Abstract
Although safety net providers will benefit from health insurance expansions under the Affordable Care Act, they also face significant challenges in the postreform environment. Some have embraced the concept of the accountable care organization to help improve quality and efficiency while addressing financial shortfalls. The experience of Cambridge Health Alliance (CHA) in Massachusetts, where health care reform began six years ago, provides insight into the opportunities and challenges of this approach in the safety net. CHA's strategies include care redesign, financial realignment, workforce transformation, and development of external partnerships. Early results show some improvement in access, patient experience, quality, and utilization; however, the potential efficiencies will not eliminate CHA's current operating deficit. The patient population, payer mix, service mix, cost structure, and political requirements reduce the likelihood of financial sustainability without significant changes in these factors, increased public funding, or both. Thus the future of safety net institutions, regardless of payment and care redesign success, remains at risk.
Background
Beginning in 2014, millions of previously uninsured Americans will gain coverage under the Affordable Care Act (ACA) (Congressional Budget Office 2013; White House 2012). Although the law substantially expands access to care for low-income patients, it also creates challenges for safety net providers — hospitals and clinics whose principal mission is serving this population. First, as patients become insured they may seek care from other providers, reducing potential revenues for safety net providers. Second, the ACA substantially reduces Medicare and Medicaid disproportionate share payments — a major source of funding for safety net providers. Third, with continued state and federal budget pressures, future reductions in Medicaid — an important source of funding for the health care safety net — are likely.
The ACA also contains programs designed to encourage improvements in care delivery and financing. For example, it encourages the formation of accountable care organizations (ACOs), in which groups of participating providers are responsible for serving defined populations of patients and may share in savings associated with improvements in care quality and efficiency (Fisher and Shortell 2010). Accountable care models, built on advanced patient-centered medical homes, have potential for improving the way health care is delivered to low-income and vulnerable populations (Eyman and Luband 2011). Some states are actively promoting the formation of accountable care organizations to serve Medicaid populations, and most efforts are in the early stages of development (Kaiser Commission on Medicaid and the Uninsured 2012).
While new models like accountable care organizations and patient-centered medical homes may help safety net providers become more efficient, challenges remain in implementing these approaches (Conway and Terrell 2010). Most safety net providers have limited resources to support the investments in personnel and information technology that are needed to redesign care (Katz 2011). Additionally, these new models of care have largely been tested with Medicare and commercially insured populations and may not be as effective with vulnerable populations, where a high incidence of poverty, serious mental illness, multiple comorbidities, and adverse social determinants like unstable housing create additional challenges. Finally, because safety net organizations serve a high proportion of Medicaid patients, they require the active support of state governments to establish new financing arrangements that will support their transition to new delivery models.
To assess the opportunities and the challenges facing safety net providers in the establishment of accountable care models, we conducted a case study of the Cambridge Health Alliance (CHA), a public integrated delivery system in Massachusetts. CHA's experience is particularly relevant because Massachusetts passed state health care reform legislation similar to the Affordable Care Act in 2006 (Raymond 2010). CHA has some characteristics that seemingly make it a best-case example of the potential for a safety net provider to build an effective accountable care organization, including extensive primary care and secondary care services; a geographically dispersed network of employed primary care providers; a fully implemented electronic medical record; data analytic capability; and prior success with implementing innovative care redesign programs (Bielaszka-DuVernay 2011; Coughlin et al. 2012. The value of these assets are tempered, however, by CHA's payer mix, case mix, and cost structure, all of which limit the resources available for transformation.
This article examines the market, policy, and organizational factors behind CHA's decision to create an ACO and assesses its key strategies, areas of success, and major challenges that it has faced thus far. The case study was conducted in late 2012 and early 2013, when CHA was at a relatively early stage of implementation; therefore there is limited data to evaluate changes in performance. Nonetheless, its experience provides insights for other safety net systems considering broad financial and delivery system changes as well as for policy makers interested in supporting access to health care services for vulnerable populations.
Methods
This case study was conducted with a mixed-methods approach that included interviews, document review, and analysis of cost, quality, and utilization data from CHA to identify the principal strategies for its ACO development as well as current progress and lessons learned. Over forty interviews were conducted with individuals representing leadership, management, and line staff. Extensive document review that traced the different phases of CHA's transformational process was conducted and key themes extracted and analyzed. Finally, cost and quality data from CHA were analyzed to assess changes in performance over time.
Policy, Market, and Organizational Factors Affecting CHA's Decision to Establish an Accountable Care Organization
Massachusetts's 2006 health reform law required all Massachusetts residents to obtain health insurance coverage and provided subsidized insurance coverage to uninsured adults with incomes below 300 percent of the federal poverty level. By 2008, nearly 98 percent of Massachusetts residents had health insurance (Raymond 2011).
Prior to health care reform, Massachusetts supported safety net providers through an uncompensated care pool that pays for care provided to uninsured residents with incomes up to 200 percent of the federal poverty level. CHA and other safety net providers received generous payments from the pool as well as enhanced Medicaid rates and a variety of supplemental payments. The 2006 health reform law restructured the pool into a new Health Safety Net Fund, changed the payment formula, and reduced funding by roughly one-third, significantly reducing payments CHA received for eligible patients (Raymond 2011). In 2009, Massachusetts, along with many other states, enacted large Medicaid cuts in response to state budget shortfalls, forcing CHA and other providers to scale back services and lay off employees (Fennimore 2009).
Although the 2006 legislation did not address the growth in health care spending, policy makers feared that continued rapid growth in insurance premiums would threaten the sustainability of the coverage expansions. The state legislature subsequently passed three successive laws to control health care spending, culminating with a July 2012 bill known as Chapter 224. This bill established a statewide target for total health spending and created a new Health Policy Commission to set cost and quality targets and monitor systemwide and individual provider performance (Gosline and Rodman 2012).
The new law also encouraged payers and providers to move away from fee-for-service and expand their use of alternative payment models. The state's private payers have already aggressively pursued payment reform, and by 2012 about one in five Massachusetts residents was enrolled in an arrangement tied to a global budget (Song et al. 2012. The state Medicaid program has lagged behind other payers, but Chapter 224 requires that it enroll 80 percent of Medicaid recipients in alternative payment arrangements by 2015.
Recognizing the need to prepare safety net providers for the coming changes in Medicaid reimbursement, Massachusetts established a Delivery System Transformation Initiative (DSTI) as part of its 2012–14 Medicaid 1115 waiver. This program provides substantial incentive payments for successful completion of specific delivery reform initiatives to help safety net providers prepare to transition from fee-for-service to alternative payment arrangements (CMS 2012).
Cambridge Health Alliance and the Case for Transformation
CHA serves a diverse population of nearly one hundred thousand patients through its network of primary care clinics, in addition to operating a large psychiatry department, two inpatient hospitals, three emergency departments, and an array of specialty departments. Despite concerns about losing patients after health care reform, CHA's service volume actually increased as demand for primary care surged and existing patients remained loyal because of geographic access, availability of culturally competent services, and familiarity with the system (Ku et al. 2011.
Like many safety net systems, CHA has a complex financial structure and faces substantial fiscal pressures. CHA's delivery system generated operating losses in eight of the past ten years, with the magnitude of losses growing after the 2006 reforms (table 1). Through 2012, CHA offset delivery system losses with earnings generated by its wholly owned insurance plan. However, it was forced to sell the plan when the state established new capital requirements that CHA could not meet. While proceeds from the sale have provided CHA with some flexibility, it must now rapidly narrow operating losses to remain financially sustainable.
Structural challenges to CHA's financial sustainability include its payer mix, service mix, and cost structure. CHA relies heavily on state and federal financial support (fig. 1). CHA's service portfolio is concentrated in medicine (particularly primary care) and psychiatry, which are poorly reimbursed compared with surgical and diagnostic services. The combination of its relatively modest commercial volume and lack of tertiary care gives CHA very limited negotiating clout with private payers (Center for Healthcare Information and Analysis 2013). Finally, it operates three small hospital campuses with twenty-four-hour emergency departments, creating high fixed overhead costs.
The challenging fiscal outlook and growing emphasis on payment reform in Massachusetts created an imperative for CHA to change both its delivery model and its financial arrangements with payers. CHA leaders believed that this was essential for continuing to effectively serve vulnerable populations while remaining financially solvent.
CHA determined that alternative payment arrangements would provide more flexibility to optimally manage patient care while controlling overall spending. Global payment would also allow CHA to capture and reinvest the gains of innovative programs like its childhood asthma program, which has reduced hospitalizations and emergency visits by 50 percent since 2002 through optimizing medication management (Bielaszka-DuVernay 2011). Beginning in 2009, CHA began a wide-ranging process to restructure its contracts, redesign its delivery model, and engage its workforce in becoming a more patient-centered organization.
Strategies for Transformation
CHA's principal strategies for transformation were to redesign care, realign payer contracts, engage its workforce in the transformation process, and pursue a tertiary care partner that would help build stronger specialty care capacity. These strategies, which are consistent with those of other leading safety net institutions (Coughlin et al. 2012 are described in greater detail elsewhere and summarized below (Hacker, Mechanic, and Santos, forthcoming).
Care Redesign
CHA's care transformation strategy revolves around establishing patient-centered medical homes (PCMH) at its twelve primary care clinics. The core of this work is the establishment of care teams that serve defined panels of patients with an emphasis on improving the health of the entire panel rather than focusing primarily on the acute needs of individual patients. Care teams are also striving to improve transitions of care from inpatient to outpatient settings. Using electronic medical record–generated patient lists, nurses now try to contact all patients who have been discharged from the hospital or emergency department within forty-eight hours to review medications and discharge instructions, and facilitate timely primary care appointments.
Two of CHA's primary care clinics achieved National Center for Quality Assurance (NCQA) Level III recognition in 2010 and have functioned as pilot sites for developing and spreading process improvements. CHA has empowered a vice president of Patient-Centered Medical Home Development to oversee practice reforms across the sites. It has established a multidisciplinary committee to analyze existing processes and establish standard workflows, and also holds monthly meetings with ten practice improvement teams across the system to share best practices. In 2013, four more of CHA's primary care clinics earned NCQA Level III recognition, with plans to achieve recognition at all clinics by 2015.
More than 20 percent of CHA's primary care patients have serious and persistent mental illness and multiple comorbidities, including substance abuse and addiction. Therefore CHA's care redesign work has emphasized the integration of primary care and mental health services. CHA has adopted a universal depression screening and a management protocol for newly diagnosed patients. Two of the PCMH sites are now implementing a more fully integrated care model for mental health and substance abuse disorders based in part on the IMPACT (Improving Mood-Promoting Access to Collaborative Treatment) evidence-based depression care model (Unützer et al. 2002; IMPACT 2012). Both sites have colocated psychiatric and other mental health providers with the primary care teams, and together they share care plans to best address patient needs.
CHA has also established a complex care management team to coordinate care for patients with spending in the top 3 percent of the population. This is a centralized team with care managers who work with primary care teams and patients to devise and implement comprehensive care plans. A major focus of this team is connecting patients to primary care and community resources like area agencies on aging, supportive housing, and other social services. CHA is also building a program of embedded complex care management teams composed of a registered nurse and a social worker in each primary care clinic.
Financial Realignment
CHA began its financial realignment in 2010 when it entered a global budget arrangement with shared risk for gains and losses for about fourteen thousand Medicaid managed care enrollees. By the end of 2013 CHA expects to have global risk or shared savings contracts covering about half of its primary care patients, including six thousand seniors in the Medicare shared savings program. Its objective is to move the bulk of its primary care patients into global payment by 2015 one insurer at a time. CHA also redesigned its physician compensation to better align with these new contracts by reducing incentives related to volume and increasing incentives for quality and population management.
Workforce Transformation
CHA has prioritized workforce training to encourage culture change and develop competencies that are needed for population management. CHA established a leadership development program that has now trained about 10 percent of senior and middle managers. Most of its primary care teams participate in patient-centered medical home learning collaboratives. Given the limited resources for wholesale workforce education, the learning collaboratives provide opportunities for staff to learn from peers in other organizations pursuing similar goals. CHA has modified its primary care clinic job responsibilities to meet the demands of team-based care. Leadership has actively communicated the goals of transformation and highlighted key accomplishments such as advances in patient experience scores.
External Partnerships
CHA sought a collaborative relationship with a tertiary care hospital to serve as a preferred referral partner and to assist CHA in developing specialty services in the neighborhoods where its patients live. CHA recently formalized a partnership with the Beth Israel Deaconess Medical Center, which will improve access to specialty care for CHA's patient population and enhance coordination between outside specialists and CHA's primary care.
Early Results
Within three years of beginning its transformation process, CHA had already achieved some improvement in quality and utilization. Its NCQA Level III recognized sites had achieved better ambulatory quality, access, and patient experience scores than its nonrecognized clinics (table 2), and inpatient utilization had declined in at least one Medicaid managed care population (fig. 2). These successes are major accomplishments in the face of the myriad financial and other obstacles that the institution has faced.
Challenges for Safety Net Providers
Despite early successes, CHA faces many challenges. Some derive from its role as a safety net provider, while others relate to the inherent difficulty of managing organizational change.
Financial Sustainability and the Marketplace
CHA faces the immediate financial challenge of reversing its large operating deficit. Its options for revenue growth are constrained by its reliance on government payers that are under immense pressure to restrain spending, and by its limited commercial market clout. Expanding primary care services into new geographies may be an option, but requires capital and an ability to recruit new physicians willing to care for safety net populations. Two strategies are promising for improving financial sustainability; effective population health management combined with its shift toward global payments and CHA's new partnership with the Beth Israel Deaconness Medical Center. But neither strategy is a panacea. Global rates will likely be based on its current fee structures, and CHA's ability to reconfigure rapidly for greater efficiency is affected by its highly unionized workforce. In addition, CHA is committed to its mission of providing culturally competent care to vulnerable populations, which may make it less attractive to commercially insured patients.
Defining the Role of the Hospital within an Accountable Care Organization
CHA's primary care network is a critical asset for an accountable care organization, and several respondents suggested that given the high cost of its hospitals, CHA consider divesting from them, reorganizing as a globally capitated ambulatory network, and referring more patients to low-cost community hospitals. This is unlikely for several reasons. First, its hospital operation still accounts for 84 percent of the system's patient revenue. Second, state supplemental payments are tied to its status as a Medicaid disproportionate share hospital, and its primary care clinics receive enhanced fees because Medicaid classifies them as hospital outpatient departments. Third, CHA's interoperable electronic medical record links its hospital and ambulatory sites, creating improved communication and coordination of care that would be difficult to replicate at the same level with external hospitals. Finally, CHA hospitals provide essential access to care in communities that would strenuously resist large cutbacks.
CHA faces a dilemma common to many accountable care organizations: should it be an integrated primary care network with an aligned hospital that it manages as a cost center, or a hospital system that draws patients from its primary care network? CHA's financial structure makes it difficult for managers not to view the hospitals as revenue centers; this is true for many safety net systems where disproportionate share funding is tied to hospital volume. Without its hospital, CHA's ambulatory network would lose access to supplemental funding and face a large decline in outpatient fees. Similarly, CHA's hospitals probably could not survive if referrals from its primary care network were significantly diminished. Without public policies that allow it to restructure without major losses of supplemental revenues, CHA will be unable to fully commit to managing its hospitals as cost centers.
Building a Culture of Action and Accountability
Like many mission-driven systems, CHA's culture has historically supported flexibility and innovation. However, many staff stated that this culture also results in slow decision making, conflicting priorities, and a lack of clarity about who is ultimately responsible for implementing changes or achieving specific goals. Some primary care site leaders noted conflicting signals from different parts of the organization regarding whether they should manage based on a fee-for-service or global payment environment, and thus whether they should focus on generating volume or on redesigning the care model. Since the move into global payment contracts is negotiated one payer at a time, CHA providers will continue to face conflicting financial incentives for the near future. This cannot be avoided but must be managed effectively. CHA has recently made changes to its ambulatory leadership structure to address this issue and is reassessing management throughout the organization.
Managing the Conflicting Demands of Multiple Initiatives
Given resource limits, CHA has aggressively pursued a range of external supports for its care redesign efforts. It participates in numerous collaboratives, federal programs like Medicare's Shared Savings and Community Care Transitions Programs, and the state's delivery system transformation waiver. These efforts require substantial staff resources and often have extensive reporting requirements. Simultaneously, it has invested in leadership development to empower providers in the transformation process. Nonetheless, many primary care practitioners and their staff feel overwhelmed in dealing with the current challenges. This is not unique to CHA, but the stress is heightened by the complex medical and social characteristics of CHA's patient population, the organization's resource limitations, and the imperative to change quickly. Despite challenges, physicians and staff are encouraged by the changes because they believe that the objective is to make CHA a better place for both its patients and its employees. As one physician told us, “We are redesigning the engine as we're flying the plane,” and that is not easy.
The Provider of Last Resort
As “providers of last resort,” safety net systems serve populations with the greatest social needs. Caring for these patients requires a unique set of skills and resources, including specialized staff such as care coordinators/outreach personnel and multilingual interpreters (Lewis et al. 2012. In the current environment providers need flexibility to manage spending, but safety net providers are constrained by their mission and by external politics. In April 2013, for example, CHA proposed eliminating eleven of its twenty-seven inpatient mental health beds for children and teens as a cost-cutting measure, but reversed the decision after the state public health department protested and the legislature came up with temporary additional funding (Conaboy 2013). Thus while health care reform requires providers to redesign care to reduce cost, safety net organizations face a dilemma. They are expected to be financially accountable while being required to maintain services that are critical for the public's health, regardless of economic viability. Their business decisions are intimately connected to the greater public good and therefore prone to public opinion and need.
Discussion
The Cambridge Health Alliance's experience provides some insight into the opportunities and challenges facing safety net providers under national health care reform. There are several questions that will ultimately determine whether CHA and other safety net providers can be successful in the postreform environment. First, can safety net providers successfully implement patient-centered medical homes and accountable care organization structures? This will depend both on internal capabilities — infrastructure, leadership, and ability to implement needed organizational changes — and on access to sufficient funding. Second, will these new models be able to meaningfully reduce spending for safety net populations while maintaining quality and access to services? Third, can safety net providers effectively navigate the competitive market either because of unique capabilities or in partnership with other organizations? Finally, can safety net providers meet their financial challenges while still maintaining their public mission? In an era of shrinking public funding, the future size and scope of the health care safety net will depend on the answers to these questions.
The establishment of an ACO built on a PCMH foundation is central to CHA's strategy. It has been relatively successful putting these models in place, although their long-term efficacy remains uncertain. Its success to date has been enhanced by its existing electronic health record infrastructure and prior work on process improvement initiatives. These efforts are further supported by the state's DSTI waiver, which will provide CHA up to $22.4 million in annual funding for three years. The Center for Medicare and Medicaid Services has granted similar waivers to several other states including California. Such targeted infrastructure funding is likely to be essential for many safety net organizations that lack reserves or sufficient operating income to fund delivery system restructuring.
But despite a relatively rapid implementation and modest improvements in spending and quality measures, these changes are unlikely to shrink CHA's operating deficit on the scale that is necessary. The many challenges facing CHA including case mix, service mix, and payer mix along with shrinking supplemental payments and continued uninsured populations (undocumented immigrants) make it likely that large-scale cost-cutting measures are inevitable. This will create pressure to scale back money-losing services like mental health care, a strategy that may be politically untenable.
CHA has chosen to affiliate with an academic medical center. This relationship has the potential to help it develop new programs and build infrastructure that will support more effective operations. However, many safety net institutions find that their payer mix, cost structure, and high degree of unionization give them limited appeal as partners for private health systems. CHA's new affiliation offers many potential benefits, but these have yet to be proved. Ultimately the resources that its academic partner will be willing to devote to CHA depend on its own fortunes in an increasingly competitive environment.
Safety net organizations are the providers of last resort, and public funding has helped them maintain this mission. Some believe that national health care reform reduces the need for safety net providers, but in fact the demand for safety net services is likely to grow (Coughlin et al. 2012. For example, in Massachusetts, emergency room visits increased after its health care coverage expansion (Kowalczyk 2010), and safety net providers continued to see the same patient populations (Ku et al. 2011. Thus while care redesign will help safety net providers improve quality and efficiency, their ability to meet their public mission will depend on public policies that support their long-term financial viability. We believe such policies are consistent with the objective of health care reform to create a more just and effective health care system (Bachrach, Braslow, and Carl 2012).
We would like to thank all the Cambridge Health Alliance staff and leaders who participated in interviews. We would also like to thank Kat Xu, MPH, for her assistance with the project.
