Lessons from the Field: Making Accountable Care Organizations Real

NIHCR Research Brief No. 2

Timothy K. Lake, Kate A. Stewart, Paul B. Ginsburg

Policy makers hope that the development of accountable care organizations (ACOs)—organized groups of physicians, hospitals or other providers jointly accountable for caring for a defined patient population—can improve health care quality and efficiency. An examination of existing provider efforts to improve care delivery illustrates that substantial financial and time investments are required to accomplish changes in care delivery, even among groups of providers affiliated with one another for many years, according to a new qualitative research study from the Center for Studying Health System Change (HSC). A common challenge for health care organizations’ efforts to improve care—similar to what the ACO concept seeks to encourage—is implementing changes with minimal disruption to patients and productivity. To achieve improvements in care delivery, the seven provider organizations studied tapped existing financial reserves or external grant funding. They also sought strong physician and organizational leadership and encouraged transparency and flexibility when making changes. As payers develop guidelines to contract with ACOs, they will need to support providers’ capability to develop and sustain improvements in care delivery, such as new health information technology and data reporting systems.


The ACO Concept

T o slow medical care cost growth and maintain or improve health care quality, policy makers, researchers and practitioners have proposed the development of accountable care organizations (ACOs). While there is no single, well-accepted definition of ACOs, there is general agreement that ACOs will constitute groups of providers—physicians, other clinicians, hospitals or other providers—that together provide care and share accountability for the cost and quality of care for a population of patients.1 Payers would contract with ACOs to care for a defined group of patients, using financial incentives to encourage ACOs to meet cost and quality goals. Ultimately, policy makers hope that ACOs will improve outcomes and reduce overuse of medical care.

Under the Patient Protection and Affordable Care Act, Congress established the “Medicare shared savings program” to develop ACOs for patients enrolled in fee-for-service Medicare.2 Under the program, participating ACOs will share in any savings with Medicare if the ACO meets quality standards and cost benchmarks. The law allows for flexibility in participating organizational structures. For example, ACOs may be comprised of physician group practices, networks of individual practices, partnerships or joint ventures between hospitals and professionals, hospitals employing professionals, and other groups deemed appropriate. Each ACO must care for at least 5,000 Medicare beneficiaries, establish appropriate management and leadership structures for clinical and administrative activities, develop processes to promote evidence-based medicine and patient engagement, report on quality and cost measures, coordinate care, and demonstrate patient-centeredness. Draft regulations with more detail on the Medicare shared savings program are expected to be released for public comment in early 2011.

Although there is considerable provider interest in developing ACOs, there has been limited experience with arrangements, organizational structures and care-delivery models similar to what ACOs might constitute. In the Medicare Physician Group Practice (PGP) demonstration, which involves 10 large physician practices across the United States, providers were offered performance bonuses based on meeting quality standards and lowering costs for fee-for-service Medicare beneficiaries. This model of incentives may be particularly relevant to incentives designed under an ACO program. However, results from the first four years of the PGP demonstration were mixed; all sites demonstrated quality improvements, but only five of the 10 practices have received performance bonuses based on savings to Medicare.3

Other recent efforts include the Brookings-Dartmouth ACO Learning Network, a forum for organizations interested in becoming ACOs to share information and obtain practical guidance; it currently counts about 60 health care organizations and health systems as members and has initiated some ACO pilots.4 Premier, a network that includes more than 2,300 hospitals, recently established ACO Readiness and Implementation Collaboratives to help member hospitals build organizational infrastructure to contract with payers based on cost and quality metrics.5


Moving ACOs from Theory to Practice

In concept, three important activities related to ACO development must occur to meet the goal of improving the quality and efficiency of care:

  • Policy makers and payers must create new payment arrangements, performance measures, reporting processes, and other programmatic features designed to hold providers accountable for improving quality and increasing efficiency.
  • Providers must decide to form an ACO to contract with payers on the basis of the cost and quality of care for a population of patients.6
  • Providers working together in ACOs must change how care is delivered to patients to improve the quality and efficiency of care.

This qualitative study focused on the third activity by examining recent efforts by selected provider organizations to undertake health care delivery reforms consistent with ACO program goals (see Data Source). Activities studied include changes in the way medical services are delivered and coordinated, such as implementing quality improvements in specific clinical areas and managing transitions of patients across care settings. The study also examined providers’ development of new infrastructure or organizational arrangements considered necessary to achieve improvements in care coordination, quality of care and efficiency. These included internal payment and financial reforms, development of health information technology (HIT), information exchange, and quality reporting.

The activities studied are not unique to these organizations, especially among other large provider organizations. But the organizations’ efforts are notable because many were pursuing multiple reforms simultaneously. At the same time, many of the changes pursued by these or other large provider organizations are unusual among the many small, independent physician practices that still provide a large portion of patient care in the United States.

The overall goal of the study was to help inform policy makers about the activities that future ACOs are likely to undertake to be successful, the challenges confronted by these organizations, and the ways those challenges were addressed.


Profiled Organizations and Market Contexts

T he study focused on the activities of seven provider organizations: the Billings Clinic, the Carilion Clinic, Physician Health Partners, ProHealth Physicians, Sharp HealthCare, UniNet and Westshore Family Medicine/Mercy Health Partners. These organizations varied in origins, size, ownership structure, financial arrangements and range of clinical services offered. They include three integrated delivery systems, one physician-hospital organization (PHO), two medical groups—one large group and one small group owned by a hospital—and one management services organization (MSO) affiliated with four independent practice associations (IPAs). Six of the seven organizations were large, with more than 250 physicians, while the one small group practice had fewer than 10 physicians, although it is affiliated with a hospital and larger primary care network (see box below for more information about the organizations).

The market environment of the organizations varied considerably, ranging from rural to urban. Some organizations had relatively little competition, while others had strong competition. The dominant type of payment across organizations also varied; some organizations accepted risk through capitated payments, while others functioned primarily in a fee-for-service environment, often with some pay-for-performance incentives.


Examples from the Field: Interventions to Improve Care

Billings Clinic, Billings, Mont., is a nonprofit health care organization that includes more than 280 multispecialty physicians and other clinicians, a hospital and other facilities. Almost all of Billings’ services are paid on a fee-for-service basis. Billings Clinic is one of the 10 practices participating in the Medicare Physician Group Practice demonstration. Interviews with Billings Clinic respondents focused on a new cancer care navigator program started in 2004 to improve cancer care coordination, patient satisfaction and outcomes.

Carilion Clinic, Roanoke, Va., includes a 600-physician multispecialty group practice and eight nonprofit hospitals. Carilion was previously a hospital-owned and -led organization called the Carilion Health System. In 2006, the Carilion Board of Directors decided that the most effective way to improve health care quality and care coordination and reduce costs was to transform Carilion into a physician-led organization. Interviews with Carilion Clinic respondents focused on development of physician payment incentives for cost and quality performance for affiliated clinicians. Carilion is one of four ACO pilots in the Brookings-Dartmouth ACO Learning Network.

Physician Health Partners (PHP), Denver, is a MSO that contracts with four IPAs—the largest includes about 180 physicians and the smallest includes about 20-25 physicians. The IPAs contract with PHP to provide IPA management, provider relations, contracting, financial and data management, and utilization and case management services. Each IPA has its own board, which makes contracting decisions and holds contracts, but the IPAs have no staff working on administrative or infrastructure-related issues. Interviews with PHP respondents focused on efforts to improve information technology infrastructure for the IPAs, including implementation and use of electronic medical records and patient registries to improve clinical integration and quality.

ProHealth Physicians, with sites throughout Connecticut, is a physician-owned, primary care organization with more than 250 clinicians in more than 75 sites across the state. Most locations are staffed by pediatricians, family practitioners and internal medicine physicians, although ProHealth also includes some specialists, such as otolaryngologists, a pediatric gastroenterologist, a sleep specialist, and various diagnostic and therapeutic services. ProHealth serves approximately 10 percent of the population of Connecticut, including Medicare and Medicaid patients, as well as patients with commercial insurance. Interviews with ProHealth respondents focused on recent HIT efforts, including electronic medical record implementation and development of a health information exchange and electronic patient registry.

Sharp HealthCare, San Diego, is a nonprofit organization with seven hospitals and other facilities and is affiliated with a 400-physician multispecialty medical group, Sharp Rees-Steely, and an IPA, Sharp Community Medical Group, with 700 physicians in private practice. Interviews with Sharp HealthCare respondents focused on several concurrent efforts to improve quality, including a disease management program for patients with congestive heart failure, efforts to improve follow-up care after hospitalization, implementation of electronic medical records, and efforts to measure and improve patient satisfaction.

UniNet, Omaha, Neb., is a PHO sponsored by Alegent Health, a hospital-based system; Creighton University Medical Center; and Creighton Medical Associates. UniNet represents more than 950 employed and independent physicians and 10 hospitals and other facilities. Interviews with UniNet respondents focused on disease management programs, a program to reduce hospital readmissions and early-stage efforts to implement electronic medical records.

Westshore Family Medicine/Mercy Health Partners, Muskegon, Mich., is an eight-physician primary care practice owned by a local hospital system, Mercy Health Partners, a part of the multistate Trinity Health System. Mercy Health Partners owns multiple physician practices in the area and has established a primary care network (PCN) and a PHO to provide billing and other administrative and logistic support to Westshore and 14 other primary care practices, although each practice operates independently. Westshore also participates in a primary care research network (PPRNet) that shares a common EMR system. Westshore implemented its EMR about 15 years ago and has used it for quality improvement. Interviews with Westshore respondents focused on efforts to improve quality of care for patients with diabetes and how the practice’s affiliations with the PCN, PHO and PPRNet help support these activities.


Care Delivery Improvements

A ll of the organizations studied were engaged in multiple efforts to improve care coordination and quality of care—activities likely to be pursued by ACOs. While interviews typically focused on one to two activities being pursued by each organization, most organizations were pursuing many changes at the same time. In some cases, respondents in the provider organizations noted the activities were undertaken explicitly in preparation for becoming an ACO. The activities generally fell into two categories: 1) interventions to improve care delivery; and 2) investments in infrastructure or other organizational changes to encourage or facilitate care-delivery improvements.

Changes in care-delivery processes. Several of the organizations were implementing programs designed to increase care coordination and improve management of patients with chronic conditions. For example, Sharp HealthCare, an integrated delivery system in San Diego, has begun using case managers who work with congestive heart failure (CHF) patients with highly complex medical and social needs. The case managers work primarily over the telephone, helping to coordinate medical care, as well as such community services as transportation. The program is designed to reduce CHF complications and resulting hospitalizations through early detection and management. Sharp also is working on reducing other types of hospitalizations through a continuity-of-care program, which involves calling patients within 48 hours of hospital discharge to ensure patients have follow-up visits scheduled and any drug prescriptions have been filled. Nurses also review discharge summaries with patients over the phone and answer their questions.

UniNet, a PHO in Omaha, has several initiatives to improve quality of care and outcomes among patients with diabetes and asthma, including telephone-based disease management services for English- and Spanish-speaking patients and group education classes. UniNet also is working with Blue Cross Blue Shield of Nebraska to implement a care transitions intervention to reduce readmissions, based on a model developed by Eric Coleman, including ongoing telephone contact with patients for a month after discharge.7 The PHO also established relationships with home health care agencies to ensure discharged patients are cared for at home when needed.

Westshore Family Medicine, part of Mercy Health Partners in Michigan, has developed a multi-faceted program to improve quality of care for patients with diabetes. The practice has hired a nurse case-manager who reminds patients to get their laboratory work done one to two weeks prior to any appointments. During the visit, the case manager meets with patients to review lab results and patients’ self-management plan, make any modifications, and refer patients to community resources as needed.

Billings Clinic, a Montana-based integrated delivery system, has developed a cancer navigation program as part of a broader initiative to develop a regional destination cancer clinic. Cancer navigators are specially trained nurses who coordinate all medical care for cancer patients, starting with efforts to make an accurate initial cancer diagnosis and development of a treatment plan. The navigators serve a central function in developing a “virtual clinic” in which patient appointments are scheduled with multiple physicians on the same day—an important feature for the many rural patients who travel considerable distances to get to Billings. The navigators accompany patients to appointments, answer patient questions, coordinate care with non-Billings providers (e.g., if patients undergo chemotherapy in their local communities), and ensure that all physicians, including patients’ primary care providers, remain informed about patients’ progress and treatments. The program was designed to improve care coordination, patient satisfaction and outcomes.

Investments or changes in infrastructure. Several organizations also were investing in care infrastructure or making other organizational changes to support care improvements. For example, Westshore Family Medicine uses a registry developed by its affiliated physician-hospital organization and a registry function in its electronic medical record (EMR) system to track patients and ensure they receive needed services. The registries are used for enhancing delivery of preventive services and for improvement in diabetes care. While implementing registry functions, Westshore empowered medical assistants and nurses to order services flagged by the registry rather than waiting for a physician to order these services.

ProHealth Physicians, a medical group operating throughout Connecticut, has developed a data warehouse that contains information from billing and practice management systems, laboratory databases, and EMRs. ProHealth uses the data warehouse for patient registry functions, such as generating monthly reports for each physician detailing which patients need what services and comparing the physician’s performance to established quality benchmarks. The monthly reports also include “recommended actions” for each patient who is not up-to-date in receiving relevant services, as reflected in the patient-registry data.

Physician Health Partners, a Denver MSO serving four IPAs, is facilitating adoption of EMRs in affiliated physician practices, including such functions as e-prescribing and access to electronic hospital discharge summaries and laboratory and radiology results. Physician Health Partners also has supported adoption of patient registries for adults and children with several chronic conditions.

Carilion Clinic in Roanoke developed a new set of financial incentives for employed physicians based on cost and quality performance, as it transitions from a hospital-led health system to a physician-led organization, similar in structure to other prominent multispecialty clinics elsewhere. The internal payment changes are seen as part of an attempt to transform the overall strategic orientation and culture of the organization toward better coordination of ambulatory care services and greater physician leadership and accountability for cost and quality performance. The goal of this transformation was to better position the organization for ongoing and expected payment changes, such those envisioned in ACOs.


Challenges to Change

E ven though each organization faced different challenges to care-delivery and infrastructure improvements, some of the challenges faced were nearly universal, including financing new efforts, addressing staff concerns and productivity problems during implementation, and developing appropriate and sustainable infrastructure to support these efforts.

Funding and the business case for change. All organizations noted that developing these ACO-like improvements required substantial investment, both in time and money. Many funded these activities through existing reserves, while others, particularly those engaged in quality measurement or HIT-investment-related activities, applied for and received grant funding. Some also noted that the incentives for changes or ability to fund investments may vary within larger organizations. For example, respondents at the Billings Clinic noted that their cancer navigator program—which is supported with general funds and not reimbursed directly by payers—might not be financially feasible in some other clinical areas with relatively low fee-for-service payment rates. None of the organizations indicated attaining a positive return on investments related to improvement activities. Although some noted ACO incentives or enhanced payments for patient-centered medical homes in the future might improve the business case for these activities, many acknowledged that the economic and market rewards may not materialize for a long time, if ever.

Resistance to change. Respondents at multiple organizations noted that people are often resistant to and fearful of change. It can be challenging for organizations to engage people who initially resist change. This was particularly true in situations requiring staff to assume new responsibilities or delegate work they previously had been responsible for. In some cases, recruiting and training staff with the appropriate skills also were a challenge. These issues are particularly relevant in organizations where clinicians are affiliated in nonexclusive ways, such as contractual or network-based arrangements.

Potential disruptions to productivity. Productivity was another key issue for these organizations. Many of the care-delivery and infrastructure improvements required changes in workflow that affected productivity of clinical and administrative staff. For example, implementing EMRs and using patient registries to ensure delivery of needed chronic disease care and preventive services often required changes in office workflow and staff responibillities. Most organizations noted loss of productivity while implementing and institutionalizing new technologies and programs. Finally, one of the larger organizations studied also noted that bureaucracy within larger health systems may impede quick implementation of new programs.

Limited infrastructure to pursue change. To improve care coordination and quality, whether through new HIT initiatives, disease management programs or corporate restructuring, all organizations required reliable data to measure and track performance. Organizations developing patient registries often noted that the registries are only as good as the data contained in them. Thus, if the data fed into the registries are inaccurate, the registries are inaccurate and not useful for population management or obtaining performance bonuses. Disease management programs also needed good data to be able to measure impacts of interventions. Other data issues mentioned included concerns about sharing patient data under federal privacy regulations. Several organizations noted they required legal advice before moving forward with care coordination efforts.


Addressing Key Challenges

T o overcome the challenges, organizations typically developed multifaceted strategies, including use of physician leaders and more open lines of communication to engage and reassure staff, enhanced financial incentives, and infrastructure support.

Enhancing physician involvement. One common feature of all of the organizations was their reliance on physician leadership and grassroots physician support to bring about organizational change. Most organizations noted the importance of physician buy in and ownership of change. A common theme during interviews was that any effort to transform a medical organization requires strong physician leadership, because physicians will listen to respected colleagues but not necessarily to non-physician administrators. This was particularly true for larger organizations and organizations where relationships may not be exclusive or employment based. For example, leaders at network-based organizations, such as Physician Health Partners and UniNet, noted that they needed to pay particular attention to outreach and communication to educate affiliated providers about the benefits of participation.

This did not imply a limited role for nonclinical leadership, but it meant that nonclinical leadership needed to partner with physicians and allow them to lead and champion projects to obtain acceptance from other clinicians. In addition, many organizations noted the importance of other clinical staff, including nurses and medical assistants, in developing new programs or interventions. Several noted that these programs required changes in these staff members’ jobs, and their buy in to the process was critical as well. For example, nurses and medical assistants may have been empowered to deliver or schedule preventive care services to patients before the patient saw the physician, where previously these services were not tracked well or required a physician order.

To obtain physician and staff support and to identify physician champions and leaders, respondents noted that transparency, open lines of communication and focus on teamwork were critical. For example, Carilion’s leadership held a series of discussions with clinical leaders throughout the organization about needed changes within the organization. These discussions preceded a public announcement in 2006 describing a broad set of organizational changes, including increasing physician leadership within the organization and changes in physician compensation to reward quality and efficiency. The public announcement was then followed by many discussions with providers both in the organization and the community.

For all of the organizations, transparency and communication also helped mitigate some of the fear of change that clinical and nonclinical staff may have had. Commitment to transparency also facilitated support among clinical and nonclinical staff because they understood the programs’ goals and steps needed to achieve these goals. Flexibility also was important; organizations noted some procedures may need to be customized to existing practice or physician workflow procedures.

Improving incentives. Organizations also found ways to align the goals of care improvement with payment and other financial incentives for affiliated clinicians. For example, several organizations working on expanding use of HIT offered stipends, grants or loans to assist practices in purchasing computer hardware and software. One of the IPAs affiliated with Physician Health Partners provided funding originally set aside for pay-for-performance bonuses to help its member practices adopt certified EMR systems. Organizations also were able to incentivize providers to improve documentation and delivery of chronic disease and preventive care services by showing how much more reimbursement they might obtain from performance-based contracts based on high quality scores. This not only encouraged clinicians to improve documentation and delivery of services but incentivized them to use data systems appropriately and further develop their infrastructure for measuring and reporting quality.

Infrastructure support for needed changes. Smaller medical practices, even those affiliated with larger organizations, have particular challenges implementing changes. These practices have fewer internal resources to support improvement activities compared to larger, highly integrated organizations. Nonetheless, smaller practices were able to engage in improvement activities through financial, technical and human capital assistance from larger affiliated organizations. For example, Westshore Family Medicine had fewer than 10 physicians but was owned by a hospital system that included a primary care network and PHO. The hospital, PHO and primary care network helped provide infrastructure support for EMRs, billing and reimbursement, and patient registry services, respectively. In addition, organizations affiliated with IPAs or PHOs noted that they provided extensive technical assistance to smaller practices when implementing HIT services, including substantial on-site assistance analyzing workflows, training staff, assessing implementation and providing ongoing support.


Policy Implications

In concept, ACOs are intended to encourage a variety of changes in care delivery across settings and providers. The seven organizations studied offer just a sample of activities that might be undertaken by new ACOs. Moreover, the clinicians within the organizations studied generally have had long-standing affiliations with one another. Yet, to have nationwide impact, effective ACOs will need to go beyond existing provider organizations, such as these, to include many physicians working in unaffiliated small practices who have not yet been engaged in efforts to improve care delivery. Change for these providers may be especially challenging.

In designing ACO policies, policy makers and payers should be aware that delivery system change is difficult and time consuming. Payer incentives for improved cost and quality outcomes alone may be insufficient for rapid improvements in care, especially for the majority of providers who do not face such incentives now.

Flexibility in ACO policies and implementation. Based on the experiences of these seven diverse organizations, policy makers may not want to be highly prescriptive about what organizational, provider composition, governance or structural features are required for participation in ACOs. Although the outcomes were not evaluated, diverse provider organizations in this study were able to implement changes designed to improve care in areas likely to be encouraged by ACO arrangements. Moreover, although the research literature indicates large, integrated provider organizations can be effective in achieving cost and quality goals, there is little evidence supporting the relative effectiveness of particular organizational structures or types of provider composition, such as the specific specialty mix of providers or the extent or types of affiliations with hospitals or other facilities.

One of the advantages of ACO-type payment reforms over reforms targeted at individual providers is increased patient sample sizes for reliably measuring performance. Less certain is the optimum size of an ACO to maximize organizational performance or efficiencies of scope or scale. There are concerns that the formation of very large ACOs may allow them to command too much market power, ultimately undermining the ability of policy makers and private purchasers to pursue cost-containment aims.8

At the same time, establishing certain requirements or incentives for the development of specific infrastructures within otherwise diverse organizations may be beneficial or even necessary for program operations. In particular, the organizations in this study viewed the development of information systems and cost and quality reporting capacities as necessary to support care-delivery reforms. At a minimum, ACO programs may need to require certain reporting capacities to measure and reward performance.

Finally, policy makers should be realistic about how quickly organizations can be expected to respond to new incentives or participate in care-delivery reforms. For example, one possibility is to develop slower vs. faster tracks, in terms of development of organizational readiness to take different levels of financial risk for cost and quality outcomes—similar to suggestions from others about testing different approaches to ACO policies.9 Generally, respondents at the seven organizations studied reported it took longer than expected to implement even relatively targeted reforms, with unexpected challenges, such as temporary interruptions to care delivery or productivity, often slowing progress. At the same time, many respondents noted that timelines should remain ambitious to maintain momentum.

Supporting leadership and communication within ACOs. Policy makers should consider initiatives to develop physician and administrative leaders within ACOs, with a particular emphasis on development of skills and approaches for enhancing communication about changes that may occur under ACOs and the reasons for those changes. All organizations studied noted that strong physician leadership and communication among all participants were critical to developing care coordination and quality improvement programs. Given natural resistance to change, it was also clear that administrative and clinical participants’ acceptance and support were facilitated by strong communication, transparency and organizations’ ability to match incentives to goals. Policy makers may want to consider sponsoring forums, not unlike the Brookings-Dartmouth ACO Learning Network, where organizations can share insights for developing strong physician leadership and strategies for effective communication to increase the likelihood of successful organizational transformation into ACOs.

Developing needed infrastructure. Transforming organizations into successful ACOs will require ongoing investment in data systems, workflow improvement, and systems of performance monitoring and communication within and across various practices and organizations. The seven organizations in this study financed these types of infrastructure development activities through existing reserves or grant funding. They considered these expenditures critical investments in the future well-being of their organizations. However, these organizations’ activities may not be sustained without fundamental changes to payment structures that allow them to recoup their investments. For example, the cost of a disease registry includes initial start-up costs to develop the database, but also requires ongoing maintenance and updates. More broadly, there are likely many organizations interested in becoming ACOs that have even less financial and human capital resources to draw on than the seven organizations studied.

The Health Information Technology for Economic and Clinical Health (HITECH) Act of 2009 provided incentives for physicians to invest and engage in meaningful use of HIT. Policy makers may want to consider similar incentive programs to encourage health care organizations to invest in other specific infrastructure needed to develop an ACO. These might include incentives or payments for certain cost and quality measurement and reporting functions, as well as development of certain evidence-based programs known to improve quality, increase care coordination or reduce costs. In addition, similar to regional extension centers sponsored within HITECH for technical support of HIT adoption and use, policy makers might consider developing technical assistance programs for developing ACO infrastructure, in addition to financial support.


Notes

1. Fisher, Elliott S., et al., “Creating Accountable Care Organizations: The Extended Hospital Medical Staff,” Health Affairs, Vol. 26, No. 1 (January/February 2007).

2. Patient Protection and Affordable Care Act (Public Law No. 111-148), Section 3022; and Newman, David, Accountable Care Organizations and the Medicare Shared Savings Program, Congressional Research Service Report to Congress, Washington, D.C. (Nov. 4, 2010).

3. Centers for Medicare and Medicaid Services, “Medicare Demonstrations Illustrate Benefits in Paying for Quality Health Care,” News Release (Dec. 9, 2010).

4. Overview of the Brookings-Dartmouth ACO Learning Network, available at: https://xteam.brookings.edu/BDACOLN/Pages/OverviewoftheNetworkMembership.aspx. (Last accessed Dec. 20, 2010.)

5. Premier ACO Collaboratives. Available at: http://www.premierinc.com/quality-safety/tools-services/ACO/index.jsp. (Last accessed Dec. 20, 2010.)

6. Most ACO program designs assume voluntary provider participation, but this may evolve over time. Depending on how ACO policies are structured, it is uncertain whether patients will make participation decisions. Some reform models assume that patients are largely left out of the decision-making process but are also entirely free to use providers of their choice within or outside an ACO. Others argue for more explicit choice and recognition on the part of patients to align themselves to an ACO, even if not locked into or enrolled in the ACO (sometimes referred to as “soft lock-in”).

7. Coleman, Eric A., et al., “Preparing Patients and Caregivers to Participate in Care Delivered Across Settings: The Care Transitions Intervention,” Journal of the American Geriatrics Society, Vol 52, No. 11 (2004).

8. Devers, Kelly, and Robert A. Berenson, Can Accountable Care Organizations Improve the Value of Health Care by Solving the Cost and Quality Quandaries? Robert Wood Johnson Foundation and Urban Institute, (October 2009); and Berenson, Robert A., Paul B. Ginsburg and Nicole Kemper, “Unchecked Provider Clout in California Foreshadows Challenges to Health Reform,” Health Affairs, Vol. 29, No. 4 (April 2010).

9. Shortell, Stephen M., Lawrence P. Casalino and Elliott S. Fisher, “How the Center for Medicare and Medicaid Innovation Should Test Accountable Care Organizations,” Health Affairs, Vol. 29, No. 7 (July 2010).


About the Authors

Timothy K. Lake, Ph.D., is a senior researcher at Mathematica Policy Research (MPR); Kate A. Stewart, Ph.D., is a researcher at MPR; and Paul B. Ginsburg, Ph.D., is president of the Center for Studying Health System Change and research director of the National Institute for Health Care Reform.


Data Source

This study examined seven provider organizations’ efforts to implement care delivery changes similar to interventions. Researchers interviewed a total of 34 people between March and May 2010 who either worked in or were affiliated with the organizations. Organizations were initially identified through expert recommendations and Internet searches and were purposefully selected to represent diverse organizations undertaking care delivery changes likely to be relevant to ACO programs. Respondents included administrative and clinical leaders, practicing clinicians and outside collaborators, such as other providers in the local community or affiliated health plan staff. Each interview was conducted using a two-person team with semi-structured interview protocols. Interview notes were summarized and jointly reviewed for completeness and accuracy. Interview notes were analyzed to identify challenges and lessons learned.

The National Institute for Health Care Reform (NIHCR) contracted with the Center for Studying Health System Change (HSC) between 2009 and 2013 to conduct health policy research and analyses to improve the organization, financing and delivery of health care in the United States. HSC ceased operations on Dec. 31, 2013, after merging with Mathematica Policy Research, which assumed the HSC contract to complete NIHCR projects.