Accountable Care Organizations: The New Player in the Health-Reform Landscape

To assist health plans as they consider and pursue ACOs, Booz & Company conducted extensive primary and secondary research, proprietary economic modeling, and in-depth interviews with academic experts, external vendors, and senior executives of payors and providers. This Perspective reports Booz & Company’s findings on the emerging ACO landscape.

Show transcript


Gary D. Ahlquist Sanjay B. Saxena, MD Brett L. Spencer, MD

Accountable Care Organizations The New Player in the Health-Reform Landscape

This report was originally published before March 31, 2014, when Booz & Company became Strategy&, part of the PwC network of firms. For more information visit

Contact Information Atlanta Ralph Alewine Partner +1-404-519-0184 Chicago Gary D. Ahlquist Senior Partner +1-312-578-4708 Brett Spencer, MD Principal +1-312-578-4643 Abhishek Gopalka Senior Associate +1-312-578-4585 New York Gil Irwin Senior Partner +1-212-551-6548 Joyjit Saha Choudhury Principal +1-212-551-6871 Szoa Geng Associate +1-212-551-6276 San Francisco Curt Bailey Partner +1-415-263-3701 Thom Bales Partner +1-415-627-3371 Paolo F. Borromeo Principal +1-415-627-3387 Sanjay B. Saxena, MD Partner +1-415-263-3729

The authors would like to thank Szoa Geng, Abhishek Gopalka, and Phil Lathrop for their valuable contributions to this Perspective.

Booz & Company


Many payors and providers are considering or actively developing accountable care organizations (ACOs) as a result of their inclusion in the U.S. healthcare reform law. At this early stage, however, the long-term success of the ACO construct is not assured: Its value and longevity will be measured by the ability of ACO partners to drive cost reduction while improving care quality and patient experience. To assist health plans as they consider and pursue ACOs, Booz & Company conducted extensive primary and secondary research, proprie­ tary economic modeling, and in-depth interviews with academic experts, external vendors, and senior executives of payors and providers. This Perspective reports Booz & Company’s findings on the emerging ACO landscape, including:
• An assessment of the fast-evolving ACO marketplace • An analysis of the impact of ACOs on payors and providers, as well as competitive implications • Capability recommendations for payors ACOs create both market risks and opportunities for payors. But for payors who are able to provide capabilities that engender accountability to patients, care redesign, and organizational alignment in ACOs, their promise outweighs the risks. These payors will be able to drive urgently needed improvements in medical value, bolster the stability of the private market, and ensure their positions and futures in the shifting healthcare marketplace.

Booz & Company



What does the ACO landscape currently look like? What models and initiatives have appeared in the marketplace? ACO Activity in the U.S. Healthcare Market There is a swelling wave of ACO activity in the U.S. healthcare market in response to the 2010 Patient Protection and Affordable Care Act (PPACA), which included ACO-related provisions and support for demonstration projects. This activity has been further encouraged by state-led reform initiatives. For example, on February 17, 2011, Massachusetts Governor Deval Patrick announced a comprehensive healthcare payment and delivery reform legislation, “An Act Improving the Quality of Health Care and Controlling Costs by Reforming Health Systems and Payments,” designed to stimulate ACO development and global payment systems.i Stakeholders throughout the U.S. healthcare system are involved in ACOs: • Virtually every major payor is either involved in, planning, or seriously considering ACOs. Many health plans are actively helping providers, especially integrated systems and primary care physician (PCP) groups, to form ACOs. • Hospitals are cautious but increasingly keen on leading the forma-

tion of ACOs to avoid the risk of being left behind. According to HealthLeaders Media’s 2011 industry survey, 74 percent of hospital chief executives say that their organizations will be part of an ACO within the next five years. In addition, 16 percent of the respondents believe they already have the components for ACOs in place and 60 percent say they will have ACO components in place within the next five years.ii • PCPs are interested in ACOs given their focus on managing across the care continuum with a strong emphasis on primary care. Specialists tend to be more wary, because ACOs encourage the shift of services from specialty to preventive care, as well as rationalize the use of expensive procedures and testing. However, even they are pursuing participation in ACOs to replace lost volume and gain a share of the anticipated savings. According to HealthLeaders Media’s 2011 industry survey, 52 percent of physician group leaders said that their organizations will be part of an ACO within the next five years and 20 percent say they already have ACO components in place.iii External vendors, such as McKesson and UnitedHealth Group subsidiary OptumInsight, are focused on assisting providers with the analytics and


Booz & Company

infrastructure required to organize and operate ACOs. Third-party administrators see a re-emerging role for themselves as medical management is pushed out to providers. The level and breadth of activity suggests that the most fundamental change that ACOs may bring in the healthcare market is a substantial increase in the levels of collaboration among payors and providers. For example, ACOs that involve the full spectrum of care delivery will require greater plan-provider cooperation to pilot new payment models. Even if fee-for-service (FFS) continues to be the primary chassis for provider reimbursement, outcomes-based rate revisions, performance bonuses, and risk sharing will become increasingly prevalent. Payors and providers will also actively collaborate in specific areas, such as care management—which is expected to become more integrated—with a risk-adjusted, outcomes-based care management fee.

Evolving ACO Models ACOs are a new and ambiguous enough construct in the healthcare landscape to have so far eluded formal definition. Essentially, they are collaborative structures populated by a set of provider partners, often in association with one or more payor partners, who agree to deliver coordinated care and meet performance benchmarks for quality and affordability in order to manage the total cost of care for their member populations. ACOs are the latest attempt to generate medical value by combining HMO-like structural components, enhanced information technology, greater financial risks and rewards for providers, and an emphasis on care redesign processes. The nascent ACO sector is still very early in its development and is changing daily. Although there are a plethora of new ACO initiative announcements, most are still in the pre-launch stage, with ongoing discussions on key aspects such as governance, patient attribution, and

payment structure. Further, their progress is impeded by a lack of formal finalized guidelines from the Centers for Medicare & Medicaid Services (CMS). As of this writing, most in-market ACOs had launched operations only in 2010 or were planning to do so in early 2011. The ACOs that are currently operational are mostly in the pilot phase and very few have reported results to date, although the results that have been reported show glimmers of promise. While the sector is newly emerging, it is safe to say that most payors and providers are either involved in one or more ACOs or are evaluating the economics and feasibility of participation. Some of these projects are more ambitious, while others are simple rebrandings of existing constructs. On the surface it may seem that there are many ACO variations emerging in the marketplace. However, our research suggests that most ACO efforts fit into one of the three forms described on the following pages.

The most fundamental change that ACOs may bring in the healthcare market is a substantial increase in the levels of collaboration among payors and providers.

Booz & Company


Model 1: Adapted Integrated Delivery System These ACOs are organized around existing integrated delivery systems that feature either a single entity that acts as payor and provider or an association of providers with multiple care settings and employed physicians already affiliated with an external payor. Thus, because a complete system is already in place, it is relatively fast and easy for these partners to reorganize into ACOs. In fact, because the ACO concept was in many ways inspired by effective integrated delivery systems, such as Geisinger Health System and Fairview Health Services, ACOs that follow this model have been able to gain a first-mover advantage in the new sector. Adapted integrated delivery system ACOs are typically strongest in the area of organizational alignment— that is, the range of partners in the ACO, the parameters that govern their relationships, and the structures and processes that support them. In most cases, significant alignment existed before the ACO initiative began. They are also experienced in

the management of medical risk and are comfortable with assuming a portion of the financial risk associated with it. The most common large-scale change that adapted integrated delivery system ACOs have undertaken is the implementation of financial incentives aimed at supporting enhanced accountability for patient outcomes. This is the area in which adapted integrated delivery system ACOs are typically lacking, and it is where participation of payors is most useful. Often, the main role of payors is to provide financial incentives in the form of performance bonuses or shared savings that support these ACOs as they seek to take responsibility of the total cost of care of their patient populations. The provider partners in these ACOs are also looking to payors to provide risk management assessment, data analytics, and, in some cases, disease management. Of course, payors also provide an established patient base. In some cases, the adapted integrated delivery systems in the ACO space have also taken steps to incorporate

elements of accountability to patients and care redesign into their operations. Based on our interviews and research, it appears that the leaders of the adapted integrated delivery system ACOs recognize the importance of these two key elements for long-term success and are planning to pursue more fundamental changes in these areas. The following are examples of adapted integrated delivery system ACOs: Norton Healthcare/Humana: The partnership between insurer Humana, Inc. and Norton Healthcare, one of the five participants in the ACO Pilot Project sponsored by The Engelberg Center for Health Care Reform at the Brookings Institution and The Dartmouth Institute for Health Policy and Clinical Practice, is a good example of this model. Norton is an integrated delivery system—a not-forprofit hospital and healthcare system that already has a 44 percent market share in its local area. Humana has a base of local employer plans to populate the ACO. Humana initiated this ACO when it began exploring the concept with

Because the ACO concept was in many ways inspired by effective integrated delivery systems, ACOs that follow this model have been able to gain a first-mover advantage in the new sector.


Booz & Company

Brookings-Dartmouth in 2008. In early 2010, it announced it would partner with Norton Healthcare, which runs five hospitals and 12 clinics in the Louisville, Ky., area. As is typical in this model, the payment structure of the ACO includes a savings plan that will be shared by the provider, payor, and employers. This initiative used prospective patient attribution (i.e., pre-defined allocation of patients to an ACO) to enroll 10,000 members for pre-existing Humana employer plans. Further, the ACO targeted the following areas for improvement and savings: • Preventive screenings, tests, and vaccinations • Chronic illnesses management • Use of generics • Assessment to determine appropriate levels of care The initiative is in its pilot phase and has not yet reported results. However, Humana has stated that it plans to use its experience with this pilot to develop additional ACOs with provider partners in other regions.iv

Additional in-market examples of adapted integrated delivery system ACOs include the following initiatives (see Exhibit 1, page 6, for details): • Blue Cross and Blue Shield of Illinois and Advocate Health System • Monarch Healthcare, HealthCare Partners, and Anthem Blue Cross and Blue Shield in Los Angeles and Orange County, Calif. Model 2: Virtually Integrated ACOs The second ACO model is composed of multiple providers organizing in association with a payor, who contributes the financial incentives that support collective accountability for patient health outcomes and the technological infrastructure used to connect the disparate providers. There are two variations of this model, which we will describe separately: one in which the providers include mainly PCPs and one in which the provider set includes a broader range of partners including specialists, hospitals, and other care settings. Primary Care–Focused Virtually Integrated ACOs: The primary care– focused virtually integrated ACO is designed between a payor and care provider(s) with an emphasis on preventive care. Often these initiatives are built on the chassis of a pre-exist-

ing patient-centered medical home (PCMH), pre-existing pay-for-performance contracts with physician groups, or other settings with an emphasis on primary care interventions. One strength of these ACOs is ease of governance, because only PCPs and payors are involved. Another strength of the model is its primary care redesign potential, due to its evolution from formats, such as PCMHs, that are characterized by strong care coordination and the extensive use of quality data. The primary weakness of this model derives from the limited range of care settings included in these ACOs. While the primary care focus means a larger slice of the savings for the ACO’s primary care providers and payor partners, the overall savings is likely to be smaller because specialists and hospitals have little or no incentive to participate. These initiatives must also devote a good deal of attention to determining how they will align their ACOs with external care settings and in developing a payment infrastructure, which typically already exists in hospitals but not in physician groups. Finally, primary care providers are typically reluctant to undertake financial risk. While many physician groups are signing contracts with

Booz & Company


Exhibit 1 Examples of Adapted Integrated Delivery System ACOs

PROVIDER STRUCTURE Large health care system: five hospitals and 12 clinics in Louisville area

PATIENT POPULATION 10,000 Humana Participating members

PAYMENT STRUCTURE Shared savings between provider, payor, and employers



EARLY RESULTS/ PROGRESS - Early stages of implementation - Identified areas of emphasis include: -  Targeting improvements in use of preventive screenings, tests, and vaccinations - Better coordination in management of chronic illnesses - Appropriate use of generics - Improved access to appropriate level of care - Announced on 10/6/2010 - Contract effective from 1/2011 to 12/2013

Norton Healthcare— Humana

- Participant in the BrookingsDartmouth ACO Pilot Project - Humana plans to continue developing more ACO models by partnering with providers in other regions

Advocate— Blue Cross and Blue Shield of Illinois (BCBSIL)

Largest Integrated PPO and health system in HMO Illinois, operating Members 200+ care sites across metropolitan Chicago, including 10 acute care hospitals, two children’s hospitals, four Level I trauma centers, a home healthcare company, and a large medical group - Monarch HealthCare: IPA with more than 2,500 independent, privatepractice physicians - HealthCare Partners: more than 1,200 employed and affiliated PCPs, and more than 3,000 employed and contracted physicians overall PPO members in Los Angeles and Orange Counties

- Limits on annual Prospective FFS rate increases with shared upside savings - Total global risk for HMO products - Shared upside and downside on per member/monthbased medical cost trend for PPO products

Savings measured by Advocate’s performance compared to other BCBSIL contracting network providers

Monarch HealthCare— HealthCare Partners— Anthem Blue Cross and Blue Shield

- Initially: Shared savings - Some form of global capitation going forward

Currently unknown

- Participant in the BrookingsDartmouth ACO Pilot Project - Led by the two physician organizations

- Five-year pilot launched in January 2011

Source: Booz & Company research, company press releases, Becker’s Hospital Review, Brookings-Dartmouth ACO Learning Network


Booz & Company

performance-linked bonuses, they are less likely to enter gain-sharing agreements that include downside risk because they control a smaller percentage of total medical costs and do not have the financial wherewithal to withstand potential catastrophic outcomes. Aetna Medicare Advantage: A good example of the primary care–focused virtually integrated ACO is Aetna’s collaboration with providers in its Medicare Advantage program. Aetna has partnered with 36 primary care practices serving 20,000 of its Medicare Advantage plan members. The pilots, which Aetna began launching in 2007, combine personalized care management with electronic patient and evidence-based treatment information to support participating PCPs. Its payment system includes performance-based incentives for doctors, as well as compensation for the additional time and resources that PCPs devote to the ACO’s Aetna provided significant assistance for care delivery redesign. It used ActiveHealth Management’s CareEngine System to identify gaps in patient care and nurse case managers to work with doctors in the primary care practices, assist in coordinating patient care, and help members set up personal health records. Aetna also helped the physician groups coordinate with a range of other providers, from specialists to hospitals to community social services, including government agencies for in-home

services, home-delivered meals, and caregiver support. The initiative has been successful. Participating providers are meeting performance goals, such as follow-up office visits within 30 days of discharge and two annual office visits for certain chronic patients. In addition, a September 2010 Commonwealth Fund study showed significant reduction in “duplicate/ unnecessary” services (members in this pilot consumed 43 percent less acute hospital care in 2010 compared to Medicare recipients not in the pilot). A Health Affairs case study also computed an estimated 22 percent reduction in medical costs due to an increase in hospice election and a reduction of acute care services use, compared to a non-participating control group. Finally, physicians participating in the program discovered time savings for themselves and their office staff.vii Aetna continues to work with a variety of providers to create ACOs based on several models, including for their commercial members. In addition, they are leveraging the capabilities of their ActiveHealth Management subsidiary to assist providers with infrastructural capabilities, such as data management and case management. Additional in-market and announced examples of primary care-focused virtually integrated ACOs include the following initiatives (see Exhibit 2, page 8, for details):

• Roanoke, Va.-based Carilion Clinic, with payors to be determined • Tucson Medical Center and UnitedHealthcare Western States Health Plan Full-Spectrum Virtually Integrated ACOs: The full-spectrum virtually integrated ACO involves collaborations between a payor and a broader spectrum of providers than just PCPs, often including specialists, hospitals, and other community care settings. Often, these initiatives are emerging in markets with progressive providers who do not necessarily want to become fully integrated but want to participate in the shared savings that ACOs promise. In this model, the role of the payor is to provide infrastructure assistance and financial incentives for accountability of performance. Payors also help delineate how payment and shared savings will be distributed among participating providers. The inclusion of a wider range of providers enables more comprehensive and effective care delivery redesign, with associated potential for higher savings. Hospitals, for example, can use a variety of additional levers to increase healthcare value that physician groups typically cannot, such as limits on unnecessary testing, development of OBS (observation) units to avoid admissions, enhanced follow-up, and

Booz & Company


Exhibit 2 Examples of Primary Care–Focused Virtually Integrated ACOs

PROVIDER STRUCTURE 600+ physicians in a multi-specialty group practice and eight not-for-profit hospitals

PATIENT POPULATION 60,000 Medicare patients

PAYMENT STRUCTURE Shared savings and financial incentives for employed physicians based on cost and quality performance



EARLY RESULTS/ PROGRESS Anthem, CIGNA, United Healthcare, and Southern Health have expressed interest in partnering

Carilion Clinic

BrookingsDartmouth to provide technical assistance for pilot set up

- Large medical UnitedHealthcenter and affiliated care— physician groups Tucson - ACO includes Medical Center about 12 employed physicians and 50-60 independent physicians

Commercial, managed Medicaid, and Medicare FFS patients

- Shared savings Retrospective - PCPs to receive monthly care coordination fee and bonus for performance based on clinical quality factors - Future might include capitation model

- Based on a PCMH pilot financed by United with major employers such as IBM and Raytheon - United provides data and IT infrastructure - DartmouthBrookings provides “external manager” payment structure design

Preliminary results: - 4.5% decline in necessary emergency department visits - 22.5% drop in unnecessary emergency department visits - Some self-funded employers have expressed interest in joining

Aetna— Medicare Advantage

Aetna-driven ACO model created with 36 participating primary care practices

20,000 Aetna Performance Medicare Advantage (quality)-linked members incentives for facilities and physicians

Currently unknown

- Uses ActiveHealth’s -  Aetna began CareEngine System testing in 2007 to identify gaps in - Nearly all participatpatient care ing medical groups - Aetna helped form met performance the care team: targets, including doctor groups follow-up office paired with special- visits within 30 ists, hospitals, and days of discharge community social and two office services visits per year for certain chronically ill patients - Reduced acute hospital care by 43% compared to unmanaged Medicare Advantage members, according to a Commonwealth Fund study - Health Affairs study revealed an increase in hospice election and a decrease in acute care services, estimated to reduce medical costs by 22% (compared to control group)

Source: Booz & Company research, company press releases, Becker’s Hospital Review, Brookings-Dartmouth ACO Learning Network


Booz & Company

ensuring that all providers have test results readily available. In addition, these initiatives operate along the entire care continuum, enabling them to take greater accountability for overall patient wellness. Finally, hospitals’ payment, reporting, and IT infrastructure better enables the transition to a fully-functioning ACO, and, typically, they also have the ability to take on greater financial risk than physician groups. However, full-spectrum virtually integrated ACOs face much larger organizational challenges. These include often difficult decisions on leadership structure, process designs to determine payment and savings distribution among participating providers, and design decisions, such as which specialists to include and how leakage will be handled. Further, although these ACOs are willing to assume a higher portion of the risk, few are interested in managing the total cost of care of patient populations. Hill/CHW/BSC: The partnership of San Ramon, Calif.-based Hill Physicians (with 520 MDs), Catholic Healthcare West (with four hospitals), and Blue Shield of California, serving the California Public Employees’ Retirement System (CalPERS), is another good example of this model. The three partners—hospital, physi-

cians, and payor—comprise an adapted integrated ACO in the Sacramento area, one of the earliest and most ambitious entrees in the sector. The initiative, the planning for which dates back to 2007, was officially announced in April 2009. Prospective enrollment was opened in fall 2009, and approximately 40,000 CalPERS members in the Sacramento region, who were already in the Blue Shield HMO and receiving primary care through Hill, joined the ACO. The pilot was launched in January 2010. The partners’ goals for the pilot include a full range of objectives, including: • Reducing the cost of healthcare trend to 0 percent in aggregate for members • Building enrollment as a means of increasing the partners’ market shares • Creating a sustainable model for continuous improvement in cost, quality, and service that can be expanded to other geographic markets To achieve these goals, the ACO pilot is employing integrated processes, data, and quality metrics, as well as

clinical best practices and member and physician engagement. Further, incentives are aligned so that each partner shares in the savings but is at financial risk if targets are not met. This pilot is one of very few that has announced results, all of which are encouraging, according to executives within the partner organizations.v Results include: • Membership growth by more than 2,500 members since open enrollment • A 30 percent reduction in re-admissions • A reduction in average length of stay by 0.72 day for all admissions • A 7.6 percent reduction in ER/ Urgent Care admissions • A 15 percent reduction in total bed days Vermont ACO: Another example of the full-spectrum virtually integrated ACO is the Vermont Accountable Care Organization, which is spearheaded by Vermont’s Health Care Reform Commission. Its participants include three payors (Blue Cross and Blue Shield of Vermont, MVP Health Care, and CIGNA) and

Booz & Company


provider partners including three community hospitals (Northeastern Vermont Regional Hospital, Southwest Vermont Medical Center, and Springfield Hospital). Other stakeholders include the state hospital association, state medical society, the business community, state health reform staff, the Vermont Department of Health, the state legislature, and the Department of Banking, Insurance, Securities, and Health Care Administration. In 2008, Vermont’s Health Care Reform Commission began examining how ACO-like models could be piloted as part of the state’s comprehensive health program. With assistance from the Dartmouth-Brookings ACO Learning Network, three pilots are now being implemented across the state. The payors in these pilots are providing financial incentives designed to support collective patient accountability, and a common payor database to determine the allocation of payments among participating providers. Global budgets for each ACO are based on historical trends and adjusted for patient mix, with cost and quality incentives for providers to drive accountability. The level of risk undertaken by providers varies based on their level of sophistication and their degree of integration. “Newly formed entities with little experience managing care or risk” have entered into shared savings agreements,

while more sophisticated providers have taken on more risk in partnerships with two-sided gain sharing, in which providers reimburse payors for a portion of above-target medical expenses but also receive a larger portion of any savings.viii To prevent underutilization of medical services, performance measurement is not only limited to medical expenses but also includes clinical process measures, outcome measures, and patient experience data. Patient attribution in the Vermont ACO pilots follow a retrospective methodology developed by Dartmouth, in which patients’ primary care physician usage patterns are analyzed. Since primary care physicians are used for patient assignment, they can only participate in one ACO, while specialists and hospitals are allowed to partner with multiple ACOs. The goal of the Vermont pilot program is to “achieve delivery system reform based on the development of a true community health system that both improves the health of the population it serves and manages medical costs at a population level,” in which providers are required to manage the “full continuum of care settings and services for its assigned patients.”ix Toward this end, the pilots are using the patient-centered medical home model as the basic building block of care coordination. Community health

teams that support approximately 20,000 patients and typically include a nurse care coordinator, mental health staff, and nonclinical community health workers are being developed. The Vermont ACO pilots are expected to become operational in 2011. Additional examples of in-market and announced full-spectrum virtually integrated ACOs include the following (see Exhibit 3 for details): • CIGNA Healthcare’s Collaborative Accountable Care initiative, which includes separate ACOs with providers in various regions of the country, including ProHealth Solutions in Connecticut, Holston Medical Group in Tennessee, and Medical Clinics of North Texas • Dartmouth-Hitchcock Health, the largest healthcare provider in New Hampshire, and Anthem Blue Cross and Blue Shield Model 3: Provider-Led ACOs The third ACO model is formed and led by provider group collectives. These ACOs are typically composed of physicians, with or without hospital participation, and often they substitute payors with third parties that provide support functions, such as middle office operations and claims. Due to the lack of payor involvement in many provider-led ACOs, payment reform is often not a highlight of


Booz & Company

Exhibit 3 Examples of Full-Spectrum Virtually Integrated ACOs

PROVIDER STRUCTURE Large primary care groups, multispecialty groups, and integrated delivery systems of physicians and hospitals

PATIENT POPULATION Members in all types of products, including PPO

PAYMENT STRUCTURE - Bonus for meeting performance targets (quality and lower costs) - Eventually expected to move toward two-sided gain sharing

PATIENT ATTRIBUTION Retrospective, based on CIGNA analysis of which PCP last treated the patient


EARLY RESULTS/ PROGRESS - CIGNA expecting savings from efficiencies - No results available yet

CIGNA Collaborative Accountable Care

- ACO enabled by CIGNA - CIGNA provides claims data and informatics to identify high-risk patients, funds nurses for outreach programs, and produces management reports - Provider practices to determine care redesign opportunities Anthem and DartmouthHitchcock to collect and analyze data to “identify and implement efficiencies and improvements in healthcare delivery”

- DHMC is New Dartmouth Hampshire’s Hitchcock only academic Medical Center medical center (DHMC)— - 900-physician Anthem Blue group practice, Cross and with strong Blue Shield background in ACOs

 urrently unknown C (Anthem members)

Dartmouth Hitchcock to manage cost and quality of care

Currently unknown

- ACO yet to be implemented - DartmouthHitchcock has earned $13M to date in the Medicare Physician Group Practice Demonstration, the model for ACOs Early results include: - Added 2,500+ members since open enrollment - 30% reduction in readmissions - Reduction in average length of stay by 0.72 day - 7.6% reduction in ER/Urgent Care admissions - 15% reduction in total bed days - Started in 2008 to “achieve delivery system reform based on the development of a true community health system that both improves the health of the population it serves and manages medical costs at a population level”

Blue Shield of California (BSC)— Catholic Healthcare West (CHW)— Hill Physicians

- CHW: large hospital system - Hill Physicians: large physician multi-specialty group practice

About 40,000 California State employees in Greater Sacramento

Shared savings

- Prospective - Initiated a program to identify and enroll patients who have gone out of network

Hill and CHW are collaborating on alternatives, including pursuing evidence-based approaches to therapy and treatments prior to recommending surgery

Vermont ACO Pilot

Three Vermontbased ACO pilots

Currently unknown

Variable shared savings model based on level of provider’s integration and sophistication: - One-sided shared savings (no risk) for new providers - Shared savings with risk for more sophisticated providers

Retrospective, using historical patient patterns

- Key stakeholders: Vermont’s three major commercial insurers, three community hospitals, a tertiary hospital, state hospital association, state medical society, business community, state health reform staff, Vermont Department of Health, the legislature, and the Department of Banking, Insurance, Securities, and Health Care Administration - Pilot created with assistance from DartmouthBrookings

Source: Booz & Company research, company press releases, Becker’s Hospital Review, Brookings-Dartmouth ACO Learning Network

Booz & Company


the structural design. As such, many of these ACOs continue to be built around an FFS chassis, and they are focused on capturing shared savings without necessarily sharing gains with payors. While most of these ACOs are based on contract arrangements among a community of providers, they can also form as the result of acquisition. For instance, in 2010, St. Joseph Mercy Health System acquired IHA, a 150-physician organization in Ann Arbor, Mich., with the express intent of forming an ACO. IHA plans to recruit independent doctors and double in size.x Unsurprisingly, these provider-led ACOs tend to focus more heavily on care delivery improvements than any of the other models. In particular, they seek to coordinate care delivery across multiple provider settings, including aggressive management of the referral process, patient information share, and outcome tracking across the care continuum. This is a major strength of providerled ACOs: Because providers tend to have better insight into care process inefficiencies and physician practice motivators, they can generate the savings needed to compensate for lost revenue and increased effort, and modify behavior accordingly. The main obstacle that provider-led ACOs face is the lack of organizational expertise and bandwidth to engage in functions that traditionally fall within the purview of payors, such as network, medical, and

financial risk management. Provider executives leading the charge on provider-led ACOs recognize the challenge of defining the correct patient population for inclusion in an ACO, as well as the challenge of evaluating care outcomes across a system of providers in an actionable manner. Significant organizational efforts in governance and partner selection are thus required for these ACOs to succeed. A few sophisticated provider-led ACOs engage third-party vendors to oversee these important functions, as is the case with United Outstanding Physicians in Michigan. United Outstanding Physicians: United Outstanding Physicians (UOP), a Dearborn, Mich.-based physician organization with approximately 1,000 private and employed physicians, launched its ACO in 2010 in partnership with Automated Benefit Services Inc., a third-party administrator that will process medical claims; Continuum Management Services, an information technology and medical management company; and CIG Corp., a financial and regulatory management company. Under the UOP ACO, physicians are individually responsible for improving the outcomes and affordability of care provided to patients. They are incented through a unique stockbased reimbursement system, with 1,020 shares of stock distributed to 720 physicians. The distribution ratio reveals an emphasis on preventive care and recognizes the greater influ-

ence and responsibility of primary care physicians in providing it: 291 primary care physicians were each given two shares of the ACO stock, while specialists were each given one share. The remainder of the 60,000 shares will be available to new physicians or retained by the company. UOP was founded on the basis of strong physician leadership to monitor and improve peer performance and has continually emphasized evidence-based medicine as a tool for improving utilization performance. Physicians who participate in the UOP ACO will have access to electronic prescribing and electronic registry technology to better track their patients. Further, the ACO expects to contract with several hospitals for inpatient care and has approached healthcare systems in its primary service area in western Wayne County, including Oakwood Healthcare, St. Mary Mercy Livonia, Garden City Hospital, Detroit Medical Center, and Henry Ford Wyandotte. Additional in-market and announced examples of provider-led ACOs include the following initiatives (see Exhibit 4 for additional details): • Hyannis, Mass.-based Cape Cod Healthcare, Inc. • Optimus Healthcare Partners, a joint venture of two independent practice associations in New Jersey: VISTA Health System and Central Jersey Physician Network


Booz & Company

Exhibit 4 Examples of Provider-Led ACOs



PAYMENT STRUCTURE Shared savings; specifics not disclosed




United Outstanding Physicians (UOP)

Detroit-based UOP patients physician organization (with more than 1,000 physicians) partnering with several local health systems for inpatient care

- United Outstanding Goal: contract with Physicians will man- Medicaid or private age the ACO and payors as an ACO doctors will own it - Designated 1,020 stocks to 720 physicians: - 291 PCPs given two shares each - Specialists given one share - Remainder shares kept for new physicians - Contracts signed with: - Third-party administrator that will administer medical claims (Automated Benefit Services) - Medical management company (Continuum Management Services) - Financial management company (CIG Corp) - Two organizations -  Objectives: to be formed: to strengthen - Physician Orgathe working nization (PO): relationships physician-owned, among local physicianphysicians and governed hospitals, and to - Physician-Hospiimprove quality, tal Organization efficiency, and (PHO): joint vencoordination of ture of the PO patient care and Cape Cod - Long-term goal: Healthcare sysevolve into an ACO tem—to pursue risk contracts with joint governance - CCNS to contribute ACO knowhow and IT infrastructure - ACO (Optimus Healthcare Partners) will enroll physicians from combined 650 membership of VISTA and CJPN before expanding to rest of New Jersey - ACO to adopt evidence-based medicine, care coordination, and whole-person orientation Goal: to provide an alternative to hospitals or larger groups for physicians who wish to remain independent

Cape Cod Healthcare

Two hospital, Currently unknown physician–owned system partnering with local Cape Cod physician community

Contract with Caritas Christi Network Services (CCNS), a large Boston-based physician network, to manage patient budgets and healthcare quality

Currently unknown

Optimus Healthcare

Two IPAs: Vista Currently unknown Health (Summit) and Central Jersey Physician Network

“New compensation models rewarding good outcomes”

Currently unknown

Source: Booz & Company research, company press releases, Becker’s Hospital Review, Brookings-Dartmouth ACO Learning Network

Booz & Company



ACOs offer both volume and margin stabilization opportunities for procedural specialists. Specialists who are facing impending volume declines, driven by more aggressive utilization controls in the post-reform environment, can participate in ACOs to gain new volume. Such participation will also help them stabilize their margins by gaining a share of the ACO savings that would likely accrue with or without their participation. In a time when shifts in the healthcare model will make volume increasingly difficult to generate, ACOs present a valuable opportunity for hospitals to preserve and expand their volume. Hospitals that take leading roles in ACOs can gain market share, enhance their negotiating positions vis-à-vis payors, and develop key capabilities such as data analytics. ACOs can also be used to support the vertical integration efforts of hospitals; they can use them as “burning platforms” to acquire physician groups and a means of establishing direct contract

with self-insured employers, cutting payors out of the process. ACOs represent an opportunity for first-movers and low-cost, high-quality providers to define their brands and build valuable equity in the local market. Such brands will be key to creating and gaining market share in the post-reform era. Finally, over the long term, ACOs can enable providers to streamline their claims processing and eliminate obstacles to reimbursement. This can improve their cash flow and reduce their dependence on revenue-cycle management vendors. Provider risks: The primary ACO risk for providers is that they will lose volume if they cannot successfully position themselves in this emerging market. The ACO market scenario is based on the assumption that the total amount of care delivered to patients will decrease as the quality of care improves. This suggests that there will be a significant amount of

What is the potential impact of the ACO construct on the competitive positions of providers and payors? ACO Impact on Providers Provider opportunities: PCPs can derive significant benefits from ACOs due to their role as lynchpins in ACO structural design. This position increases the negotiating power of PCPs relative to specialists, hospitals, and payors. In addition, ACOs create the opportunity for PCPs to gain reimbursement or subsidies for medical management work, which many physicians already do or would like to do but cannot afford to undertake.


Booz & Company

unused capacity in the health system (e.g., specialists, inpatient settings). As a result, providers that are late or unsuccessful in pursuing ACOs could potentially lose even more volume as they lose market share to first-movers and low-cost, high-quality providers. Another major potential challenge for providers is the risk that ACOs may not even provide the economic returns necessary to support their participation. For instance, Steven Lieberman, a Brookings Institute visiting scholar, concluded that the shared savings plan in the Brookings-Dartmouth ACO

model, which is being piloted in five sites, may be inadequate. Lieberman calculated that the shared savings could constitute a significant portion of the primary care budget, but as little as 2 percent of the total hospital and specialist budget—a percentage that may not be enough to compensate them for the revenue reduction from decreased utilization and infrastructure investments associated with ACOs.11 There are also potential drawbacks in the various ACO payment models. Shared savings is particularly effective as a means to flatten the cost curve

in areas with high rates of healthcare inflation. However, in areas that are already relatively efficient or have lower rates of inflation, such saving may not materialize (see Exhibit 5).xii Further, since the U.S. healthcare system is largely built around a volume-based business model, there is a risk that ACOs and their associated care redesign strategies, such as prevention and lifestyle management, could reduce the economic viability of hospitals and large provider groups. “Most hospitals are carrying huge fixed cost overhead and we’re talking

Exhibit 5 Varying Impact of Shared Savings Models

Healthcare Spending

High-Cost, High-Inflation Regions

Low-Cost, Low-Inflation Regions

No reward for lower spending than other regions if it is higher than previous levels




Source: Center for Healthcare Quality & Payment Reform

Booz & Company


about reducing the patient population,” said one provider executive. “[It’s like] asking them to root for a weak flu season as opposed to for a strong one… it’s a sea change.”xiii In an ACO environment, providers will have to shift from a volume-based to a margin-based business model. One reason some ACO pilots have focused on PCPs is that these providers are in some ways better positioned to make this shift due to the fact that they currently provide services such as preventative care and disease management, albeit sub-optimally, with no reimbursement. Hospitals and other large providers will need to actively pursue ways to develop more dynamic modes of operation in which they can scale up or down based on demand and provide the appropriate level of care based on patient needs. Finally, to successfully develop and operate ACOs, providers will need to lead change within the organization and across stakeholders; drive the commitment to reduce costs, improve quality, and accept accountability for results; and identify, invest in, and develop a host of new capa-

bilities required to successfully plan and operate ACOs. Until ACOs are fully defined and their guidelines are established, the full extent of these capabilities is not known. Even when the capabilities are known, providers typically do not possess and have little experience with them (see “ACO Capabilities Needed by Providers”). ACO Impact on Payors Payor opportunities: First and foremost, ACOs offer payors an opportunity to reduce medical costs from more efficient utilization, lower unit costs, and improved outcomes. With the increasing willingness of providers to accept two-sided gain sharing, ACOs also offer payors an opportunity to transfer some of the financial risk of care when incurred expenses exceed agreed-upon total cost-of-care amounts. Although payors will continue to bear the burden of insurance risk, providers in ACOs (and especially those in an adapted integrated delivery model) are assuming an increasing share of the clinical risk. Second, although helping to establish ACOs will likely cause payors’ admin-

istrative costs to rise in the short-term, over the long-term, there may be an opportunity to reduce these dramatically. Incremental savings could accrue from consolidated billing based on the collectivization of providers and treatments that currently require separate payments. And eventually, if the shift to global payment is successful, ACOs could help eliminate a major portion of back office functions. Third, ACOs offer payors an opportunity to capture new revenue streams by packaging and selling medical management and reporting capabilities (especially for care delivered outside the ACO network). These include opportunities to create new businesses to assist ACOs with population health management, overall administration, and turnkey solutions. Payor risks: The primary ACO risk for payors relates to their relationships with providers and the associated threat of marginalization in the marketplace. Since payors will compete to contract with ACOs that offer the lowest total cost of care, it will be more difficult for all payors to


Booz & Company

ACO Capabilities Needed by Providers To successfully develop and operate ACOs, providers will require five broad sets of capabilities. These include: Technological infrastructure capabilities, including tools for outcome measurement and reporting, baseline databases from which to evaluate savings, analytics capable of mining clinical data for care delivery redesign insights, health information technologies supporting electronic medical and patient health records, and claims processing. Risk management capabilities, including population health management and actuarial cost assessments, savings/income distribution infrastructure, and predictive trend analysis and indicators. Care coordination, support, and wellness capabilities, including integrated care management, care leakage management, patient engagement tools (especially for non-compliant patient populations), and care coordinator/ navigator provisioning. Sales and marketing capabilities, including ACO product development, payor and employer-facing sales force development and management, and promotions and brand management. Finance and business development capabilities, including contract negotiation skills, and compensation design and management.

maintain relationships with low-cost, high-quality providers. This risk could be exacerbated if patient attribution processes and regulations prevent physician groups from aligning with more than one ACO, in which case payors that are unable to secure key partnerships could be cut out of low-cost ACOs or, in cases in which a provider group dominates a specific market, the ACO market itself. To the extent that ACOs drive quasiprovider consolidation, they are also likely to increase the negotiating leverage of providers. That said, this risk is expected to be countered to some extent by the expansion of government oversight on the part of the Federal Trade Commission and Department of Justice, which will likely monitor rates closely for power abuses associated with provider consolidation. The central role that providers can play in establishing ACOs can also create a risk for payors. Providers who take a lead in ACOs could gain an advantage by, for instance, setting reimbursement standards and dictating the amount of payor infrastructure that is used

Booz & Company


versus outsourced in ACOs. This threat is increased in markets where there is already a substantial level of care delivery integration and payors are poorly positioned to provide the “connective tissue” for ACOs. However, the worst-case scenario would entail provider-led ACOs disintermediating payors entirely. Large, self-insured employers, for example, could join ACOs directly, obviating the need for payors. If the health insurance exchanges mandated by the reform act are successful and they actively

set the terms of benefit packages, they could contract directly with potentially lowest-cost ACOs, leaving out traditional payors. Finally, provider-led ACOs could eliminate external payors by developing or outsourcing their medical management and administration functions, and using alternate payment methodologies. In fact, specialized vendors are already seeking to assist providers with the technological infrastructure, risk-management, and medical management capabilities they will need to operate ACOs (see “External Vendors in the ACO Space”).

In summary, ACOs represent a new basis of competition for payors and providers. As ACOs gain scale— indeed, if they are to successfully gain scale—they will accelerate the shift to a healthcare system that runs on a margin-based financial model in which both payors and providers will prosper by becoming ever more collaborative in their drive to improve the quality of care and reduce its total cost. These goals will be supported by a high-quality, cost-effective care delivery model that is patient-centric in its approach (see Exhibit 6).

Exhibit 6 Competitive Implications of ACOs


Revenue - Focus on higher volumes in an FFS environment Competition - Payors and providers debating approvals and negotiating contract rates, with general lack of trust Institution-centricity - Limited exchange of data; disjointed patient outreach State-of-the-art care - Marketing differentiation built on newest technology and latest procedure, even if little clinical difference

Margin - Focus on higher profitability (e.g., by reducing “controllable costs”) Collaboration - Payors and providers working together to create true healthcare value, with mutual “skin in the game” Patient-centricity - Full transparency; increased patient engagement High quality, cost-effective care - Marketing differentiated on quality of care supported by results at a better price

Source: Booz & Company


Booz & Company

External Vendors in the ACO Space A review of the market reveals the following external players offering solutions in the ACO space: Ingenix, a subsidiary of UnitedHealth Group: Ingenix offers “comprehensive data analytics and reporting capabilities to deliver insights that inform strategies to improve patients’ health outcomes” to providers. In February 2011, Boston-based Beth Israel Deaconess Physician Organization announced it was using the Ingenix Impact Suite software in a pilot program that will integrate performance measurement data from its 1,800 providers and physicians and analyze it for insights aimed at improving medical outcomes. McKesson Corporation: McKesson provides advice and technology solutions designed to support payment reform strategies. The company, which is actively engaged with a number of providers in the ACO space, is seeking to offer comprehensive solutions for provider-based ACOs, including data and analytics, support for health information exchanges, and patient engagement support. Treo Solutions: Troy, N.Y.-based Treo Solutions, in association with 3M Health Information Systems, offers an ACO solution that includes analytics, strategies, and technologies to redesign payment structures. Its aim is to help align payment to quality outcomes, support the move to patient-centered episode and bundled payments, and achieve accountable care. Zynx Health: Zynx Health, a subsidiary of Hearst Corporation, is a developer of evidence- and experience-based clinical decision support solutions designed to improve the quality, safety, and efficiency of patient care. In October 2010, Zynx announced the launch of a pilot program with five healthcare provider organizations, including Cedars-Sinai Medical Center in California, Memorial Hermann Healthcare System in Texas, Northwest Hospital & Medical Center in Washington, St. Joseph Health System in California and Texas, and WellStar Health System in Atlanta. The pilots will use the company’s service and software solutions to assess existing clinical processes and workflows, determine how they can become more efficient and effective, and identify potential barriers to clinical decision support adoption.

Booz & Company



ties: capabilities that support accountability to patients; capabilities that support care redesign; and capabilities that support organizational alignment. “Accountability to patient” capabilities: To succeed, ACOs must demonstrate their value to patients. HMOs lost favor in many markets in large part because they were perceived by consumers as sub-optimal health plan options. In some cases, this perception was based on false information; to this day, some HMOs enjoy higher patient satisfaction and achieve better outcomes than their PPO counterparts. Yet, as the health insurance market generally becomes more retail-oriented, patient perception will become even more important in determining the success—and failure—of insurance products. For this reason, it is crucial that any health plan embarking on the development of an ACO emphasize the importance of accountability to patients and communicating the benefits of ACOs to consumers. Doing so is critical both to the success of the ACO and, by extension, to the success of any strategy built around ACOs. So how should accountability to patients be pursued? There are three potential means for establishing accountability to the patient: 1.  Incorporate performance on patient experience metrics in financial incentive structures and creatively pursue additional ways to understand and incorporate patient input into ACO design and associated product development.

What issues must payors consider and what capabilities must they develop to capture the opportunities in ACOs? Most payors, including national and regional plans and the Blue Plans, are already moving to capture the opportunities and counter the competitive challenges inherent to ACOs. They are navigating the changing basis of competition that ACOs represent by pursuing increasingly collaborative initiatives with providers. They are preparing for the shift to margin-based financial model, as well as the increased negotiating pressure that ACO development could bring to bear, by selecting provider partners who can utilize significant market shares and a deep understanding of local markets to generate meaningful savings. They are also striving to deliver strong technological and data analytics capabilities to the ACO space to support a primary care– driven delivery model that is focused on accountability to patients and to counter the risk that someone else will set the standards in this emerging sector. Three ACO Capability Sets To further these efforts, payors should consider the following issues and develop three sets of ACO capabili-

In this nascent stage of development, most ACOs are not yet adequately measuring patient experience. While most ACO pilots cite improved patient experience as a program objective, few feature experience metrics in a prominent way. For example, only one of the approximately 100 performance metrics proposed by the National Committee for Quality Assurance (NCQA) for ACO accreditation directly measures patient experience. xiv Without targeted assessment, many dimensions of the patient experience that lie outside quality and affordability metrics will be unaccounted for, including experience of facilities and processes (e.g., care environment, wait time, administrative efficiency), emotional support (e.g., communication, empowerment, compassion), and overall brand impression (e.g., brand loyalty, willingness to refer). 2.  Dramatically increase the transparency and ease of use of consumer-facing provider evaluation systems. Payors usually have a significant market advantage in this regard due to the breadth and depth of the provider and patient data they possess. The development and expansion of existing transparency systems should start immediately so that consumers have time to become familiar with them before ACOs become more dominant in the marketplace. Toward this end, payors should actively market the transparency tools and the benefits and successes of ACOs. If ACOs do achieve the superior outcomes that their framers envision, effective marketing coupled


Booz & Company

with robust and accessible transparency systems will enable patients to become immediately aware of ACOs’ successes and proactively choose high-performing ACOs over other options. 3.  Balance the benefits of a closed network with patient desire for open networks. A third important element of accountability to patients is establishing and maintaining the balance between closed-network structures that facilitate performance measurement and opennetwork access that patients prefer. This aspect has become an important consideration as CMS debates the tradeoffs of prospective patient assignment to ACOs versus retrospective attribution. Notwithstanding the question of attribution, to avoid the pitfalls that weakened the reputation and appeal of HMOs, ACOs should allow patients to access providers outside the ACO and use other benefit design strategies as an alternative to closed networks. Payors can employ a broad range of benefit design strategies. These

strategies include differential copayments or coinsurance for care inside and outside the ACO, the addition of auxiliary services (such as a populated and comprehensive Patient Health Record) for those who receive the majority of their care within the ACO, and the broader use of valuebased benefit design, such as lower co-payments for evidence-based services that promote efficiency (e.g., X-rays versus CT scans as an initial diagnostic tool.) Value-based benefit design can also help tackle the ongoing challenge that physicians face in navigating the sometimes conflicting goals of meeting patient expectations and delivering evidence-based care. The rise of directto-consumer advertising and online information has led to an increasingly informed and, in many cases, demanding patient population. However, patients are less informed about the risks of overutilization. A renewed emphasis on shared decision making, and a commitment on the part of caregivers to communicate the relative risks and benefits of different treatment meth-

ods, will be essential in avoiding the perception among patients that increased efficiency means less optimal care. Care redesign capabilities: To achieve better outcomes, ACOs must significantly redesign care processes, including workflow design, communication pathways, team structures, and the relationships among providers, payors, and patients. A fundamental re-architecting of care delivery is necessary to generate efficiencies for providers and reduce their costs and, in turn, yield savings for payors. Specific examples of care redesign include new scheduling and access arrangements, development of teambased care, and quality improvement activities to reduce errors and readmissions (see Exhibit 7 for additional examples). Considering the central importance of care redesign in enabling the success of ACOs from the perspectives of payors and providers, too little attention is being paid to it in current initiatives. Among those providers who are actively experimenting

Exhibit 7 Selected Examples of Care Redesign

SELECTED EXAMPLES OF CARE REDESIGN Process Improvements - New scheduling and access arrangements - New ways of bringing evidence to the point of care - Institution of more point-of-care services Teamwork Focus - Development of team-based care - New coordination arrangements with other parts of the healthcare system - Data and information sharing across provider types Waste Reduction - Changes in practice management to eliminate redundancies - Quality improvement activities to reduce errors and readmissions - Implementation of EBM guidelines to establish protocols for appropriate use

Source: Booz & Company

Booz & Company


with care redesign, most are in the early stages of this work, such as creating quality documentation and procedural checklists in order to standardize care delivery and establish a baseline for measurement and future improvement. As yet, few providers are taking more ambitious steps, such as redesigning workflows and engaging patients in their own care. The extent of the change that must occur ensures that care redesign will not be easy. In pursuit of care redesign, ACOs will require partnerships across payors, hospitals, physicians, and patients. We recommend that payors work collaboratively with providers on three key actions to support care redesign: 1.  Development of a strategic plan for care redesign Payors, providers, and patient representatives should work in concert to prioritize and plan care redesign initiatives. This planning should include scheduling of specific care redesign activities, and the identification and adoption of methods for evaluating the impact of care redesign initiatives. 2.  Structuring and implementation of financial incentives for improving care practices

In the first stage of ACO development, financial incentives should directly reward and/or fund care redesign activities. Over time and as outcome measures become more sophisticated, incentives can support care redesign indirectly by rewarding performance. 3.  Provision of enabling capabilities for care redesign such as data analytics, wellness programs, identification of redundant or unproductive processes, and care coordination support Payors will need to provide sophisticated data to help providers pinpoint process inefficiencies and identify the causal relationships between procedures and outcomes. Further, they will often need to reconfigure these data for use in specific provider settings. One provider interviewee expressed frustration at the slow pace of this work to date: “We’ve been trying for two years to get the data [from our payor] that we need to improve critical processes… They don’t yet have the capability to adapt data systems to share the relevant information.” Particularly in cases where an ACO does not include the full spectrum of providers, payors can also play a rol