Week2part3

By Carrie H. Colla, Valerie A. Lewis, Stephen M. Shortell, and Elliott S. Fisher

First National Survey Of ACOs Finds That Physicians Are Playing Strong Leadership And Ownership Roles

ABSTRACT The extent to which physicians lead, own, and govern accountable care organizations (ACOs) is unknown. However, physicians’ involvement in ACOs will influence how clinicians and patients perceive the ACO model, how effective these organizations are at improving quality and costs, and how future ACOs will be organized. From October 2012 to May 2013 we fielded the National Survey of Accountable Care Organizations, the first such survey of public and private ACOs. We found that 51 percent of ACOs were physician-led, with another 33 percent jointly led by physicians and hospitals. In 78 percent of ACOs, physicians constituted a majority of the governing board, and physicians owned 40 percent of ACOs. The broad reach of physician leadership has important implications for the future evolution of ACOs. It seems likely that the challenge of fundamentally changing care delivery as the country moves away from fee-for-service payment will not be accomplished without strong, effective leadership from physicians.

T hrough the Centers for Medicare and Medicaid Services, the Afford- able Care Act established a new, voluntary federal program that en- couraged the formation of account-

able care organizations (ACOs). ACOs aregroups of providers that arecollectively held responsible for the care of a defined population of patients. The core idea behind having insurers contract with ACOs is to create financial incentives for physicians and other provider organizations in an ACO to both improve the quality of care that their patients receive and reduce cost growth. Commercial insurers and state Medicaid agen- cies have begun to adopt contracts that are simi- lar to those in the new federal Medicare ACO program.1

The transition from current fee-for-service practice to the population-based care models en- visioned for successful ACOs will require a num- ber of major and potentially disruptive changes to current practice. These changes include incor-

porating evidence-based information more quickly into patient diagnosis and treatment, engaging patients and families more directly in care, delivering higher levels of team-based care, and providing more-transparent cost and quality data to physicians and other providers.2

An ACO is responsible for the broad continu- um of its patients’ care, regardless of whether or not it provides a particular service. As a result, ACOs may focus on developing and managing relationships with organizations such as post- acute care facilities, health departments, and community social service organizations.3,4

A major question is whether emerging ACOs will have the clinical and managerial leadership needed to navigate this transition.5–7 Physician leadership will be particularly important in motivating the implementation of quality im- provement and cost reduction programs by en- suring greater commitment to the ACO’s overall mission.8,9

The transition to an ACO is almost certain to

doi: 10.1377/hlthaff.2013.1463 HEALTH AFFAIRS 33, NO. 6 (2014): 964–971 ©2014 Project HOPE— The People-to-People Health Foundation, Inc.

Carrie H. Colla (carrie.colla@ dartmouth.edu) is an assistant professor at the Geisel School of Medicine, Dartmouth College, and the Dartmouth Institute for Health Policy and Clinical Practice, in Lebanon, New Hampshire.

Valerie A. Lewis is an assistant professor at the Geisel School of Medicine and the Dartmouth Institute for Health Policy and Clinical Practice.

Stephen M. Shortell is a professor at the School of Public Health and director of the Center for Healthcare Organizational and Innovation Research, both at the University of California, Berkeley.

Elliott S. Fisher is a professor at the Geisel School of Medicine and director of the Dartmouth Institute for Health Policy and Clinical Practice.

964 Health Affairs June 2014 33:6

ACOs & Medical Homes

by Rachel McCartney on June 6, 2014Health Affairs by content.healthaffairs.orgDownloaded from

have a major impact on many if not most aspects of physicians’ lives, including their incomes, de- gree of autonomy, work environments, and clin- ical routines.10,11 Physicians’ buy-in to these changes is likely to be critical. Previous research has shown that involving physicians in the gov- ernance of provider organizations improves communication and builds trust by assuring practicing physicians and clinical staff that their professional values are represented when key organizational decisions are made.12–14 Physician governance also assures patients that their needs will be considered, along with those of the orga- nization.15

In addition, physicians and policy makers have expressed concern that hospitals, because of their managerial strengths and resources, might quickly dominate the leadership of ACOs, thus accelerating the trend toward hospitals’ employ- ment of physicians and, in a more pessimistic view, interfering inappropriately in the physi- cian-patient relationship.8 Robert Kocher and Nikhil Sahni have argued that physicians should pursue ACO leadership in the initial wave of ACO development to preserve their long-term inter- ests.10 These authors believe that whoever dom- inates the formation of ACOs at the start is likely to continue to do so—a phenomenon called “path dependence.” Despite the potential importance of engaging

physicians in ACO leadership, little information is available about the leadership and manage- ment structures of ACOs.16 In this article we ana- lyze new national survey data on ACOs. We ex- plore the extent to which physicians are engaged in the leadership of emerging ACOs, including whether ACOs identify themselves as physician led, have boards that are run by physicians, and are physician owned. We also examine how physician-led ACOs compare to other ACOs in terms of structure, size, and services provided. And we examine the implications of leadership types for ACO capabilities and the future of the ACO model.

Study Data And Methods Overview The first wave of the National Survey of Accountable Care Organizations was fielded from October 2012 to May 2013. The design of the survey was based on published frameworks for evaluating ACOs,17–23 interviews with early ACO leaders, qualitative work with multiple ACOs,24–26 and a review of questions from exist- ing surveys.3,27,28

The survey included questions regarding ACOs’ contracts with payers, organizational components, capabilities, and activities. Cogni- tive testing on a sample of questions was com-

pleted by executives representing seven ACOs. Representatives of nineteen ACOs completed pi- lot testing of the full survey. The survey was completed by the person in the

ACO who was most knowledgeable about its con- tracts and activities. Most of the respondents were ACO executives, including CEOs, executive directors, and chief medical officers. The survey was offered either online or via telephone: 98 percent of the respondents completed it online. The survey was approved by the Institutional

Review Board at Dartmouth College. Participants We defined an ACO as a group of

providers that are collectively held accountable for the total cost and quality of care for a defined patient population. We identified likely ACOs that had been established by August 2012 through multiple sources: Participants in Medi- care ACO programs (Shared Savings and Pio- neer) were identified through public documents. Participants in state Medicaid ACO programs were identified through publicly available an- nouncements and communication with state Medicaid offices. And commercial payer ACOs were identified from diverse sources, including provider surveys that identified ACOs, participa- tion in ACO collaboratives (such those run by Premier or the Brookings-Dartmouth Learning Network), published case studies on ACOs, cer- tification by the National Committee for Quality Assurance, and public announcements of ACO contracts. We were broadly inclusive of potential ACOs in

our initial population. In all, 292 organizations were deemed possibly eligible and were invited to participate in the survey. Estimates of the number of ACOs currently in

existence vary widely, as do lists of the attributes that define an ACO.We used a set of preliminary screening questions to identify and exclude from our pool of invited organizations those that did not meet our strict criteria for an ACO: responsi- bility to a payer for both total cost of care and quality of care, and a contract in place to develop an ACO with at least one public or private payer. Thus, we excluded organizations that had de- clared themselves to be ACOs but did not have an ACO contract, organizations operating under a pay-for-performance approach that had no re- sponsibility for the total cost of care, and organ- izations operating under simple capitation that had no significant payments based on quality performance. Of the 292 potential participants, 203 com-

pleted the screening questions. Thirty were inel- igible to participate; the remaining 173 ACOs completed the full survey. This resulted in a re- sponse rate of 70 percent, based on the method-

June 2014 33:6 Health Affairs 965

by Rachel McCartney on June 6, 2014Health Affairs by content.healthaffairs.orgDownloaded from

ology of the American Association for Public Opinion Research.29

The response rate was higher among ACOs that had contracted with Medicare (81 percent) and lower among ACOs with Medicaid contracts (48 percent).We tested for nonresponse bias by comparing the distribution of Medicare ACOs in our sample with the distribution of all Medicare ACOs across the organizational categories devel- oped by Zirui Song and Thomas Lee.30 We found that the distribution of our sample across the categories was very similar to that of all Medicare ACOs on key variables such as involvement of physician group practices, hospitals, and safety- net clinics (see Appendix Exhibit 1).31 We thus opted not to develop survey weights to account for nonresponse. Outcomes Our analysis of the survey results

was largely organized around what type of lead- ers ACOs said that they had, as represented by responses to the question: “Which of the follow- ing best describes the organization of your ACO?” The possible answers were physician led; hospital led; jointly led by physicians and hospital; state, region, or county led; coalition led; led by a federally qualified health center; and some other arrangement, with space provided for the respondent to specify the arrangement. We first completed a factor analysis of re-

sponses that we hypothesized would be indica- tive of physician leadership in ACOs, including answers to questions on governance, ownership, leadership, and influence. Based on that analy- sis, we created a composite measure of physician leadership. However, we found that 98 percent of the ACOs that we considered to have strong physician leadershipaccordingto thiscomposite measure had identified themselves as physician led. Based on these results, we used the single question on leadership to identify physician- led ACOs. We initially divided the remaining ACOs into

subcategories, but we found similarities across ACOs jointly led by physicians and hospitals and ACOs in the remaining subcategories. Thus, we compared ACOs that identified themselves as physician-led with all other types of ACOs, using bivariate models with two-sample comparison of means tests. We measured a number of care management

and coordination capabilities, using behavioral- ly anchored response categories—that is, we an- chored a quantified scale with specific narrative examples of advanced, moderate, and poor per- formance.We report the proportion of ACOs that reported having the most advanced capabilities. We created an index of health information tech- nology (IT) development from twelve questions on health IT capabilities and use (Appendix

Exhibit 2).31 An ACO was considered to have “ad- vanced” health IT if it had complete or near- complete capabilities on eight of the twelve measures. Limitations Our study had some limitations.

The National Survey of Accountable Care Organ- izations addressed a broad range of questions to only one person at each ACO; therefore, the re- sponses reflect the views and knowledge of that individual. We believe that for each ACO sur- veyed, we identified the respondent who was the most knowledgeable about the organization. However, that person’s responses might not be representative of those of other people in leader- ship roles in the ACO or—importantly—of front- line clinicians and staff within the ACO. Com- pared to executives at other ACOs, executives at physician-led ACOs may respond differently to questions to demonstrate their managerial ap- titude. We used behavioral anchoring of responses to

guard against subjective interpretation of survey questions. However, it is possible that different types of respondents (such as physicians versus administrators) interpreted the questions differ- ently. In addition, despite testing for nonresponse

bias using publicly available data on Medicare ACOs, it was difficult to determine the presence of nonresponse bias in organizations with com- mercial contracts. This is because there were no publicly available data on ACOs with commercial contracts with which to compare our sample. Given our high response rate, however, the im- pact of any such bias should be small. A final limitation is that we relied on self-

identification to classify ACOs as being physi- cian-led. We provide some data that appear to be consistent with respondents’ self-identifica- tion, and we completed a factor analysis that confirmed the distinction between physician- led ACOs and other ACOs in almost all cases. Nonetheless, we acknowledge our reliance on the respondents.

Study Results Leadership Physicians are playing a strong role in the leadership of the first wave of ACOs. Fifty- one percent of the respondent ACOs identified themselves as physician led, and another 33 per- cent reported that they were jointly led by hos- pitals and physicians (Appendix Exhibit 3).31

Only 3 percent reported being led by hospitals alone; the remaining 13 percent wereled by other entities. Physicians constituted a majority of the gov-

erning boards of 78 percent of all responding ACOs, 94 percent of physician-led ACOs, and

ACOs & Medical Homes

966 Health Affairs June 2014 33:6

by Rachel McCartney on June 6, 2014Health Affairs by content.healthaffairs.orgDownloaded from

65 percent of ACOs led jointly by hospitals and physicians. Physicians also werea majority of the boards in most of the other types of ACOs. Overall, physicians owned the equipment

and employed the staff in 40 percent of ACOs. Physicians owned 62 percent of physician-led ACOs, compared to 16 percent in all other ACOs (p < 0:001). Furthermore, 58 percent of both physician-led ACOs and other ACOs reported “extensive and active involvement of clinicians in ACO discussions and decision making.”

Organizational Components Compared to other ACOs, physician-led organizations were less likely to include a hospital and to be part of an integrated delivery system (Exhibit 1). Phy- sician-led ACOs were also less likely to include a federally qualified health center or a rural health clinic, but they were as likely as other ACOs to include a nursing home. Physician-led ACOs had fewer primary care

physicians and specialist clinicians but included more individual medical groups than other ACOs (Exhibit 1). The larger number of individual medical groups may indicate that at least some of these ACOs consisted of independent practice associations or included a network model in which physicians owned their individual practi- ces. In fact, 37 percent of physician-led ACOs consisted solely of physician practices, com- pared to only 6 percent of ACOs with other leadership types (p < 0:001; data not shown). Physician-led organizations also had fewer pa- tients per contract than other ACOs (Exhibit 1). Similar patterns emerged when we compared

the services provided by physician-led ACOs and those provided by other ACOs. Physician-led ACOs were less likely than other ACOs to directly provide services across the continuum of care, including emergency and postacute services (such as rehabilitation, skilled nursing, and home health care) and services that are especial- ly important in the care of special or vulnerable populations, such as behavioral health or hos- pice care (Exhibit 1).

Contracts Physician-led ACOs were more likely than other ACOs to have a Medicare con- tract but less likely to participate in the Pioneer program, which is designed for large organiza- tions that are prepared to take on financial risk (Exhibit 1). Two-thirds of physician-led ACOs were participating in the Medicare Shared Sav- ings Program, and few were taking on downside risk. Fifteen percent of physician-led ACOs were participating in the Advance Payment Program, which is part of the Shared Savings Program but provides start-up capital to physician-based and rural providers. One concern about the ability of physician-led

organizations to become ACOs is that they might

have difficulty securing capital to fund the tran- sition.10 A quarter of the ACOs in our sample reported that securing sufficient funds to launch an ACO was very challenging, but there was no difference on this measure between physician- led and other ACOs. Significantly fewer physician-led ACOs were

participating in contracts that included down- side risk, compared to ACOs with other leader- ship structures (Exhibit 1). In both groups, near- ly all of the ACOs currently bearing risk had had previous experience with risk-based contracting, through either bundled payment initiatives or capitation. Physician-led groups were slightly less likely than other ACOs to have had experi- ence with risk-based contracts. Capabilities Physician-led ACOs were as like-

ly as ACOs with other leadership types to report having advanced caremanagement and health IT capabilities (Exhibit 2). However, fewer than half of all ACOs had these advanced capabilities. In the two cases where the difference between physician-led and other ACOs was significant, a higher percentage of physician-led ACOs re- ported having the capabilities. Physician-led ACOs areleading in outpatient care management and health IT. However, they are lagging in their ability to manage care across settings (transi- tions and readmissions), which is consistent with the fact that physician-led ACOs are less likely to include hospitals and postacute care providers. Monitoring And Reporting Quality And Fi-

nancial Results About half of both physician- led and other ACOs reported having the ability to monitor systemwide quality performance met- rics and provide meaningful and timely feedback to clinicians (Exhibit 3). Only a third of both physician-led and other ACOs monitored com- prehensive and timely financial performance rel- ative to benchmarks. Timely financial data that allow ACOs to gauge their performance against a benchmark may be important in the organiza- tions’ ability to stay within a global budget and be eligible to receive shared savings. The lack of such data was a complaint among organizations that participated in a Medicare ACO demon- stration.32

Our survey also asked about challenges to ACO implementation. Developing health IT infra- structure was the most frequently cited chal- lenge, and half of the respondents reported that it was very challenging. As noted above, most ACOs reported being

unable to monitor financialperformance relative to benchmarks. However, physician-led ACOs were significantly more likely than other ACOs to measure and report financial performance at the practice and clinician levels (Exhibit 3).

June 2014 33:6 Health Affairs 967

by Rachel McCartney on June 6, 2014Health Affairs by content.healthaffairs.orgDownloaded from

Outlook On The ACO Model Fifty-four per- cent of leaders at physician-led ACOs said they believed that at least half of US patients will be covered by ACO-like contracts in the next five

years, but only a third of leaders at other ACOs had the same belief (p ¼ 0:010; Appendix Exhibit 4).31 About three-quarters of both groups believed that more than half of the patients in their market will be covered by ACO-like con- tracts. Respondents in physician-led ACOs be- lieved that the ACO model will spread quickly and were optimistic about its potential to im- prove quality. However, fewer respondents were optimistic about the model’s potential to reduce cost growth. This is consistent with previous re- search on performance in the Physician Group Practice Demonstration33 and the initial year of the Pioneer ACO program.34

Discussion In spite of initial concerns about potential hos- pital dominance,8,10 physicians are at the fore- front of leadership in the early implementation of the ACO payment and delivery model. In the first national survey of both public and private ACOs, we found that over three-quarters of ACOs were either physician led or jointly led by physi- cians and hospitals, and that physicians consti- tuted a majority of the governing boards of 78 percent of ACOs. Active leadership by physicians has been

shown to be critical to the success of efforts to change physician practice14,35 and to help over- come potential resistance from physicians, pa- tients, and other groups to new financial mod- els.15,36 Thus, these findings suggest that emerging ACO governance structures offer the promise of continued support and performance improvement. Physician-led ACOs differ from other ACOs in

key ways. Physician-led ACOs are less likely to include hospitals and other types of providers, but they are more likely to include physician groups—and, if they do include physician groups, to have a larger number of the groups (Exhibit 1). Despite having different leadership structures and offering fewer services, physi- cian-led ACOs have care management and health IT capabilities that are similar to those of oth- er ACOs. Because they are less likely to include hospitals

or postacute care facilities, physician-led organ- izations may face greater challenges than other ACOs in managing transitions between settings of care and managing hospital-based care, if it is provided by hospitalists who do not have a for- mal relationship with the ACO. It is possible, therefore, that ACOs with other leadership types may be better equipped to coordinate care through the participation of organizations across the care continuum. Physician-led ACOs are also less likely than

Exhibit 1

Organizational Characteristics Of Accountable Care Organizations (ACOs), By Leadership Type

Type of ACO leadership

Characteristic Physician led (n=89; 51%)

Other (n=84; 49%)

All (N=173)

Organizational structure

Include a hospital Percent of ACOs 41 87**** 63 No. of hospitals 2.1 4.7** 3.4

Include a medical group Percent of ACOs 93 87 90 No. of medical groups 23.3 13.8 18.8 No. of primary care clinicians 156.2 204.4 179.1

Include a specialist group Percent of ACOs 49 64 56 No. of specialist groups 13.0 11.6 12.3 No. of specialist clinicians 215.7 268 240.6

Include an FQHC or RHC Percent of ACOs 15 43**** 28 No. of FQHCs or RHCs 0.3 2.0*** 1.1

Include a nursing home Percent of ACOs 20 25 22 No. of nursing homes 1.6 1.5 1.5

Belong to an integrated delivery system Percent of ACOs 44 64** 54

Services provided (% of organizations)

Emergency 30 78**** 54 Rehabilitation 26 59**** 42 Behavioral health 28 57**** 42 Skilled nursing 10 21** 15 Pediatric 55 64 59 Palliative or hospice 27 58**** 42 Home health or visiting nurse 17 49**** 32 Pharmacy 20 38*** 29

No. of assigned patients in largest contract (% of organizations)

Fewer than 5,000 1 0 0 5,000–10,000 40 30 36 10,001–20,000 31 18 25 20,001–50,000 19 37** 27 More than 50,000 8 15 11

Type of contract (% of organizations)

Any Medicare 75 57** 66 Pioneer 9 21** 15 Shared Savings Program 51 37**** 44 Advance Payment Program 15 0**** 8

Medicaid 13 38**** 25 Private insurer 45 57 51 Experience with risk Current contract 20 40*** 29 Previous experience 72 86** 79

SOURCE Authors’ analysis of data from the National Survey of Accountable Care Organizations. NOTES “Other ACOs” are led by hospitals and physicians jointly (33 percent); coalitions (6 percent); hospitals (3 percent); state, region, or county (1 percent); federally qualified health center (FQHC) (1 percent); and other (5 percent). All numbers are mean. Advance Payment Program ACOs are part of the Medicare Shared Savings Program. Significance denotes difference from physician-led ACOs. RHC is rural health clinic. **p < 0:05 ***p < 0:01 ****p < 0:001

ACOs & Medical Homes

968 Health Affairs June 2014 33:6

by Rachel McCartney on June 6, 2014Health Affairs by content.healthaffairs.orgDownloaded from

other ACOs to offer services that traditionally are segregated from medical care, such as pharmacy or behavioral health services. This may make it more difficult for physician-led ACOs to track medication compliance, for example, or follow up on mental health referrals or discharges. The networks in physician-led ACOs may be limited because physicians are ill equipped to develop and manage relationships with multiple organ- izations—a crucial goal of ACOs to ensure smooth care transitions and continuity of care. Many stakeholders22,37 have expressed opin-

ions about what ACOs should or will look like. To our knowledge, however, this article is the first to present national data on physician lead- ership in the first wave of ACOs and on the rela- tionship between leadership and organizational characteristics, capabilities, and views of leaders on the future of accountable care.

Our findings suggest that there are diverse types of leadership and paths of development for ACOs, as the Centers for Medicare and Med- icaid Services intended when it designed its ACO program.9 Because it is not yet clear which char- acteristics and capabilities are important for an ACO’s performance, policy makers and payers may need to continue to provide different types of support and contracts to encourage the con- tinued development and success of physician- led ACOs. An ACO’s leadership and organizational struc-

ture will have implications for strategies to lower cost growth. The organizationalstructure of phy- sician-led ACOs is more focused on outpatient care, and these ACOs may reduce the use of ser- vices provided in settings outside of the ACO, such as inpatient or postacute care.38,39 For exam- ple, in the Physician Group Practice Demonstra-

Exhibit 2

Care Management And Technology Capabilities Of Accountable Care Organizations, By Leadership Type, 2012–13

0 10 Percent of ACOs reporting

20

All primary care providers attest to meaningful use by 2013

Has advanced IT capabilities

Has fully developed program to assess and reduce hospital readmissions

Routinely assesses inappropriate use of the ED and uses these data to reduce use

Has systems in place to ensure smooth transitions across care settings

Actively engages in programs to reduce hospital admissions for ambulatory care–sensitive conditions

Has comprehensive chronic care management in place

Comprehensive previsit planning, medication management, and reminders for preventive care are conducted

30 40 50

All ACOs Physician-led ACOs Other ACOs

SOURCE Authors’ analysis of data from the National Survey of Accountable Care Organizations, October 2012–May 2013. NOTES For an explanation of “other ACOs,” see notes to Exhibit 1. “Meaningful use” of electronic health records is defined by the Health Information Technology for Economic and Clinical Health Act of 2009. Difference between physician-led ACOs and other ACOs is significant (p < 0:05) in the “advanced IT capabilities” and “comprehensive previsit planning” categories. ED is emergency department. IT is in- formation technology.

Exhibit 3

Capabilities Of Accountable Care Organizations Related To Monitoring And Reporting, By Leadership Type, 2012–13

0 20 Percent of ACOs reporting

40

Monitors systemwide quality performance and provides meaningful and timely feedback to clinicians

Monitors comprehensive and timely financial performance relative to benchmarks

Measures financial performance at the practice level

Measures financial performance at the clinician level

Reports financial performance at the practice level

Reports financial performance at the clinician level

60 80

All ACOs Physician-led ACOs Other ACOs

SOURCE Authors’ analysis of data from the National Survey of Accountable Care Organizations, October 2012–May 2013. NOTE Dif- ference between physician-led ACOs and other ACOs is significant (p < 0:05) in all cases except the “Monitors comprehensive and timely financial performance” and “Monitors systemwide quality performance” categories.

June 2014 33:6 Health Affairs 969

by Rachel McCartney on June 6, 2014Health Affairs by content.healthaffairs.orgDownloaded from

tion, a pilot of the ACO concept, one of the effects was to reduce spending on acute hospital care.33,40 This could have adverse consequences for patients if they did not receive necessary care. Physician-led organizations may be better suited to address the overuse of health care services than other ACOs, if physician-led ACOs do not own or receive revenue from providers of dis- cretionary services.41,42

Physician-led ACOs may struggle with chal- lenges that hospital-led ACOs could find easy to overcome. For example, to achieve coordina- tion across providers, ACOs require clinical, ad- ministrative, and technological cooperation not only among physicians but also with other pro- viders beyond the walls of typical physician prac- tices.10 Direct management of an ACO may create technical and professional challenges for physi- cians: ACOs require capital investment, assump- tion of financial risk, and the ability to distribute gains or losses. These tasks have historically proved difficult for physician groups to man- age.10 However, our research shows that physi- cian-led ACOs are investing in health ITcapabili- ties at or above the level of other ACOs.

Conclusion The broad reach of physician leadership has im- portant implications for the future evolution of ACOs. It seems likely that the challenge of fun- damentally changing care delivery as the country moves away from fee-for-service payment will not be accomplished without strong, effective leadership from physicians. The factors that con- tribute to successful ACOs are likely to be multi- ple. Previous work has highlighted a range of factors that are vital to success, including the complexity of the intervention; the commitment of all levels of management and front-line deci-

sion makers to implementing it; and the correct mix of skills, peer opinion leaders, incentives, and adequate communication among key deci- sion makers.35

Research from the managed care era showed that when physicians are expected to serve both their patients and the financial interest of their organization, physicians’ involvement in leader- ship is essential.15 The same research showed that physicians must work collaboratively with other groups to develop a balanced set of ac- countability approaches that can appropriately meet the interests of payers, accreditation bod- ies, patients, and community groups. It is possible that the physician-led model of

accountable care may face greater challenges than other ACOs in integrating care across di- verse health care providers, which is required for effective care coordination. It is also possible that physician-led ACOs may have a shorter reach of influence than other ACOs into some settings of care that are important to both the quality and the cost of care. Participating physicians may need to give up

some autonomy for an ACO to be successfully implemented. Encouraging the use of team- based care and standardized processes may be difficult for physician-led ACOs to the extent that physicians resist these changes. Further research is needed to see how practi-

ces evolve into ACOs, how ACOs attempt to achieve cost savings while maintaining and im- proving quality, and whether patient experience and overall population health is differentially improved across leadership styles and organiza- tional structures of ACOs. Organizations that are physician led may need to learn from ACOs that are led by a coalition or jointly by hospitals and physicians how to achieve the integration neces- sary to coordinate care across settings. ▪

The authors are grateful to Harold Sox for comments on the article. This research was supported by grants from the Commonwealth Fund. The funding source did not play a role in the design and conduct of the study or the preparation of the article.

NOTES

1 Lewis VA, Colla CH, Carluzzo KL, Kler SE, Fisher ES. Accountable care organizations in the United States: market and demographic factors associated with formation. Health Serv Res. 2013;48(6 Pt 1):1840–58.

2 DeVore S, Champion RW. Driving population health through account- able care organizations. Health Aff (Millwood). 2011;30(1):41–50.

3 National Study of Physician Organi- zations [home page on the Internet]. Berkeley (CA): University of Cali- fornia, Berkeley; [last updated 2012 Oct 9; cited 2014 Apr 17]. Available from: http://nspo.berkeley.edu/

4 Shortell SM. Bridging the divide between health and health care. JAMA. 2013;309(11):1121–2.

5 Institute of Medicine. Crossing the

quality chasm: a new health system for the 21st century. Washington (DC): National Academies Press; 2001.

6 Singer S, Shortell SM. Implementing accountable care organizations: ten potential mistakes and how to learn from them. JAMA. 2011;306(7): 758–9.

7 Jacobson PD, Jazowski SA. Physi-

ACOs & Medical Homes

970 Health Affairs June 2014 33:6

by Rachel McCartney on June 6, 2014Health Affairs by content.healthaffairs.orgDownloaded from

cians, the Affordable Care Act, and primary care: disruptive change or business as usual? J Gen Intern Med. 2011;26(8):934–7.

8 Crosson FJ. The accountable care organization: whatever its growing pains, the concept is too vitally im- portant to fail. Health Aff (Mill- wood). 2011;30(7):1250–5.

9 Berwick DM. Launching accountable care organizations—the proposed rule for the Medicare Shared Savings Program. N Engl J Med. 2011; 364(16):e32.

10 Kocher R, Sahni NR. Physicians versus hospitals as leaders of ac- countable care organizations. N Engl J Med. 2010;363(27):2579–82.

11 Reid RJ, Coleman K, Johnson EA, Fishman PA, Hsu C, Soman MP, et al. The Group Health medical home at year two: cost savings, higher patient satisfaction, and less burnout for providers. Health Aff (Millwood). 2010;29(5):835–43.

12 Rehm JL, Alexander JA. New trends in governing board–medical staff relations. Trustee. 1986;39(7):15–8.

13 Shortell SM. Effective hospital- physician relationships. Ann Arbor (MI): Health Administration Press Perspectives; 1991.

14 Weiner BJ, Shortell SM, Alexander J. Promoting clinical involvement in hospital quality improvement ef- forts: the effects of top management, board, and physician leadership. Health Serv Res. 1997;32(4): 491–510.

15 Shortell SM, Waters TM, Clarke KB, Budetti PP. Physicians as double agents: maintaining trust in an era of multiple accountabilities. JAMA. 1998;280(12):1102–8.

16 Addicott R, Shortell S. Collaborative governance through accountable care organizations: recommenda- tions for policy and practice. Health Care Manage Rev. Forthcoming.

17 Berenson RA. Shared Savings Pro- gram for accountable care organi- zations: a bridge to nowhere? Am J Manag Care. 2010;16(10):721–6.

18 Fisher ES, Shortell SM, Kreindler SA, Van Citters AD, Larson BK. A framework for evaluating the for- mation, implementation, and per- formance of accountable care or- ganizations. Health Aff (Millwood). 2012;31(11):2368–78.

19 Lewis VA, Larson BK, McClurg AB, Boswell RG, Fisher ES. The promise and peril of accountable care for vulnerable populations: a framework for overcoming obstacles. Health Aff

(Millwood). 2012;31(8):1777–85. 20 Muhlestein D, Croshaw A, Merrill T,

Peña C, Brent J. The accountable care paradigm: more than just managed care 2.0 [Internet]. Salt Lake City (UT): Leavitt Partners; 2013 [cited 2014 Apr 17]. Available from: http://leavittpartners.com/ wp-content/uploads/2013/08/ Accountable-Care-Paradigm.pdf

21 Noble DJ, Casalino LP. Can ac- countable care organizations im- prove population health? Should they try? JAMA. 2013;309(11): 1119–20.

22 Shortell SM, Casalino LP. Health care reform requires accountable care systems. JAMA. 2008;300(1): 95–7.

23 Shortell SM, Casalino LP, Fisher ES. How the Center for Medicare and Medicaid Innovation should test ac- countable care organizations. Health Aff (Millwood). 2010;29(7):1293–8.

24 Kreindler SA, Larson BK, Wu FM, Carluzzo KL, Gbemudu JN, Struthers A, et al. Interpretations of integra- tion in early accountable care or- ganizations. Milbank Q. 2012; 90(3):457–83.

25 Larson BK, Van Citters AD, Kreindler SA, Carluzzo KL, Gbemudu JN, Wu FM, et al. Insights from transfor- mations under way at four Brookings-Dartmouth accountable care organization pilot sites. Health Aff (Millwood). 2012;31(11): 2395–406.

26 Kreindler SA. The politics of patient- centred care. Health Expect. 2013 May 22. [Epub ahead of print].

27 American Hospital Association. Care coordination survey. Chicago (IL): AHA; 2012.

28 American Medical Group Associa- tion. Accountable care organization readiness assessment [Internet]. Alexandria (VA): AMGA; 2010 [cited 2014 Apr 17]. Available from: http:// www.amga.org/wcm/PI/Collabs/ ACO/assessTool.pdf

29 American Association for Public Opinion Research. Standard defini- tions: final dispositions of case codes and outcome rates for surveys [In- ternet]. Deerfield (IL): AAPOR; [re- vised 2011; cited 2014 Apr 17]. Available from: http://www.aapor .org/AM/Template.cfm?Section= Standard_Definitions2&Template=/ CM/ContentDisplay.cfm &ContentID=3156

30 Song Z, Lee TH. The era of delivery system reform begins. JAMA. 2013; 309(1):35–6.

31 To access the Appendix, click on the Appendix link in the box to the right of the article online.

32 Wilensky GR. Lessons from the Physician Group Practice Demon- stration—a sobering reflection. N Engl J Med. 2011;365(18):1659–61.

33 Colla CH, Wennberg DE, Meara E, Skinner JS, Gottlieb D, Lewis VA, et al. Spending differences associat- ed with the Medicare Physician Group Practice Demonstration. JAMA. 2012;308(10):1015–23.

34 CMS.gov [Internet]. Baltimore (MD): Centers for Medicare and Medicaid Services. Press release, Pioneer Accountable Care Organi- zations succeed in improving care, lowering costs; 2013 Jul 16 [cited 2014 Apr 17]. Available from: http:// www.cms.gov/Newsroom/Media ReleaseDatabase/Press-Releases/ 2013-Press-Releases-Items/2013-07- 16.html

35 Greenhalgh T, Robert G, Macfarlane F, Bate P, Kyriakidou O. Diffusion of innovations in service organiza- tions: systematic review and recom- mendations. Milbank Q. 2004; 82(4):581–629.

36 Levinson W, Gorawara-Bhat R, Dueck R, Egener B, Kao A, Kerr C, et al. Resolving disagreements in the patient-physician relationship: tools for improving communication in managed care. JAMA. 1999;282(15): 1477–83.

37 Fisher ES, Shortell SM. Accountable care organizations: accountable for what, to whom, and how. JAMA. 2010;304(15):1715–6.

38 Ackerly DC, Grabowski DC. Post- acute care reform—beyond the ACA. N Engl J Med. 2014;370(8):689–91.

39 Mechanic R. Post-acute care—the next frontier for controlling Medi- care spending. N Engl J Med. 2014;370(8):692–4.

40 Colla CH, Lewis VA, Gottlieb DJ, Fisher ES. Cancer spending and ac- countable care organizations: evi- dence from the Physician Group Practice Demonstration. Healthcare. 2013;1(3):100–7.

41 McWilliams JM. Accountable care organizations: a challenging oppor- tunity for primary care to demon- strate its value. J Gen Intern Med. 2013 Nov 20. [Epub ahead of print].

42 Anderson GF. Leadership in creating accountable care organizations. J Gen Intern Med. 2011;26(11): 1368–70.

June 2014 33:6 Health Affairs 971

by Rachel McCartney on June 6, 2014Health Affairs by content.healthaffairs.orgDownloaded from

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.