Health Care Reform and Care at the Behavioral Health—Primary Care Interface

The 2010 Patient Protection and Affordable Care Act has the potential to effect a major transformation in how health care is delivered in the United States. These changes will be driven, in part, by a series of demonstration projects and initiatives for reorienting health services to increase provider accountability and strengthen the role of primary care. If successful, these new programs are likely to be expanded within Medicaid and Medicare and eventually become a model for care delivery throughout the health care system.

This article describes how two such demonstration programs, one addressing primary care in the general medical sector and the other supporting improved primary care in specialty mental health settings, might lay the groundwork for improvements in care at the primary care-behavioral health interface. We begin with a brief overview of the clinical processes and organizational strategies that have been demonstrated to improve care at this interface, provide an overview of the two demonstration programs, and then consider supporting factors—new financing models, better quality indicators, enhanced health information technology, and technical support for local sites—that could help ensure that these new initiatives translate into improved care.

Models for improving care

To improve outcomes, health reform initiatives need to support the evidence-based clinical processes that have been documented as improving clinical outcomes. Multiple randomized controlled trials have found that team-based interventions improve quality of care for and outcomes of common mental health and substance use disorders in primary care ( 1 , 2 ) and the delivery of primary medical care in specialty behavioral settings ( 3 ). A recent Agency for Healthcare Research and Quality synthesis found that integration, defined as sharing of treatment decision making and the colocation of primary care and mental health specialists, was not in and of itself predictive of improved outcomes but that together, the elements in these models consistently resulted in improved quality and outcomes of care ( 4 ). Within the broader array of services delivered in these models, key "active ingredients" that would need to be supported include systematic screening and use of qualified care managers ( 5 ).

These clinical approaches can be delivered through a variety of organizational and structural relationships, including colocation of services, referral approaches, and partnerships between general health care providers and mental health and substance abuse treatment providers. No particular organizational approach guarantees or precludes these process elements of care. However, clinical integration is generally easier to support in structured organizational models than in more loosely organized referral relationships. The 2006 Institute of Medicine report ( 6 ) on improving the quality of care for mental health and substance use conditions recommended that sites should "transition along a continuum of evidence-based coordination models … adopt[ing] models to which they can most easily transition from their current structure, that best meet the needs of their patient populations, and that ensure accountability." The demonstration projects outlined below and other elements of health reform have the potential to promote a spectrum of organized models of care—and with them the opportunity to support evidence-based clinical models of care improvement.

Overview of demonstration projects

Medical home initiatives

Despite the growing number of individuals treated for common mental disorders in primary care, a considerable literature has demonstrated continued quality deficits in those settings ( 4 , 7 ). These gaps, in part, reflect problems in the broader primary care system. Most medical practices in the United States still do not have the infrastructure or capacity to implement evidence-based, organized approaches to care delivery ( 8 ). Thus improving the structure of the primary care system overall could have considerable benefits for the treatment of mental and substance use disorders in general medical settings in the United States ( 9 ).

Among the most promising strategies for revitalizing and redefining the primary care system is the patient-centered medical home ( 10 ). This model, which was originally developed for children with chronic illnesses in the 1960s ( 11 ), was reconfigured in recent years by major purchasers, health plans, and primary care organizations, working as the Patient-Centered Primary Care Collaborative ( 12 ). The model draws on Wagner's ( 13 ) chronic care model, which describes the environmental, structural, and community characteristics needed for multidisciplinary teams to work with patients in improving illness management. Medicare, large health plans, and state Medicaid agencies are currently conducting demonstration projects to test new payment methods (a combination of fee-for-service payments, monthly care management fees, and bonuses) on quality and costs of the patient-centered medical home model ( 14 ). Some of these demonstration projects explicitly include mental health and substance use conditions. For instance, the State of Oregon has recently adopted standards and measures for patient-centered primary care homes that include the following measure under the standard for care coordination: "When I need to see a specialist or get a test, including help for mental health or substance use problems, help me get what I need at your clinic whenever possible and stay involved when I get care in other places"( 15 ).

The Patient Protection and Affordable Care Act includes provisions for patient-centered medical home projects within both Medicare and Medicaid. Within Medicare, these programs will be implemented in the new Center for Medicare and Medicaid Innovation, which will test innovative payment and service delivery models designed to reduce expenditures while preserving or enhancing the quality of care ( 16 ). For Medicaid enrollees, the legislation proposes a new state plan option to permit enrollees with at least two chronic conditions, or at least one serious and persistent mental health condition, to designate a provider as a health home. States are expected to design and implement care models, track costs and avoidable hospitalizations, implement information technology, and monitor and report on quality and outcomes of care.

Primary care colocation grants

Persons with serious mental disorders treated in the specialty mental health sector face challenges in accessing appropriate primary medical services ( 17 ). This poor quality of care may, in part, contribute to excess rates of medical morbidity and mortality among persons with serious mental disorders ( 18 ). For this population, "specialty medical homes," located in community mental health settings, may provide a strategy for delivering integrated, high-quality care ( 19 ).

In 2009, in response to growing concerns about the problem of morbidity and mortality among mental health consumers, the Substance Abuse and Mental Health Services Administration issued the first set of awards for a new grant program to provide community mental health organizations with funding to provide primary care services and wellness and prevention services to their clients, either directly or via partnerships. A total of 13 sites were funded in 2009; these sites are using a variety of strategies, including colocation of services and partnership models to improve primary care for their clients. Eight more sites are slated to be funded in 2010.

Under health reform, Congress will expand this program considerably, with $50 million in funding for the current fiscal year. Although this grant program will provide funding only for a small proportion of community mental health providers in the United States, it will make it possible to identify and understand a series of best practices for specialty mental health homes that can subsequently be implemented more broadly. The program evaluation will provide data about the implementation, clinical outcomes, and sustainability of these programs in real-world community settings.

Key elements needed to ensure success of these initiatives

An initial evaluation of the National Demonstration Project, a patient-centered medical home project sponsored by the American Academy of Family Physicians, recommended several supporting elements that would be essential for successfully implementing future medical home projects ( 20 ). These elements included establishment of appropriate financing models, development of appropriate quality and accreditation metrics, adaptation of health information technologies, and implementation of appropriate technical support. These echo key components described as essential for supporting quality improvement efforts in general medical populations ( 21 ) and for persons with mental or substance use disorders ( 6 ). In the section below, we discuss how each of these features is important for the demonstration projects to be successful in improving care at the primary care-behavioral health interface.

Implementing new financing models

Accountable care organizations. The Patient Protection and Affordable Care Act has provisions for organizing hospitals, specialists, and primary care providers as accountable care organizations—collectives of providers that would take responsibility for a group of patients. Under most accountable care organization models, providers are paid bonuses based on their ability to meet quality goals and contribute to reduced costs.

Psychiatrists, like other specialists, view the possibility of joining accountable care organizations with some caution, given uncertainty about who will oversee them (for example, hospitals or primary care practices) and concerns over possible loss of revenue compared with current fee-for-service payment schemes ( 22 , 23 ). However, membership could also support development of the new service models, new financing models, and the measurement and quality improvement infrastructure, which has been difficult to achieve in the current system. They could provide the opportunity for mental health and substance abuse treatment providers to integrate vertically with other components of the health care system, contribute to achieving cost and quality targets, and share in the payment methods being discussed in relationship to accountable care organizations (such as fee-for-service plus shared savings, episode or case rates, and pay for performance).

Accountable care organizations and patient-centered medical homes can be mutually reinforcing, with accountable care organizations providing an organizational environment to support patient-centered medical homes and patient-centered medical homes allowing accountable care organizations to optimize quality and efficiency of care ( 24 ). They could provide economies of scale for solo practitioners as well as community-based mental health and substance abuse treatment providers, allowing them to develop virtual patient-centered medical homes ( 25 ). Accountable care organizations would not guarantee integration in and of themselves, but they could provide a structure in which integrated models could be supported and incentives for integration provided.

Both similarities and differences exist between these approaches and the 1990s managed care experience, and lessons learned from those experiments should be applied to these new models. During the 1990s, managed behavioral health care was largely operated separately from general health insurance managed care programs, an arrangement that provided expertise in managing mental health care but raised potential challenges in coordination with general medical care ( 26 ).

In contrast, accountable care organizations would include persons with general medical conditions and those with mental health conditions in the same risk pools. Thus, although these organizations could provide incentives for better coordination of care, they might also divert resources away from populations with mental disorders and other complex comorbid conditions. Because of the high costs in the Medicaid program associated with comorbid mental health and substance use disorders ( 27 ), these populations could become targets of cost savings for accountable care organizations, as they have been under Medicaid disease management programs. More generally, pay-for-performance approaches should be applied with caution to mental health and substance use conditions, pending better indicators, risk adjustment models, and capacity to establish accountability across multiple providers and systems of care ( 28 ).

State financing innovations. Current state initiatives may also provide models for these organizational and financing approaches to supporting improved care at the primary care-behavioral health interface. In the Community Care of North Carolina (CCNC) project, Medicaid enrollees receive health care and care management through local networks made up of physicians, hospitals, social service agencies, and county health departments. Preliminary evidence suggests that these programs may help improve quality of care for chronic medical illnesses and save costs ( 29 ). The CCNC project is a primary care case management model that could be used as a prototype for accountable care organizations under health reform.

Although the CCNC itself was not designed as an integration initiative, in the past several years four CCNC networks have worked with state and regional mental health authorities to pilot a model for integrating mental health and primary care. Recently, the CCNC system began a gain-sharing demonstration with Medicare, designed to better serve persons dually eligible for Medicare and Medicaid. In the demonstration, the CCNC networks will expand current care coordination efforts for the Medicaid population to dually eligible persons and, over time, to the Medicare-only population as well. The CCNC networks will receive a per-member-per-month fee to cover care management, care transitions, and colocation of mental health services. Medicare savings beyond an established threshold will be shared with the networks and reinvested ( 30 ). Planned expansion of integrated services through CCNC-employed mental health and substance abuse treatment staff may further assist primary care practitioners in meeting the expectations for medical home management of chronic health conditions, including mental health and substance use conditions.

A key financing approach for the patient-centered medical home is a monthly care management fee paid per enrollee per month in addition to fee for service. There is an opportunity to build on this idea by combining it with a unique financing model for integrated care now under way in Minnesota. More than 90 clinics have participated in an initiative known as DIAMOND (Depression Improvement Across Minnesota, Offering a New Direction), based on the IMPACT model (Improving Mood—Promoting Access to Collaborative Treatment) ( 31 ). By providing an organizational and financial framework to support this evidence-based approach to depression management, the DIAMOND program has been able to demonstrate initial outcomes that are superior to usual depression treatment given to patients in primary care.

Behind the clinical statistics, the DIAMOND project is applying the concept of an all-payer case rate for depression care. Minnesota health plans are paying a monthly per-person case rate to participating clinics for a bundle of services—including a depression care manager and consulting psychiatrist—under a single case rate billing code. For some of the participating plans in Minnesota, the case rate payments are being made from the health care side of the plan, rather than the mental health side, so that any cost savings can accrue to the health plans ( 32 ). Combining the DIAMOND payment model with the patient-centered medical home care management monthly fee could facilitate the adoption of collaborative care models for common mental disorders in primary care.

Standardization of billing codes. Other changes under health reform may provide greater financial viability for integrated models of care and evidence-based strategies for quality improvement. The legislation promises to bring more standardization to Medicaid (for example, eligibility thresholds, essential benefits, and minimum payment rates to primary care providers), and it could include a requirement for state plans to incorporate the current CPT (current procedural terminology) codes that support integration (for example, Health and Behavior 96150 series and Screening and Brief Intervention 99408 and 99409) and eliminate frequently described barriers to billing (such as same-day billing prohibitions). Medicaid and Medicare demonstration projects should provide a setting in which to assess the practicality and use of these changes and the more widespread use of bundling models such as those used in the DIAMOND project, which could be valuable for improving quality and increasing incentives for coordination of care across providers.

Quality metrics

Broadening the range of quality measures. Rigorous quality assessment standards are essential for the successful implementation and evaluation of demonstration projects and other changes occurring under health reform. However, quality metrics for mental health and substance use disorders are generally more limited than those for other chronic conditions ( 7 ). The National Committee for Quality Assurance is seeking to expand its quality indicators for mental health and substance use conditions. Implementing quality measures for serious mental illnesses is of particular importance for evaluating Medicaid programs and other public-sector entities under health reform.

As demonstration projects and broader reform efforts move forward, it will be important to develop and measure indicators not only for individual general medical and mental health conditions but also for the key processes associated with clinical integration—effective communication (transfer of information across providers), coordination (shared understanding of goals and roles), and continuity of care (uninterrupted delivery of services across levels of care) ( 33 ). However, there are no validated measures of coordination or clinical integration that can be used for assessing quality of care of persons with mental and substance use disorders ( 34 ). Demonstration projects for patient-centered medical homes and accountable care organizations could provide a laboratory in which to develop and test candidate measures of clinical integration that could subsequently be included in efforts to implement these models more widely.

Other quality assessment organizations will also need to be engaged in these quality assessment and improvement efforts. The Physician Quality Reporting Initiative was established in 2007 to assess quality of care among physicians; it provides incentive payments to physicians for reporting data quality measures for Medicare beneficiaries ("pay for reporting") ( 35 ). Physicians can receive a bonus payment of 2% based on their total Medicare Part B payments if they select at least three quality measures and report data for those measures on at least 80% of applicable patient encounters. However, mental health has limited representation in these measures; of 179 indicators, only four are related to mental health (depression screening, evaluation, suicide assessment, and acute medication treatment).

The National Quality Forum, which collects and certifies quality measures from a range of sources, is working on a consensus development project funded by the Department of Health and Human Services to develop a more robust set of outpatient indicators for mental health, including serious and persistent mental illnesses ( www.qualityforum.org ). Candidate measures include management of common medical comorbidities, preventive medical services, and enhanced clinical outcomes of medical illnesses, as well as measures of coordination, such as documentation of communication by an outpatient mental health clinician to the patient's primary care clinician ( 36 ). In 2007 the National Quality Forum issued a set of evidence-based practices for the treatment of substance use conditions and is working on approaches to measuring continuing care management for those conditions. These mental health and substance use measures can be used as potential candidates for development and specification by the National Committee for Quality Assurance.

Expansion of accreditation and certification programs. The National Committee for Quality Assurance should also be supported in expanding its accreditation and certification programs to include more robust quality measures. New draft certification standards for the patient-centered medical home include references to integration of mental health and substance use screening and brief treatment. The managed behavioral health organization accreditation process needs to be strengthened to incorporate expanded quality indicators, including measures of coordination with general health care.

Supporting health information technology

Health information technology is a central feature facilitating quality improvement and better integration of services ( 37 ). In its patient-centered medical home certification standards, the National Committee for Quality Assurance includes multiple information technology features, including patient tracking and registries, electronic prescribing, and test tracking. However, mental health and substance abuse treatment systems have historically lagged behind other areas of medicine in the development and standardization of these information technology tools. Furthermore, regulatory barriers have limited the exchange of information between primary care and mental health and substance abuse treatment settings.

The Patient Protection and Affordable Care Act explicitly requires that information technology be a part of medical home demonstration projects, and it will also be critical in facilitating the success of integration efforts. In developing these technologies, standardized templates for electronic medical records and personal health records should include the data elements needed to manage and coordinate general medical care and mental health and substance abuse care. These systems need to be carefully designed to ensure that critical information on health status and services can be extracted for measuring service patterns and performance.

The Health Information Technology for Economic and Clinical Health Act authorizes roughly $36 billion for health information technology. Most of the funds are expected to be distributed between 2011 and 2016 as adoption incentives through Medicare and Medicaid to qualified health care providers who adopt and use electronic medical records in accordance with the act's requirements. Although mental health and substance abuse treatment providers are not eligible for these funds, legislation has recently been introduced to include them as qualified health care providers in this grant program ( 38 ).

Building capacity for technical assistance

Implementing these demonstration projects will be complex, and states and local sites will require considerable technical support if these projects are to be successful. An initial evaluation of the National Demonstration Project, a two-year patient-centered medical home practice transformation project sponsored by the American Academy of Family Physicians, described the challenges in transforming the organizational cultures and physician practice patterns in the 36 participating sites ( 20 ). The authors described this process as a highly local developmental one requiring both top-down leadership and bottom-up engagement with physicians and other clinicians. To help current primary care practices successfully transition to medical homes, they recommended that technical support be tailored to characteristics of practices and organizational readiness.

For the new demonstration projects, technical assistance roles will similarly require a grounding in evidence-based approaches to integration practices along with a knowledge of how these clinical models work in local settings. It will require maintaining an inventory of evidence-based approaches to integrated care and to measurement and quality improvement and developing and disseminating standardized templates for electronic health records, personal health records, and the registry. Expertise will be needed not only from content experts and researchers but also from quality improvement organizations with experience in driving large-scale practice change. Practice management experts will need to work with sites in demonstration projects to make these programs financially sustainable after grants end and with other sites that do not have specific funding to underwrite quality improvement efforts.

Conclusions

These two demonstration projects—patient-centered medical home demonstration projects and expansion of the Substance Abuse and Mental Health Services Administration primary care project in community mental health centers—offer considerable potential to improve care at the primary care-behavioral health interface. Given the complexity of the problems underlying poor-quality care in safety-net settings, the success of these efforts will hinge on the ability of clinicians, managers, and policy makers from various agencies to work across traditional organizational boundaries. In anticipation of the implementation of health reform legislation, a number of these collaborations have already begun. The Health Resources and Services Administration will be cofunding the technical assistance center for primary care colocation grants with the Substance Abuse and Mental Health Services Administration. The Assistant Secretary for Planning and Evaluation is cosponsoring the evaluation of that program, along with the Substance Abuse and Mental Health Services Administration, and working with the National Committee for Quality Assurance to develop new quality metrics that can be used in evaluating the impact of health reform. In addition, the Agency for Healthcare Research and Quality and the National Institute of Mental Health are working together in setting a research agenda for mental health information technology and comparative effectiveness research that will further inform these health reform efforts.

As these interagency collaborations move forward, many of the same elements demonstrated to be essential for improving clinical quality of care—a clear locus of accountability, long-term follow-up, effective communication, and rigorous monitoring and feedback—will also be essential to ensure that these demonstration projects, and health reform more generally, fulfill their potential to improve care at the primary care-behavioral health interface.

Acknowledgments and disclosures

The authors received financial support for this work from Mathematica Policy Research, Inc., and the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services.

The authors report no competing interests.