care and share in the savings that can
be realized through improvements in the quality and efficiency of care. Health care
providers receive a portion of the resulting savings in the form of bonuses for achieving quality
targets and reducing overall spending growth for a defined group of patients. A portion of the
savings are retained by health plans resulting in lower premiums for private insurance and
reduced spending under public programs.
The ACO model creates incentives for providers to emphasize primary care, prevention, and
adherence to evidence-based guidelines. This can take the form of a medical home model,
disease management (DM) program, or coordination of care for patients with multiple health
conditions. It is also compatible with value-based benefit design that encourages the use of
cost-effective services by reducing the copayments associated with that care. In addition, the
ACO model reduces incentives to acquire new technology that would add to service volume
without improving efficiency and quality, which would reduce potential bonus payments.
ACOs are a new approach to health care delivery that has yet to be demonstrated. The
Brookings Institution will be jointly sponsoring a pilot ACO project with the Dartmouth Institute
for Health Policy and Clinical Practice beginning in 2010. The Patient Protection and Affordable
Care Act requires a pilot demonstration of the ACO model for Medicare and Medicaid, similar
to the “voluntary” model described below. The bill also allows providers who voluntarily meet
quality standards to share in cost savings that result.
There are several varieties of the ACO model that have been discussed. In general, the ACO
proposals give providers an opportunity to share in savings from efficiencies without actually
being “at risk” for patient costs. However, the concept could be modified so that providers are
capitated for at least some portion of the care provided to patients, which should enhance
savings. Providers could also accept bundled payments for individual episodes of care which
could be awarded to ACO-like organizations on the basis of competitive bidding.
POLICY OPTION
A recent report from the Medicare Payment Advisory Commission (MedPAC) outlined two
alternative ACO models, including a “mandatory” model and a “voluntary” model. Under the
mandatory model, all physicians are assigned to an ACO that includes at least one hospital,
while the voluntary model gives providers the option of forming an ACO. The voluntary model
—18—
3URPRWLQJ$FFRXQWDEOH&DUH2UJDQL]DWLRQV$&2VFRQWLQXHG
has limited savings potential because the bonus must be high enough to attract providers to
the program. By comparison, the mandatory model can set lower bonus levels and can actually
impose penalties for poor performance. In both the mandatory and voluntary models, patients
have the option of obtaining care from any provider, including those not participating in the ACO.
In this analysis, we estimated the effect of implementing these two versions of the ACO model
in New York, including one that is mandatory and one that is voluntary.
To illustrate the savings potential of a mandatory ACO, we estimated the cost impact of
adopting such a program that applies to all individuals and all providers in the State of New
York. Existing networks of providers would be permitted to form their own ACO. We assume
that the State then creates ACOs throughout New York from those providers who have not
formed an ACO of their own. We assume that each ACO includes at least one hospital and
enough physicians to have a panel of at least 10,000 patients. We also assume that physicians
are allocated to ACOs in a way that includes a cross-section of specialist providers.
We assumed that savings would be the same as the estimated reduction in utilization under
the Independent Practice Association (IPA) model HMO, estimated to be approximately 4%.16
IPA HMOs are similar to ACOs in that they are associations of providers who are paid on a
fee-for-service (FFS) basis, but have a potential bonus structure that is similar to the ACO
model. However, we reduced these savings by one-third to reflect that the ACOs cannot control
out-of-ACO utilization as can an HMO. Unlike HMOs, the ACO cannot require patients to use
only providers participating in the ACO, which diminishes the ACO’s ability to reduce costs.
We assume no changes in spending for people already in HMO plans, because their health care
providers already operate under a bonus structure similar to the ACO.
Based upon available research, we also assume that ACOs can slow the rate of growth
in health care costs. For example, incentives for providers to obtain new technology such
as imaging equipment are expected to be reduced in an ACO model. Studies have shown
that health spending growth is slowed as HMO market share increases.17 Based upon these
studies, we assumed that the annual rate of growth in health spending for affected groups
would be reduced by approximately 0.6 percentage points per year. We assume that these
savings phase in over a period of four years.
ESTIMATED EFFECTS
We estimate that the total amount of health spending in New York affected under the ACO
model would be approximately $1.1 trillion over the 2011 through 2020 period. This total
includes covered health spending for acute care for all State residents covered under FFS18
insurance over that period (excludes long-term care services). If the policy were applied to all
payers, we estimate that under the assumptions listed above, the mandatory ACO model could
16 Miller, R.H., and Luft, H.S., “Managed Care Plan Performance Since 1980: A Literature Analysis,” Journal of the American Medical Association,
Vol. 271, No. 19, May 18, 1994, pp. 1512-1519.
17 Robinson, J.C., “HMO Market Penetration and Hospital Cost Inflation in California,” Journal of the American Medical Association,
266 (20 November 1991): 2719-23.
18 Managed care organization costs were excluded due to their inclusion of “managed” practices.
—19—
reduce spending by $49.8 billion over that 10-year period (Figure 2).19 We estimate that under
the assumptions listed above, the voluntary ACO model could reduce spending by $10.7 billion
over that 10-year period.
FIGURE 2. Potential and Actionable Savings under the ACO Model for
New York: 2011–2020 (billions) *
MANDATORY ACO MODEL VOLUNTARY ACO MODEL
POTENTIAL SAVINGS ACTIONABLE SAVINGS POTENTIAL SAVINGS ACTIONABLE SAVINGS
2011 $0.86 $0.25 $0.18 $0.05

2012 $2.01 $0.58 $0.43 $0.12
2013 $2.79 $0.81 $0.59 $0.17
2014 $3.46 $1.00 $0.74 $0.21
2015 $4.26 $1.24 $0.91 $0.26
2016 $5.16 $1.50 $1.10 $0.32
2017 $6.16 $1.79 $1.31 $0.38
2018 $7.26 $2.11 $1.55 $0.45
2019 $8.49 $2.46 $1.81 $0.52
2020 $9.35 $2.86 $2.10 $0.61
2011–2020 $49.80 $14.59 $10.71 $3.11
* “Potential” savings are the amounts that could be saved if all public and private payers were to adopt these programs. “Actionable” steps include
savings that could be realized through State action under Medicaid or State and local government worker health benefits programs.
Source: The Lewin Group estimates. Numbers may not add to totals due to rounding.
DISCUSSION
ACOs represent an emerging concept, thus there are no data available for estimating its likely
impact. Consequently, as discussed above, we have estimated potential savings based upon the
experience in IPA HMOs. IPAs are similar in structure to ACOs in that providers continue to be
paid on a FFS basis, typically with a program of withholds and bonuses similar to that proposed
for ACOs. While IPAs could form the basis of ACOs, we have not identified any components
of the ACO model that would make it more effective than an IPA HMO, and in fact our savings
assumption is less than that observed for IPAs. Therefore, we assume no additional savings for
individuals now receiving care through an HMO.
The effect of promoting the development of ACOs in New York without Federal intervention
would be limited because the State cannot require Medicare or Employee Retirement Income
Security Act (ERISA) plans to adopt such reforms. Thus, to estimate actionable savings we
assume that the State requires adoption of ACOs for all State and local workers not already in
an HMO, and the FFS Medicaid population. We acknowledge that this requires sufficient ACO
capacity, which could take a considerable amount of time to develop. We assume no change
for integrated delivery systems, such as HMOs, because they are already serving people under
19 The percentage savings increase at an increasing rate because we include a reduction in the rate of growth in health spending. So savings
result from both: 1) adopting more efficient medical practice; and 2) a reduced rate of increase from year to year, reflecting reduced
incentives to acquire new technology.
—20—
capitated health plans. The PPACA signals increasing Federal interest in the ACO concept by
authorizing shared savings with ACOs under Medicare beginning in 2012, and including funding
for a Pediatric ACO demonstration under Medicaid.
With both mandatory and voluntary ACO models, the potential to control costs is diminished in
cases where patients assigned to the ACO use providers that are not participating in that ACO.
For example, a patient seeing an ACO participating primary care provider may still access a
non-participating specialist who may use a hospital outside the ACO. However, it is possible that
incentives, such as premium discounts or co-pay penalties, could reduce non-ACO utilization.
Integrated delivery systems, such as HMOs, are able to direct patients to use only providers in
the plans’ networks, but there is nothing in the ACO model that restricts access for patients
who wish to use non-ACO providers. This lack of control over patient access limits the cost-
savings potential for the ACO under both the voluntary and mandatory models. In addition, while
we assumed a phase-in period of four years for this scenario, it should be noted that some
hospitals are better equipped than others to join ACOs and may be able to do so quickly, while
others would be far more challenged.
The ACO scenario overlaps with others that we have modeled, including Modernizing Primary
Care and Bundled Payments, all of which result in savings through improved efficiency in the
delivery of care. It may be possible to develop a combined approach that would include both
ACOs and a payment reform, such as bundled payments, as ACOs work to decrease the number
of episodes while bundled payments work to reduce costs within each episode. However,
because both options improve efficiency in the system, savings are not necessarily additive.
—21—
Modernizing Primary Care
SPENDING (2011–2020) (2011–2020)
SUBCATEGORY (billions) (billions)
(2011–2020)
(billions)
$ % $ %
Mandatory Medical Home $1,620.97 $33.66 2.1% $10.99 0.7%
Voluntary Medical Home $1,620.97 $9.11 0.6% $2.25 0.1%
Care Coordination * $1,620.97 $0 0.0% $0 0.0%
Advanced Disease Mgmt. $1,620.97 $10.74 0.7% $3.30 0.2%
Pay for Performance $1,620.97 $1.33 0.1% $0.46 0.0%
Baseline spending includes total primary and acute care spending for all State residents.
* Based on the findings of the Medicare Coordinated Care Demonstration, we assumed that widespread adoption of this version of the coordinated
care model would not result in net savings, although quality may improve.
BACKGROUND
3
roposals to improve primary care are intended to promote prevention and management
of care that will minimize costly avoidable complications. A common focus of these
initiatives is to coordinate care for people with chronic conditions, which accounts for
75% of all health spending according to the available evidence.20 A variety of approaches to
modernizing primary care have been developed and/or implemented nationwide by both public
and private payers, all of which have overlapping goals. These include medical homes, care
coordination, disease management programs, and pay-for-performance (P4P) programs.
The medical home initiative would identify a primary care provider as a patient’s “medical
home,” who would be paid to provide coordinated evidence-based primary care through a team
of providers, including physicians, physician assistant, nurse practitioners, nurses, behavioral
health professionals, and others. Disease management includes programs designed to ensure
that patients with chronic health conditions are treated according to evidence-based guidelines.
Coordinated care is targeted at people with multiple chronic health conditions who are often
in the care of several specialists at once.21 Under a coordinated care model, the primary care
provider coordinates the care provided by these multiple specialists to avert negative outcomes,
such as drug-to-drug interactions and duplicative tests and services. Of course, this
approach overlaps with medical homes and DM.
Under the P4P model, physicians are rewarded for keeping people healthy and for obtaining
favorable health outcomes. P4P can also be structured to reward physicians if they reduce
costs. Under this program, data are collected on providers and compared with various
performance benchmarks to measure the relative quality of patient health, outcomes, and
costs. Providers who exceed these performance benchmarks receive a bonus. Thus, P4P
20 Congressional Budget Office. High-cost Medicare beneficiaries. A CBO Paper. May 2005. Web site: http://www.cbo.gov/showdoc.
cfm?index=6332&sequence=0.
21 Thorpe (2007), “Potential Savings Under the AdvaMed Plan Associated with Health Reforms Focusing on Chronic Care Management, Prevention
and Health Information Technology” found at: http://www.advamed.org/NR/rdonlyres/03AE0ADD-3472-4F29-BC58-32EC0575AB67/0/
healthreformsavingsthorpeFINAL.pdf.
—22—
0RGHUQL]LQJ3ULPDU\&DUHFRQWLQXHG
relies upon a bonus system to change physician incentives while the medical home, DM, and
coordinated care programs pay providers to provide evidence-based primary care services.
POLICY OPTION
Medical Homes. As a relatively new concept, there are several different definitions of
what a medical home is, but they all include the concept of coordinated care provided by
a team of professionals and led by a primary care provider. New York State is currently
working to implement medical home incentives as part of the Medicaid program. We
examined two versions of the medical home model, including the “mandatory” model and
the “voluntary” model. Under the mandatory model, all patients select a medical home
provider. Patients are required to access all medical care from or on the referral of medical
home providers. The providers receive a fee to provide preventive and primary care services
for their patients. We assume that, under a mandatory medical home model, savings
would be in proportion to savings documented for Primary Care Case Management (PCCM)
programs under Medicaid (based on a study of savings under a Medicaid PCCM program in
Iowa by Momany et al.).
Under a voluntary medical home model, patients are encouraged, rather than required, to
access all care through the medical home provider, who is paid a fee to provide these services.
Patients have the option of participating in the program. For those who do, copayments are
eliminated for care provided through the medical home as an incentive to participate. Patients
may see specialists without referral but would have a copayment. Little data are available on
the potential impact of the voluntary model on health care costs. For illustrative purposes, we
assume that its cost savings effects are approximately half as great as under the mandatory
model. We also assume that voluntary participation would phase-in over time.
For both the mandatory and voluntary options, we assume that all primary care providers
would be willing to participate as medical home providers, although it would not be mandatory.
Care Coordination. Approximately 75% of all health spending is attributed to people with
chronic health conditions. Many of these patients have multiple health conditions and are often
in the care of several specialists at once.22 Under the coordinated care model, the primary
care physician coordinates the care provided by these multiple specialists to avert negative
outcomes, such as drug-to-drug interactions and duplicative tests and services. Physicians are
paid a fee for providing these care coordination services. Based on the findings of the Medicare
Coordinated Care Demonstration, we assumed that widespread adoption of this version of the
coordinated care model would not result in net savings, although quality may improve.
Disease Management. Blue Cross and Blue Shield of Minnesota (BCBSMN) implemented
a program that uses predictive modeling to identify people with multiple chronic conditions
who can be expected to require relatively high levels of medical care. These individuals
are then asked to enroll in an expanded DM program covering 17 chronic conditions, in which
22 Thorpe, “Potential Savings Under the AdvaMed Plan Associated with Health Reforms Focusing on Chronic Care Management, Prevention and
Health Information Technology,” (2007), found at: http://www.advamed.org/NR/rdonlyres/03AE0ADD-3472-4F29-BC58-32EC0575AB67/0/
healthreformsavingsthorpeFINAL.pdf.
—23—
they receive proactive interventions designed to prevent conditions from becoming acute. 23
For purposes of this analysis, we estimated the effects of applying the BCBSMN advanced
DM approach to the New York population based upon the Minnesota experience.
Pay for Performance (P4P). We estimated the impact of a P4P program for physician care
based upon the results of the CMS Physician Group Practice (PGP) Demonstration. In this
demonstration, 10 large physician groups were selected to receive bonus payments for realizing
savings for the Medicare patients they serve. Savings were split between the providers and
the program, with providers receiving 80% and the program retaining 20%. Each of the groups
included had at least 200 physicians, and included freestanding group practices, faculty group
practices, and physician networks including small and individual practices.24
ESTIMATED EFFECTS
We assume that the universe of health spending potentially affected by changes in primary care
includes benefits payments for all primary and acute care services covered under Medicare,
Medicaid, and private insurance. These include doctor office visits, inpatient care, hospital
outpatient care, and emergency room care for these payers. This excludes spending under
these programs for nursing homes, home health, public health, and medical non-durable goods
other than prescription drugs (e.g., aspirin, cough syrup, bandages, etc.).
Medical Homes. We estimate that the mandatory medical home model could reduce health
spending in New York State by $33.7 billion over the 2011 through 2020 period. Figure 3 shows
how net savings are distributed by payer.
FIGURE 3. Potential Savings from Adopting a Mandatory Medical

Home Program for All Payers in New York (billions) *
FEDERAL STATE AND LOCAL PRIVATE TOTAL
YEAR HOUSEHOLDS
GOVERNMENT GOVERNMENTS EMPLOYERS SAVINGS
2011 $0.05 $0.02 $0.03 $0.01 $0.11
2012 $0.77 $0.35 $0.48 $0.21 $1.80
2013 $1.05 $0.47 $0.64 $0.27 $2.44
2014 $1.37 $0.61 $0.82 $0.35 $3.15
2015 $1.52 $0.68 $0.89 $0.38 $3.46
2016 $1.67 $0.74 $0.96 $0.41 $3.78
2017 $1.84 $0.81 $1.03 $0.44 $4.13
2018 $2.03 $0.89 $1.11 $0.48 $4.51
2019 $2.24 $0.97 $1.20 $0.51 $4.92
2020 $2.45 $1.06 $1.29 $0.55 $5.35
2011–2020 $14.99 $6.61 $8.45 $3.61 $33.66
* We estimated savings occurring among all three payer groups in both fee-for-service and managed care plans.
Savings in managed care plans are estimated at half the rate of fee-for-service.
23 Gold, W. & Kongstvedt, “How Broadening DM’s Focus Helped Shrink One Plan’s Costs,” Managed Care Magazine, November 2003.
24 Kathleen Sebelius, Secretary HHS, “Physician Group Practice Demonstration Evaluation Report to Congress,” US Department of Health and
Human Services, 2009.
—24—
We estimate that the voluntary medical home model could reduce health spending in
New York State by $9.1 billion over the 2011 through 2020 period. Figure 4 shows how these
savings are distributed by payer.
FIGURE 4. Potential Savings from Adopting a Voluntary Medical

Home Program for All Payers in New York (billions) *
YEAR FEDERAL GOVERNMENT STATE AND LOCAL GOVERNMENTS PRIVATE EMPLOYERS HOUSEHOLDS TOTAL SAVINGS
2011 -$0.25 -$0.12 -$0.16 -$0.07 -$0.60
2012 $0.08 $0.04 $0.05 $0.02 $0.19
2013 $0.21 $0.10 $0.13 $0.06 $0.49
2014 $0.35 $0.16 $0.21 $0.09 $0.81
2015 $0.42 $0.19 $0.25 $0.11 $0.96
2016 $0.49 $0.22 $0.28 $0.12 $1.11
2017 $0.57 $0.25 $0.32 $0.14 $1.27
2018 $0.65 $0.28 $0.36 $0.15 $1.44
2019 $0.74 $0.32 $0.40 $0.17 $1.63
2020 $0.83 $0.36 $0.43 $0.19 $1.81
2011–2020 $4.09 $1.79 $2.26 $0.97 $9.11
* We estimate savings occurring among all three payer groups in both FFS and managed care plans.
Savings in managed care plans are estimated at half the rate of FFS.
Care Coordination. An evaluation of the Medicare Coordinated Care Demonstration concluded

that there is no evidence to suggest that this coordinated care model would reduce program
expenditures. Therefore, we assumed that widespread adoption of this version of the coordinated
care model would not result in net savings, although quality may improve.
Disease Management. We estimate that the DM model could reduce health spending in New
York State by $10.7 billion over the 2011 through 2020 period. Figure 5 shows how these savings
are distributed by payer.
FIGURE 5. Potential Savings from Adopting an Advanced Disease

Management Model in New York by Payer (billions) *
YEAR FEDERAL GOVERNMENT STATE AND LOCAL GOVERNMENTS PRIVATE EMPLOYERS HOUSEHOLDS TOTAL SAVINGS
2011 -$0.01 -$0.01 -$0.02 -$0.01 -$0.06
2012 $0.05 $0.06 $0.11 $0.05 $0.26
2013 $0.09 $0.10 $0.18 $0.08 $0.45
2014 $0.18 $0.18 $0.34 $0.15 $0.85
2015 $0.24 $0.24 $0.45 $0.19 $1.13
2016 $0.28 $0.28 $0.53 $0.22 $1.32
2017 $0.33 $0.33 $0.61 $0.26 $1.53
2018 $0.36 $0.36 $0.64 $0.27 $1.64
2019 $0.39 $0.38 $0.68 $0.29 $1.75
2020 $0.43 $0.41 $0.72 $0.31 $1.87
2011–2020 $2.36 $2.33 $4.24 $1.81 $10.74
* Based upon the available research, we estimate savings occur only in private health plans, including managed care plans under Medicare and
Medicaid. Savings in managed care plans are estimated at half the rate of FFS.
—25—
Pay for Performance (P4P). We estimate the amount of savings resulting from a physician
P4P program for New York for 2011 through 2020. We estimate total savings of $6.6 billion
(Figure 6). The participating physicians could receive bonus payments of $5.3 billion (i.e.,
80%) with payers retaining $1.3 billion. While we assume that savings to payers represent net
reductions in total health spending, physicians are assumed to retain the bonus payments as
income. Thus, the program could reduce health spending in New York by $1.3 billion over the
2011 through 2020 period.
FIGURE 6. Potential Savings from Adopting a Physician Pay-for-

Performance Program in New York by Payer (billions) *
STATE AND
FEDERAL PRIVATE GROSS NET SAVINGS
YEAR LOCAL HOUSEHOLDS
GOVERNMENT EMPLOYERS SAVINGS BY PAYERS
GOVERNMENTS
2011 $0.08 $0.04 $0.05 $0.02 $0.20 $0.04

2012 $0.19 $0.09 $0.12 $0.05 $0.44 $0.09
2013 $0.25 $0.11 $0.15 $0.07 $0.58 $0.12
2014 $0.27 $0.12 $0.16 $0.07 $0.62 $0.12
2015 $0.29 $0.13 $0.17 $0.07 $0.67 $0.13
2016 $0.31 $0.14 $0.18 $0.08 $0.71 $0.14
2017 $0.34 $0.15 $0.19 $0.08 $0.76 $0.15
2018 $0.37 $0.16 $0.20 $0.09 $0.82 $0.16
2019 $0.40 $0.17 $0.21 $0.09 $0.88 $0.18
2020 $0.43 $0.19 $0.23 $0.10 $0.94 $0.19
2011–2020 $2.94 $1.30 $1.67 $0.72 $6.63 $1.33
* We estimate savings occurring among all three payer groups in both FFS and managed care plans. Savings in managed care plans are estimated at
half the rate of FFS.
The estimates presented above are the potential savings from adopting these primary care
programs assuming that all payers in the State adopt these programs. In fact, the State
legislature has little control over most health spending in the State, as states are prohibited
from regulating private self-funded health plans under the Employee Retirement Income
Security Act ERISA, which encompasses approximately 40% of private coverage. In addition,
as a Federal program, Medicare benefits are not subject to state control.
For illustrative purposes, we estimated the impact of these programs assuming that the
State requires use of these primary care models in all State-sponsored health plans.
—26—
Figure 7 presents our estimates of total potential savings (as estimated above) and the amounts
that could be saved through “actionable” steps available to the State legislature.
FIGURE 7. Potential and Actionable Savings under Alternative Primary Care Models
for New York: 2011–2020 (billions) *
MANDATORY VOLUNTARY ADVANCED DISEASE PHYSICIAN PAY FOR PERFORMANCE
MEDICAL HOME MEDICAL HOME MANAGEMENT POTENTIAL SAVINGS ACTIONABLE SAVINGS
POTENTIAL ACTIONABLE POTENTIAL ACTIONABLE POTENTIAL ACTIONABLE TOTAL SAVINGS NET TOTAL SAVINGS NET
SAVINGS SAVINGS SAVINGS SAVINGS SAVINGS SAVINGS SAVINGS OF BONUS SAVINGS OF BONUS
2011 $0.11 -$.20 -$0.60 -$0.84 -$0.06 -$0.03 $0.20 $0.04 $0.07 $0.01
2012 $1.80 $0.46 $0.19 -$0.29 $0.26 $0.08 $0.44 $0.09 $0.15 $0.03
2013 $2.44 $0.71 $0.49 -$0.08 $0.45 $0.14 $0.58 $0.12 $0.20 $0.04
2014 $3.15 $0.99 $0.81 $0.15 $0.85 $0.27 $0.62 $0.12 $0.21 $0.04
2015 $3.46 $1.12 $0.96 $0.25 $1.13 $0.35 $0.67 $0.13 $0.23 $0.05
2016 $3.78 $1.26 $1.11 $0.36 $1.32 $0.40 $0.71 $0.14 $0.25 $0.05
2017 $4.13 $1.41 $1.27 $0.48 $1.53 $0.47 $0.76 $0.15 $0.26 $0.05
2018 $4.51 $1.56 $1.44 $0.61 $1.64 $0.50 $0.82 $0.16 $0.28 $0.06
2019 $4.92 $1.74 $1.63 $0.73 $1.75 $0.54 $0.88 $0.18 $0.31 $0.06
2020 $5.35 $1.93 $1.81 $0.88 $1.87 $0.58 $0.94 $0.19 $0.33 $0.07
2010–2020 $33.66 $10.99 $9.11 $2.25 $10.74 $3.30 $6.63 $1.33 $2.28 $0.46
* “Potential” savings are the amounts that could be saved if all public and private payers were to adopt these programs. “Actionable” savings include
those that could be realized through State action under Medicaid or state and local government worker health benefits programs.
DISCUSSION
Each of these scenarios assumes that a sufficient number of primary care physicians and
practitioners will be available to provide the medical home and DM services. Workforce could
be a significant barrier for the Medicaid population due to the relatively lower payment levels
for physician services under the program. Although, this barrier will be somewhat mitigated by
the Patient Protection and Affordable Care Act (PPACA), which increases Medicaid primary care
rates to Medicare levels for two years. The policies described in this proposal should dovetail
with overall State health care workforce policy. PPACA also makes investments in workforce
and primary care access, which could help to bolster State workforce needs.
We considered forming a new model of primary care that includes elements of the medical
home, DM, coordinated care, and the P4P program. However, there is such overlap among
these approaches that it is not clear how one would integrate them. These approaches tend
to be substitutes for each other because they are designed to induce the same changes in
medical practice.
—27—
For example, if we were to establish a medical home model together with a P4P program, most
of the savings achieved through the medical home would be shared with the physician (i.e.,
80%), thus reducing overall savings to consumers. In addition, it would make no sense to pay
providers a fee to provide DM services when they are also receiving a fee for medical home
services because the medical home already emphasizes DM.
Consequently, payment and incentive structures under these programs would need to be rethought
if they were integrated into a single program. Finding a way to integrate multiple approaches to
improve savings potential will be a major task in designing a new model of primary care.
CMS will initiate a demonstration of the medical home concept that will begin in 2010 for Medicare
patients. It may be several years before we have evidence on its cost impacts, and then only for
the aged and disabled population. New York State is also working to implement medical homes,
including a statewide Medicaid program and a multi-payer pilot in the Adirondacks. However, the
multi-payer pilot is in its very early stages and the start of the Medicaid program is pending CMS
approval of a State Plan Amendment. PPACA also provides opportunities to advance medical
homes, including through developing community health teams and establishing health homes
that would receive enhanced Medicaid payment for treating those with chronic conditions. Clearly
the evidence base for cost impacts under medical homes is still developing, and costs are likely
to increase in the early years due to care management fees. However, savings are ultimately
expected to more than offset these costs over time.
—28—
Expanding Palliative Care
BASELINE SPENDING POTENTIAL SAVINGS (2011–2020) ACTIONABLE SAVINGS (2011–2020)
(2011–2020) (billions) (billions) (billions)
$ % $ %
$235.38 $11.93 5.1% $11.93 5.1%
Baseline spending includes cost of hospitalizations for people over age 18 with seven to 30 day stays.
BACKGROUND
$
study by Hogan et al. estimates that approximately 27.4% of health spending under
Medicare occurs at end of life.25 This has stimulated interest in palliative care
consultation programs designed to allow the patient greater control over the kind of
care they receive when seriously ill or dying. According to Morrison, palliative care programs
have been shown to improve physical and psychological symptom management, caregiver well-
being, and family satisfaction. They have also been shown to result in lower health spending.26
Palliative care is a process in which patients and families meet with the physician to identify
patient goals at a given stage of illness and to adopt a treatment plan that is consistent
with these goals. The process typically results in “advance directives,” which identify the type
of life prolonging care that a patient is willing to receive. For example, a patient may specify
a “do not resuscitate” order or other instruction to limit life-prolonging care. Palliative care
also includes developing a plan for “pain management” and other instructions designed
to improve quality of life. Palliative care is often provided in hospitals, but can be provided in
the community as well. Palliative care teams are multi-disciplinary, including, for example,
palliative care physicians, specialists, nurses and nurse aides, social workers, therapists,
and spiritual advisors.
POLICY OPTION
Palliative care consultation programs have been shown to reduce costs for both decedents
and survivors of high-cost illnesses, while improving the quality of care provided at end
of life. 27 One study of the impact of palliative care on costs was conducted by Morrison. 28
The study analyzed data from eight hospitals with established palliative care (PC)
programs for the years 2002 through 2004. The study compared costs for patients receiving
PC and those receiving usual care (UC), which it standardized for severity of illness and
demographic characteristics. After adjusting for the cost of the PC team (less revenues),
net savings were approximately 13.4% for patients discharged alive and 19.1% for those who
died in hospital. These savings include the cost of the PC team less revenues for providing
25 Hogan, C., ”Medicare Beneficiaries Costs of Care In the Last Year of Life,” Health Affairs, Vol.20, No. 4, 2001.
26 Morrison, S.R., et al, “Cost Savings Associated with U.S. Hospital Palliative Care Consultation Programs,” Archives of Internal Medicine,
168(16), 2008.
27 L. Gelfman et al., “Does Palliative Care Improve Quality? A Survey of Bereaved family Members,” NIH Public access manuscript, February 15, 2007.
28 Morrison, S.R., et al, “Cost Savings Associated with U.S. Hospital Palliative Care Consultation Programs,” Archives of Internal Medicine,
168(16), 2008.
—29—
([SDQGLQJ3DOOLDWLYH&DUHFRQWLQXHG
these services. 29 These results are consistent with studies showing that those who receive
hospice care have lower average expenditures at end of life than Medicare beneficiaries who
do not use hospice. 30
The Morrison study indicates that among live discharges with a length of stay (LOS) of
seven to 30 days, approximately 6.8% received palliative care consultation. 31 Approximately
50.5% of patients with that LOS who died in the hospital received PC. There were several
limitations to the Morrison study. First, the study documented reduced costs within an
admission, but did not show savings over time as a result of data limitations. Second, while
the study documented savings, it did not show that these savings actually translated into
savings to the payer.
Meier documented the impact of palliative care on costs and patient satisfaction at Mount
Sinai Hospital in New York City. 32 The study showed cost savings for Medicare beneficiaries
in 2001 of $757,500 per year for patients with a length of stay greater than 14 days, and
$455,900 per year for patients with a LOS greater than 28 days. When costs were compared,
savings were approximately 4.3 times the cost of providing palliative consultation, which is
a return on investment of 4.3:1. The study also reports a high level of satisfaction for patients
and family members.
We estimated potential savings resulting from expanded use of PC through two paths:
1) expanding the PC program in hospitals that already have a program; and 2) establishing
a PC program in those hospitals that do not already have one. Specifically, we assumed
that all hospitals in the State are required to adopt a PC program. We also assumed that
hospitals are required to obtain certification that PC is offered to all patients treated for
chronic illness. Thus, we modeled both the increase in the number of hospitals with PC
programs and an increase in the proportion of patients in hospitals that currently receive
these consults.
A recent study showed that approximately 57% of hospitals in New York have a palliative care
program. 33 We assumed that hospital adoption of palliative care programs under current law
will increase at the same rate observed over the 2000 through 2008 period. We also assumed
that, under this policy option, the percentage of patients receiving these services in hospitals
that have already established a PC program would increase from a current level of 2.5% to
9.5%. This is based upon the Morrison research suggesting that the optimum level is 10% of
29 We calculated the adjustment based upon data provided by the author on savings, PC team costs, and revenues for PC services
for a typical 400 bed hospital.
30 M.B.Buntin and H.Huskamp, “What Is Known About the Economics of End-of-Life Care for Medicare Beneficiaries?”
Gerontologist, (2002); 42(3): p.9.
31 The Morrison sample included 43,973 patients with live discharges with an LOS of 7 to 30 days, of which 2,966 received PC consultation (6.8%). Of
the 4,726 patients who died in hospital with an LOS of 7 to 30 days 2,388 received PC consultation (50.5%).
32 D. Meier, “Palliative Care Improves Quality and Reduces Costs in Hospital and Community Settings,” Director, Hertzberg Palliative Care Institute.
33 “America’s Care of Serious Illness: A State-by-State Report Card on Access to Palliative Care in Our Nation’s hospitals,”
Center to Advance Palliative Care and National Palliative Care Research Center, September 2008.
—30—
all discharges, which we discounted to 9.5% as the “achievable” rate of PC consults under
the assumption that the optimum level would not be fully realized. We assume that this
increase would phase in over the first four years of the requirement.
We estimated potential savings to hospitals under this policy option using the New York State
hospital discharge data, which record information for each hospital discharge in the State
including hospital charges, LOS, and type of discharge (i.e., live, death). In keeping with the
Morrison analysis, we selected hospital stays of seven to 30 days. We estimated costs for each
discharge using the cost to charge ratio reported for each hospital taken from the Medicare
cost report data for New York hospitals. 34
We also estimated reductions in spending for physician care other than that provided by
hospital staff (which is included in hospital costs), including costs for the attending physician
and physician consults while in the hospital. Unfortunately, the New York hospital discharge
data do not include these physician costs. For this study, therefore, we estimated physician
costs while in the hospital based upon the Medical Expenditure Panel Survey (MEPS) data,
which include the cost of physician charges associated with each hospital admission. 35
We assumed that savings in physician care would be in proportion to the Morrison estimates
of savings in hospitals.
We then calculated hospital cost savings using the percentage savings estimated by Morrison
for people receiving PC. As discussed above, these estimates of percentage savings from
PC consultation include savings net of the cost of providing the PC services.
ESTIMATED EFFECTS
We estimated the cost of hospitalizations for people over the age of 18 with seven- to 30-day
stays directly from the New York Hospital Discharge data. 36 We adjusted these data to 2011
based upon CMS projections of health care cost growth. Using this approach, we estimated
that hospital inpatient revenues in New York will be $41.1 billion in 2011, of which approximately
$17 billion is attributed to people over the age of 18 with a hospital stay of seven to 30 days.
Using the assumptions discussed above and detailed in the Technical Appendix, we estimate
that this policy option could reduce health spending by $11.9 billion over the 2011 through 2020
period. This includes savings in hospital and physician services less the cost of implementing
the PC program. We estimated the distribution of these savings across payers in proportion
34 Using these assumptions, Hospital stays with a LOS of 7 to 30 days account for approximately 41.4% of hospital costs in New York.
35 The MEPS reports spending by geography region only and does not permit state-level estimates. Thus, we based these estimated physician costs
as a percentage of hospital costs in the MEPS data for the north-east region of the country.
36 The New York discharge data include data for each hospital discharge in the state which provides information on patient age, length of stay
and total hospital charges. These data enable us to identify discharges for people over the age of 18 with a stay of 7 to 30 days by type of discharge
(live or dead). Hospital charges reported in these data were converted to costs based upon the hospital cost to charge ratio for each hospital.
—31—
to the source of payment reported in the New York Hospital Discharge data for the proportion
of hospital admissions that we assume would receive PC (Figure 8). 37
FIGURE 8. Savings from Requiring Hospitals to

Have Palliative Care Programs by Payer: 2011–2020 (millions)
FEDERAL STATE AND LOCAL PRIVATE TOTAL
YEAR HOUSEHOLDS
GOVERNMENT GOVERNMENTS EMPLOYERS SAVINGS
2011 $34 $5 $5 $4 $48

2012 $295 $45 $41 $34 $415
2013 $501 $76 $70 $58 $704
2014 $737 $111 $102 $85 $1,035
2015 $929 $140 $129 $107 $1,305
2016 $1,065 $161 $148 $122 $1,496
2017 $1,129 $170 $157 $130 $1,586
2018 $1,196 $180 $166 $138 $1,681
2019 $1,267 $191 $176 $146 $1,780
2020 $1,341 $202 $186 $154 $1,883
2011–2020 $8,493 $1,282 $1,181 $977 $11,933
DISCUSSION
Our estimates suggest that savings could be increased significantly by expanding the use of
PC consulting in hospitals that already have a program. As discussed in the introduction to
this paper, savings to providers or insurers are not savings to consumers unless the savings
are passed back in the form of lower prices and/or premiums. For example, under this
scenario, hospital cost savings associated with utilization reductions would have to lead to
reduced DRG payments which ultimately lead to reduced premiums. For the purposes of these
analyses, we assume that savings are passed back, although, as referenced previously, this is
not automatically the case per Morrison’s study.
In this scenario, we assume that all hospitals in the State are required to have a palliative
care program as a condition of licensure by the State and that hospitals are required to obtain
certification that PC is offered to all patients treated for chronic illness. We also assume that
PC programs would meet quality guidelines. (We note that national certification standards for
PC programs are reportedly under development, but not yet available.)
This policy would apply to all care provided in these hospitals regardless of payer source.
The service would be provided by the hospital regardless of whether health plans provide
separate payments for palliative care services. Thus, we assume that all adult patients with
37 As discussed above, we assumed that savings would occur only among patients with a LOS of 7 to 30 days. Of these, we assumed that 6.8% of
patients with live discharges would have been selected for palliative care consultation, and that 50.5% of patients who die in hospital would have
received palliative care consultation.
—32—
a hospital stay of seven or more days would be eligible to receive PC services regardless of
payer. We note that a current barrier to the expansion of PC programs is that, under current
reimbursement policy, only physicians and nurse practitioners can bill directly for services,
potentially limiting reimbursement and expansion of PC programs.
The New York State Health Foundation has funded a program with the Center to Advance
Palliative Care (CAPC) to broaden the reach of existing PC programs in New York State
hospitals. In its research, CAPC found that many of the existing PC programs in New York
hospitals are small in scale (fewer than 300 beds) and would be less likely to establish a PC
program than larger hospitals.
Palliative care can be provided in the community as well. The Kaiser Permanente Palliative
Care Project is a multidisciplinary care management approach for home-based end-of-
life care and treatment. The program is designed to facilitate the transition from acute to
palliative care during the last 12 months of life with the goal of improving quality of life
through the provision of symptom control, pain relief, emotional and spiritual support, and
patient education.
A study of this program showed that those enrolled in PC care averaged a 45% decrease in
costs as compared to UC patients. 38 This reduction reflects significantly fewer emergency
department visits, hospital days, skilled nursing facility days, and physician visits than those in
the comparison group. The study also found that PC patients had increased satisfaction with
services after 60 days of enrollment. Patients enrolled in the PC program were more likely to
die at home than UC patients.
Palliative care does not always show savings. One study of dementia patients found no savings
from palliative care for these patients. 39 The author theorizes that patients with advanced
dementia present unique barriers to palliative care providers, such as perceived uncertainty of
prognosis and difficulty assessing comfort level. However, it is likely that PC programs result
in more coordinated, higher value care, and that PC teams pull treatment options together and
help patients and families make the decision best for them. We also note that palliative care
savings would likely overlap with estimated savings under the long-term care scenarios that
were modeled, including both the mandatory managed care for dual eligibles and rebalancing
long-term care in New York Medicaid.
38 R.D. Brumley, S. Enguidanos, and D.A. Cherin, “Effectiveness of a Home-Based Palliative Care Program for End-of-Life,” Journal Palliative
Medicine, (2003); 6(5): p. 715-24.
39 J. Ahronheim et al. “Palliative care in Advanced Dementia: A Randomized Controlled Trial and Descriptive analysis,” Journal of Palliative Medicine,
Volume 3, Number 3, 2000.
—33—
Implementing Mandatory Managed Care for
Medicaid Dual Eligible Population
$ % $ %
$496.85 $10.76 2.2% n/a n/a
Baseline spending includes combined Medicare and Medicaid spending for dual eligibles.
BACKGROUND
'
ual eligibles are individuals who secure eligibility for both Medicare and Medicaid.
This subgroup tends to frequently use health care services. In 2008, dual eligibles
represented 10% of all national health spending, 40% of Medicaid spending, and 25%
of Medicare spending.40 The vast majority of dual eligibles’ spending occurs in traditional FFS
coverage environments, which can lead to poorly coordinated care and unnecessarily high costs.
The large size of New York’s dual eligible population (approximately 650,000 persons, currently)
and the very high per capita costs in this sector create opportunities for large-scale savings to
occur via a well designed and operated coordinated care program.
POLICY OPTION
This scenario estimates the impact of enrolling New York’s Medicaid/Medicare dual eligibles
into a fully integrated coordinated care setting. Under this model, dual eligibles would be
mandatorily enrolled into capitated managed care organizations (MCOs). MCOs would receive
the Medicare funding, pool them with Medicaid funds, and assume full financial risk for the
entire Medicaid and Medicare benefits package for their enrolled dual eligibles. Overall savings
achieved would split 50-50 between the Federal government and New York State. Capitation
rates would be established for this program in a manner that builds in net savings (which
requires a different payment methodology on the Medicare “side” than is currently used in the
Medicare Advantage program). 41
Several changes in Federal law and program rules would be needed to implement this bold
policy option. Federal law does not currently allow for mandatory enrollment of dual eligibles
into Medicare managed care, and CMS would need to grant waiver authority to allow New York
State to mandate enrollment of dual eligibles into Medicaid managed care.
The offsetting costs of using the capitated model involve the administrative cost allocations
that must be made to the participating MCOs, and the “profit” allocations that must be
made in order for health plans to take on the considerable financial risks that capitation
contracting entails. The Lewin Group estimates a 5% share of monthly MCO dual eligibles’
premiums are needed to cover the MCOs’ administrative costs, and that a 2% risk margin
40 Internal Lewin analysis.
41 MedPAC, Report to the Congress: Medicare Payment Policy; March 2010, p. 266.
—34—
,PSOHPHQWLQJ0DQGDWRU\0DQDJHG&DUHIRU0HGLFDLG'XDO
(OLJLEOH3RSXODWLRQFRQWLQXHG
(or profit margin) is needed. The 5% figure is extremely low (on a percentage basis) when
viewed alongside most administrative allocations that occur in the Medicaid managed
care and Medicare managed care arenas, but it is quite a large allocation when viewed on
a PMPM basis.
The low percentage allocation is made feasible by two dynamics. First, the nursing home costs
for institutionalized dual eligibles significantly increase the monthly capitation rate that the
MCOs require to serve the full spectrum of dual eligibles—but institutionalized persons cause
very little increase in administrative costs (essentially writing a monthly “rent check”). The
MCOs’ care management efforts will occur despite whether a person is institutionalized, and in
fact, are likely to be greatest for persons who are not institutionalized.
Second, the mandatory enrollment nature of the initiative also lowers administrative cost
requirements given that marketing expenses are essentially removed from the program.
Available dollars are thus focused on “serving” rather than on “selling.” In contrast, extensive
marketing costs are needed in the Medicare Advantage program to try to attract dual eligibles
to enroll, and these marketing efforts typically yield only modest enrollment levels since the
dual eligibles have first dollar coverage and no financial reason to enroll in an MCO.
The Medicaid long-term care cost figures for 2008 were derived from a data set provided to
The Lewin Group by New York State 42 and the Medicaid Statistical Information System (MSIS)
Medicare costs for New York’s dual eligibles were estimated using other CMS data sources.
ESTIMATED EFFECTS
Figure 9 presents the estimated statewide medical costs by year and category of service if all
New York dual eligibles were transitioned into a fully integrated capitation model of coverage
beginning in 2011. These estimates assume that the coordinated care program is implemented
beginning in 2011. Skilled nursing facility cost savings will be very low initially (given that
the opportunities for discharge are limited). However, nursing home savings are expected
to accumulate favorably over time, as nursing home diversion opportunities are identified and
successfully implemented. An estimated 25% savings is assumed on the nursing home costs that
are deemed “impactable” with already institutionalized persons and future persons who enroll
in Medicaid after already becoming institutionalized (and “spending down”) not being deemed as
“impactable.” In the initial year of implementation (2011), only 5% of dual eligibles’ nursing home
costs are deemed to be impactable, although this figure gradually increases to 46% as of 2020.
Net costs for adult day care and personal care are assumed to be unaffected by the coordinated
care program. There may be substantial opportunities to reduce existing usage of these
services, but the shift away from nursing homes could increase the usage of these community-
based services.
42 State Medicaid long-term care service use and payment data for 2000 and 2008.
—35—
FIGURE 9. Net Savings from Statewide Implementation

of Coordinated Care Approach (millions)
COST CATEGORY 2011 2011–2020
Nursing Home Transition Waiver $0.0 $0.2
Skilled Nursing Facility $100 $8,215
Adult Day Care $0.0 $0.0
Personal Care $0.0 $0.0
Home Health -$20 -$1,806
Long Term Home Health -$11 -$934
Assisted Living -$2 -$131
Subtotal Medicaid LTC from data file sent to The Lewin Group $68 $5,345
Total, All Other Medicaid (based on MSIS data) $810 $11,734
Medicaid Total $878 $17,078
Medicare Parts A & B $1,176 $17,042
Medicare Part D $582 $8,436
Medicare Total $1,759 $25,478
Duals Total (Medicaid + Medicare) $2,636 $42,556
Managed Care Organization Administration & Profit $2,216 $31,801
Net Savings From Model $420 $10,755
Percentage Savings From Model 1.2% 2.2%
Source: The Lewin Group estimates. Savings estimates assume that 100% of dual eligibles are enrolled in
the coordinated care program. Numbers may not add to totals due to rounding.
Statewide implementation of an ambitious, fully integrated model for dual eligibles is unlikely
since managed care plans may not adequately serve all regions of the State. Figure 10
demonstrates the impact of limiting the program to various proportions of New York’s dual
eligible population. For example, if 30% of New York’s dual eligibles were enrolled in this
model, it is estimated that total savings over the first 10 years of implementation (State and
Federal savings combined) could be approximately $3.2 billion.
FIGURE 10. Net Savings across Medicare and Medicaid Pooled Funds from
Coordinated Care Approach: Less than Statewide Implementation (millions)
PERCENTAGE OF STATEWIDE DUALS PROJECTED TOTAL 10 YEAR TOTAL,
ENROLLED IN COORDINATED CARE PROGRAM SAVINGS, 2011 2011–2020
10% $42 $1,076
20% $84 $2,151
30% $126 $3,227
40% $168 $4,302
50% $210 $5,378
60% $252 $6,453
70% $294 $7,529
80% $336 $8,604
90% $378 $9,680
100% $420 $10,755
Source: The Lewin Group estimates.
—36—
DISCUSSION
The estimated savings are primarily driven by mandatory enrollment of dual eligibles in
coordinated care, which also allows for shared savings between the Federal government and
New York State. The model assumes MCOs would serve as the entity pooling the Medicaid
and Medicare funds, assuming full financial risk, and ensuring coordinated care for enrolled
dual eligibles.
The net savings estimates shown in the previous tables take into consideration the savings
the MCOs will create in acute care and long-term care spending, as well as the offsetting
impacts of paying for the MCOs’ administrative efforts and providing a 2% profit margin. It is
important to note that funding for Medicare and Medicaid is traditionally completely separate.
Without a shared-savings approach, the program administrative costs would largely be borne
by Medicaid, limiting the total $10.8 billion in savings to Medicaid savings of approximately
$1.2 billion over 10 years, while Medicare would realize nearly $9.6 billion in savings. There are
other potential approaches to improved care coordination for dual eligibles, but all approaches
that require state administration would require shared-savings between Federal and state
governments to be cost effective and feasible.
Making MCO enrollment mandatory for Medicare beneficiaries would, of course, also require
the consent of Federal policymakers and a change in Federal statute, and thus presents a
potential barrier to implementation. Another possibility might be an automatic enrollment
with the option for enrollees to opt-out, though this introduces potential for adverse selection
and reduced savings. The need for Federal consent, combined with the need to share savings
with Medicare, mean that none of the potential savings for this option is truly “actionable”
by New York State alone. However, CMS has appeared to be increasingly receptive to coordinated
models, and the PPACA includes the creation of a new office dedicated to coordination for dual
eligibles between the Federal government and states.
New York is well positioned to implement this option using MCOs because it could build
on programs that are already in place. For example, New York Medicaid already contracts
with Special Needs Plans (SNPs) under the Medicare Advantage and Medicaid Advantage
Plus programs, providing support to approximately 6,000 individuals. Another approximately
30,000 dually eligibles receive their care through partially capitated managed long-term
care plans or the Program of All-inclusive Care for the Elderly (PACE). 43 However, while
these programs set the stage for true integration, enrollment is currently voluntary and
only a small share of those eligible have enrolled. To maximize cost savings, existing models
will likely need to allow for mandatory enrollment and true financial integration and
shared savings.
43 New York State Department of Health, Monthly Medicaid Managed Care Enrollment Report (May 2010), http://www.health.state.ny.us/health_care/
managed_care/reports/enrollment/monthly/. Includes 5,639 Medicaid Advantage enrollees; 531 Medicaid Advantage Plus enrollees; 3,405 PACE
enrollees; and 26,797 Managed Long-Term Care enrollees.
—37—
PPACA authorized changes to Medicare Advantage payments, making Special Needs Plans
an uncertain basis for ongoing care integration. Alternatively, the State could implement this
approach without relying on Medicare managed care plans to assume the Medicare risk and
coordinate Medicare benefits. For example, the State Medicaid program could become the
integrated entity, in which case it would receive the Medicare funding, pool it with Medicaid
funds, and assume full financial risk for the entire Medicaid and Medicare benefits package
for their enrolled dual eligibles. New York State would need Federal approval to receive the
Medicare funding and act as the integrated entity.
—38—
Adopting Bundled Payment Methods
PERCENTAGE OF COSTS POTENTIAL SAVINGS (2011–2020) ACTIONABLE SAVINGS (2011–2020)
BASELINE SPENDING
REPRESENTED BY (billions) (billions)
(2011–2020) (billions)
BUNDLED CONDITIONS
$ % $ %
26% $133.00 $6.30 4.7% $1.56 1.2%
100%* $511.50 $24.20 4.7% $6.20 1.2%

Baseline spending includes spending for all services that would be included under the bundled payment program, such as hospital services,
physician services performed during the hospitalization and post-acute care provided within 30 days of hospitalization for the conditions and
procedures listed in Figure 11.
* Estimated savings assumes 100% of services requiring hospitalization are bundled. This was extrapolated based on the fact that the modeled
conditions currently account for approximately 26% of spending for all services that would be bundled. This assumes that the same level of savings
can be achieved for all bundled services.
BACKGROUND
&
urrent payment methods tend to reimburse hospitals on a per discharge basis and
physicians on a per service basis. Although hospitals have a financial incentive to
manage a patient’s care during the hospital stay, there is little or no incentive to manage
a patient’s condition following discharge. Physicians are usually paid for each service they
provide, thus there is little financial incentive for them to manage utilization of ancillary tests
and services for a patient while in the hospital or to manage a patient’s care after discharge
from the hospital to help avoid preventable readmissions.
Bundled payments would consist of a single payment for all services provided to a patient
during an episode of care for hospital and physician care, while the patient is in the hospital and
in the post-acute phase, regardless of the amount of services provided. These bundled payment
rates should be adjusted for the severity of the patient’s illness to account for the additional
resources required to treat sicker patients. The payments can be made to the hospital or some
managing entity that then distributes the funds to all the parties involved.
Bundled payment programs have been implemented in Medicare and private insurance.
An evaluation of the Medicare Participating Heart Bypass Demonstration found that it saved
Medicare an average of 15% for bypass surgery patients in demonstration hospitals (with
inflation adjustment.) 44 The Geisinger Health System in Pennsylvania implemented a bundled
payment program for non-emergency coronary artery bypass graph (CABG) procedures
(ProvenCare) in 2006, which has been shown to save 5%.
Since bundled payments are usually based on an average of what it currently costs to provide the
full range of services for an episode, hospital and physician groups could profit if they can reduce
adverse events and control costs to below the average. Groups that cannot lower costs will lose
money under the system. Thus, the bundled payment method provides incentives for hospitals
and physicians to work together to manage a patient’s care throughout an entire episode of care.
This may require major changes in physician practice patterns and hospital operations as well as
investments in different strategies to reduce adverse events and preventable readmissions.
44 J. Cromwell, D.A. Dayhoff, et al, “Medicare Participating Heart Bypass Demonstration: Final Report,” CMS, (1998)
—39—
$GRSWLQJ%XQGOHG3D\PHQW0HWKRGVFRQWLQXHG
POLICY OPTION
In this analysis, public and private health plans would make prospective payments for entire
episodes of care, potentially encompassing inpatient care, physician services while hospitalized,
and post-acute care services including short-term skilled nursing facility (SNF) and home
health care. By offering bundled payments, this reform would provide an opportunity for
hospitals, physicians, and other health care providers to benefit from reducing complications
and hospital readmissions and allow for more flexibility in allocating resources.
We selected specific types of conditions and procedures to be included in the bundled payment
model, based on groups of closely associated diagnosis-related groups. The basis for the
savings estimates in this analysis is from evaluations of prior programs that bundled payments
for heart bypass surgery and assumes that similar levels of savings (as a percentage of
baseline spending) can be achieved for the other services that we included, such as pregnancy,
mental health, and substance abuse. However, actual achievable savings may be higher or
lower, which will effect the overall cost savings estimates.
FIGURE 11. Services selected for Bundled Payment Analysis

ALCOHOL AND DRUG ABUSE
ALCOHOL & DRUG DEPENDENCE W REHAB OR REHAB/DETOX THERAPY
ALCOHOL ABUSE & DEPENDENCE
COCAINE ABUSE & DEPENDENCE
DRUG & ALCOHOL ABUSE OR DEPENDENCELEFT AGAINST MEDICAL ADVICE
OPIOID ABUSE & DEPENDENCE
OTHER DRUG ABUSE & DEPENDENCE
CARDIOLOGY—MEDICAL
ACUTE MYOCARDIAL INFARCTION
CARDIOLOGY—SURGICAL
CARDIAC DEFIBRILLATOR & HEART ASSIST IMPLANT
CARDIAC VALVE PROCEDURES W CARDIAC CATHETERIZATION
CARDIAC VALVE PROCEDURES W/O CARDIAC CATHETERIZATION
CORONARY BYPASS W CARDIAC CATH OR PERCUTANEOUS CARDIAC PROCEDURE
CORONARY BYPASS W/O CARDIAC CATH OR PERCUTANEOUS CARDIAC PROCEDURE
PERCUTANEOUS CARDIOVASCULAR PROCEDURES W AMI
PERCUTANEOUS CARDIOVASCULAR PROCEDURES W/O AMI
DIGESTIVE SYSTEM—MEDICAL
MAJOR ESOPHAGEAL DISORDERS
MAJOR GASTROINTESTINAL & PERITONEAL INFECTIONS
NON-BACTERIAL GASTROENTERITISNAUSEA & VOMITING
OTHER ESOPHAGEAL DISORDERS
DIGESTIVE SYSTEM—SURGICAL
MAJOR SMALL & LARGE BOWEL PROCEDURES
Figure 11 continued on next page X
—40—
FIGURE 11. Services selected for Bundled Payment Analysis (continued)

FEMALE REPRODUCTIVE SYSTEM
UTERINE & ADNEXA PROCEDURES FOR NON-MALIGNANCY EXCEPT LEIOMYOMA
UTERINE & ADNEXA PROCEDURES FOR NON-OVARIAN & NON-ADNEXAL MALIG
UTERINE & ADNEXA PROCEDURES FOR OVARIAN & ADNEXAL MALIGNANCY
HIP & KNEE REPLACEMENT
HIP JOINT REPLACEMENT
KNEE JOINT REPLACEMENT
INFECTIOUS DISEASES
SEPTICEMIA & DISSEMINATED INFECTIONS
MENTAL DISEASES
MAJOR DEPRESSIVE DISORDERS & OTHER/UNSPECIFIED PSYCHOSES
PREGNANCY
CESAREAN DELIVERY
VAGINAL DELIVERY
VAGINAL DELIVERY W COMPLICATING PROCEDURES EXC STERILIZATION &/OR D&C
VAGINAL DELIVERY W STERILIZATION &/OR D&C
RESPIRATORY SYSTEM
ASTHMA
BRONCHIOLITIS & RSV PNEUMONIA
OTHER PNEUMONIA
We assume hospitals would be paid a single bundled rate that covers hospital inpatient stays
and any readmissions occurring within 30 days of admission. The payment amount would also
include post-acute rehabilitation, skilled nursing, home health services, hospital outpatient
services, and payment for the attending physician and all consults. Under this program, we
assume that bundled payment rates are set at the average current spending for the bundle of
services. This is different from other bundled payment programs that set the rates equal to that
average cost less some allowance for expected reductions in adverse events.
We estimated savings from this program assuming that hospitals and physicians are able to
reduce costs for all these services similar to what was found under the Medicare Participating
Heart Bypass Demonstration, which we estimate was approximately 15% and Geisinger’s
ProvenCare model, which showed savings of 5%. For this analysis, we assumed the more
conservative 5% overall cost savings and assumed that these savings would be passed on to
insurers, and eventually to consumers, in the form of lower payment rate increase for these
services over time. We assume that it would take approximately three years for the full effect of
these savings to materialize. Our estimates were derived using New York State Medicaid claims
data for 2008, a proprietary claims database from Ingenix, Inc. with approximately 1 million
UnitedHealth Group covered lives in New York State in 2008, and New York State inpatient
hospital discharge data for 2007.
—41—
ESTIMATED EFFECTS
For this option, we estimated savings under two scenarios. The first scenario (Figure 12)
reflects the impact if all payers (Medicare, Medicaid, State, and local governments, commercial
health plans, and private ERISA plans) changed their payment methodology.
FIGURE 12.
Estimated Potential Savings by Stakeholder Group 2011–2020 (millions)
(assumes bundled payments are implemented across all payers) *
STATE AND
FEDERAL PRIVATE STATEWIDE
YEAR LOCAL HOUSEHOLDS
GOVERNMENT EMPLOYERS HEALTH SAVINGS
GOVERNMENTS
2011 $91 $29 $32 $47 $199
2012 $215 $69 $77 $112 $473
2013 $253 $81 $91 $132 $557
2014 $270 $86 $96 $140 $592
2015 $287 $92 $102 $149 $630
2016 $306 $98 $109 $159 $673
2017 $326 $104 $117 $170 $717
2018 $348 $111 $124 $181 $764
2019 $370 $118 $132 $193 $813
2020 $394 $126 $141 $205 $866
2011–2020 $2,860 $913 $1,022 $1,489 $6,284
* Estimates assume that episode savings are similar for Medicare and Medicaid patients and similar to privately insured patients within the same
initial hospitalization diagnosis-related group.
Recognizing that New York State can only influence—but not compel—changes in Medicare and
private ERISA plans, the second option (Figure 13) reflects the savings that would accrue from
payment methodology changes in the State’s Medicaid program, as well as the State’s employee
benefit program. This table shows that the Federal government could save $537 million over
the 10-year period due to the 50% Federal matching rate for Medicaid. The State and local
governments could see savings of $913 million due to savings to the Medicaid program and
employee health benefit programs. Households with a family member covered through the State
or local government employee benefit programs could see approximately $109 million in savings.
—42—
FIGURE 13. Estimated Actionable Savings by Stakeholder Group 2011–2020 (millions)

(assumes bundled payments are implemented for Medicaid, and State and
local government health benefit programs)
FEDERAL STATE AND LOCAL STATEWIDE HEALTH
YEAR HOUSEHOLDS
GOVERNMENT GOVERNMENTS SAVINGS
2011 $17 $29 $3 $50
2012 $40 $69 $8 $117
2013 $48 $81 $10 $138
2014 $51 $86 $10 $147
2015 $54 $92 $11 $156
2016 $57 $98 $12 $167
2017 $61 $104 $12 $178
2018 $65 $111 $13 $190
2019 $69 $118 $14 $202
2020 $74 $126 $15 $215
2011–2020 $537 $913 $109 $1,560
DISCUSSION
In this study, we estimated the effect of implementing a bundled payment system for conditions
that together represent approximately 26% of hospital costs in the State. These include
the most commonly provided hospital services and the services that will be included in the
upcoming CMS bundled payment demonstration. In fact, bundled payments could be devised
for all services involving a hospitalization. If bundling reduces spending by 5%, as we assumed
above, total savings could reach $24.2 billion, assuming bundled payments are implemented
across all payers, and $6.2 billion if implemented for Medicaid and public employees over the
2011 to 2020 time period. However, it would take some time to develop payment systems for
each individual hospital service.
The model assumes payment bundles cover 30-day episodes. But, longer episodes may be
more appropriate for some conditions, such as diabetes.
The bundled payment program creates incentives for hospitals and physicians to reduce the
number of tests and other services a patient receives during an episode period. There are
concerns, therefore, about the quality of care that a patient receives when incentives are to
provide fewer services. On the other hand, bundled payments may incentivize quality care to
prevent costly readmissions and adverse events. To ensure that improved quality of care is not
jeopardized, the bundled payment model could be implemented with a P4P program to ensure
that evidence-based standards of care are met.
This analysis assumes that the bundled payment model is implemented statewide and for
every hospital. Implementation of a bundled payment model may be easier in hospitals that
already have collaborative relationships with physicians and other post-acute providers, such
as integrated delivery systems. Implementing bundled payments may be more difficult for other
hospitals that would need to ensure access to a network of physician specialties to perform all
—43—
the services within the bundle. The hospital, or some managing entity, would need to establish
contracts for the billing and distribution of the payments for all the parties involved. Thus, it
may not be possible to implement the program at every hospital at the start, and may take
several years for all hospitals to develop the required infrastructure.
The New York State Medicaid program has been moving in the direction of bundled payments
and, in December 2008, implemented hospital outpatient prospective payment for outpatient
surgeries and visits. This payment methodology—also known as ambulatory patient
groups—is a type of bundled payment system for hospital services provided during a single
outpatient visit, which is similar to diagnosis-related groups for inpatient care. However,
these payment systems do not incentivize hospitals and physicians to work together to
create efficiencies and manage a patient’s care throughout a complete episode of illness.
Bundling payments for inpatient episodes of care will further promote efficiencies and
management of patients’ care.
Implementing this type of payment methodology for employee benefit systems will require
significant changes in the structure of the State’s employee benefit program. Currently,
New York State and New York City contract separately for managed care organization,
physician, pharmacy, behavioral health, and hospital services for State and City employees.
An effective inpatient episode of care payment methodology will require this contracting
approach to be modified.
The PPACA authorizes bundled payment demonstrations for Medicare and Medicaid. These
demonstrations provide New York State providers and insurers with additional opportunities
to test the effectiveness of bundling payments to better integrate care, improve quality, and
reduce health care expenditures.
—44—
Imposing a Tax on Sugar-Sweetened Beverages
$ % $ %
$136.30 $5.63 4.1% $5.63 4.1%
Baseline spending includes only health spending for treating chronic diseases related to obesity, such as Type 2 diabetes, cardiovascular disease,
certain types of cancer, musculoskeletal disorders, etc.
BACKGROUND
2
ne out of every three Americans is obese, with one out of the remaining two being
overweight. 45 The percentage of adults in New York State who are overweight or
obese increased from 42% in 1997 to 60% in 2008, and obesity among children and
adolescents has tripled in the past three decades. Low-income populations, those with low
educational attainment, and communities of color experience higher rates of obesity.
Obesity causes serious health problems such as Type 2 diabetes, heart disease, high blood
pressure, high cholesterol, cancer, and osteoarthritis. The Surgeon General estimated that
obesity is associated with 112,000 deaths each year, and poor diet and physical inactivity cause
up to an additional 365,000 deaths per year. 46
Obesity also imposes a serious and substantial economic burden. According to a 2009 study,
the nation could be spending up to $147 billion (in 2008 dollars) per year in treating health
conditions caused by obesity. 47 This is more than 10% of all medical spending. We estimate
that New York State will spend approximately $9.9 billion on adult obesity-related health
problems in 2011 and $136.3 billion over the 10-year period from 2011 to 2020. This accounts
for approximately 5.6% of statewide health spending with approximately 80% of this cost paid
through publicly funded health care programs, such as Medicare and Medicaid.
Consumption of sugar-sweetened beverages (SSB) has drastically increased since the

early 1970s and is likely contributing to the growing obesity epidemic in the U.S. 48 The link
between obesity and SSB has been well documented in scientific literature but, despite
growing concern, consumption of SSB remains high. The potential benefits from reduced
consumption of SSB include decreased incidence of chronic diseases and hence decreased
health care costs. 49 The Centers for Disease Control and Prevention (CDC) recommends
reducing the amount of SSBs as one way for people to reduce their intake of added sugars
and help manage their weight.
45 Flegal KM, Carroll MD, Ogden CL, Curtin LR. Prevalence and trends in obesity among US adults, 1999-2008. JAMA 2010 January 20;303(3):235-41.
46 US Department of Health and Human Services. Surgeon General’s call to action to prevent and decrease overweight and obesity. US Department of
Health and Human Services 2001;Available at: URL: http://www.surgeongeneral.gov/topics/obesity/calltoaction/CalltoAction.pdf.
47 Finkelstein EA, Trogdon JG, Cohen JW, Dietz W. Annual medical spending attributable to obesity: payer-and service-specific estimates. Health Aff
(Millwood ) 2009 September;28(5):w822-w831.
48 Popkin BM. Patterns of beverage use across the lifecycle. Physiol Behav 2010 January 4.
49 Jacobson MF. Liquid Candy: How Soft Drinks are Harming Americans’ Health. Washington, D.C.: Center for Science in the Public Interest; 2005.
—45—
,PSRVLQJD7D[RQ6XJDU6ZHHWHQHG%HYHUDJHVFRQWLQXHG
POLICY OPTION
For this policy option, we assume that a 0.9 cents per ounce excise tax is imposed on sugar-
sweetened soft drinks (this equates to approximately an 18% sales tax) in New York State in
addition to the current State and county sales taxes (these sales taxes total 8.875% in New
York City and average approximately 8% across the rest of the State). The objective of the tax
is to increase the price of sugar-sweetened soft drinks relative to alternative beverages, which
creates incentives for people to purchase healthier alternatives. This proposal is similar to
an item in Governor Paterson’s proposed 2010–2011 budget, which included an excise tax on
non-diet sodas and fruit drinks that contained less than 70% natural fruit juice. Under an excise
tax, manufacturers could spread the increased cost of the tax across all their products, which
would eliminate the intended effects of creating a higher price for sugar-sweetened soft drinks
relative to healthier alternatives. Therefore, under this policy option, manufacturers would be
required to apply the increased cost of the excise tax across only sugar-sweetened soft drinks.
This model examines the potential impact of a policy option designed to reduce obesity and
related health costs. A series of modeling assumptions are required to estimate health care
cost reductions, associated with policy options designed to reduce intake of soft drinks or trans
fats. This study reviewed existing scientific evidence on price elasticity of demand for SSB,
SSB consumption patterns among U.S. adults, the relationship between energy balance and
body weight, associations between body weight and health outcomes, and associations between
health conditions and health care costs.
Our analysis suggests that the excise tax will translate into a 16.7% price increase in SSB.
Based on a conservative price elasticity of - 0.8, consumption of SSB on average is estimated
to drop by 13.4%. (The elasticity is a measure of consumer sensitivity to price. An elasticity
of - 0.8 means that a 1% increase in price is associated with a reduction in use of 0.8%). Under
the assumption that this energy intake reduction is not compensated for through other food
sources, it will lead to an energy deficit of 23 calories on a daily basis, that is, 13.4% of the total
estimated average daily energy intake from sugary drinks of 175 calories.
We based our estimate on available epidemiological literature, as well as additional simulation

analyses, to understand how the dynamic relationship between disease incidence (new cases) and
prevalence rates (accumulation of existing cases) determines the full potential of body weight loss
that could be realized at a population level. We found that in the first year immediately after the
policy change, only approximately 8% and 3% of the potential health benefit can be realized for pre-
Medicare and Medicare populations, respectively.50 This is because the population is starting to
depart from the current health state and moving toward a healthier equilibrium. At the tenth year,
up to 62% and 30% of the potential benefits would have been realized for each population segment,
respectively. By the 20th year, approximately 80% and 60% of full potential can be achieved.
There is no distinction between potential and actionable savings. New York State can fully
implement this policy action.
50 Analysis of the data indicates that reduction in obesity has a greater affect on reducing prevalence of diseases in the younger population under age
65. Disease prevalence among the aged population is due to factors other than just obesity. Thus, reducing obesity alone in the aged population
would have less of an impact on disease prevalence because these other factors would still exist.
—46—
ESTIMATED EFFECTS
We estimated the change in health spending by stakeholder group in New York due to the
reduced medical costs associated with the SSB tax. Figure 14 shows that the excise tax on
sugary drinks will result in a $5.6 billion reduction in New York State health spending over
the 2011 to 2020 period. We assume that the excise tax is fully implemented in 2011. Minimal
savings of approximately $79 million could be realized during the first year, but the annual
savings could grow to approximately $1.2 billion by 2020.
FIGURE 14. Estimated Health Care Savings by Stakeholder in New York State Due to
0.9 Cents per Ounce Sugar Sweetened Beverage Excise Tax, 2011–2020 (millions)*
FEDERAL STATE AND LOCAL PRIVATE TOTAL STATEWIDE
YEAR HOUSEHOLDS
GOVERNMENT GOVERNMENTS EMPLOYERS HEALTH SAVINGS
2011 $39 $24 $6 $10 $79
2012 $79 $49 $12 $20 $159
2013 $124 $77 $19 $30 $249
2014 $175 $108 $26 $42 $352
2015 $227 $142 $34 $55 $458
2016 $287 $179 $42 $69 $577
2017 $356 $221 $51 $84 $712
2018 $428 $264 $60 $100 $852
2019 $509 $313 $69 $117 $1,008
2020 $600 $366 $80 $136 $1,183
2011–2020 $2,823 $1,743 $401 $663 $5,629
* Tax revenue collected from the sugary drink tax, as well as the cost to households to pay the tax, are excluded from this table.
DISCUSSION
Conservative assumptions were used throughout the analysis; hence the reported potential benefits
of improved diet are likely understated. To investigate the impact of modeling assumptions, we also
conducted a first order sensitivity analysis on the major parameters employed. For example, if price
elasticity of -1.0 was used in the model instead of -0.8, so that a 1% increase in price resulted in a
1% decrease in consumption, rather than 0.8, the total economic benefit would be 25% greater.
Similarly, the amount of the excise tax would affect the statewide health care savings estimates.
Although the relationship between the size of the tax and the reduction in statewide health care
costs are not perfectly linear, a rule of thumb would be that statewide health spending could
be reduced by $312 million (over 10 years) for each percent of tax, which is calculated as total
statewide health savings ($5.629 billion) divided by 18.
Similar to the Governor’s proposal, we designed this option using an excise tax on manufacturers
instead of a sales tax. The objective of the tax is to increase the price of sugar-sweetened soft
drinks relative to alternative beverages, which creates incentives for people to purchase healthier
alternatives. The use of a sales tax would enable the state to target the exact products to tax.
However, people purchasing sugar-sweetened drinks with food stamps are exempt from the
—47—
sales tax. Also, a sales tax is not shown on the posted price of the product, so the difference in
prices between sugar-sweetened soft drinks and alternatives are not apparent at the time the
consumers are making their decisions. These factors lessen the intended effects of the tax.
An excise tax on manufacturers of sugar-sweetened beverages is assumed to be an increase

in their production costs which would be passed on to consumers as increased prices for
their products. The increased prices would affect the posted price of the product and people
using food stamps would not be exempt from the excise tax. However, manufacturers could
spread the increased cost of the tax across all their products, which would eliminate the
intended effects of creating higher prices for sugar-sweetened soft drinks relative to healthier
alternatives. Therefore, under this policy option, manufacturers would be required to apply the
increased cost of the excise tax across only sugar-sweetened soft drinks.
We assume overweight and obese individuals respond to potential SSB taxes similarly.
However, the tax is regressive, and it is very likely to disproportionately affect those who are
low-income and obese as suggested by Brownell et al. In addition, earmarking the SSB tax
revenue for investments in obesity prevention could directly benefit these populations. Both
features make a tax potentially even more effective in fighting obesity among such population
segments that bear the most health disparity.
In this study, we assume physical activity and diet are constant while SSB consumption is not
constant. We make these assumptions to tease out the net effect of introducing SSB taxes.
However, in the real battle against the obesity pandemic, there is no silver bullet. Research
shows that 80% of obesity is caused from excess calorie intake and 20% is the result of physical
inactivity. Interventions that target physical activity and diet have to be combined to maximize
the sustained weight loss. This signifies the importance of directing the revenue generated
from an SSB tax to other types of weight loss programs to achieve the best possible results.
Reduced excess body weight and corresponding disease risk will prolong life. Prolonged life
will lead to additional health care spending. It remains unclear to what extent prevention can
generate cost savings while curbing the health impact from chronic conditions.51, 52
Children are not included in this study because their health risks of chronic conditions remain
low, even while the rates of chronic diseases (e.g., Type 2 diabetes and hypertension)—much
of which can be attributed to obesity and overweight—have been growing among adolescents
at a significant rate. Moreover, the adolescent population has been the focus of research and
intervention and prevention studies because other studies have shown that unhealthy body
weight in childhood will adversely affect health and significantly increase the risk for adulthood
obesity.53, 54, 55 The reduction of unnecessary energy intake from SSB among youth would have a
long-term effect on public health that is not captured in this analysis.
51 Russell LB. Preventing chronic disease: an important investment, but don’t count on cost savings. Health Aff (Millwood) 2009 January;28(1):42-5.
52 Goetzel RZ. Do prevention or treatment services save money? The wrong debate. Health Aff (Millwood ) 2009 January;28(1):37-41.
53 James J, Kerr D. Prevention of childhood obesity by reducing soft drinks. Int J Obes (Lond) 2005 September;29 Suppl 2:S54-S57.
54 Harnack L, Stang J, Story M. Soft drink consumption among US children and adolescents: nutritional consequences. J Am Diet Assoc 1999
April;99(4):436-41.
55 Grimm GC, Harnack L, Story M. Factors associated with soft drink consumption in school-aged children. J Am Diet Assoc 2004 August;104(8):1244-9.
—48—
This study only projects health care savings within 10 years, hence excluding children leads to
a small downward bias in total savings; however, if we were to model lifetime savings, then it
would be imperative to include children since they will ultimately define the health profile for
the population when they enter their prime age in 30, 40, and 50 years from now. The health
promotion initiatives they receive during their childhood will perceivably benefit their entire
lifespan and cumulate into sizable economic benefits.
One policy advantage of this option is the fact that it requires relatively little implementation
effort aside from garnering sufficient support among lawmakers. Estimated savings would
presumably occur as a consequence of the increased cost of SSBs.
—49—
Expanding Hospital Pay for Performance
$ % $ %
$162.03 $3.85 2.4% $1.31 0.8%
Baseline spending includes total inpatient spending for services assumed to be included under pay-for-performance programs, beginning with acute
myocardial infarction, community-acquired pneumonia, isolated coronary artery bypass graft, heart failure, and hip and knee replacement surgery in
2011, and increasing to include services accounting for 42% of inpatient spending by 2020.
BACKGROUND
8
nder the current fee-for-service payment systems, providers are often paid the same
amount for the same service regardless of the quality of the care provided. In addition,
there is little economic incentive under the current system for providers to improve
their quality processes. Pay for performance uses financial incentives to promote quality
improvements using a variety of clinical processes of care and health outcomes quality measures.
This proposal would provide hospitals with bonus payments for improvements in the quality of
care for patients receiving selected services based upon quality measures of health outcomes
for these services. Under such a system, data would be compiled by providers for their patients
that receive selected health services. The results would be adjusted for the severity of illness
and compared with benchmark measures of outcomes for these services. Hospitals with
favorable results would be rewarded with higher payments.
The Patient Protection and Affordable Care Act includes a provision under Medicare to
implement a P4P program, beginning in FY2013. This program is to be implemented in a budget
neutral manner nationwide, and will be redistributive so that penalties for poor performance
will effectively fund bonuses to the high performers. Under this program CMS will withhold 1%
of payments in FY 2013, growing to 2% by 2017.
POLICY OPTION
Studies of P4P programs have concluded that the incentives worked to improve quality. 56
However, the results are mixed on the effect that these changes in quality have on costs. It is
widely believed that improvements in quality will actually reduce costs by eliminating avoidable
complications, reducing readmissions and reducing hospital length of stay. The notion that
improvements in quality ultimately reduce costs is consistent with the total quality measurement
(TQM) model used in other industries, which stresses reduced production costs through
improved quality. However, there is limited evidence that this actually occurs in health care.
Premier, Inc., which participated in the CMS Hospital Quality Incentive Demonstration (HQID),
has estimated the savings realized as a result of quality improvements made in response to
the HQID bonuses. Premier estimated savings per patient from 2003 to 2006 that ranged from
56 Lindenauer, P., “Public Reporting and Pay for Performance in Hospital Quality Improvement” New England Journal of Medicine”, 356; 5,
February 1, 2007.
—50—
([SDQGLQJ+RVSLWDO3D\IRU3HUIRUPDQFHFRQWLQXHG
$463 for knee replacement surgery to $1,599 for heart attack patients. It bases these savings
estimates on the reductions in hospital readmissions and length-of-stay for the procedures
studied.57 Based on these data, we estimated that every 1% improvement in quality translated to
a 0.3% reduction in costs.
In this option, we assume that HQID is implemented for hospitals in New York. We assume that
the P4P program initially includes the five medical procedures covered under HQID (accounting
for approximately 14% of hospital costs, including acute myocardial infarction, community-
acquired pneumonia, isolated coronary artery bypass graft, heart failure, and hip and knee
replacement surgery) and is expanded to additional services over time so that by 2020, the
services covered under the P4P program account for approximately 42% of total hospital costs,
which would be three times the initial amount.58
The program provides financial incentives for hospitals to demonstrate higher quality
in the conditions covered by the program. Bonuses will be structured the same as those
in the HQID demonstration program and will depend upon attaining a minimum level of
performance. The incentive system would reward hospitals in the 90th percentile with
a 2% bonus on their payment for the clinical area. Hospitals that achieve the 80th percentile
would receive a 1% bonus. The program is not designed to penalize hospitals that fall into
the bottom of the distribution.
Several key assumptions were made to estimate savings under this option. The first was
determining how much quality could be improved under a P4P program. We based this
assumption on data from CMS Hospital Compare, which showed that quality scores for
New York hospitals improved dramatically even though no incentive payments were tied to
performance, but only that the information would be publically reported. Second, a 2007 study
by Lindenauer found that annual quality improvement due solely as a consequence of the
financial rewards increased quality by 2.3% to 3%.59
The next assumption was to determine the extent to which these improvements in quality
impact costs. The studies we reviewed showed conflicting results, but several did indicate that
quality improvements can be associated with lower costs. The Premier data, described above,
provided an estimate of potential cost savings.
For this analysis, we assume that savings generated by reduced hospital readmissions are
realized by payers and public programs through reduced utilization of these hospital services.
The remaining net savings for hospitals are assumed to be passed on to payers and public
programs through lower negotiated rates and lower diagnosis-related group payments over
time. Hospitals are assumed to keep the bonus payments as profits. Finally, commercial
payers are assumed to pass their share of savings on to employers and consumers in the
form of lower premiums.
57 “Performance Pays: Reliable Care Costs Less and Saves Lives,” Press Conference Webcast, Orlando FL, June 2006.
58 This would include only approximately 6% of all hospital conditions or procedures (28 diagnostic related groups).
This would require implementing approximately three new conditions per year in the P4P program.
59 Lindenauer, P., “Public Reporting and Pay for Performance in Hospital Quality Improvement” New England Journal of Medicine”,
356; 5, February 1, 2007.
—51—
ESTIMATED EFFECTS
Assuming the program is implemented throughout New York for all payers, including
Medicare, Medicaid, commercial insurers, and other government programs (i.e., TRICARE
and Workers’ Compensation), we estimate that potential total gross savings under the
P4P program could be $4.8 billion over the 2011 to 2020 period (Figure 15). These savings
would be offset by the costs to administer the program and the incentive payments to high
achieving hospitals. We estimate that the net reduction in health spending in New York
could potentially be $3.9 billion over this period.
FIGURE 15. Estimated Net Potential and Actionable Savings for New York Hospitals
Under the P4P Model 2011–2020 (millions)
TOTAL GROSS SAVINGS ADMINISTRATIVE COSTS INCENTIVE PAYMENTS NET SAVINGS
YEAR
POTENTIAL ACTIONABLE POTENTIAL ACTIONABLE POTENTIAL ACTIONABLE POTENTIAL ACTIONABLE
2011 $44 $10 $0 $0 $17 $4 $27 $6

2012 $74 $19 $8 $2 $22 $6 $45 $11
2013 $135 $38 $16 $5 $27 $8 $91 $26
2014 $221 $67 $25 $8 $33 $10 $163 $49
2015 $334 $106 $35 $11 $40 $13 $260 $82
2016 $462 $152 $45 $15 $48 $16 $369 $121
2017 $607 $205 $56 $19 $56 $19 $494 $167
2018 $769 $265 $68 $23 $65 $23 $636 $219
2019 $952 $333 $81 $28 $76 $27 $795 $278
2020 $1,156 $410 $95 $34 $88 $31 $974 $346
2011–2020 $4,753 $1,605 $427 $144 $472 $155 $3,854 $1,307
Under the actionable savings scenario, we assume that the State would require the use
of this P4P program in all State and municipal employee health benefits programs and the
Medicaid program. Hospitals would be required to accept the program as a condition of
serving people covered under State programs. In the wake of the New York Attorney General
lawsuit regarding “preferred providers,” legally tenable methods for steering volume must
be considered. To the extent possible, value-based benefit designs would be considered as
a means to encourage consumers to use medical services more efficiently.
Under this scenario, we estimate the P4P model could produce actionable gross savings
of $1.6 billion and net savings of approximately $1.3 billion over the 2011 to 2020 period
assuming that savings are realized for only Medicaid and State and municipal employee health
programs. Figure 16 shows actionable savings to the Federal government of $523 million
under the program as a result of the 50% Federal matching rate for Medicaid. The State and
local governments could see actionable savings of $729 million, and households could see
approximately $55 million in savings.
—52—
FIGURE 16. Net Potential and Actionable Savings Under the P4P Model by
Stakeholder Group 2011–2020 (assumes program implemented for Medicaid and State
and Local Government employees only) (millions)
FEDERAL STATE AND LOCAL PRIVATE STATEWIDE HEALTH
HOUSEHOLDS
GOVERNMENT GOVERNMENT EMPLOYERS SAVINGS
YEAR
POTENTIAL ACTIONABLE POTENTIAL ACTIONABLE POTENTIAL ACTIONABLE POTENTIAL ACTIONABLE POTENTIAL ACTIONABLE
SAVINGS SAVINGS SAVINGS SAVINGS SAVINGS SAVINGS SAVINGS SAVINGS SAVINGS SAVINGS
2011 $15 $2 $3 $3 $3 $0.0 $5 $0.4 $27 $6

2012 $24 $4 $7 $7 $5 $0.0 $9 $0.6 $45 $11
2013 $49 $10 $15 $15 $10 $0.0 $17 $1.3 $91 $26
2014 $86 $19 $28 $28 $19 $0.0 $30 $2 $163 $49
2015 $135 $33 $46 $46 $30 $0.0 $48 $4 $260 $82
2016 $190 $48 $68 $68 $44 $0.0 $67 $5 $369 $121
2017 $253 $67 $93 $93 $59 $0.0 $90 $7 $494 $167
2018 $324 $88 $122 $122 $76 $0.0 $114 $9 $636 $219
2019 $403 $112 $155 $155 $95 $0.0 $142 $11 $795 $278
2020 $491 $140 $192 $192 $117 $0.0 $173 $14 $974 $346
2011–2020 $1,970 $523 $729 $729 $457 $0.0 $697 $55 $3,854 $1,307
DISCUSSION
One of the key assumptions for this policy option was the extent to which these improvements
in quality impact costs. The studies we reviewed showed conflicting results. For example, a
study of the tradeoffs between hospital costs and quality showed that each 1% increase in the
quality indicators studied was associated with a 1.34% increase in costs.60 A more recent study
found that low-cost hospitals had modestly lower quality scores on process of care measures
for patients with myocardial infarction and congestive heat failure.61
However, there are a number of studies demonstrating that lower-cost hospitals often rate
high on quality. A study by Fisher showed that “better quality hospital care does not cost
more.” 62 Other studies have examined the relationship between hospital quality and cost, and
found that even after risk adjustment for severity of illness, many hospitals exhibit high-quality
outcomes, yet are low-cost providers. 63 Another study of the relationship between quality and
costs for radical prostatectomy cases showed that quality was driven by factors under the
control of physicians, such as operating room time, units of blood transfused, anesthesia time,
and surgical time.64 For this reason, we feel it is reasonable to assume that hospital quality
improvements can be associated with lower costs.
60 Morey, R.C., et al, “The Trade-off Between Hospital Cost and Quality of Care,” Medical Care, Vol. 30, No 8, August 1992.
61 Ashish K., “Measuring Efficiency: The Association of Hospital Costs and Quality of Care,” Health Affairs, Vol. 28, No 3, 2009.
62 Presentation by Fisher R.L., Academy Health services research and health policy meeting, 2001.
63 Jiang, H. J., et al, “Factors Associated with High-Quality/Low-Cost Hospital Performance,” Journal of Health Care Finance, 2006; 32(3), 2006.
64 Benoit R.M., et al, “The Relationship Between Quality and Costs: Factors that Affect the Hospital Costs of Radical Prostatectomy,” Prostate
Cancer and Prostatic Diseases (2001).
—53—
A second key assumption used for this analysis is that savings generated by hospitals under
this policy option are assumed to be passed on to payers and public programs through lower
negotiated rates and lower diagnosis-related group payments over time. Although hospitals
may reduce their costs, these savings cannot be counted as reduced health spending unless
payers extract these savings from hospitals through lower payment levels and payers pass
these reduced costs on to consumers. However, hospitals are assumed to keep the P4P bonus
payments as profits.
The HQID demonstration included approximately 230 hospitals throughout the nation,
including 17 hospitals in New York State. However, management commitment, available
infrastructure, and physician practice patterns for these participating hospitals may be
different from many of the other hospitals in New York State, which may ultimately impact
results in both quality improvements and cost reductions. Improvement in quality across
hospitals has been attributed to implementing quality initiatives and providing awareness of
guidelines among staff, improved hospital processes to create better support for meeting
guidelines, improved staff documentation of procedures, and developing staff practices that
are more consistent with guidelines. 65
65 Laschober, M. et al.,” Hospital Response to Public Reporting of Quality Measures”, Health Care Financing Review, Spring 2007.
—54—
Realizing Administrative Simplification through
Health Information Technology
SPENDING (2011–2020) (billions) (2011–2020) (billions)
SUBCATEGORY
(2011–2020)
(billions) $ % $ %
Standardized Quality Requests $1.57 $0.65 41.5% $0.65 41.5%
Standardized Credentialing/
$6.18 $0.92 14.9% $0.92 14.9%
Verification
Baseline spending includes all billing and insurance-related costs.
BACKGROUND
:
e estimate that total billing and insurance-related (BIR) costs in New York
will be approximately $18.2 billion in 2011 (Figure 17). This total includes the cost
of administering all transactions between insurers and providers, including
billing, payment, confirmation of patient eligibility, utilization review functions, and
certification. BIR accounts for approximately 12.5% of all health spending for New York
residents covered under these programs (excluding long-term care). For commercially
insured people, BIR represents approximately 22% of total premiums. 66
Reducing administrative BIR costs that do not contribute value to patient care has been seen
as a way to significantly reduce the costs of health care to patients with little or no impact
on the quality of care delivered.
FIGURE 17. Estimated Billing and Insurance-Related Costs for New York in 2011 (billions)
Hospital BIR ($5.0)

Commercial Insurers ($6.0)
27.5% 33.0 %
Physician BIR ($4.2)

23.1 % Physician Coverage ($3.0)
16.4 %
Note: See Technical Appendix for a detailed

Total Billing and Insurance-Related accounting of these costs.
Costs in New York: $18.2 billion. Source: The Lewin Group estimates.
66 James G. Kahn, Richard Kronick, Mary Kreger and David N. Gans, The Cost Of Health Insurance Administration In California: Estimates For
Insurers, Physicians, And Hospitals, Health Affairs, 24, no. 6 (2005): 1629-1639.
—55—
5HDOL]LQJ$GPLQLVWUDWLYH6LPSOLILFDWLRQWKURXJK+HDOWK
,QIRUPDWLRQ7HFKQRORJ\FRQWLQXHG
POLICY OPTION
The main focus of this set of policy options is to reduce the administrative burden on health care
providers and insurers by reducing BIR costs through standardization and improved use of health
information technology, including:
• Streamline and Standardize the Growing Demand for the Collection and Reporting of
Clinical Information for Quality Measures (Standardized Quality Requests). Under this
option, health plans, private accrediting organizations and government agencies that
require reporting of quality data will have standardized reporting requirements.
• Consolidate and Streamline the Process for Establishing the Professional Competency
and Scope of Practice Credentials of Health Care Providers (Standardized Credentialing/
Verification). Using a single standardized process for accreditation and licensing statewide
could reduce costs for physicians, hospitals, and payers without compromising quality.
These include credentialing requirements for the Centers for Medicare and Medicaid
Services, the Joint Commission on Accreditation of Healthcare Organizations (JCAHO), and
other credentialing entities.
We originally planned to include an option that would promote standardized electronic

systems to facilitate transactions between providers and health plans for eligibility billing
and authorization for health services. However, the Patient Protection and Affordable
Care Act recently signed into law by President Obama includes a series of administrative
simplification requirements designed to automate these transactions through health
information technology. Because Federal law will advance these standards, we did not
include this as a third option.
Using published studies, we estimate that hospitals in New York will spend approximately
$1.6 billion over 10 years on quality data reporting and analysis, of which 62% will be
attributed to Hospital Quality Alliance (HQA)-approved measures for CMS and JCAHO. 67
Under the Standardized Quality Requests scenario, these data would be reported in a
standardized format and other quality reporting efforts would be required to accept these
data. Implementation of this option could reduce the cost of reporting for non-CMS/JCAHO
quality programs by approximately two-thirds, resulting in savings of $652 million over 10
years, with a phase-in period of four years.
Based upon administrative data we estimate that total spending for accreditation and
certification in New York will be approximately $500 million in 2011 under current
law, including provider and insurer costs. We assumed savings of 17% based upon the
performance of an automated credentialing system developed by the Council for Affordable
Quality Healthcare (CAQH) in areas where it is currently used. We assume that savings would
phase in over a period of four years.
67 This reflects the current degree of standardization for quality measure reporting for the New York State Coalition of Prepaid Health Services
Plans (PHSP).
—56—
ESTIMATED EFFECTS
We estimate that the administrative standardization options discussed above would reduce New
York State health spending by nearly $1.6 billion over the 2011 through 2020 period. Of these,
$1.1 billion would be attributed to providers and approximately $427 million would go to health
plans (Figure 18).68
FIGURE 18. Combined Savings of Administrative Simplification Options

in New York State: 2011–2020 (millions)
COMBINED IMPACT OF ADMINISTRATIVE SIMPLIFICATION OPTIONS SAVINGS BY SECTOR
YEAR STANDARDIZED
STANDARDIZED TOTAL PROVIDER PAYER
CREDENTIALING /
QUALITY REQUESTS SAVINGS SAVINGS SAVINGS
VERIFICATION
2011 $33 $21 $55 $45 $10
2012 $38 $44 $82 $62 $20
2013 $44 $69 $113 $81 $32
2014 $50 $96 $146 $102 $45
2015 $57 $101 $158 $111 $47
2016 $65 $106 $171 $122 $49
2017 $74 $112 $186 $134 $52
2018 $84 $117 $202 $147 $55
2019 $96 $123 $219 $162 $57
2020 $109 $130 $239 $179 $60
2011–20 $652 $919 $1,571 $1,144 $427
In Figure 19, we present estimated savings by payer group. Savings are allocated to payer
groups in proportion to the distribution of affected administrative costs by payer. Savings for
people with employer coverage are allocated to households and employers in proportion to the
portion of the premium paid by these two groups. These estimates reflect that most of these
savings would be for people with private health insurance.
FIGURE 19. Changes in New York Health Spending due to the Administrative Standardization
Options by Payer Group: 2011–2020 (millions)
FEDERAL STATE AND LOCAL PRIVATE HOUSEHOLD NET CHANGE IN
YEAR GOVERNMENT GOVERNMENT EMPLOYER OUT-OF-POCKET NEW YORK HEALTH
SAVINGS SAVINGS SAVINGS AND PREMIUMS SPENDING
2011 $6 $9 $26 $14 $55
2012 $9 $14 $39 $21 $82
2013 $12 $19 $54 $28 $113
2014 $16 $24 $70 $37 $146
2015 $17 $26 $75 $40 $158
2016 $18 $28 $82 $43 $171
2017 $20 $31 $89 $46 $186
2018 $22 $33 $96 $50 $202
2019 $24 $36 $105 $55 $219
2020 $26 $39 $114 $60 $239
2011–20 $170 $260 $749 $393 $1,571
68 The distribution of savings between payer and provider is based upon the sum of administrative costs for these groups. All savings for
standardized reporting of quality data in are attributed to providers.
—57—
DISCUSSION
Due to multiple and inconsistent regulations set by JCAHO, Medicare, and Medicaid, there are
different standards for administrative transactions. These standards need to be consistent to
reduce the number of transactions and associated costs. The Patient Protection and Affordable
Care Act addresses the need for an automated system that will simplify the process of
communicating transactions between providers and health plans, and requires the Secretary
of HHS to establish a consensus-based process to develop additional standards and processes
that can reduce the cost of these transactions. The law defines the following objectives for the
processes to be developed:
• enable determination of an individual’s eligibility and financial responsibility for specific

services at the point of care;
• include a comprehensive system requiring minimal paper and other communications;

• provide for management of a timely process for claims processing and denials including
adjudication and appeals;
• provide a standard formatting of data with sufficient detail to administer service eligibility and
prohibits additional conditions; and
• require timely updates of information to ensure that the database includes the most recent
information.
The Lewin Group estimates that the administrative simplification provisions of the PPACA
will result in BIR savings in New York of approximately $10.1 billion over the 2011 through
2020 period. These include savings to physicians of $5.6 billion and savings of $2.4 billion
for hospitals. Savings to insurers would be $2.2 billion. Of course, for savings to be realized,
providers and payers both must participate in the use of a simplified system, which may require
mandates and/or penalties to apply to both.
It is important to recognize that provider administrative savings are not savings to

consumers unless they are passed back to consumers in the form of lower charges for
health services. For example, savings to physicians from a standardized certification
process are a windfall savings to the provider unless these savings are somehow negotiated
back by payers resulting in lower premiums. In the previous table we assume that savings
are passed back to payers.
—58—
For the most part, these administrative simplification steps could be implemented
throughout most of the New York health care system by the State legislature. Private plans
could be required to participate, including health plans serving Medicaid recipients. The
legislature could also require use of these systems in the State’s fee-for-service Medicaid
program. Private plans would have input to the design of the process, but would be required
to use the data provided through the system. While the State cannot require self-funded
plans to use the system, these plans are likely to voluntarily accept these processes if they
can reduce costs while maintaining quality objectives.
The major barrier to implementation will be for the Medicare program. Like all insurers, the
Medicare program requires accreditation/certification of providers according to its own rules.
The Federal government also sets its own standards for quality data reporting and cannot
be required to participate in the automated verification processes. It is possible to design the
standardized process so that it is, to the extent practical, consistent with Federal standards.
Therefore, we assume that all of the savings estimated above, including those related to
Medicare, are achievable with State action.
—59—
Rebalancing Long-Term Care
$ % $ %
$147.28 $1.02 0.7% $1.02 0.7%
Baseline spending includes all Medicaid long-term care service costs, including both nursing facility and community-based long-term care.
BACKGROUND
1
ew York State’s Medicaid program plays a fundamental role in providing long-term
care in both residential and community-based settings for a large population of
beneficiaries with extensive functional and cognitive impairments, and behaviorally
and medically complicated needs. An estimated 247,000 Medicaid beneficiaries receive
long-term care services each month through 12 distinct programs. In 2007, Medicaid
spending for this population’s long-term care was roughly $12.3 billion, approximately one-
quarter of all Medicaid spending in the State. 69 In 2007, 53% of long-term care spending
was for nursing homes and the remaining 47% for community-based services (such as home
health, adult day, and TBI waiver-related services). It is difficult to estimate the proportion
of all long-term care spending for which Medicaid accounts, as there are no state-level
estimates of all sources of long-term care spending. But, a large share of spending is borne
by the Medicaid program.
Currently, no one organizational unit within New York has responsibility for long-term care
budgets, policies, or programs. The lack of a single responsible entity contributes to difficulties
in viewing long-term care as a system of services and supports through which consumers and
their families will move as their needs and challenges change with time. It also affects the
State’s ability to create easy and seamless consumer access to the full range of long-term care
services, and creates issues with overall program monitoring and cost management.
Going forward, New York’s Medicaid long-term care program faces considerable challenges:
heightened fiscal and demographic pressures, scarcity of affordable housing, and other
external factors. Therefore, New York needs a clear vision for the Medicaid long-term care
system of the future, and needs to design a systematic implementation action plan to serve the
fast growing population in need of long-term care services.
POLICY OPTION
To address these issues, we assume that New York can undertake the following steps, which
are recommended as a package rather than individual options:
• Enhance the Aging and Disability Resource Centers (ADRCs) throughout the State to add more
proactive intervention in critical pathways to institutions consistently across the State and
create a single point of entry into the State’s long-term care system.
69 Alene Hokenstad, Meghan Shineman, and Roger Auerbach, “An Overview of Medicaid Long-Term Care Programs in New York”,
http://www.uhfnyc.org/publications/880507.
—60—
5HEDODQFLQJ/RQJ7HUP&DUHFRQWLQXHG
• Complete the development of a single, standardized, automated assessment, service plan,

and authorization process that could be applied to both nursing home and community
residents (allowing for comparisons across groups).
• Develop data capability and report production to inform day-to-day program management and
to provide information for decision makers’ long-range planning.
• Develop rate and fee systems for both institutional and community-based long-term care
services that work, in concert, to promote the overall goals of the State to provide services in
the least expensive and most appropriate settings.
• Institute an aggressive diversion and transition program to avoid institutionalization and assist
nursing home residents wishing to return to the community. This approach would include:
• identifying current and soon-to-be nursing home residents who could be appropriately
cared for in a home or community-based setting;
• reimbursing providers for these kinds of assessments; and

• diverting a significant portion of nursing home residents into home-based long-term care.
We first developed a “baseline” based upon the status quo of service use for nursing
homes and community-based settings in 2000 and 2008 (base year).70 Alternative modeling
scenarios included: 1) using the change in nursing facility use rates experienced in New
York from 2000–2008 to trend forward from 2008; 2) trending forward from 2008 based on
nursing facility use rates of another state (Vermont) with a similar nursing facility use rate in
1995 as New York, but that made greater progress in offering more home- and community-
based options; 71 and 3) trending forward from 2008 based on the national average trend in
nursing facility use rates. Our modeling assumes no change in bed supply.
Our long-term care analysis also estimates potential savings based on use and spending
projections (assuming there are no changes to the system compared to the mix of services
expected under a reformed system). The projections account for changes in population,
changes in potential need, changes in the mix of service, as well as program implementation
cost. We assume that savings phase in over a period of four years.
ESTIMATED EFFECTS
We estimate that if New York State can modernize its long-term care system by taking the
recommended actions, the State will realize financial savings. As depicted in Figure 20, under
the two scenarios, New York State could save in a range of $0.3 billion to $1.02 billion for
Medicaid over the 10-year period from 2011 to 2020.
70 Data source: State Medicaid long-term care service use and payment data for 2000 and 2008 provided by New York State.
71 Vermont has worked to implement the majority of the five-step package.
—61—
FIGURE 20. Projected Change in Spending under Assumed Scenarios

(2011–2020) (millions)
BASELINE ADMIN ASSUMPTION - VERMONT ASSUMPTION - NATIONAL AVG.
LTC COST, COST SAVINGS SAVINGS
YEAR EXCLUDING TO CARRY OUT BASELINE (NET OF REFORM BASELINE (NET OF REFORM
ADMIN LTC LTC COST IMPLEMENTATION LTC COST IMPLEMENTATION
COST REFORM COST) COST)
2011 $14,190 $92 $14,091 $7 $14,123 -$24
2012 $14,299 $94 $14,168 $38 $14,210 -$4
2013 $14,409 $95 $14,244 $70 $14,298 $16
2014 $14,520 $97 $14,321 $102 $14,386 $37
2015 $14,632 $99 $14,398 $134 $14,474 $58
2016 $14,768 $101 $14,533 $134 $14,613 $54
2017 $14,905 $103 $14,668 $134 $14,753 $49
2018 $15,044 $105 $14,805 $134 $14,895 $45
2019 $15,184 $106 $14,943 $134 $15,037 $40
2020 $15,325 $108 $15,083 $134 $15,182 $35
Total
$147,277 $1,000 $145,255 $1,021 $145,971 $305
(2011–2020)
Source: The Lewin Group estimates of long-term care service use and cost savings under two different assumptions for New York State. The baseline
LTC cost information is from The Lewin Group’s estimates of 2008 New York Medicaid claims. Numbers may not add to totals due to rounding.
DISCUSSION AND POLICY IMPLICATIONS

Our savings model accounted for administrative costs to carry out the suggested action
plan, including enhancing the ADRCs throughout the State, developing a standardized
assessment tool, and developing and producing reports to assist decision-making. Consistent
with the budget data submitted by ADRCs across the country, we assume $4 per person to
operate ADRCs statewide in 2008, increasing with inflation to cover all of the overhead and
implementation expenses of the single entry point and regular data reporting. We did not
deduct the cost of developing an assessment tool because the State has nearly completed this
activity. ADRCs comprise a collaborative effort between the Administration on Aging and the
Centers for Medicare and Medicaid Services to streamline access to long-term care services.
With its latest ADRC grant, New York plans to strengthen the promotion of NY Connects, which
is a State-funded program to establish county-level, consumer-centered access points to
information and assistance for all individuals in need of long-term care services, regardless
of age, income, or payment source.
—62—
The outlined policy options constitute the next logical steps for New York State’s Medicaid
program, and the State has already begun to implement several of them. For example,
New York State adopted ADRCs and recently launched a single standardized automation
system. New York State has also implemented a small nursing home diversion program.
While the general approach to rebalancing is typically the same across states, some,
such as Washington and Oregon, have approached rebalancing aggressively and are much
further along, resulting in a more significant shift toward community-based settings. Our
conservative savings estimates are based on the fact that, while New York State may be able
to achieve similar success over the long-term, it is currently positioned to continue taking
the first steps toward rebalancing. The PPACA will also help to advance the rebalancing of
long-term care through new opportunities, including the Community First Choice Option
which creates a Medicaid State Plan option to offer community-based attendant services and
supports with a 6% increase in the Federal matching rate.
Because this option is focused on the State Medicaid program, we consider all of the savings
under this option to be “actionable” by the New York State. While these actions could potentially
carry over to other populations, our savings estimates are for Medicaid only. To accrue the
savings estimated under this option, New York State needs to craft a strategic agenda to
overcome the systematic barriers related to inadequate information collected at points of entry,
payment methodology, effective care management, and affordable housing. Realizing these
efficiencies requires developing an information infrastructure to support real time analysis and
decision-making, as well as a fee system that encourages services provided to individuals in the
least expensive and most appropriate settings. In addition, to make these changes sustainable,
the State should focus on community resource issues, especially affordable housing, to enable
people to use more in-home community-based services and to relieve the barriers to the
expansion of Medicaid reimbursed assisted living.
—63—
Using Alternative Delivery Systems
$ % $ %
$24.30 $0.35 1.4% $0.35 1.4%
Baseline spending includes outpatient spending for low acuity services and immunizations.
BACKGROUND
7
he number of retail clinics operating in the United States has grown from 200 in 2006
to approximately 1,100 by the end of 2008. However, few retail clinics (seven CVS
Minute Clinics and five Duane Reade clinics) operate in New York State even though
nurse practitioners and advanced nurse practitioners, which primarily staff retail clinics,
have the ability to see patients and prescribe medications independently of physicians.72
Retail clinics have certain attributes that make them attractive to consumers, including
convenience (most offer evening and weekend hours and appointments are not necessary),
easy access, quick service, lower costs, and transparent pricing (most post a list of services
with prices).
Nationally, the number of clinics declined in 2009 due primarily to the economic downturn
and many close during summer months because the conditions they treat, such as
influenza, tend to be seasonal. However, many are expanding their scope of services to
include preventive, screening, and lab services to maintain year-round operations. Industry
forecasts project retail clinic sites to grow steadily over the next five years to approximately
3,200 sites by 2014.73 Thus, there are industry plans to continue to grow the number of retail
clinics in operation.
Studies have shown that costs of care for episodes initiated at retail clinics were substantially
lower than those of matched episodes initiated at physician offices, urgent care centers, and
emergency departments. According to Mehrotra, episodes initiated at retail clinics were $110
compared to $166 for physician offices, $156 for urgent care centers and $570 for emergency
departments.74 Thygeson et al found similar price differences between retail clinics compared
to urgent care settings, physician offices, and emergency departments.75
Based on a review of the literature, we identified nine conditions that are commonly treated in
retail clinics. These conditions include the following: sinusitis; bronchitis; pharyngitis; other
upper respiratory infections; immunizations; inner ear infections; swimmer’s ear; conjunctivitis;
and urinary tract infections.
72 National Association of State Legislatures, “Retail Health Clinics: State Legislation and Laws”, November 2009.
73 Deloitte Center for Health Solutions, “Retail Clinics: Update and Implications”, August 2008.
74 A. Mehrotra, “Comparing Costs and Quality of Care at Retail Clinics With That of Other Medical Settings for 3 Common Illnesses,”
Annals of Internal Medicine, (September 2009)
75 Marcus Thygeson, Krista A. Van Vorst, Michael V. Maciosek, and Leif Solberg, “Use and Costs of Care in Retail Clinics Versus Traditional
Care Sites,” Health Affairs, (September/October 2008).
—64—
8VLQJ$OWHUQDWLYH'HOLYHU\6\VWHPVFRQWLQXHG
POLICY OPTION
The intent of this policy option is to promote the growth of retail health clinics throughout
the State and to encourage patients to use retail clinics for low acuity conditions instead of
traditional settings. We assume that retail clinics take the following steps to promote increased
utilization of alternative delivery models in New York:
• negotiate acceptable rates with Medicaid MCOs;

• promote training programs for nurse practitioners (NPs)/physician assistants (PAs) in New
York, thus increasing the supply of alternate provider types;
• change policies to reduce any statutory and regulatory barriers to entry for alternative
providers (including scope of practice regulations for NPs/PAs); and
• provide tax incentives for large employers to open workplace clinics.

The literature is limited as to how particular policy options or regulations effect the growth
of retail clinics, especially in New York State. Thus, for illustrative purposes, we assume that
these steps increase the penetration of retail clinics in New York to the same levels and over
the same timeframe as seen in Minnesota to estimate potential savings. However, the actual
impact of these policy options may be very different.
Moving the treatment of these minor conditions from physician offices, urgent care centers,
emergency departments, and hospital clinics to retail clinics could reduce overall spending for
these services. We based our cost savings estimates on the Mehrotra study 76 finding described
above because it measured cost differences for a complete episode of care compared to the
Thygeson et al. study, which looked at price differences per visit. Mehrotra did not estimate
the cost difference for immunizations performed at retail clinics compared to other settings.
Therefore, we estimated the savings per immunization to be a similar percentage savings as
determined for the minor conditions above.
The savings estimates presented in the Mehrotra study were based on payments from a private
health plan. However, physician payment rates from Medicare are substantially less than rates
paid by private insurers and therefore savings would be smaller. Medicaid physician payment
levels in New York are substantially lower than Medicare levels. Thus, the Medicaid payment
rates would be less than the prices charged by the retail clinics. For this analysis we assume
that Medicaid managed care plans would negotiate payment rates with the clinics at levels they
currently pay primary care providers. The State would not see savings for Medicaid enrollees
substituting retail clinic services for physician office visits, but could see savings from enrollees
using retail clinics instead of the emergency departments and urgent care clinics.
Our estimates were derived using New York State Medicaid claims data for 2008 and a
proprietary claims database from Ingenix, Inc. with approximately 1 million UnitedHealth Group
76 A. Mehrotra, “Comparing Costs and Quality of Care at Retail Clinics With That of Other Medical Settings for 3 Common Illnesses,”
Annals of Internal Medicine, (September 2009).
—65—
covered lives in New York State in 2008. We estimated the number of visits for the specific
conditions listed above for New York residents using these data. We estimated 6.5 million
immunizations and 8.8 million low-acuity visits in 2008.
ESTIMATED EFFECTS
Over the 10-year period from 2011 to 2020, we estimate this policy could reduce State health
spending by $345.4 million (Figure 21). The Federal government could realize savings of $26.9
million from reduced physician spending for Medicare, Medicaid, and the Civilian Health and
Medical Program of the Uniformed Services (CHAMPUS). New York State government and local
governments could see savings of $42.9 million due to reduced costs for Medicaid and employee
benefit programs. Similarly, private employers could see savings of $136.7 million in reduced
benefit costs, which we assume would eventually be passed on to workers in the form of higher
wage increases. Households could see savings of $138.9 million in reduced out-of-pocket
spending and reduced premium costs, including the employee’s share of employer premiums.
FIGURE 21. Estimated Savings from Expanding Retail Clinics by

Stakeholder Group 2011–2020 (millions)*
STATE AND CHANGE IN
FEDERAL PRIVATE
YEAR LOCAL HOUSEHOLDS HEALTH SPENDING IN
GOVERNMENT EMPLOYERS
GOVERNMENTS NEW YORK STATE
2011 $0.6 $0.9 $3.0 $3 $8

2012 $1 $2 $6 $6 $16
2013 $1 $3 $10 $10 $25
2014 $3 $4 $14 $14 $35
2015 $3 $5 $15 $15 $37
2016 $3 $5 $16 $16 $40
2017 $3 $5 $17 $17 $42
2018 $4 $6 $18 $18 $45
2019 $4 $6 $19 $19 $48
2020 $4 $6 $20 $20 $51
2011–2020 $27 $43 $137 $139 $345
* Assumes full phase in by 2014. Future health spending inflation is based on CMS National Health Expenditure Projections, which is estimated to
increase at approximately 6% per year.
DISCUSSION
The savings estimates presented in this section are highly dependent on the growth in
retail clinics in New York State and the growth in retail clinics in New York is dependent on
reimbursement levels from private insurers, Medicare, and Medicaid, as well as other barriers
that may limit the business model for retail clinics. However, the literature is limited as to how
particular policy options or regulations effect the growth of retail clinics.
—66—
Certain barriers may exist in New York that have prevented the growth of retail clinics
compared to other states that have seen very dramatic growth. The higher price for labor, space
and capital that clinics would face in New York City relative to other states and the rest of New
York State could deter national companies from expanding retail clinics into New York. The
relatively low Medicaid reimbursement levels in New York, especially New York City, could
also deter clinic expansions. However, Medicaid payment rates for primary care will increase
in 2013 and 2014 under the Patient Protection and Affordable Care Act. PPACA also raises
Medicare primary care reimbursement rates. These provisions could make expansion in
New York City more attractive.
The clinic’s ability to establish good relationships with the physicians and hospitals in the
community is important so that patients who require follow-up care and who are diagnosed
with more serious conditions can receive referrals for the appropriate levels of care, and so
that there is no duplication of services provided. In January 2010, Duane Reade announced
plans to add another 20 clinics over the next year or so, which illustrates the possibility
of growth of retail clinics in New York. The New York State Health Foundation is supporting
a study by the Manhattan Institute to better understand the factors contributing to and
inhibiting growth of retail clinics in New York State.
The above estimates were based on a very limited number of services that are commonly
provided by retail clinics. However, there have been recent trends to expand the scope of
services provided by retail clinics to include preventive care, laboratory services, wellness
programs, and chronic disease management services.77 If retail clinics can expand their
scope of services, the potential for even greater savings exists.
77 Deloitte Center for Health Solutions, “Retail Clinics: Update and Implications”, August 2008.
—67—
&RQFOXVLRQ
+
ealth care spending in New York State is at an unsustainable level, and is expected
to continue growing at a rapid pace if current trends are allowed to continue.
Health care costs are squeezing the budgets of government, employers, and
households. The enactment of the Patient Protection and Affordable Care Act will
further increase the number of New Yorkers accessing the health care system, increasing cost
pressures and the need to find more efficient, higher value approaches to health care delivery.
While this landmark law will provide subsidies to make coverage and care more affordable,
additional reforms will be required to fundamentally address mounting health care costs and
sustain coverage expansion.
Concerns about the rising cost of health care would be less pressing if there was unambiguous
evidence that greater spending meant better health outcomes or a higher quality of care. The
evidence, however, suggests that the nation’s increasing spending on health care may not be
improving the quality of that care or health outcomes.
While the challenge is great, opportunities and options do exist to bend New York State’s cost
curve. This report outlines a range of approaches that over the next decade could achieve
tens of billions of dollars in savings while also improving quality. These opportunities exist
across the health care delivery continuum. Some, such as implementing bundled payments,
would work to fundamentally restructure provider payments. At least one would reduce
administrative inefficiencies in health care transactions (administrative simplification through
health information technology) and another would tax consumption of sugar-sweetened
beverages to reduce harmful health outcomes. Nearly all of the options would incentivize better
coordination of care and move New York State’s health care system toward providing the right
care to the right patients in the right setting.
A cohesive combination of these approaches will be needed to address the negative incentives
of the existing health care system. With so many of the approaches working to achieve the same
goal of more integrated care, however, there are interactions among them.
Implementing these policy options separately or as a package would slow the growth of health
care spending in New York State, in a way that does not destabilize the foundation of the delivery
system. Unlike across-the-board rate reductions in provider payments that do not address root
causes, these 10 approaches could achieve transformative and sustainable health care savings
while increasing quality and access to health care services.
Successful achievement of real efficiencies alongside improved quality and access will
require a concerted effort not to shift costs to other providers and payers. Simply moving
costs elsewhere within the health care system will not fundamentally address the factors
contributing to rising health care costs or generate sustainable net savings.
—68—
&RQFOXVLRQFRQWLQXHG
These 10 options highlight specific opportunities to begin tackling health care spending in
New York State. The time to start implementing payment and delivery system reforms is
now. Current trends forecast no end to health care cost growth and State fiscal gaps. Federal
health care reform raises the urgency that New York State and providers tackle health care
costs while implementing expanded coverage provisions. Fortunately, it also offers new
opportunities, including Federal partnerships and demonstrations, to advance many of the
policy options presented in this paper. The time is right for New York State to move beyond the
status quo and toward a more efficient, effective, and higher value health system.
—69—
Improving the State of
New York’s health
VOICE: 212-664-7656
FAX: 646-421-6029
MAIL: 1385 Broadway,
23rd Floor
New York, NY 10018
WEB: www.nyshealth.org

NYSHealth Lewin Group

Uploaded by

Document Information

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

NYSHealth Lewin Group

Uploaded by

Copyright:

Available Formats

%HQGLQJWKH

Promoting Accountable Care Organizations (ACOs) 18

Modernizing Primary Care 22

Expanding Palliative Care 29

Implementing Mandatory Managed Care For Medicaid Dual Eligible Population 34

Adopting Bundled Payment Methods 39

Imposing A Tax On Sugar-Sweetened Beverages 45

Expanding Hospital Pay For Performance 50

Realizing Administrative Simplification Through Health Information Technology 55

Rebalancing Long-Term Care 60

Using Alternative Delivery Systems 64

Significant responsibility for identifying and implementing cost-saving opportunities is expected

4 New York State Division of the Budget. 2010-11 Executive Budget.

POLICY SCENARIOS DESIGNED TO HELP CONTAIN COSTS AND

THE 10 SCENARIOS INCLUDE:

• Expanding Palliative Care. Require hospitals to establish a palliative care program to

• Realizing Administrative Simplification through Health Information Technology. Implement

FIGURE ES-2. Summary of Projected Savings by Policy Scenario, 2011–2020 (billions)

9 This definition of spending is termed “health services and supplies.

FIGURE 1. Projected Spending in New York by Type of Service...

Home Health ($17.2) Administration ($11.6)

Nursing Home ($18.2) Hospital ($66.5)

Worker’s Comp ($2.4) MediGap ($1.3)

Total Spending = Non-Group ($3.7)

Ten Options to Bend the Cost Curve

10 New York State Division of the Budget. 2010-11 Executive Budget.

14 Commonwealth Fund State Scorecard on Health System Performance, 2009.

Voluntary $1,096.56 $10.71 1.0% $3.11 0.3%

2011 $0.86 $0.25 $0.18 $0.05

FIGURE 3. Potential Savings from Adopting a Mandatory Medical

FIGURE 4. Potential Savings from Adopting a Voluntary Medical

Care Coordination. An evaluation of the Medicare Coordinated Care Demonstration concluded

FIGURE 5. Potential Savings from Adopting an Advanced Disease

FIGURE 6. Potential Savings from Adopting a Physician Pay-for-

2011 $0.08 $0.04 $0.05 $0.02 $0.20 $0.04

FIGURE 8. Savings from Requiring Hospitals to

2011 $34 $5 $5 $4 $48

40 Internal Lewin analysis.

FIGURE 9. Net Savings from Statewide Implementation

100%* $511.50 $24.20 4.7% $6.20 1.2%

FIGURE 11. Services selected for Bundled Payment Analysis

FIGURE 11. Services selected for Bundled Payment Analysis (continued)

2011 $91 $29 $32 $47 $199

2012 $215 $69 $77 $112 $473

2013 $253 $81 $91 $132 $557

2014 $270 $86 $96 $140 $592

2015 $287 $92 $102 $149 $630

2016 $306 $98 $109 $159 $673

2017 $326 $104 $117 $170 $717

2018 $348 $111 $124 $181 $764

2019 $370 $118 $132 $193 $813

2020 $394 $126 $141 $205 $866

2011–2020 $2,860 $913 $1,022 $1,489 $6,284

FIGURE 13. Estimated Actionable Savings by Stakeholder Group 2011–2020 (millions)

Consumption of sugar-sweetened beverages (SSB) has drastically increased since the

We based our estimate on available epidemiological literature, as well as additional simulation

An excise tax on manufacturers of sugar-sweetened beverages is assumed to be an increase

2011 $44 $10 $0 $0 $17 $4 $27 $6

2011 $15 $2 $3 $3 $3 $0.0 $5 $0.4 $27 $6

Hospital BIR ($5.0)

Physician BIR ($4.2)

Note: See Technical Appendix for a detailed

We originally planned to include an option that would promote standardized electronic

FIGURE 18. Combined Savings of Administrative Simplification Options