Strategies for Integrating and Coordinating Care for Behavioral Health Populations: Case Studies of Four States

This report describes the different strategies that four states--Louisiana, North Carolina, Tennessee and Vermont--are utilizing to improve the integration and coordination of care for individuals with mental health and substance use disorders. These strategies include introducing new managed care arrangements and care coordination mechanisms, enhanced payment to providers to deliver comprehensive services, and the full integration of physical and behavioral health services in managed care contracts. [87 PDF pages]

"

Abstract

Several states are implementing strategies to overcome the historical lack of alignment in the financing and delivery of physical health, behavioral health, and supportive services for individuals with behavioral health conditions. These strategies include introducing care coordination mechanisms, enhanced payment to providers to deliver comprehensive services, and the full integration of physical and behavioral health services in managed care contracts. This project conducted an environmental scan and discussions with state officials and other stakeholders in Louisiana, North Carolina, Tennessee and Vermont to describe each state's approach to integrating and coordinating care for behavioral health populations. The case studies describe each program's goals, financing, covered services, care coordination and integration mechanisms, information systems and data collection infrastructure, and quality monitoring practices. The case studies also provide some insights into implementation successes and challenges and identify opportunities for further research.

Acronyms

The following acronyms are mentioned in this report and/or appendix.

Executive Summary

Individuals with serious mental illnesses and other chronic behavioral health conditions need access to a comprehensive array of physical health, behavioral health, and other supportive services. Yet few of these individuals receive this type of care. Recent research suggests that less than 5 percent of Medicaid beneficiaries with schizophrenia or bipolar disorder receive continuous medications, minimal medication monitoring, psychosocial services, and preventive physical health care (Brown et al. 2012).

Fragmentation in the financing and delivery of services leads to gaps in the quality of care. Some states, however, are implementing programs intended to better integrate and coordinate the delivery of comprehensive services for individuals with behavioral health conditions through a variety of financing mechanisms and delivery models. For instance, some states are integrating behavioral and physical health benefits into their Medicaid managed care contracts or are incorporating other supportive services or care coordination efforts into these arrangements (Greenberg 2012; Kim et al. 2012; Hamblin, Verdier, and Au 2011). Other states are adopting enhanced primary care case management (PCCM) programs that provide payments to providers and other incentives and tools to ensure that individuals receive comprehensive services. More information on how states are financing and organizing the delivery of services will help federal and state policymakers in their efforts to improve the integration and coordination of care.

Under contract to the Office of the Assistant Secretary for Planning and Evaluation, Mathematica Policy Research conducted case studies of four state programs in which different strategies are being used to improve the integration and coordination of care for adults with behavioral health conditions. Data for the studies came from an environmental scan and discussions with state officials and other stakeholders. The four study states and their programs are:

• Louisiana: Several state agencies pool funding into a contract with one managed care organization (MCO) to manage the delivery of specialty mental health and substance abuse services for Medicaid and non-Medicaid populations.

• North Carolina: Enhanced PCCM is used to coordinate services for Medicaid beneficiaries and to support primary care providers' (PCPs') ability to function as medical homes for individuals with behavioral health conditions.

• Tennessee: All Medicaid MCOs are responsible for physical and behavioral health benefits, and the state has recently integrated long-term care services into its managed care contracts.

• Vermont: As part of a statewide multipayer initiative, the state is working to transform primary care practices into patient-centered medical homes (PCMHs) that provide mental health services and support community health teams (CHTs).

As summarized in Table ES.1, these state programs harness different funding streams and use a variety of strategies to organize and deliver care. To some extent, each program reflects the unique state environment in which it was developed. As one program representative noted, there is not a one-size-fits-all approach to improving the integration and coordination of care. Some states tailored their programs to existing programs and infrastructure; others opted for wholesale system reform.

Despite these differences, some features are common to all four states. An important component of each program is to connect individuals with an array of state- and community-funded social services such as housing assistance and employment services. The use of information systems are critical components as well. Several programs are either providing an electronic health record (EHR) platform and/or encouraging providers to use EHRs and other information technologies, such as web portals and registries, to share patient information, coordinate care, and inform clinical decision-making. The states are also using information from these systems to monitor the quality of care. Finally, each program has employed a variety of quality-improvement strategies in order to help meet program goals.

The case studies are a useful snapshot of states' activities, but further research could focus on the implementation and effectiveness of specific program components; this information could help policymakers implement similar programs elsewhere in the nation. For instance, qualitative data could tell us more about the structures and processes of care in these four programs, and about the implementation successes and challenges. States are using claims data and EHRs to develop quality monitoring infrastructures, which could be used to examine the impact of these programs on service utilization and costs. Future evaluations of these programs must take into account the specific context in which they were implemented.

I. Introduction

Individuals with serious mental illnesses (SMI) and other chronic behavioral health conditions require a comprehensive array of physical, behavioral, and other supportive services in order to live independently in the community. Recent research suggests that less than 5 percent of Medicaid beneficiaries with schizophrenia or bipolar disorder receive regular medications, medication monitoring, psychosocial services, and any preventive physical health care during the year (Brown et al. 2012). Many of these individuals suffer from chronic physical health conditions, including diabetes and cardiovascular disease, but fail to receive adequate care (De Hert et al. 2011).

The historical lack of alignment in the financing and delivery of physical and behavioral health care as well as other supportive services for individuals with behavioral health conditions has contributed to gaps in the quality of their care and to their use of costly services. Indeed, individuals with SMI have high rates of emergency department (ED) visits and inpatient hospitalizations (Durden et al. 2010; Greenberg 2012). Research has found that individuals with mental illness are one of the costliest groups of Medicaid recipients (Kronick et al. 2009). Among Medicaid beneficiaries with chronic physical conditions, health care costs for those with a mental illness are as much as 75 percent higher than for those without a mental illness (Boyd et al. 2010).

Concerns regarding the quality and costs of care for this population have prompted Medicaid programs, managed care organizations (MCOs), and state and county mental health agencies to seek better strategies for financing and delivering services that integrate and coordinate physical and behavioral health care as well as other supportive services (Greenberg 2012; Kim et al. 2012). There is currently tremendous variation in the financing arrangements and delivery models used to provide care for this population. In many states and communities, physical health care and behavioral health care are provided in different service settings that receive reimbursement through distinct financing arrangements. While some state Medicaid programs provide both physical and behavioral health services using a fee-for-service model, many states contract with MCOs to provide physical and/or behavioral health services and/or managed behavioral health organizations (BHOs) to provide behavioral health services. Some states also seek to improve the coordination of services through enhanced primary care case management (PCCM) programs that provide enhanced payments to providers and other incentives and tools to ensure that individuals receive comprehensive services. Finally, certain supportive services that are often necessary for this population--including peer support, employment assistance, and housing and transportation services--are often provided outside of the auspices of state Medicaid programs, managed care arrangements, or state or county behavioral health agencies.

Each of these financing arrangements and delivery models has both strengths and limitations. As policymakers continue to look for ways to improve care for this population, they need detailed information about how states and communities are aligning the financing and delivery of services to strengthen the integration and coordination of physical, behavioral, and other supportive services for individuals with behavioral health conditions. To provide such information, the Office of the Assistant Secretary for Planning and Evaluation (ASPE) contracted with Mathematica Policy Research to conduct case studies of the financing arrangements and delivery models use to provide behavioral health services for Medicaid beneficiaries in four states: Louisiana, North Carolina, Tennessee, and Vermont. The case studies profile and describe the different mechanisms for coordinating and integrating services, including the financing of those services, the data infrastructure and information systems used, and the quality monitoring practices. These case studies were not intended to evaluate the effectiveness or outcomes of the programs; rather, they sought to profile each program and provide information that other state Medicaid programs and mental health agencies could use to inform the design of their own services. These case studies may also provide a foundation for further research focused on these programs.

Through document review and discussions with stakeholders in each state, the case studies sought to answer the following overarching questions:

• What are the goals of each program and in what context did each develop?

• How and to what extent are physical health, behavioral health, and other supportive services (housing, transportation, employment supports, and so on) covered and/or coordinated within each program?

• How is each program financed? Does the program draw from different funding sources and/or pool funds from multiple state agencies or payers?

• What are the intended outcomes? How are the quality of care and outcomes measured and monitored?

• What information systems and data infrastructure support the program?

Chapter II provides a brief summary of the methods used for the case studies. The subsequent chapters describe the key features of each state program. Appendix A contains a profile of each program.

II. Methods

The preparation of this report took place in four phases: (1) consultation with experts to identify states with unique or promising programs for financing and delivering coordinated or integrated behavioral health, physical health, and other supportive services; (2) an environmental scan and literature review to gather information about programs; (3) discussions with stakeholders in selected states; and (4) data analysis and presentation of findings in case studies.

A. State Identification

To identify promising states for this project, we contacted Medicaid and behavioral health experts at Mathematica Policy Research, the Center for Health Care Strategies, Technical Assistance Collaborative, National Association of County Behavioral Health and Developmental Disability Directors, and the Substance Abuse and Mental Health Services Administration (SAMHSA). We specifically asked these experts to help us identify states with innovative models for integrating or coordinating physical health, behavioral health, and supportive services, as well as states that provide these comprehensive services using unique managed care arrangements or financing strategies. Our goal was to identify states that formed a diverse group in terms of their Medicaid managed care arrangements, geography, and experience integrating/coordinating services.

B. Environmental Scan

After our experts had identified 11 state programs meeting our requirements, we conducted an environmental scan to gather additional information about their key features. We developed a search strategy to query PubMed, Google Scholar, and Google using key search strings, as well as the Centers for Medicare and Medicaid Services (CMS) waiver and state plan amendment (SPA) databases and the CMS Managed Care Summary (see the appendix for additional details on search methodology). We also searched the websites of MCOs and state Medicaid and behavioral health agencies to find information about the characteristics of each program. We created an Excel database to summarize and organize information relevant to the overarching research questions. Based on a review of this information, ASPE selected programs from four states--Louisiana, North Carolina, Tennessee, and Vermont--for case study because their features aligned with our study criteria and as a group exhibited the desired diversity. Some states meeting the criteria were excluded because there were already comprehensive published reports describing their programs, and ASPE wanted the case studies undertaken as part of this project to present new information.

C. Discussions with Stakeholders

For each of the four programs selected for case study, we held up to four discussions with state officials and managed care representatives to gather information. Discussions followed guidelines developed for each state and type of respondent. We followed up with stakeholders by email to clarify and gather additional information as needed.

D. Data Analysis and Summary of Findings

For each state program, we divided our data analysis into topic areas reflecting the research questions. For each topic area, we reviewed all notes from discussions with key informants as well as the literature sources identified during the environmental scan. The research team met to discuss state program findings, triangulate information across respondents, and clarify information that was unclear. Based on this data analysis and following a common outline developed in collaboration with ASPE, we summarized our findings in this report.

 

III. Louisiana Behavioral Health Partnership

A. Program Overview

The Louisiana Behavioral Health Partnership (LBHP) oversees statewide specialty mental health and substance abuse services for Medicaid-eligible and non-Medicaid-eligible adults and children. LBHP is led by the Office of Behavioral Health (OBH) and includes Medicaid, the Department of Children and Family Services (DCFS), the Department of Education (DOE), the Office of Juvenile Justice (OJJ), and Magellan Health Services as partners. State agencies that participate in LBHP pool funding into a single contract with Magellan to manage all behavioral health services. The pooling of funding is intended to leverage state general funds (SGFs) to appropriately obtain federal Medicaid matching funds, thereby freeing state funding to expand the delivery of services to non-Medicaid-eligible populations. This new arrangement represents a major transformation in the organization and reimbursement of services in Louisiana. LBHP is working to rapidly build the provider network and increase the availability of services, while also encouraging the integration of mental health and substance abuse services and the use of evidence-based and promising practices. Physical health services and behavioral health services delivered in primary care are covered through separate managed care arrangements. As described below, LBHP is undertaking a range of activities to develop the infrastructure necessary to expand the mental health and substance abuse treatment capacity within the state and to improve the quality of behavioral health care. Appendix Table A.1 summarizes the key features of the program.

B. State Context

Louisiana has undertaken several reforms in the organization and financing of behavioral health services during the past decade (DHH 2012b). In 2003, the state Department of Health and Hospitals (DHH) ceased providing substance abuse services through Medicaid and began providing them through the Office of Addictive Disorders in response to concerns about fraud and abuse (DHH 2012b). From 2003 to 2005, DHH also worked to strengthen requirements for mental health provider accreditation and revise service definitions.

In 2005, the Office of Mental Health and Medicaid developed a state-owned and state-operated administrative services organization responsible for overseeing the behavioral health provider network and managing the quality of behavioral health services. Experience with the state-operated administrative services organization led to legislation in 2007 that allowed DHH to establish a single statewide management organization (SMO) to manage the delivery of all behavioral health services. During this same period, the Office of Mental Health and Office of Addictive Disorders were merged to create the OBH and OBH began consolidating state-operated substance abuse and mental health clinics with the goal of creating integrated care settings to address co-occurring disorders.

In 2009, the state began developing a comprehensive system of care (CSoC) for children and youth with behavioral health challenges. A planning process brought together state agencies (OBH, DCFS, DOE, OJJ), advocates, and families and led to the idea of combining funding from state agencies to provide behavioral health care for children and adults. Pooling funding from across state agencies was intended to leverage SGFs to obtain Medicaid financing. Using this pooled funding approach, OBH contracted with Magellan to become the SMO and manage all behavioral health services for Medicaid-eligible and non-Medicaid-eligible adults and children effective March 1, 2012. It is important to note that LBHP includes Magellan as the SMO but the LBHP refers to a broader initiative of behavioral health care reforms and activities in Louisiana.

C. Program Goals

LBHP is undertaking several simultaneous system reform efforts, as described in the request for proposals (RFPs) that OBH issued in 2011 to procure an SMO. The reforms have the following goals:

• Implement a CSoC for children/youth and their families/caregivers, using a family-driven and youth-driven practice model, providing wraparound facilitation by child and family teams, and offering family and youth supports, with overall management of these services by the SMO.

• Improve access to, and quality and efficiency of, behavioral health services for children not eligible for the CSoC, and for adults with SMI and addiction disorder, through management of these services by the SMO.

• Transition behavioral health service delivery and operations from OBH operated regions to independent local governing entities (LGEs).

• Integrate mental health and addictions care through combining the former Office of Mental Health and Office of Addictive Disorder into OBH, under one assistant secretary.

• Seamlessly coordinate behavioral health services with the comprehensive health care system without losing attention to the special skills of behavioral health professionals.

• Advance a system of person-centered care that promotes resilience and recovery.

• Implement best practices and evidence-based practices that are effective and efficient as supported by the data from measuring outcomes, quality, and accountability.

• Leverage SGFs to appropriately obtain Medicaid financing.

The contract between OBH and Magellan is a core component of achieving these system reform efforts. Magellan is at-risk for adult behavioral health services and manages child and adolescent behavioral health services, which are reimbursed on a non-risk basis. Magellan and OBH are working together to expand the provider network and to monitor and improve quality. The remainder of this chapter focuses on the structure and financing of the arrangement with Magellan for the adult population. (LBHP is undertaking a range of activities focused on children and youth, specifically those with a serious emotional disturbance, but these are not the focus of this report.)

D. Program Financing and Contracting

The state operates the managed care arrangement with Magellan under the authority of: (1) a 1915(b) waiver for a prepaid inpatient health plan with mandatory enrollment and selective services contracting; (2) a 1915(c) home and community-based waiver; and (3) a 1915(i) state plan option for adult mental health rehabilitation services for adults with SMI. As mentioned above, funding for this program comes from Medicaid and from the pooled funding of several state agencies. The latter source includes SGFs and funding from the Substance Abuse Prevention and Treatment Block Grant. While these funds are pooled into the contract with Magellan, the grant funding is used for its intended population(s).

Although actual cost data were not publically available at the time this report was written, OBH projected per member per month (PMPM) costs for 1915(b) beneficiaries of $56.39 in the first year of the program and $60.90 in the second year of the program (DHH 2011).

E. Covered Populations and Services

Populations. Magellan manages behavioral health services for adults enrolled in Medicaid (including those dually eligible for Medicare) and for non-Medicaid adults eligible for OBH services. OBH projected that 164,360 non-disabled adults and 133,050 disabled adults would be eligible for 1915(b) services under the Magellan contract in the second year of the program (DHH 2011).

There is not a separate process for enrolling adults into services managed by Magellan. Rather, adult Medicaid beneficiaries automatically have their behavioral health care managed by Magellan. Individuals who are not currently enrolled in Medicaid or who are not eligible for Medicaid can go directly to a provider. The provider is then responsible for contacting Magellan to determine if the individual is eligible for behavioral health services and to obtain any necessary authorizations for provider services. Alternatively, individuals in need of behavioral health services can contact Magellan directly, as described below, to determine their eligibility for services and be connected with a provider.

Services. Medicaid and non-Medicaid-eligible adults have access to a range of services. Coverage for services differs across Medicaid subpopulations; notably, those eligible for 1915(i) services (those who meet the federal definition for SMI) have access to a broader array of rehabilitation services and case-conferencing services than the medically needy or other Medicaid adults. For non-Medicaid-eligible adults, Magellan is not responsible for covering inpatient psychiatric stays in general hospitals, but it does cover stays in psychiatric hospitals.

Table III.1 summarizes which services Magellan manages by covered populations.

In addition to inpatient care, Magellan manages a range of outpatient and rehabilitative services. These services, explained below, have specific eligibility and prior-authorization requirements, as described in the service definition manual available on the LBHP website (DHH 2012a).

• Case-conferencing refers to face-to-face meetings between providers to discuss treatment or treatment plans. Only licensed mental health practitioners (LMHPs), advanced practice registered nurses (APRNs), and psychiatrists can receive reimbursement for case-conferencing. (LMHPs are licensed by the state and typically include psychologists, clinical social workers, licensed professional counselors, marriage and family therapists, addiction counselors, and APRNs.) Case-conferencing is intended to coordinate treatment across agencies rather than to pay for treatment team meetings within an agency.

• Psychosocial rehabilitation is designed to eliminate functional deficits and interpersonal or environmental barriers associated with mental illness so that the individual can remain in the community. These services can be delivered by providers who have at least a high school diploma.

• Community psychiatric support and treatment (CPST) is intended to help individuals achieve their goals and live independently through individual supportive counseling, solution-focused interventions, and assistance with daily living skills. Practitioners with a master's degree can provide any CPST service, while those with a bachelor's degree (or equivalent) can provide only some CPST services. Peer specialists can also provide some CPST services.

• Crisis intervention services are intended to ameliorate psychiatric emergencies through preliminary assessment, resolution of immediate problems, and referral and linkage to appropriate community services. Some crisis intervention services can be provided by an individual with at least an associate's degree in human services, while others can be provided only by LMHPs.

Magellan does not currently provide supportive housing services, but it anticipates taking over the management of a supportive housing program from OBH in 2013. This program will provide supportive housing for approximately 3,300 individuals, primarily in areas that were affected by Hurricane Katrina and Rita that were original funded by federal legislation for Gulf Opportunity Zone redevelopment. Magellan will employ "tenant service managers" to assist program participants with daily living skills so they can maintain their housing and live independently. OBH and Magellan hope to bring managed care practices to the supportive housing model in order to increase its efficiency and to better coordinate other behavioral health services that can help individuals maintain their housing. Via the Magellan contract, Medicaid and non-Medicaid funding may be used for various services for the population in supportive housing.

Physical health services and pharmacy benefits for Medicaid beneficiaries are delivered through separate managed care arrangements, known collectively as Bayou Health. At the time of this study, Magellan and the physical health plans did not share data on consumers, and there were no formal mechanisms for coordinating physical and behavioral health services. Nevertheless, care managers from Magellan and the physical health plans do interact with each other to coordinate care. Given that Magellan is at-risk for behavioral health services, Magellan care managers have an incentive to coordinate with physical health providers and supportive services providers outside of Magellan's financial responsibility so that costly hospital psychiatric stays can be avoided. One of Magellan's current quality-improvement efforts seeks to strengthen care coordination with physical health providers, and eventually information may be shared between plans. OBH and Magellan are also planning to allow physical health providers to securely download patient information from Magellan's databases to facilitate better care planning and coordination.

Building the provider network. OBH and Magellan are working to expand the provider network, in part by allowing licensed professional counselors and licensed clinical social workers to receive Medicaid reimbursement via Magellan under the terms of the state plan option for 1915(i) services. Magellan credentials providers every three years in accordance with National Committee for Quality Assurance (NCQA) standards. Depending on licensure and education level, providers can be credentialed independently, or they can receive reimbursement working as part of a credentialed organization. Providers who had a Medicaid provider identification number at the time OBH contracted with Magellan did not have to actively enroll in the Magellan network. Rather, they were automatically enrolled and put through the credentialing process. Other providers who already had a contract with Magellan but did not serve Medicaid beneficiaries had to amend their contract with Magellan to bill for Medicaid and non-Medicaid adults.

All providers now must submit claims to Magellan. Previously, many behavioral health providers were not accustomed to billing Medicaid or other insurers. For behavioral health services provided after March 1, 2012, providers must submit a claim via the Clinical Advisor electronic health record (EHR), described below, or use an electronic claims submission available on Magellan's website, or submit a paper claim. Magellan only accepts claims from Magellan credentialed providers. Therefore, behavioral health services delivered in primary care would be submitted to the physical health plans.

Navigating services. There are several mechanisms to help consumers navigate services. Magellan staffs a toll-free number that consumers can call 24 hours a day, seven days a week, to speak with a care manager about accessing behavioral health services. The care manager conducts an initial assessment of the consumer's needs and eligibility for services and then connects the consumer with service providers. Magellan incurs a financial penalty for not answering calls within 30 seconds and for losing calls (calls terminated by the consumer waiting in queue). Magellan also maintains a website (http://www.MagellanofLouisiana.com) with providers' locations and contact information.

For adults diagnosed with SMI and eligible for 1915(i) services, Magellan contracts with clinicians who conduct independent assessments and develop treatment plans, which include recommendations for length and type of treatment. These assessments are not conducted by treatment providers but by independent assessors who have a contract with Magellan specifically for this purpose. For other adults, any provider in the Magellan network can conduct an assessment and develop the treatment plan. Magellan reviews the assessments for 1915(i) services, as well as treatment plans for other adults, and then authorizes care for a maximum period of three months. After three months, the provider must seek reauthorization. Consumers can receive up to five diagnostic assessments, 24 outpatient psychotherapy sessions, and 12 medication management sessions per year without needing prior-authorization (Magellan Health Services 2012).

F. Quality Monitoring and Incentives

LBHP developed a comprehensive quality strategy to guide the measurement and oversight of the program (see DHH 2012b). The quality strategy describes specific goals, measures, and reporting processes. OBH and Magellan are collecting a wealth of information that can be used to examine trends in the utilization, quality, and outcomes of services. Here we briefly summarize the approach to monitoring and reporting of quality measures and describe some of the incentives in place to encourage use of evidence-based and promising practices.

OBH and Magellan established the Inter-Departmental Monitoring Team (IMT) to develop and implement a plan for monitoring and improving the quality of behavioral health care. All LBHP state agencies participate in the IMT, which incorporates the input of Magellan and consumers. Consumer input is gathered through public forums, state and regional advisory councils, analysis of grievances, and a consumer satisfaction survey. In addition, an external quality review organization (EQRO) conducts an annual independent review to ensure that Magellan complies with federal Medicaid managed care regulations and to validate the results of performance measures and performance-improvement projects (PIPs) (DHH 2012b).

OBH and the IMT identified three overarching quality goals for the Medicaid program: (1) foster individualized behavioral health services for adults, youth, and families through increased access to a fuller array of evidence-based in-home and community services that promote hope, recovery, and resilience; (2) improve quality by establishing and measuring outcomes; and (3) manage cost through effective use of state, federal, and local resources. Progress toward these goals is tracked using several measures that map to more specific quality objectives, and performance on several quality measures is tied to financial incentives or penalties for Magellan (see Table III.2).

Several mechanisms are in place to encourage the use of evidence-based or promising practices. As part of its quality strategy, Magellan created a clinical advisory committee of network providers to annually review practice guidelines and offer advice on implementing and monitoring them. The service definition manuals contain several evidence-based practices for which providers receive higher reimbursement. In addition, the performance measures monitor the percentage of adult high service users (those having two or more inpatient admissions or four ED visits in a year) who enroll in Assertive Community Treatment (ACT) or psychosocial rehabilitation programs.

G. Information Systems and Data Infrastructure

Magellan uses data systems to handle prior-authorizations, claims processing, and tracking of patient care. Because Magellan is responsible for managing the majority of public behavioral health services for the state, these data systems give it the unique ability to track the utilization of services and monitor the quality of care for a large population of Medicaid and non-Medicaid-eligible adults. The state did not develop any data systems specifically for LBHP.

Magellan and OBH encourage providers (though they do not require them) to use the Clinical Advisor EHR. The Clinical Advisor system captures information about clinical interactions with patients (patient progress notes, treatment plans, etc.) and generates claims so that Magellan can reimburse providers. Thus, this system is able to capture information that can be used for quality monitoring. The system can also be used to schedule visits. The eligibility systems from several state agencies (Medicaid, OBH, DCFS) feed into Clinical Advisor to populate the system with unique beneficiary identification numbers that are used to track patients and monitor their care. Although the Clinical Advisor system interfaces with other eligibility, authorization, and claims systems, it was not specifically designed as a care management tool. Because it does not require the installation of desktop software, it is more easily accessed via the web. While some providers and hospitals have their own EHRs, some behavioral health providers are beginning to use Clinical Advisor.

H. Impact of the Affordable Care Act

At the time of this report, Louisiana is not planning to expand Medicaid eligibility under the Affordable Care Act (ACA). Nonetheless, the stakeholders we spoke with were optimistic that Louisiana was making strides to develop the provider network and infrastructure needed to meet any increase in the demand for behavioral health services that might arise out of health care reforms. In particular, stakeholders noted that Magellan now serves as the central point of entry into behavioral health services, an arrangement that would help newly eligible beneficiaries or others seeking care to navigate services and locate providers, and that would help coordination of services for individuals receiving care.

I. Successes and Challenges

The stakeholders we talked with during this study noted that the new managed care arrangement encountered several successes and challenges.

They reported the following key successes:

• Strengthening the provider network. OBH and Magellan have worked to increase the number of credentialed providers who can receive Medicaid reimbursement, and they continue to expand the network.

• Developing a central point of entry into behavioral health services. Consumers have several tools for accessing services, and Magellan care managers can help consumers navigate services.

• Developing a data warehouse that enables OBH and Magellan to track service utilization and monitor care. This common system allows Magellan to implement a range of quality measures.

• Emphasizing evidence-based and promising practices. OBH and Magellan have taken steps to strengthen service definitions, disseminate practice guidelines, and put quality measures and incentives in place that encourage the delivery of evidence-based care, including ACT and psychosocial rehabilitation.

They noted several challenges that LBHP is currently working to overcome:

• Coordinating physical and behavioral health services. There has not yet been any data sharing between Magellan and the physical health plans, nor have formal mechanisms for coordinating care yet been established. OBH and Magellan are undertaking quality-improvement initiatives to potentially share data between Magellan and physical health plans in order to facilitate better care monitoring and coordination.

• Orienting behavioral health providers to Medicaid billing practices. Prior to this new managed care arrangement, many behavioral health providers did not submit claims to Medicaid. OBH and Magellan are working to ensure that behavioral health providers become more comfortable submitting complete and accurate claims to receive reimbursement.

• Encouraging the use of a common EHR. While the introduction of the Clinical Advisor system has potential to facilitate monitoring and to improve the quality of care, not all providers have adopted this system. OBH and Magellan continue to work to encourage the system's use.

 

IV. Community CARE of North Carolina

A. Program Overview

Community Care of North Carolina (CCNC) is a statewide population management and care coordination infrastructure founded on a primary care medical home model. As an enhanced PCCM program, CCNC aims to improve the cost-effectiveness and quality of care for Medicaid recipients with chronic illness through leadership by local clinicians and a strong emphasis on care coordination, disease and care management, medication management, and quality-improvement (Kaiser Commission on Medicaid and the Uninsured 2009). CCNC's central office works with 14 regional networks, care managers, and CCNC-affiliated primary care practices (primary care medical homes) to coordinate services for Medicaid recipients and to connect them with a broad range of state and community-funded social services. Though nearly all full-benefit Medicaid recipients are eligible to enroll, CCNC focuses the majority of its care management and coordination on individuals with chronic illness.

Individuals with behavioral health needs, particularly those with comorbid physical conditions, are the focus of CCNC's behavioral health program. The program aims to facilitate integration of primary care and behavioral health care by supporting primary care providers (PCPs) in becoming the medical home for enrollees with mild to moderate behavioral health issues typically served in the primary care system as well as those with SMI typically served in the specialty behavioral health system. Historically, coordinating care for individuals with SMI has been challenging in North Carolina because the state's systems governing physical health care and mental health care are distinct. Local Management Entities (LMEs) have traditionally managed the delivery of specialty mental health services for Medicaid beneficiaries while CCNC has managed physical health services. As a result, achieving mental and physical health care integration for individuals with SMI requires close collaboration and communication between the LMEs and CCNC. The LMEs and the local CCNC networks have been working to achieve this aim; however, the state's current conversion of all LMEs to Managed BHOs has complicated these efforts.

B. State Context

Program background. CCNC evolved from a small medical homes program (the Wilson County Health Plan) in one rural county in the early 1980s. The medical homes program connects each patient with a PCP who leads a health care team in addressing all of the patients' health needs. The program was a joint partnership between the state Medicaid agency (the Division of Medical Assistance [DMA]), the Office of Rural Health, Research, and Development, and the North Carolina Foundation for Advanced Health Programs, Inc., with a private health care philanthropy (the Kate B. Reynolds Health Care Trust) providing the funding. The program's aim was to encourage physician participation in Medicaid, thereby improving access to care and reducing reliance on ED utilization. The program expanded in 1989 to become a statewide PCCM program (Carolina Access) which added a PMPM payment to PCPs to fund care coordination activities. In 1998, the current CCNC program was officially piloted in seven rural counties in response to a North Carolina Department of Health Services directive calling for the state to improve Medicaid access and quality and to lower costs. The CCNC pilot added several new elements to Carolina Access: (1) regional physician networks; (2) population management tools; (3) care management and clinical support; and (4) data and feedback (CCNC 2013b). The CCNC model worked well in North Carolina, a predominantly rural state, because it was adaptable to both urban and rural contexts. The program has also experienced high participation rates by PCPs, indicating their satisfaction with the program. In contrast, the state's comprehensive MCOs had difficulty penetrating the state's rural markets and eventually voluntarily withdrew from the Medicaid program.

CCNC was expanded statewide in 2001. Originally, CCNC worked only with non-disabled adults and children and focused on single chronic illnesses such as asthma and diabetes. However, it soon became clear that most cost and quality issues were related to treatment of multiple chronic illnesses. In 2005 the state expanded CCNC's role to the aged, blind, and disabled (ABD) populations, including full dual eligible beneficiaries. Also in 2005, in response to a growing need for data and informatics, CCNC's leadership shifted from the state agency to a newly formed not-for-profit organization acting as a central office. CCNC's behavioral health program was added in 2010 as part of an effort to improve quality of care and reduce health care expenditures for individuals with behavioral health care needs, including those with SMI.

Basis in other models. CCNC's founders initially looked to California's county-organized health system model to gather ideas. In the early 1980s, when the Carolina Access founders were developing the program, California was the only state undertaking a similar project to coordinate care using county-based entities. CCNC's behavioral health program was not modeled on that of other states but developed organically in response to the cost and quality issues related to treating the SMI population. CCNC realized that in order to address these issues, it needed to develop better ways to engage mental health providers.

Partnering agencies and organizations. CCNC's development was made possible through multiple state and private partnerships. The North Carolina Medicaid agency (DMA in the Department of Health and Human Services) was a founding partner and currently funds and oversees CCNC's contract. The Office of Rural Health and Community Care (ORHCC), another founding partner, currently works with CCNC on specified initiatives and recently provided funding for a chronic pain initiative. ORHCC also contracts with CCNC for the use of its "provider portal" (see Section G), allowing PCPs to use this tool to manage uninsured patients in addition to the Medicaid patients that CCNC manages. The North Carolina Foundation for Advanced Health Programs, Inc., another early partner, has sponsored much of CCNC's piloting and testing. Most recently, the foundation has partnered with CCNC on the Integrated, Collaborative, Accessible, Respectful, and Evidence-based (ICARE) Partnership to educate providers on care integration (North Carolina Foundation for Advanced Health Programs 2013). The Kate B. Reynolds Health Care Trust was CCNC's first funder, providing six grants totaling $1.6 million during the program's early development (CCNC 2013b). CCNC also works with the North Carolina Division of Mental Health, Developmental Disabilities, and Substance Abuse to coordinate policies and care management procedures, particularly in light of the state's shift to behavioral managed care.

C. Program Goals

CCNC has the following goals:

• Work together with providers and other partners to improve access to, quality of, and cost-effectiveness of care for Medicaid recipients with chronic illness.

• Produce savings for the state through improvements in health care quality and efficiency.

• Achieve these aims by: (1) establishing a medical home for patients; (2) emphasizing leadership of local primary care physicians in disease management and quality-improvement activities; (3) building local provider networks capable of managing recipient care; (4) establishing collaborative relationships with all providers; and (5) centering care around the patient (CCNC 2013a).

D. Program Financing and Contracting

Financing. CCNC and its predecessor program, Carolina Access, were initially funded through a 1915(b) managed care waiver. However, as of April 1, 2009, CCNC's core programs are now funded through a SPA (#NC-09-006). Medicaid (DMA) is the only source of funding for the program, though CCNC receives state and federal grants for specified initiatives. CCNC is working to expand services to other payers in the future (including Medicare and private insurers) and is currently piloting this approach in seven counties.

DMA funds CCNC's work through a PMPM fee. This fee is paid for each enrolled individual, not just those for whom CCNC actively coordinates care. DMA pays CCNC $12.85 PMPM for each ABD enrollee. A portion of the PMPM fee is earmarked to fund CCNC's behavioral health program. DMA pays CCNC a lower PMPM fee for each non-ABD enrollee (ranging from $0.24 to $9.01, depending on recipient category). CCNC, in turn, passes along a portion of the PMPM fee to the local CCNC networks to fund their activities. CCNC's two largest but most critical costs have been related to training and the development of its data and information systems.

To fund physicians' participation in CCNC's disease management and care coordination work, DMA pays each CCNC-affiliated practice $5.00 PMPM for ABD enrollees and $2.50 PMPM for non-ABD beneficiaries. DMA reimburses physicians for medical services on a fee-for-service basis, whereas the state is currently transitioning to a managed care carve-out for all mental health services (Shipman 2012).

Contracting arrangements. CCNC's central office holds the contract with DMA. The central office retains funds to carry out program-wide responsibilities that include informatics, analytics, program development, training, government relations, and marketing. CCNC's central office subcontracts with the 14 regional CCNC networks to carry out local services, namely care management and practice support. In turn, CCNC-affiliated practices (primary care medical homes) contract with both DMA and separately with the regional CCNC network (McCarthy and Mueller 2009).

CCNC network responsibilities. Under the terms of its contract, each regional CCNC network is responsible for managing enrollees' care, including linking them to a primary care medical home, providing disease and care management services, and implementing quality-improvement initiatives (Kaiser Commission on Medicaid and the Uninsured 2009). All CCNC networks are non-profit organizations, either independent 501(c)(3) organizations or part of existing community organizations, such as academic medical centers, federally qualified health centers (FQHCs), or public health departments. All local care management staff (including care managers, behavioral health coordinators, and network psychiatrists for the behavioral health program) are employed directly by the local network. Each network has its own leadership, including a board of directors, medical management committee, executive director, and a medical director. The board of directors includes representatives from the participating provider and community groups. At a minimum, each board must have a representative from the medical, community hospital, health department, and social services organizations from each of the counties the network covers. Most boards also include representatives from an LME, academic medical center, Area Health Education Center, or other health organization. The medical management committee includes representatives from the network's primary care medical homes. CCNC's statewide clinical advisory board, comprising the elected medical directors from each network, meets regularly to decide on new quality-improvement and disease management initiatives and to select clinical quality measures to be tracked across practices.

E. Covered Populations and Services

Covered populations. All North Carolina Medicaid recipients with full benefits--including full dual eligible--are eligible to enroll in CCNC, with the exception of nursing home residents. As of March 2013, 1.3 million Medicaid recipients were enrolled in CCNC--over 75 percent of the state's Medicaid recipients. Therefore the demographic and other characteristics of CCNC's enrolled population largely reflect those of the broader Medicaid population. As of 2010, nearly a third of CCNC's adult non-dual eligible ABD enrollees--21,070 of 72,297--had a serious and chronic mental illness (Treo Solutions 2011).

Enrollment process. Individuals can sign up for CCNC at the time of Medicaid enrollment. All fully eligible North Carolina Medicaid recipients except full dual eligibles must choose a Medicaid PCP at Medicaid enrollment. Recipients may choose a CCNC-affiliated PCP or a non-affiliated PCP. Patients who do not choose a PCP are automatically assigned to a CCNC-affiliated PCP, but can later "opt out" and choose a non-CCNC-affiliated PCP if they wish. Though full duals are not required to choose a Medicaid primary care medical home, since they receive almost all of their primary care services from Medicare, they have the option to enroll in CCNC for Medicaid services not covered by Medicare.

Outreach. Outreach and educational efforts generally take place locally during the enrollment process. At this time, county eligibility workers guide Medicaid recipients through the process of choosing a medical home and educate them about the purpose of a medical home. Individuals enrolling electronically do not receive this educational piece. CCNC is seeking to increase outreach and education for potential enrollees. For example, CCNC has begun working with the state on an opt out process for full duals and institutionalized who were previously not eligible for CCNC. As part of this process, which has resulted in an opt out rate of only 10 percent, CCNC sends a letter to these individuals explaining the program and giving individuals 30 days to opt out before being automatically enrolled in the program.

Covered services. CCNC is financially responsible for providing care management services and care coordination for enrolled patients. This involves connecting patients with the medical, behavioral health, and local social services that they need. However, CCNC is not financially responsible for, nor does it directly provide, the medical, behavioral, or social services that it coordinates. DMA reimburses providers directly for medical services on a fee-for-service basis, whereas for behavioral health services, the state is currently transitioning to a managed care carve-out (Shipman 2012). Available social services vary by locale, but typically include housing assistance, heating assistance, food assistance, vocational rehabilitation, and educational supports. Since care managers work locally, they become very knowledgeable about what social services and supports are available, and connecting patients with these services is an important part of their work.

Coordination of services. CCNC's care management and coordination work relies heavily on its data and informatics center. This infrastructure allows CCNC to identify the patients in greatest need of care management and coordination, known as "priority patients" (see Section G). Priority patients are typically individuals with chronic conditions who are not experiencing optimal care patterns (for example, have been hospitalized or use the ED frequently), or who are outliers for cost of care based on clinical risk grouping. CCNC also identifies patients for care management based on provider referral. Each local network and its care managers work with the identified priority patients to make sure they are getting the care, medications, social services, and other resources they need. Typically care managers will work consistently with the same practices to facilitate continuity of care; a larger practice may have its own care manager, while smaller practices usually share a care manager.

Care managers, typically registered nurses or social workers, initiate the follow-up and make contact with practices concerning their panel of patients. CCNC's networks currently have a total of 600-800 care managers statewide. Care managers undertake a wide range of activities, including: (1) helping patients access needed care and coordinate services; (2) conducting patient education and follow-up to promote treatment adherence and support lifestyle changes; (3) conducting home visits (for example, to assess medication adherence); (4) arranging follow-up medical appointments, transportation services, and access to community-based social services; (5) managing care transitions; and (6) working with hospitals on discharge planning. To provide consistent guidance to care managers statewide, CCNC network leaders and program staff developed the Standardized Care Management Plan, which offers benchmarks and guidelines for care management activities (McCarthy and Mueller 2009).

Integrating care for individuals with SMI. Coordinating care for individuals with SMI in particular has been a challenge in North Carolina because the state's systems for physical health care and mental health care are distinct. Mental health services traditionally have been managed by LMEs, which are local government agencies responsible for managing, coordinating, facilitating, and monitoring mental health and substance abuse services (including maintaining adequate behavioral health provider networks) within a certain geographic area. The LMEs are governed by the Division of Mental Health, Developmental Disabilities, and Substance Abuse Services (though behavioral health services for Medicaid recipients are funded by DMA). CCNC has traditionally managed the coordination of physical health services, which are governed by DMA. As a result, achieving mental and physical health care integration for individuals with SMI requires close collaboration between these two systems.

CCNC and the LMEs traditionally have used a four-quadrant model to determine whether patients should be managed by CCNC, the LME, or both. The model groups patients into four categories, those with: (1) low behavioral and physical health needs; (2) high behavioral health needs and low physical health needs; (3) low behavioral health needs and high physical health needs; and (4) high behavioral health and physical health needs. CCNC works closely with the LMEs to manage patients with both high behavioral health and physical health needs. To manage these patients, the state expects care managers from CCNC and the LME-MCOs to work together to create a "person-centered plan" which identifies the services and supports needed to help an individual achieve his or her goals and live independently. Both sets of care managers are expected to stay engaged to make sure the consumer gets all necessary physical and behavioral health services. CCNC is primarily responsible for managing the needs of consumers with high physical health needs but low behavioral health needs. CCNC will refer consumers with high behavioral health needs but low physical health needs to the LME case managers, and will stay engaged throughout to ensure that the individual receives necessary specialty behavioral health services. CCNC's patient portal (see Section G) allows all providers and case managers (including behavioral health providers and LME care managers) to view service utilization information for individual consumers, including their use of behavioral health services and prescription drugs. CCNC estimates that 80 percent of specialty behavioral health providers currently use the patient portal to aid in clinical decision-making.

As of June 2013, the state is transitioning all LMEs into 11 separate full-risk managed BHOs known as LME-MCOs. For several reasons, this change has added a layer of complexity to CCNC's efforts to integrate physical and mental health services for consumers with SMI. First, consumers' need for mental health services will need to undergo a utilization review by the LME-MCO which can delay treatment. Second, each LME-MCO has a different set of policies regarding eligibility for care management. In some cases, the LME-MCO may not agree with CCNC that a consumer needs to be assigned a care manager. Prior to the transition, CCNC and the LMEs generally agreed on the criteria for selecting patients for care management.

In response to the state's conversion of LMEs into Behavioral Health MCOs, CCNC is implementing several activities to address these challenges. In one county, CCNC's local network is collaborating with the local LME-MCO to integrate care management services for consumes with high physical health and behavioral health needs. CCNC is looking to expand this model to other LME-MCOs in the future if CCNC can obtain their buy-in. In addition, CCNC's informatics staff are building reports for the LME-MCOs that will provide them with a "preferred patient list" a list of high-risk patients with behavioral health needs that are likely to require management--for example, those who have had high ED or hospital utilization for these issues. Each CCNC network meets at least monthly with its local LME-MCO counterpart to discuss communication, referrals, and coordination as well as to address individual cases. At the agency level, CCNC meets monthly with DMA and the Division of Mental Health, Developmental Disabilities, and Substance Abuse to address broader issues, for example creating a universal referral form for all case managers and providers to use.

Behavioral health program. Through its behavioral health program, CCNC aims to facilitate integration of primary care and behavioral health care. CCNC does this by supporting PCPs in becoming the medical home both for enrollees with mild to moderate behavioral health issues typically served in the primary care system as well as those with SMI, particularly those with physical comorbidities, that are typically served in the specialty behavioral health system (CCNC 2013e). To achieve this aim, CCNC trains and supports the primary care medical homes in treating stable behavioral health conditions and substance abuse and helps PCPs feel more comfortable treating people with SMI. The program includes a strong educational component--for example, CCNC holds lunchtime trainings at practices on topics such as screening for substance abuse.

The behavioral health program is directed from the central office by a psychiatrist who leads a team comprised of a second psychiatrist and associate director, a behavioral health pharmacist, and a behavioral health care coordination program manager (CCNC 2013e). In addition, each network has its own psychiatrist and behavioral health coordinator who work on integrating care at the local level. The network psychiatrists: (1) develop collaborative relationships with LMEs; (2) identify best practices in screening and psychopharmacology for use in provider networks; and (3) facilitate engagement with community psychiatrists and key stakeholders. The behavioral health coordinators: (1) identify enrollees requiring care management; (2) help enrollees navigate the mental health and substance abuse systems; (3) employ motivational interviewing with enrollees to encourage self-management; and (4) assist PCPs in managing behavioral health needs. In addition, CCNC has incorporated behavioral health flags into its electronic care management tool. For example, the tool flags emergency room visits for mental health or psychiatric medication prescriptions to help identify members needing care management (Hamblin, Verdier, and Au 2011).

Specific behavioral health initiatives include:

• Integrated Care/Co-Location. CCNC is working to create a service delivery system that closely coordinates behavioral and physical care. The model involves a team-based approach in which physical and behavioral health providers partner to facilitate the direction, treatment, and follow-up of psychiatric disorders in the primary care setting. The model is appropriate for treating mild to moderate psychiatric disorders or maintaining the treatment of severe psychiatric disorders that have been stabilized. For individuals with severe psychiatric disorders that have not been stabilized, CCNC's care managers work with the LME care managers to create a person-centered plan and to ensure that the individual is getting the specialty behavioral health services that they need. If any physical comorbidities are present, CCNC's care managers work to assess care needs and arrange treatment for these conditions. Though the model implies that services are centered on the primary care setting, the important factor is not where the services are delivered, but how--there must be close coordination and collaboration between physical and behavioral service providers resulting in a seamless continuum of care for the patients (CCNC 2013e).

• Adult Safety with Antipsychotic Prescribing (ASAP). North Carolina Medicaid has initiated a prior-authorization policy for prescription of second-generation antipsychotics for off-label use (for example, to treat insomnia, anxiety, or primary treatment of depression). Under this initiative, the prescribing physician must obtain prior-authorization from a contractor of North Carolina Medicaid before the prescription can be dispensed. The initiative aims to reduce the inappropriate use of antipsychotics and reduce prescription drug costs for the state (CCNC 2013e).

• The North Carolina ACCEPT Project. This program employs educational campaigns targeting psychiatric professionals to encourage them to change prescribing trends for specific diagnoses including sleep disorders, depression, and treatment-resistant depression. The goal of the program is to move toward evidence-based and cost-effective prescribing practices, thereby improving psychiatric care and producing cost savings for the state. This program, funded by CCNC, is a partnership with the state's four academic medical centers (CCNC 2013e).

• Depression Toolkit for Primary Care. The CCNC Depression Toolkit was designed to help PCPs access practical, evidence-based tools to help them successfully treat adult major depressive disorder (MDD). The kit includes implementation recommendations, an algorithm to help with the initial assessment of MDD severity, a corresponding recommended treatment approach, screening tools, medication recommendations, etc. The kit also includes a guide to help PCPs decide when a referral for psychiatric care is indicated (CCNC 2013e).

Network pharmacist program. CCNC has implemented a pharmacy program which aims to create a medication management infrastructure that improves care outcomes while reducing total health care costs, not just prescription drug costs. Due to the high cost of many behavioral health medications, this program is particularly relevant for those with behavioral health needs. The program places a pharmacist within each CCNC network to aid the care management process. The pharmacists help physicians create and manage drug regimens for patients with chronic illness, perform medication reconciliation assessments, educate community pharmacists on Medicaid and CCNC pharmacy initiatives, and serve as a general resource for prescription drug and policy information. In addition, CCNC has implemented a number of specific initiatives through this program, such as the Prescription Advantage List (PAL). The list, which is optional for providers to use, ranks drugs within therapeutic categories (by highest frequency and opportunity to impact quality and cost) to encourage the use of less-expensive drugs, including generics and over-the-counter medications, whenever appropriate. CCNC providers receive quarterly feedback on a PAL scorecard showing the percentage of prescribed PAL drugs and the use of over-the-counter medications for their enrolled population (McCarthy and Mueller 2009).

Eligibility for providers. Any licensed Medicaid PCP can become a CCNC-affiliated primary care medical home. In order to qualify, providers must agree to actively participate in CCNC's care coordination and disease management initiatives, refer patients to CCNC for care management as needed, and offer after-hours care (24 hours a day, seven days a week) to reduce unnecessary ED utilization.

Interaction with other federal demonstrations. North Carolina has several federal demonstrations that interact with CCNC's work. First, the state has a five-year Medicare Quality Demonstration (646). The goal of this project is to improve the quality of care and patient outcomes for both dual eligibles and Medicare-only beneficiaries by using the CCNC model to address gaps in care, quality, and efficiency (CCNC 2013c). The program allowed CCNC to access and incorporate Medicare claims data into its informatics center to better manage and serve its dual eligibles as well as the newly enrolled Medicare-only population. North Carolina has a number of FQHC Advanced Primary Care Practice demonstrations, all of which are CCNC primary care medical homes. CCNC is working closely with these practices on implementing their demonstrations, primarily on incorporating practices' clinical and claims data into the state's Health Information Exchange, which CCNC runs. Access to these data in a clinically useable format will give the practices a platform to carry out necessary clinical and care improvement activities. North Carolina was recently approved for a CMS Demonstration to Integrate Care for Dual Eligible Individuals, which will use CCNC's medical homes model to coordinate care across primary, acute, behavioral and long-term supports and services for dual eligible individuals, incorporating a unique financing arrangement. Finally, effective October 1, 2011, CMS granted North Carolina approval for a SPA to implement the Health Homes provision through Section 2703 of the ACA. This SPA grants North Carolina an enhanced federal match (90 percent) for health home services for eligible Medicaid beneficiaries. To be eligible, beneficiaries must have two qualifying chronic conditions or one qualifying chronic condition and risk for a second. However, the SPA excludes mental illness and substance abuse disorders from this program (HHS, CMS 2012).

F. Quality Monitoring and Incentives

Quality measurement and improvement is an integral part of CCNC's work, and most of CCNC's quality monitoring is self-initiated. CCNC tracks 28 quality measures at the program, network, and practice level. Most measures are related to chronic diseases (including diabetes, asthma, heart failure, and hypertension) and disease prevention (such as cancer screening for adults). Two are specific to behavioral health, but not to adults: baseline glucose and baseline lipids in children prior to initiation of antipsychotics, then upon follow-up. CCNC also measures rates of preventable ED use and hospitalizations. The current set of quality measures was developed by a work group that included local clinicians and representatives of all 14 CCNC networks, who met over the course of a year for in-depth review of candidate measures. Quality measures are reviewed on an annual basis, and final measures are approved by vote of the CCNC clinical advisory board.

Allowing the local networks and PCPs to have ownership of quality- improvement efforts is an important part of CCNC's quality-improvement approach. Each network is responsible for piloting potential solutions and monitoring implementation, which is led by local physicians. Networks voluntarily share best practices solutions with other networks.

CCNC produces electronic quarterly reports at the practice level that compare quality measures for each practice over time and with other practices. These reports also list the patients for whom quality measures were not met so that the practice can put systems in place to better serve these patients. To evaluate cost savings, North Carolina's DMA contracts with an actuarial firm to evaluate whether CCNC is achieving projected cost savings targets.

CCNC has several mechanisms in place to promote evidence-based practices and to help providers avoid harmful practices. CCNC's provider portal (see below) calculates patient medication adherence and helps prevent medication errors. It also generates clinical care alerts that indicate, for example, whether patients with chronic illness have received the tests recommended by clinical care guidelines. In addition, two of the behavioral health program initiatives (the North Carolina ACCEPT project and ASAP [see Section E]) are designed to promote evidence-based, cost-effective prescribing practices and prevent patient harm. CCNC is also working with the state to implement an incentive-based payment system for PCPs, which would award higher payments to those who are providing evidence-based care.

G. Information Systems and Data Infrastructure

CCNC's work relies heavily on data and information systems at the program, network, and provider levels. At the program level, CCNC uses Medicaid claims data, real-time hospital data, and other clinical information from provider EHRs, along with proprietary risk-adjustment software developed by Treo Solutions, Inc., to identify "priority patients"--those with preventable hospitalizations or those who are outliers for cost of care based on condition severity. Priority patients are communicated to the local networks and are targeted for care management. Network care managers use CCNC's Care Management Information System to access patient information, document the patient care plan and note progress toward goals, and access screening tools. This system is available to CCNC's care managers as well as to local health department care managers. At the provider level, CCNC has built a provider portal, a web-based secure site that displays patient service and medication use across care settings. This tool is accessible to all providers and case managers--including PCPs, behavioral health providers, hospitals, and the CCNC and LME care managers--and enables them to deliver more targeted and appropriate care. Providers can't input data into this system, but they can use it to see what services and medications their patients have been using. Providers are highly encouraged to make use of the provider portal (and to use EHRs), but they are not required to do so. Networks and physician practices can also use the provider portal to generate demographic, cost, utilization, and quality monitoring reports on the population of patients they are responsible for as a means of informing quality-improvement activities.

Concerns regarding privacy and data sharing. In constructing its informatics center, the time and resources CCNC required to address legal concerns related to data sharing, data protection, and compliance with state and federal laws were greater than anticipated. CCNC's informatics center must adhere to Health Insurance Portability and Accountability Act provisions, the Health Information Technology for Economic and Clinical Health Act of 2009, and federal regulations related to the disclosure of substance abuse information. North Carolina also has state laws governing the confidentiality of mental health, developmental disabilities and substance abuse information as well as the confidentiality of information related to HIV and other communicable diseases. These steps required the involvement of legal counsel from multiple stakeholder groups and a thorough analysis of statutory authority and contractual relationships among CCNC, state agencies, LMEs, providers, and local health departments (CCNC 2013f). In addition, CCNC has invested many resources into establishing the required contracts, supports, and safeguards and conducting the necessary training to ensure that users have access only to permissible data.

H. Impact of the Affordable Care Act

Program representatives do not anticipate substantive changes to CCNC as a result of the ACA. North Carolina will not expand Medicaid to childless adults in 2014. As a result, CCNC's enrolled population will not change in size or demographic. North Carolina is not currently utilizing Section 1915(i) or the Health Home Medicaid options to expand home and community-based services for individuals with SMI; however, the state is looking into this as a possibility.

I. Successes and Challenges

Program successes. Over its history, CCNC notes a number of accomplishments. First, the program has achieved widespread engagement of PCPs; currently, 90 percent of primary care services are delivered to Medicaid beneficiaries via CCNC-affiliated primary care medical homes. In addition, CCNC has successfully built upon the existing health care infrastructure in North Carolina rather than completely revamping the way care is financed and delivered.

Evaluations of the program suggest it has resulted in both improved care and cost savings (Kaiser Commission on Medicaid and the Uninsured 2009). An external actuarial analysis by Mercer, Inc., estimated that, compared with historical fee-for-service costs, CCNC's care management and quality-improvement activities in 2006 saved the state between $154 and $170 million (Steiner et al. 2008). This figure grew to $194 million by 2009 (CCNC 2013d). The largest savings accrued from reduced ED, outpatient, and pharmacy costs. A separate external evaluation by the University of North Carolina at Chapel Hill compared outcomes and costs for individuals enrolled in the CCNC pilot versus the Carolina Access (PCCM) program in 2000-2002. The study estimates that during this period, the CCNC's asthma management program saved $3.5 million and the diabetes management program saved $2.1 million, largely the result of lower ED and hospital use (Ricketts et al. 2004).

Challenges encountered. As noted previously, North Carolina is currently implementing a managed care carve-out for all behavioral health services. Eleven separate full-risk MCOs will be implemented throughout 2013, each with different policies and procedures. This change has posed a challenge for CCNC in managing care for enrollees with SMI. The LME-MCOs have adopted a utilization review process which can delay treatment. In addition, the LME-MCO's criteria for selecting patients for care management now differs from CCNC's criteria. These issues complicate the hand-off between the CCNC care managers and the LME-MCO care managers; CCNC is currently working with the LME-MCOs to improve this process.

Lessons learned. CCNC representatives feel that there is no one-size-fits-all program. A successful program must be tailored to the unique needs and goals of each state and target population. Program representatives recommended moving beyond the traditional health care delivery model and reorganizing care delivery to center around the patient. CCNC has developed an alternative, primary care intervention approach in which mental health issues are caught (and milder issues treated) in the primary care setting. In this way, many patients can receive needed behavioral health care in a setting that is familiar and comfortable.

 

V. Tenncare

A. Program Overview

TennCare is the state of Tennessee's Medicaid program. In operation since 1994, the program provides health services for nearly 1.2 million adults and children and is the only Medicaid program in the nation to enroll all of its members in managed care. The state began integrating behavioral health into its managed care contracts in 2007 and completed the process in 2009. All physical and behavioral health services, including addiction and substance abuse services, are covered by three MCOs; supportive housing and supported employment services are also covered for patients with SMI. Continuing its efforts to implement a fully integrated service delivery model and to ensure that TennCare members receive all their health care services in a coordinated and cost-effective manner, Tennessee most recently integrated long-term care services into its MCO contracts through its CHOICES program. Tennessee also requires MCOs to be NCQA-certified and uses a health home model in which all enrollees are matched with a PCP who provides patient-centered care. Appendix Table A.3 summarizes the key features of the program.

B. State Context

TennCare aims to "demonstrate that the state can use managed care principles to serve Medicaid enrollees, as well as some individuals who are not Medicaid-eligible, without compromising quality of care and without spending more than the State would have spent had it continued its fee-for-service program" (Bureau of TennCare 2012b, 5). In 1996, behavioral health services for TennCare members were carved out and BHOs contracted directly with the Bureau of TennCare to manage these services. A primary aim of the carve-out was to provide services for a priority population that included adults with SMI.

By 2007, Tennessee decided that the carve-out model was not working. State officials saw a range of problems in this arrangement: the classification system was not serving members well, services for members were not coordinated, the Bureau of TennCare had to mediate disputes between the BHOs and MCOs about which organization was responsible for what services, and providers were not satisfied. Separating physical and behavioral health services, the Bureau concluded, interfered with providing comprehensive and cost-effective care for its enrollees. Tennessee therefore ceased providing behavioral health services through the BHOs and instead required its existing MCOs to provide these services.

An integrated managed care model meant that TennCare members would be able to access behavioral health services based on medical necessity. Integration also simplified contract agreements, since the state no longer had to contract with both a BHO and MCO, and it alleviated "turf wars" over which conditions were covered as physical or behavioral health. Finally, it ensured that TennCare members received comprehensive, coordinated care through a fully integrated service delivery system (James 2011).

In 2007, the Bureau of TennCare awarded regional contracts to three MCOs and began integrating behavioral health services into its MCO contracts. The MCOs accepted full-risk for all services, and the new contracts established an integrated medical and behavioral health care system for members. United and Amerigroup began serving the Middle Tennessee region in 2007, and United and BlueCare (Volunteer State Health Plan of Tennessee) began serving the West Tennessee region in 2008 and the East Tennessee region in 2009. In late 2009, behavioral health services for TennCare Select enrollees (a group including foster children and children receiving SSI benefits) were transferred from the BHO to Volunteer State Health Plan, which began operating statewide. The state had created a fully integrated delivery system for medical and behavioral health services.

Before the integration, the mental health benefit carve-out was managed by the Tennessee Department of Mental Health and Substance Abuse Services (TDMHSAS). The Department maintained oversight for the first year of integration and transitioned oversight to the Bureau of TennCare in 2008. Today, the Bureau and TDMHSAS collaborate in community and state work groups to discuss various projects related to reviewing services and needs in the community.

Continuing its quest to promote improved coordination of care for the whole person, Tennessee launched the CHOICES program in 2010. CHOICES integrated long-term care services into the MCO contracts. The MCOs began to offer new community-based alternatives to eligible individuals who would otherwise require Medicaid-reimbursed care in a nursing facility. With the implementation of CHOICES, the MCOs in Tennessee became responsible for the coordination of all medical, behavioral, and long-term care services provided to their members. The only remaining carved-out services are dental and pharmacy.

C. Program Goals

TennCare's goals are to:

• Assure appropriate access to care for enrollees.

• Provide high quality care to enrollees.

• Ensure enrollees' satisfaction with services.

• Improve health care for enrollees.

• Use a managed care approach to provide services to Medicaid state plan and demonstration eligibles at a cost that does not exceed what would have been spent in a Medicaid fee-for-service system.

• Ensure that health plans maintain stability and viability while meeting all contract and program requirements (Bureau of TennCare 2012b).

D. Program Financing and Contracting

Funding mechanism. TennCare operates under a Section 1115(a) demonstration waiver from CMS. The current demonstration, TennCare II, has been in existence since 2002 and expires in June 2013. The state has requested a three-year extension for the program under Section 1115(e). When TennCare II was last extended, in 2010, the extension included several new amendments, among them approval for the implementation of the CHOICES program outlined by the Tennessee General Assembly's Long-Term Care and Community Choices Act of 2008. The long-term care benefits were added to the existing TennCare II benefit package of primary, acute, and behavioral health services. Tennessee is now one of a limited number of states in the country to deliver managed Medicaid long-term care and the only state to do so in a manner that does not require enrollees to change their MCO (Bureau of TennCare 2012c).

MCO contracting. The three MCOs participating in TennCare are fully at-risk for all services. The MCO contracts began in 2008 and expire in 2014. According to TennCare representatives, the state primarily looked at quality over cost containment when considering proposals from MCOs; it wanted to know how the MCO would manage the population and what experience it had in the past with service integration. Contracts carefully define each requirement of the MCO and include appropriate reporting and monitoring processes to ensure compliance. Staggering the implementation of the integration (from 2007 to 2009) also helped with both the procurement (the state had a chance to refine the RFPs) and with the implementation (allowing a thoughtful, focused approach).

The MCOs are allowed to subcontract for the management of behavioral health services, but subcontractors are required to operate on site within the MCOs' offices to ensure coordinated management across services. One MCO is currently subcontracting with a vendor to provide management of behavioral health services.

The CHOICES program involves contracts with MCOs and contracts with nursing facilities and home and community-based service providers. There is a blended capitation rate with built-in assumptions regarding expected utilization and level of care provided. There are different capitation rates for duals and non-duals.

Costs of integration. TennCare is fully funded by Medicaid, with 65 percent of the cost of medical services for TennCare enrollees funded by the Federal Government and 35 percent funded by the state (Bureau of TennCare 2012d). Stakeholders from the Bureau of TennCare and an MCO indicated that integration has not given rise to any unanticipated costs. Because members with SMI receive frequent and high-cost services, the Bureau initially provided an add-on payment to MCOs based on the number of members with SMI enrolled in the MCO. However, the assessment process for identifying members with SMI proved costly and failed to capture all members with SMI. The Bureau altered their arrangement to designate members with SMI as priority based on diagnosis information from claims data and added the payments into the capitation rate. The MCOs also hosted provider forums when integration first started, but these were at no additional cost to TennCare. No additional funds were paid to the MCOs for start-up preparation.

The most recent data available from the Bureau of TennCare for PMPM costs are from state fiscal year 2011. State officials we engaged during our project reported that the statewide average for MCO acute care (no CHOICES, no pharmacy) was $275.59 PMPM. For the disabled population specifically, the cost was $896.57 PMPM, and for the general Medicaid population (Temporary Assistance for Needy Families [TANF] and related groups), the cost was $199.88.

E. Covered Populations and Services

Eligible population. All Medicaid beneficiaries in Tennessee are enrolled in the integrated MCOs, including individuals with SMI. As of April 2013, there were approximately 1.2 million Tennesseans enrolled in TennCare, of whom 120,000 had SMI diagnoses. In one MCO (Volunteer State Health Plan), 85,000 of its 432,000 members (nearly 20 percent) were diagnosed with SMI. As of July 2013, approximately 31,974 individuals were enrolled in CHOICES; approximately 15,000 of those enrolled met the criteria for SMI (Bureau of TennCare, 2012e).

Any TennCare member is eligible for behavioral health services. Adults eligible for TennCare include participants in the state's TANF program, pregnant women, single parents or caretakers of a minor child, SSI eligibles and related groups, and individuals in institutional placements or receiving home-based services as alternatives to institutional care.

There are three groups of TennCare recipients who are eligible for the CHOICES program: (1) those who receive nursing home care; (2) those who receive home care instead of nursing home care (including adults who have physical disabilities); and (3) those who receive home care because they do not qualify for nursing home care but who are at-risk for nursing home care (including adults with disabilities) (Bureau of TennCare 2013a). To enroll in CHOICES, an individual must also qualify for Medicaid long-term care. To receive home care through the program, the cost of home care must not be more than the cost of nursing home care.

Enrollment process. Upon enrollment in TennCare, each member is matched with a PCP. All members who require behavioral health services may receive them if medically necessary. TennCare members can access behavioral health services in a number of ways. One MCO mentioned relying on the PCPs' assessment to determine whether members need case management or behavioral health services. The PCP can educate members about the appropriate provider to see. Members can also access community mental health centers or contact behavioral health providers directly. At this time, none of the MCOs requires members to obtain a PCP referral for behavioral health services. Members can seek services from any licensed or credentialed professional in the network of their TennCare MCO.

MCOs also identify members whom they want to target for more complex MCO case management. As an MCO representative explained, a member with SMI should ideally be in active mental health case management. Otherwise, the patient may disappear for a while, self-medicate, and then present at an emergency room. Rather than this type of costly and inappropriate care, the MCOs want individuals with SMI to have ongoing treatment plans. When a member accesses a community mental health center, the provider determines whether the member needs mental health case management. The MCO will monitor the case management to make sure that it is appropriate and that the services are being provided by licensed or credentialed professionals. One MCO reported that it uses available claims data to identify patients who have been discharged from inpatient care and thus may warrant additional assistance and follow-up through its discharge planning process.

Covered services. TennCare has broad benefits with no limits for treatment, except for home health and private duty nursing services for adults (Bureau of TennCare 2012b). Only licensed or credentialed providers are covered. Covered services include primary care, behavioral health, addiction and substance abuse services (services provided by methadone clinics are not covered), long-term care, home and community-based services, transportation, and supported housing and supported employment services under psychiatric rehabilitation services. The supported housing benefit offered by TennCare refers to services provided at facilities that are staffed 24 hours a day, seven days a week; there are associated mental health staff supports for priority enrollees who require treatment services and supports in a highly structured setting. The facilities are for people with SMI and are not residential treatment facilities. Supported housing is intended to prepare individuals for more independent living in the community while providing an environment that offers appropriate mental health supports, including psychosocial rehabilitation (Bureau of TennCare 2013b). TennCare does not cover room and board for supportive housing.

The key component of the CHOICES program is care coordination, which includes transition and diversion programs to support home or community-based care and which offers more consumer choices. The CHOICES program allows consumers to hire non-traditional providers such as family members, friends, or neighbors; offers more residential care choices, including family care homes; and provides improved access to assisted care facilities (Bureau of TennCare 2012c).

Care coordination. The state and MCOs are seeking to integrate physical and behavioral services and to coordinate care for TennCare members. Currently an internist is available at one behavioral health site and a few community mental health centers, and a behavioral health specialist is available weekly at certain PCP offices. To help its members navigate services, one MCO said its first line of support is customer service. The organization has an on-site call service with positive response times that helps members find a new PCP or make appointments. The call center also makes calls to patients to remind them of appointments.

Outreach efforts. Representatives from one MCO described a few of their outreach services for the SMI population, a group that the MCO perceives to be underserved. The plan sees a high percentage of elderly patients with SMI whose needs have not been addressed. Representatives attribute this problem to the generational stigma attached to mental health care. They have found that telephone outreach and postcards do not work for this population, and they have learned through community focus groups that these individuals rely on their religious community before the behavioral health community. The plan also understands that many individuals do not want to talk to someone whom they do not know. It is working on creating relationships with individuals in the CHOICES program population with care coordinators and are developing tool kits for religious leaders. Thus if someone in a congregation presents with mental health issues, the leader will have access to a list of helpful services.

Other outreach efforts by TennCare MCOs include building relationships in the community and showing providers where to direct patients if they need help with behavioral health issues. Partnering with the Tennessee chapter of the American Academy of Pediatrics, one health plan is helping to train PCPs and pediatricians in how to talk to families about behavioral health needs and how to recognize the symptoms and warning signs of behavioral health problems; it has produced a series of training videos for that purpose. Representatives of this MCO say that because people are unlikely to trust an insurance company, they are training and sharing their knowledge with members of the community who already have a trusted voice. In their view, effective outreach is hands-on and local.

F. Quality Monitoring and Incentives

State role in quality monitoring. TennCare's Division of Quality Oversight seeks to ensure that TennCare members have access to timely, appropriate, and high quality health care services and experience optimal health outcomes. The MCOs are monitored through rigorous reporting, site visits, conference calls, and meetings (Bureau of TennCare 2013a). Tennessee requires MCOs to report on quality measures, and the MCO contracts require hundreds of deliverables related to quality reporting. Quality measures include the following:

• Health care effectiveness. In 2006, Tennessee became the first state in the nation to require its MCOs to be NCQA-certified. The state also began requiring that all MCOs report annually on the full set of Healthcare Effectiveness Data and Information Set (HEDIS) measures. HEDIS allows the MCOs to be measured on standardized, evidence-based performance measures; the HEDIS scores are compared to national averages and published. With the implementation of behavioral health integration, Tennessee began reporting on behavioral health measures in HEDIS in 2009. The state was not able to do this when health services were divided between MCOs and BHOs, because HEDIS is specific to MCOs. The HEDIS measures related to behavioral health include antidepressant and attention deficit hyperactivity disorder (ADHD) medication compliance and follow-up after hospitalization for mental illness. The MCOs are required to contract with an NCQA-certified HEDIS auditor to validate their processes in accordance with NCQA requirements.

• Consumer experience. Tennessee MCOs are required to contract with an NCQA-certified vendor to conduct annual Consumer Assessment of Healthcare Providers and Systems (CAHPS) surveys. The MCOs must submit both their HEDIS and CAHPS results to TennCare, NCQA, and the state's EQRO (described further below).

• Inpatient stays and readmissions. To monitor inpatient stays, TennCare has separate reporting requirements that mirror HEDIS; these focus on the length and number of inpatient psychiatric hospital stays. To monitor readmissions, the state looks at initial appointment timeliness and whether/when patients receive treatment after a hospital discharge.

• Patient access. The state imposes standards for the numbers and types of providers participating in the MCOs' networks. An MCO cannot participate in TennCare unless its network is of a certain size. As required by their contracts, the MCOs must demonstrate their ability to provide all contracted services on a timely basis and ensure accessibility to services. The Bureau of TennCare routinely evaluates provider networks and requests a corrective action plan when it identifies non-compliance (Bureau of TennCare 2013a).

• Performance-improvement. MCOs are contractually obliged to conduct two clinical and three non-clinical PIPs relevant to the enrollee population. One of the two clinical PIPs must be relevant to one of the behavioral health disease management programs for bipolar disorder, major depression, or schizophrenia. Two of the three non-clinical PIPs must be in the area of long-term care.

Table V.1 describes goals for some of the performance measures outlined above and the progress the state has made toward reaching those objectives.

Federal role in quality monitoring. In addition to the requirements imposed by the state, there are federal requirements for quality monitoring. The Balanced Budget Act of 1997 requires that an EQRO independently review TennCare's health plans. The EQRO conducts federally mandated activities such as performance measure and PIP validation and also administers specific requests by the Bureau of TennCare, including an annual HEDIS/CAHPS report, impact analysis report, and provider data validation. The EQRO reviews each health plan individually. Recent EQRO reports related to behavioral health services included the 2010 PIP Validation Report, which evaluated a PIP on follow-up for children prescribed ADHD medication, conducted by Amerigroup; a PIP on behavioral health postpartum depression screening, conducted by Volunteer State Health Plan; and a PIP on improving compliance with continuing treatment for MDDs, conducted by UnitedHealthcare (Bureau of TennCare 2012b).

Pay-for-performance quality incentive payments. Since 2006, TennCare has offered pay-for-performance quality incentive payments to its MCOs. In 2010, TennCare began offering quality incentive payments for three behavioral health HEDIS measures. MCOs are eligible for incentive payments if they demonstrate significant improvement from the baseline for the specified measures, or if they meet a specific goal. Significant improvement is determined by using NCQA's minimum effect size change methodology. In 2010, Amerigroup met the criteria for four quality incentive payments, and Volunteer State Health Plan met the criteria for nine quality incentive payments (Bureau of TennCare 2012c).

Tennessee's MCOs also have incentive programs for providers such as pay-for-performance programs customized to the provider. Providers' raises are tied to specific performance metrics. For behavioral health providers, measures that monitor outcomes are included in their contracts with the MCOs. Providers that do not meet the required metrics are paid less, and those with negative or unsafe outcomes have their contract terminated.

G. Information Systems and Data Infrastructure

With a managed care model in place long before behavioral health integration, Tennessee did not need to require any new information systems. The MCOs already had developed systems for eligibility, administrative, and claims data. The MCOs were also used to submitting their data to the state. The only change for the state has been that instead of receiving data from both MCOs and BHOs, all the data now come from MCOs. EHRs are not required by the state or MCOs, but many providers now use them. Most providers also bill their MCOs electronically, but again, this is not required. According to one MCO representative, sharing information between plans and providers can result in better patient care, although sharing data does necessitate addressing privacy concerns.

H. Impact of the Affordable Care Act

Whether Tennessee will opt into the federal Medicaid expansion remains uncertain. TennCare representatives said that if TennCare does expand its covered population, they do not expect there to be an impact on benefits that are covered. Projected numbers for the "eligible but not enrolled" population for fiscal year 2014 are about 46,000; the "newly eligible" population if TennCare chooses to expand will be approximately 145,000 (Gordon, Long, and Dungan 2013). The state anticipates that the expanded population will be similar to adult populations that have been served in the past under the demonstration, and it does not anticipate making any major changes to its goals, objectives, or performance measures. However, since the population will be different from the current population and will include members who were previously uninsured, it might be necessary to formulate additional goals and objectives (Bureau of TennCare 2012b).

I. Successes and Challenges

Program successes. Both the Bureau of TennCare and the MCOs consider the integration efforts a success. The bureau believes that it has implemented a fully integrated health care delivery system that provides comprehensive care for the whole person. There is no thought of returning to the earlier BHO model. The earlier model did not encourage providers to see patients frequently. The state's quality and cost outcomes are moving in the right direction, and TennCare members are receiving more services now than they did before the integration took place. Effectiveness of care measures, including the treatment of behavioral health conditions and medication management, have continued to show improvement. The proportion of TennCare CHOICES members receiving home or community-based care has also improved (Bureau of TennCare 2012b).

Representatives from both TennCare and the MCOs assert that costly treatments have been reduced for both behavioral and physical health services. Specifically, MCOs have seen a reduction in inpatient utilization, especially for long-term stays. The MCOs have taken advantage of the supportive housing benefit to move patients out of hospitals and into community-based alternatives. MCOs have reported their inpatient and subacute numbers have gone down, and their outpatient numbers have gone up.

State MCOs and providers have embraced integration. Integration was made easier because the incentives of the state, the providers, MCOs, and beneficiaries were aligned: everyone desired fewer ED visits and hospitalizations. The stakeholders with whom we spoke said that providers and patients have also welcomed the integration. Many members are more comfortable accessing behavioral health services through their PCP than in some other way and are happy to be getting comprehensive care.

Challenges encountered. Although TennCare providers have in general embraced integration, some behavioral health providers were initially apprehensive that the MCOs would not understand behavioral health care. Further, representatives from an MCO indicated that some PCPs were afraid to screen for behavioral health issues because they were unsure how to proceed if the screen was positive. Concerns from both the provider and MCO perspective were soon alleviated as integration moved forward successfully.

The Bureau of TennCare said that getting behavioral health providers to adjust to a different type of management style was challenging. Although they were used to a managed care model, their payment methodologies had been "grant-like" under the state's BHO model. The community mental health centers, in particular, were used to delivering services under the old payment system and had to be taught the skills needed under the new system. The state worked closely with them to educate them on issues such as billing for services, submitting claims in a timely manner, getting prior-authorization, and so on. As one respondent said, "The community mental health centers were not a very sophisticated group of providers from a business standpoint. The behavioral health providers all have more of a fee-for-service atmosphere now and were ultimately supportive of the integration."

To address challenges at the provider level, MCOs have implemented various initiatives. Specifically, Volunteer State Health Plan said that it has conducted trainings for providers on crisis services. It also has a PCP referral line that allows the PCP to speak with clinical staff. Nurses will call members and make appointments for them. The MCO also has psychiatric consultations available for the provider and is currently pursuing the outreach initiative with the local American Academy of Pediatrics mentioned above. The MCO is also focusing on expanding behavioral health in primary care, especially in rural areas.

Advice for other states. Representatives from the Bureau of TennCare said that in the past year, over half of states have requested guidance from the bureau on various topics, including integration. They acknowledged that integration may have been easier in Tennessee because the state had experience with managed care for physical health services. But overwhelmingly, they recommended that other states should integrate their behavioral and physical health programs.

TennCare representatives said that working with providers is critical, and that states must remember that an integrated health care system is a partnership. Before they implement anything states must understand providers' point of view; states must also help providers to get where they need to be after the implementation. Although members of the mental health advocacy community might worry that individuals will not receive behavioral health services in an integrated system, the respondents said that this is not the case. They advised bringing the behavioral health community into the conversation and emphasized the importance of a detailed and hands-on approach to monitoring the MCOs and providers. When the integration started, the bureau closely monitored the numbers and types of claims coming in, and it has continued to do so. The Bureau keeps close tabs on what providers are doing.

The stakeholders also advised states to think carefully about the MCO procurement process and about implementation. Providers should know what they are bidding on. Contracts with MCOs should be detailed, with each requirement carefully defined. Educating providers, particularly the community mental health centers, is important, and should be undertaken even on tasks such as filling out claims. Community mental health centers should be aware of the opportunities that integration opens up (for example, case managers in the centers can now help people get their physical health services, too).

Regarding data and quality initiatives, TennCare representatives said that prompt access to reliable encounter data is very important because hard data make it possible to correct misinformation based on anecdotes. Quality requirements should be spelled out in the MCO RFP, and an independent review such as those conducted by an EQRO can go a long way in dispelling stakeholder concerns. Different types and levels of incentives and sanctions can be used to ensure compliance. The respondents also advised states to consider using a state-level satisfaction survey (for example, CAHPS) since these will allow the state to track satisfaction over time.

 

VI. Vermont Blueprint for Health

A. Program Overview

Vermont Blueprint for Health (Blueprint) is a public-private patient-centered medical home (PCMH) program for improving health, wellness, and disease prevention. It aims to transform service delivery at the local level by helping local health and social service providers coordinate resources. Blueprint encompasses a variety of initiatives and has continued to expand in scope since it was first implemented in 2006. The two major reforms to date have been: (1) to incentivize and equip all willing PCPs to transition to NCQA-certified PCMHs; and (2) to implement locally developed multidisciplinary community health teams (CHTs), which expand the capacity of PCMHs, coordinate care and social service linkages for patients, and provide more intensive services to those with the most complex needs. Blueprint also supports patient self-management workshops, multiple services for providers to transition to PCMHs, and a health information technology (HIT) infrastructure that facilitates targeted outreach, population management, continuous quality-improvement, and evaluation. Other Blueprint services target specific populations. Multi-insurer payment reforms have accompanied these practice reforms. Medicaid, Medicare, and the state's largest commercial insurers make quality incentive payments to PCMHs based on their NCQA score, and they also contribute to funds for core members of CHTs.

Although Blueprint is not specifically focused on behavioral health, CHTs typically support patients with complex needs, including patients with SMI. In addition, Blueprint seeks to increase the capacity of PCPs to care for mental health and substance abuse conditions within their practices by equipping them to screen for, provide basic treatment for, and monitor common mental health conditions. Blueprint also seeks to strengthen collaborations with behavioral health providers. Finally, it has expanded implementation of Wellness Recovery Action Planning (WRAP) workshops and is in the process of extending the health home concept to providers of mental health and addiction treatment services.

B. State Context

Blueprint for Health is a major component of Vermont's health care reform strategy and has received much legislative support. Vermont initiated Blueprint as a chronic care initiative to curtail escalating health care costs. Policymakers learned that managing chronic diseases, such as diabetes and coronary artery disease, could help curb cost growth, and that the primary care setting was well positioned for delivering chronic disease management. Blueprint quality incentive payments and practice reforms were first codified in 2006 as part of Vermont's comprehensive health reform legislation (Vermont Act 191, An Act Relating to Health Care Affordability for Vermonters; Watkins 2012). As stated in the legislation, the goal of Blueprint was to "achieve a unified, comprehensive, statewide system of care that improves the lives of Vermonters with or at-risk for a chronic condition." The 2006 legislation required a pilot of early Blueprint activities, mainly related to diabetes care, which was first implemented in 2007 in three communities (HHS, AHRQ 2012).

Blueprint's focus expanded beyond chronic care management with legislation in 2007 (Vermont Act 71), and in 2008 more pilot testing was begun in the same three communities. These pilots included core Blueprint elements in place today, including PCMHs, CHTs, HIT, and financial contributions from Medicaid and major commercial insurers. The three pilot areas covered a population of approximately 60,000 patients, or about 10 percent of the state's 630,000 residents (HHS, AHRQ 2012; Watkins 2012).

In 2010, the state legislature (through Vermont Act 128) mandated statewide expansion of Blueprint reforms, requiring the state division responsible for implementation, the Department of Vermont Health Access (DVHA), to enroll all willing PCPs by October 2013 (HHS, AHRQ 2012). In addition to managing Blueprint, DVHA, within the Agency of Human Services, is designated as the state's Medicaid MCO and contributes funding as the Medicaid payer (Vermont Agency of Human Services 2011). (There are no behavioral health carve-outs under the MCO arrangement, although long-term care services are separately managed. Vermont's Medicaid MCO does not influence the structure of the Blueprint for Health, however, so is not discussed in this case study.)

State legislation in 2008 (Vermont Act 204) required all major commercial insurers in the state to contribute toward the Blueprint payment reforms (HHS, AHRQ 2012; Watkins 2012). Yet the state still had to obtain insurers' support. State administrators explained the potential cost savings of the Blueprint model and engaged insurers in further discussions to structure payment reforms. The small size of the state limits the number of private insurers, which helped Vermont to achieve support from all payers.

Multi-insurer involvement is one of the reasons Blueprint is considered a public-private partnership. Blueprint also receives input and support from other state agencies, private partners, and local participants; in some cases, partnerships are formal. State legislation established the Blueprint Executive Committee, which is composed of commissioners of DVHA, Department of Health, Department of Mental Health, and Department of Information and Innovation, as well as other government officials and non-governmental stakeholders (DVHA 2013a). There is also a Mental Health and Substance Abuse Advisory Committee, which includes leadership from the Department of Mental Health, the Department of Health's Alcohol and Drug Abuse Programs, and numerous treatment and advocacy organizations (DVHA 2012). In addition, a non-profit organization, Vermont Information Technology Leaders, has been involved in implementing the state's HIT infrastructure (DVHA 2013b). Since Blueprint is intended to be a comprehensive integration strategy at the community level that can be adapted to local needs and resources, local-level participants also influence Blueprint activities. Local participants include PCPs, community mental health clinics, human service agencies, housing agencies, area agencies on aging, and other local service providers. Various public and private stakeholders are also involved in the self-management workshops.

C. Program Goals

Since its inception as a chronic care model, Blueprint has expanded into a coordinated, statewide "agent of change" for reforming health care, wellness, and disease prevention efforts. It has three overarching aims: improve health care, improve population health, and reduce health care costs. Recognizing that health and wellness must be addressed in both medical and non-medical settings, the Blueprint framework was designed with the following goals (DVHA 2010):

• Establish a service continuum across disciplines that are not usually well integrated, such as physical health care, mental health and substance abuse services, social and economic services, housing, and public health services. In particular, Blueprint aims to better integrate mental health and addictions treatment with primary care.

• Improve access to well-coordinated preventive health services, centered on the needs of patients and families.

• Eliminate payment as a barrier for patients and families.

• Demonstrate the financial sustainability of Blueprint reforms. The payment reforms are investments on the part of the insurers and the state, which the state hopes will lead to reduced health care spending.

D. Program Financing and Contracting

The central Blueprint payment reforms--payments to incentivize quality (rather than volume) and enhance capacity through core CHT members--are financed through payments made to practices from the state's public and private insurers. Practices receive these payments on top of payments for health care provision.

Quality incentive payments for PCMHs. All large private insurers (Blue Cross Blue Shield of Vermont, MVP Health Care, and Cigna Health Care), as well as Medicaid and Medicare, contribute toward PCMHs' quality payments at the same PMPM rate, based on the number of active patients enrolled in their insurance plan (that is, those seen by the practice in the last 24 months). Quality payments range from $1.40 to $2.50 per patient, depending on the PCMH's NCQA score. For example, if a PCMH enrolled in Blueprint achieved the highest NCQA score and saw 1,000 Medicaid patients in the past 24 months, Medicaid would pay the provider $2,500 on top of contracted health care payments for a given month, or $30,000 over a year. Payments are disbursed from insurers directly to providers on a monthly or quarterly basis. Practices are rescored every three years using up-to-date NCQA standards. Patients are attributed to a practice and insurer through an all-payer claims database (DVHA 2013c; personal communication with state Blueprint staff June 3, 2013).

Cost for Core CHT members. All insurers also share the costs for "core" CHT members. Core members are those who work directly with PCMHs. A local multidisciplinary group of medical, behavioral health, and non-medical stakeholders determines the composition of the team, and an existing administrative entity (such as a hospital or FQHC) in each community hires the team members in order to avoid establishing a new administrative layer (DVHA 2010). Core CHT members cost $70,000 (for one full-time equivalent worker) per 4,000 patients (DVHA 2013b). Four of the insurers each contribute approximately 22 percent of the total core CHT costs, and the fifth insurer, which is smaller, contributes about 11 percent.

As shown in Table VI.1, Medicaid, Medicare, and other Vermont state agencies have made additional investments to extend services to targeted populations and to facilitate implementation of Blueprint practices. In addition, synergies between Blueprint initiatives, complementary state efforts--such as the Vermont Chronic Care Initiative (VCCI) and the Vermont Health Information Exchange (VHIE), described below--and local social service providers defray the costs attributable to Blueprint alone.

Financing mechanisms. According to Blueprint staff, the state exercised multiple mechanisms to enable Medicaid and Medicare to contribute funds. Vermont's 1115(a) waiver (known as the Global Commitment to Health) authorizes the use of Medicaid funding for Blueprint services. In addition, Vermont has a pending SPA to exercise the Medicaid Health Home option under the ACA, which establishes the Hub and Spoke model (described below) for those with opioid addiction and which in the future may be used to establish health homes for other populations. Vermont's participation in CMS's Multipayer Advanced Primary Care Practice Demonstration authorizes the use of Medicare funds for Blueprint.

E. Covered Populations and Services

Enrollment process. There is no formal enrollment process for patients; rather, PCPs enroll into Blueprint. All types of PCPs are eligible to enroll, but they must meet NCQA standards to be recognized as a PCMH and receive quality incentive payments. Vermont recognizes multiple provider specialties as PCPs, including internal medicine, general medicine, family medicine, pediatrics, and naturopathic medicine. Vermont also recognizes nurse practitioners and physician assistants as PCPs.

As of December 2012, 102 PCPs had achieved NCQA recognition and another 24 PCPs were preparing for recognition, together representing about two-thirds of Vermont's PCPs. Over 420,000 patients were associated with the recognized practices, or approximately two-thirds of the state's population. Also in December 2012, 89 full-time equivalent core CHT members were working with the practices (DVHA 2013b). With NCQA's recent release of standards for specialty practices, Blueprint is developing plans to expand payment and practice reforms to specialty mental health and addiction treatment facilities through the Hub and Spoke model and other efforts. While PCPs are not resisting Blueprint enrollment, a state staff member we spoke with for this study noted that small practices in particular may be less inclined to enroll due to difficulty meeting NCQA's PCMH standards.

Patients served. All Vermont residents are eligible for care management from CHTs and for self-management workshops--at no cost to them or their provider. (Patients may incur a copay established by their insurance contract for a medical appointment, but they do not incur copays for assistance provided by CHT members.) However, CHTs tend to target certain types of patients, such as those with diabetes, mental illness, or co-occurring conditions. According to a state staff member, a typical patient referred to the CHT might have both diabetes and depression and need additional support to get the diabetes under control. Another patient might have a mental health or substance abuse issue and could benefit from more treatment, follow-up, or care coordination than a practitioner can provide during a typical appointment. In addition to identifying patients via referrals from practitioners, CHTs also target certain patients based on panel management or population management, a process whereby CHTs examine clinical data to identify and target specific subgroups (for example, women over age 50) that could benefit from outreach and intervention, then implement an intervention protocol established by the practice, and later follow up with patients to influence adherence. Other services under the Blueprint framework that support specific populations include Support and Services at Home (SASH), which helps Medicare patients live at home, and Hub and Spoke, a health home model designed to curb opioid addiction.

Patient services. For patients, Blueprint's most fundamental reforms have been the transition of PCPs to NCQA-certified PCMHs and the establishment of CHTs. Each is described below.

• Patient-centered medical homes. NCQA's standards for PCMHs, listed in Table VI.2, establish the services that patients seen by PCMHs should receive. PCMHs offer improved access for patients, more communication and follow-up, consistent care based on national guidelines for prevention and control of chronic diseases, improved coordination of care and linkages with other services (medical, behavioral health, and social and economic), and resources to enable patients to better manage their own care (NCQA 2011). Vermont has shaped the PCMH services available to patients through the Blueprint framework of CHTs, self-management workshops, and data systems that make population management and continuous quality-improvement possible (data systems are described below).

• Community health teams. CHTs perform a range of functions: they coordinate service linkages for vulnerable participants across medical and non-medical service settings, help treat mild depression and anxiety, support patient self-care through one-on-one interactions and workshops, track care and conduct more intensive and individualized follow-up than what a practice can typically provide to increase the likelihood that patients adhere to treatments or referrals, and conduct population management and engage the general population in preventive health care. CHT members come from nursing, social work, nutrition, psychology, pharmacy, administrative support, and other backgrounds. Core CHT members tend to work within a practice (or split their time across multiple practices) and meet regularly with functional team members who represent local service providers to form a continuum of care (DVHA 2013b). A Blueprint staff member added that team members might also meet to figure out how to serve a particular patient without duplicating their efforts.

Benefit of PCMHs and CHTs for patients with mental illness. Some aspects of Blueprint's service reforms were designed specifically to improve care for patients with mental illness. Through PCMHs and CHTs, Blueprint aims to increase the capacity of PCPs to treat common mental health illnesses and addictions and coordinate care with specialists. Blueprint practices implement standardized screening and treatment protocols for conditions such as depression, anxiety, ADHD, and substance abuse; monitor the impact of care; and consult with specialized psychiatric or other mental health professionals.

Self-management workshops also have the potential to be an important resource for individuals with mental illness. As of February 2012, six workshops had been conducted on topics such as self-management of chronic disease, diabetes, chronic pain, tobacco cessation, and mental illness. The chronic disease workshop and the variations for diabetes and chronic pain are based on the Stanford Chronic Disease Self-Management Program, created by Dr. Kate Lorig, and are led by trained and certified peers. The mental illness workshop uses the WRAP curriculum, which offers information and teaches skills to individuals with depression and anxiety. WRAP aims to shift the focus in mental health care from symptom control to prevention and recovery. Participants create recovery plans and identify activities and resources to help them maintain their well-being. All of the self-management workshops encourage patients to set goals, create self-care plans, and solve personal obstacles. WRAP workshops meet four hours a week for six weeks (DVHA 2013b).

According to a state staff member, Blueprint added WRAP to its menu of self-management workshops in 2012 after one CHT went seeking a mental health treatment program for patients experiencing coverage and access barriers with specialty providers (including long wait lists). Vermont's Department of Mental Health first introduced WRAP in 1997, and the curriculum has since expanded to other states with support from SAMHSA, although with inconsistent funding over the years. Under the Department of Mental Health, WRAP had been implemented in community mental health centers. By including WRAP as a self-management workshop, Blueprint shifts outreach for WRAP to a primary care setting, enables use of Medicaid funding, and expands the number of workshops available statewide.

Other Blueprint and state initiatives targeted toward specific patient groups. Blueprint encompasses two other initiatives of note that support specific patient groups, the Hub and Spoke program and SASH, and it intersects with a third state initiative that does so, VCCI. These are described below. Targeted services are likely to expand in the future as the state considers a Medicaid SPA to extend the health home concept to other sectors, including the long-term care system and specialty mental health clinics.

• Hub and Spoke. This is the state's first major effort to extend the health home concept and Blueprint payment reforms to mental health and addictions centers. The focus is on treating opioid addiction, a rising public health and fiscal concern in Vermont. "Hubs" are regional specialty mental health and addiction treatment centers, which provide intensive treatment to patients, and consultation to and coordination with practices (the "spokes"). Vermont plans to have all practices in which physicians prescribe buprenorphine become spokes. Patients are stabilized in hubs and then referred to spokes, which provide less intensive treatment and which can refer patients back to hubs for episodic care as necessary (such as for a relapse), thus facilitating a continuum of treatment. This model also helps the state integrate into the broader mental health and physical health care settings two medication treatment approaches (methadone and buprenorphine) that had been separated due to different funding streams and regulations. Implementation efforts began in 2013 (DVHA 2013b).

• Support and Services at Home. SASH is a Blueprint initiative funded through Medicare. SASH teams help Medicare beneficiaries living in subsidized housing or elsewhere in the community age safely at home by coordinating health care and long-term care. SASH also seeks to reduce Medicare expenditures. Services focus on care coordination, education and coaching to support self-care, and transition support after a stay in a hospital or rehabilitation facility. Team members, who may work for local housing organizations, help participants develop an individualized healthy aging plan and meet their aging goals. SASH teams also identify goals that are common across participants to create a community-level healthy aging plan. Team members include a coordinator and wellness nurse, as well as representatives from local home health agencies, area agencies on aging, mental health providers, and other professionals. SASH began as a pilot in 2009; as of January 2013, there were 26 teams across most regions of the state (Cathedral Square Corporation 2013; DVHA 2013b).

• Vermont Chronic Care Initiative. Although VCCI is a distinct state initiative separate from Blueprint, its goals overlap with Blueprint's goals of improved chronic care management and Medicaid cost reduction. VCCI staff also collaborate with core CHT members as functional team members. VCCI targets Medicaid patients whose costs are in the top 5 percent, or whose utilization patterns (for example, hospitalizations) indicate risk of high costs. VCCI staff, typically nurses or case workers, provide case management and care coordination, coaching and health education to promote self-care, and education to improve patients' communication with their health care providers (DVHA 2013b, 2013d). Staff are collocated in practices or medical facilities.

Supports for providers. In addition to increasing the capacity of practices through CHTs, the Blueprint framework provides various supports to practices. These supports, listed below, can help practices meet PCMH standards, and more generally help the state achieve its goal of creating a more coordinated and patient-centered system of care (DVHA 2013b):

• Evaluation Quality Improvement Program (EQuIP) practice facilitators help practices achieve NCQA recognition. Generally, EQuIP facilitators help practices institute quality-improvement mechanisms. They can teach practices to use the HIT infrastructure and improve their management of chronic conditions, immunizations, preventive services and screenings, and access to care (such as same-day appointments and avoidance of EDs). Support from facilitators can continue after NCQA recognition. Facilitators come from social work, nursing, patient advocacy, and other disciplines. In 2012, 13 practice facilitators helped around 90 practices achieve NCQA-PCMH recognition.

• Learning collaboratives bring together staff (physicians, nurses, office managers) from multiple practices to improve care. In 2012, learning collaboratives focused on medication-assisted treatment for opiate addiction and on asthma treatment. Generally, the practices agree to collect common data on quality measures, identify and test practice improvements, and share information about practice changes.

• Training in shared decision-making (SDM) is available for PCP staff, EQuIP facilitators, and CHT members. SDM workshops focus on empowering patients to have more informed and productive conversations with providers. The workshops are provided through a partnership with Health Dialog and the Foundation for Informed Medical Decisions, which together developed the SDM tools. State legislation required the Blueprint state team to test SDM as a pilot program and to use a nationally certified intervention model.

• HIT infrastructure, including the central clinical registry, offers work flow tools and reports to aide in population management and planning. Blueprint also assists practices in using the HIT infrastructure via Blueprint Sprint teams (see Section G).

F. Quality Monitoring and Incentives

The Blueprint framework includes incentives and mechanisms for practices and CHTs to monitor and improve quality on an ongoing basis, and for the state to conduct system-level evaluations.

Encouraging evidence-based care and continuous quality-improvement. The chief way in which Vermont has encouraged evidence-based care is by expanding NCQA-PCMH certification statewide. Practices must be able to measure and improve performance to achieve PCMH recognition and receive quality payments. Linking payments on a sliding scale to NCQA scores incentivizes high performance. Blueprint supports for providers, including EQuIP practice facilitators, learning collaboratives, and the central clinical registry, create capacity within a practice to meet this PCMH standard. In addition, a Blueprint staff member reported that Blueprint is facilitating quality-improvements in two ways: (1) by generating a series of "practice profiles" based on an all-payer insurance claims database, to help practices benchmark their health care utilizations, use of preventive health care screenings, other care practices, and costs; and (2) by conducting annual reviews of clinical charts focused on key issues, such as the degree to which recommended diabetes care is being implemented, or whether patient self-management plans are present and followed up on.

Evaluation strategies. Evaluation strategies include examining the following performance categories: health care expenditures, health care utilization, the quality of health services, patient health outcomes, and patient experience of care. According to Blueprint staff, Vermont is preparing to assess the extent to which the health of the population is improving (for example, "Have hospitalization rates decreased?"), costs have declined, and patients are using preventive care and engaging in self-management workshops. Furthermore, the state hopes to understand the relationship between performance measures, such as the relationship between NCQA scores, patient participation in Blueprint initiatives, health care utilization patterns, and costs. Vermont has been systematically building data sources to form queryable databases so state staff and others can mine them for evaluation. Performance measures will derive from an all-payer claims database, the central clinical registry, the NCQA scoring database, and NCQA's PCMH patient experience survey (Table VI.3).

G. Information Systems and Data Infrastructure

HIT infrastructure. Vermont has expanded its existing investments in HIT through Blueprint. The two core pieces of infrastructure are the VHIE and the central clinical registry; the latter was developed for the Blueprint framework. VHIE stores demographic and clinical data from practices' and hospitals' EHRs. The VHIE, in turn, interfaces with the central clinical registry.

To support Blueprint's PCMH practice reforms, the state contracted with Covisint DocSite to establish and support the central clinical registry. The registry enables practices to conduct panel management, generate reports, and track their efforts with individual patients in a user-friendly and web-based environment. The registry also offers work flow and planning tools, and it includes an activity tracker so CHT members can track their referrals, actions taken on patients' behalf, and outcomes. A goal is to have the registry become an integrated health record used by the different individuals and organizations who work with a patient. There were 363 licensed registry users at the end of 2012 (DVHA 2013b). Basic access to the registry is free for practices and CHTs; an enhanced version is available for a fee.

The registry is an important tool for sharing information about care and prescriptions received in multiple locations, which may be especially helpful for treating patients with mental illness. For example, a CHT member working within a PCMH may be able to learn if a patient has visited a hospital or a partner at a specialty mental health clinic since the patient's last visit to the PCMH (if this information is available in the registry), and may follow up with the specialty provider or patient to offer additional support.

System and reporting requirements. Blueprint does not require practices to use EHRs or to use the central clinical registry. A state objective was to refrain from having practices report more data than what is required as part of their normal course of care. The only required reporting is for practices to record their staff so the state can administer quality incentive payments. However, use of EHR systems is widespread due in part to PCMH standards for tracking care and conducting panel management, and in part to state efforts to encourage PCPs to adopt EHRs. Most PCPs in Vermont now use EHRs; practices that do not (and even those that do) can enter data directly into the registry. CHT members can also input data directly into the registry to track information on the patients they work with.

Data quality efforts. To standardize the data in the central clinical registry, Blueprint created and routinely updates a data dictionary and condition measure set that is based on national guidelines for preventive health maintenance and treatment of chronic conditions. To support integration of EHRs, VHIE, and the registry, Blueprint created Blueprint Sprint teams, composed of representatives from Blueprint, Covisint, Vermont Information Technology Leaders, and the practices. In 2013, teams continued to work with practices to improve data quality and transmission, both of which have been a major challenge (DVHA 2013b).

H. Impact of the Affordable Care Act

Vermont has a pending SPA to exercise the Medicaid Health Home option under the ACA to establish the Hub and Spoke model for those with opioid addiction. At the time of this report, state staff were also designing health home models under the Blueprint framework for other populations, including adults with SMI. Another aim of the ACA, to expand the Medicaid-eligible population, is unlikely to influence the Blueprint program. Vermont's Global Commitment to Health, a Section 1115(a) demonstration waiver effective October 2005, enabled the state to achieve universal health care coverage. Vermont staff are, however, attempting to access ACA funding available to other states that are expanding insurance coverage, and to ensure that existing beneficiaries do not lose benefits.

I. Successes and Challenges

Successes. A Blueprint staff member we spoke with for this study highlighted several major successes or advantages to Blueprint's approach:

• The CHT concept has been a major success of Blueprint. The CHTs have increased the ability of PCPs to work with the broader community of local health, housing, and human services. The network of CHTs and local partners can be used as a vehicle to implement state or locally based initiatives that align with Blueprint's goals. As one state Blueprint staff member described it, CHTs have been "magnets for [instituting] practical, workable supports and services."

• The comprehensiveness and flexibility of the Blueprint framework are key advantages. Blueprint developers were careful not to dictate how the program should operate locally in order to give community service providers the flexibility to adapt the framework to their local needs and resources. This approach also helps communities take ownership of their programs and prevents situations in which state-level decision-makers fail to foresee circumstances on the ground.

• Multipayer involvement is a "great strength" of Vermont's approach. While recognizing multipayer participation is not necessary for a successful case management model, state staff noted that multipayer involvement (in conjunction with care standards set by the state) enables PCPs to offer the same services to all patients, regardless of their coverage.

• Promising preliminary results suggest health care expenditures and ED visits decreased for patients in pilot Blueprint sites relative to similar patients in a matched comparison group, or increased at slower rates. For instance, annual per capita expenditures for Medicaid patients in the earliest pilot sites were $623 more than the comparison group in 2007 (at baseline) and $197 less than the comparison group in 2011 after three years of operation (statistical significance is not reported; DVHA 2013b). An earlier report released by Onpoint Health based on data from 2008 through 2010 also suggests favorable trends in expenditures and utilization (Finison 2012).

Challenges. The same Blueprint staff member who noted the program's successes also described two main implementation challenges, both related to the HIT infrastructure:

• Implementing the data systems has been, by far, the biggest challenge to fulfilling the vision of Blueprint. "The whole vision around having the providers document information during their normal course of care through [EHR systems], which connect through interfaces and populate a registry, is far more complex than ever imagined in the beginning." The challenges have largely resulted from a lack of data standards across disparate systems. Vermont practices use EHR systems supplied by more than 30 different vendors, and standardization in how information is recorded in EHRs, and transmitted and received by other data systems, is lacking. A second challenge with EHRs is that they typically require practitioners to enter information into open-ended text fields, which cannot easily or readily be used in analysis. Both challenges should become less of a concern as more standards for EHRs are set at the federal level.

• Creating an integrated community record for any staff member who views, uses, or adds to a case has presented issues. While maintaining patient privacy was not cited as a key challenge, managing the consent and authorizations for multiple users affiliated with different service systems was reported to be challenging.

Advice to other states. A Blueprint staff member advises other states to think not only about health care financing reforms, but also about the practice reforms needed on the ground to support a holistic reform strategy. For instance, states should consider what quality-improvement and measurement mechanisms and supports providers would need in order to access the information necessary to monitor and improve their care.

 

VII. Conclusions

States are using an array of strategies to address the fragmented delivery of services for individuals with behavioral health needs. This report highlights the approaches used by four states--Louisiana, North Carolina, Tennessee, and Vermont--to coordinate and integrate the delivery of physical health, behavioral health, and other services and supports for individuals with behavioral health conditions. This report is not an exhaustive review of all state programs that are attempting to improve the coordination and integration of services for individuals with behavioral health needs. Previous reports have identified several other states and communities that are adopting such innovative strategies (Greenberg 2012; Hamblin, Verdier, and Au 2011). Rather, these case studies are intended to highlight some of the key features of selected state programs and to inform the efforts of other states and policymakers working to improve care for this population. We did not attempt to evaluate the effectiveness or costs of these programs, and we spoke with only a limited number of officials in each state. Thus, this report does not necessarily offer the programs it describes as models, or maintain that they are effective or should be replicated in other states.

While each state program has unique elements and must be understood within its own particular policy context, we did observe some common features and similarities across programs described below.

A. Program Goals

The programs highlighted in these case studies share similar goals; they seek to improve the accessibility, efficiency, quality, and cost-effectiveness of health care for Medicaid beneficiaries (and in some states, non-Medicaid beneficiaries) with behavioral health conditions. While the specific financing mechanisms and delivery strategies that states used to achieve these goals varied, programs did share some common elements, including strengthening case management and care coordination across different types of providers, promoting evidence-based and best practices, and using data to measure and improve quality of care.

B. Creative Financing of Services

The states included in these case studies creatively used a range of funding mechanisms in an attempt to better coordinate and integrate care for individuals with behavioral health conditions. Some states have pooled funding across state agencies or health care payers. Other states have relied exclusively on federal waivers to allow Medicaid to fund services. States have also relied on state, federal, and private grants to get their program off the ground or to fund specific initiatives. Louisiana used several Medicaid managed care waivers, home and community-based services waivers, and a state plan option to finance a CSoC. North Carolina used grant funding from a private health care philanthropy to fund its first program, the Wilson County Health Plan, from which CCNC evolved. The state later used a Section 1915(b) waiver to create CCNC; eventually, the program was written into the Medicaid state plan via an amendment. Vermont and Tennessee both used Section 1115(a) demonstration waivers to fund their programs, while Vermont also used the Medicaid Health Home state plan option to finance Blueprint. In sum, these states have harnessed the available funding streams in an effort to move toward a service delivery system that overcomes historical silos of care.

C. Covered Populations and Services

The programs highlighted in this report each chose to include large segments of the state's population. In some states, both Medicaid-eligible and non-Medicaid-eligible adults have been included in statewide system redesign efforts; other states have targeted efforts to Medicaid beneficiaries only. One state, Tennessee, chose to integrate the full range of physical, behavioral, and long-term services and supports under a single managed care contract, whereas other states chose to retain separate managed care contracts for physical and behavioral health services but have mechanisms and efforts underway to strengthen care coordination.

Some states are undertaking efforts to broaden the types of services and supports available to individuals with behavioral health conditions. An important component of each program involves connecting consumers with a variety of state-funded and community-funded social services such as housing assistance and employment services. Both Louisiana and Tennessee are working to incorporate managed care techniques into the provision of supportive housing. TennCare offers a supportive housing benefit to people with SMI who would otherwise not be able reside in their communities. In Louisiana, Magellan anticipates assuming responsibility for the management of a supportive housing program, which will serve approximately 3,300 individuals.

D. Data for Quality Monitoring and Improvement

The collection of data and the use of information systems are critical components of the programs examined. While none of the programs requires providers to use EHRs, several of them are working to encourage providers to use EHRs and they are using that information to monitor the quality of care. In addition, programs are developing novel ways of allowing providers to share information. Two programs (Vermont and North Carolina) have implemented their own secure web-based systems whereby medical, behavioral, and other providers can access utilization and medication data on each consumer to aid in making clinical decisions and track outcomes. Louisiana is working to develop a similar system. Each program has employed a variety of quality-improvement strategies in order to help meet program goals and ensure that beneficiaries receive high quality care. While these states have ongoing quality monitoring and evaluation efforts, they offer many opportunities for further evaluations that would examine how financing strategies and system redesign efforts impact the quality, outcomes, and costs of care. Some of the programs may lend themselves to quasi-experimental evaluation designs because they are implementing the program in certain regions or staggering implementation, which would allow for more rigorous testing of certain program components.

E. Importance of State Context

To some extent, each program reflects the unique state environment in which it was developed. As one program representative noted, there is not a one-size-fits-all approach to improving the integration of care for this population. In some cases, states tailored their programs to adapt to existing state programs and infrastructure; in other cases, states opted for wholesale system reform, which may have been spurred by other incremental policy changes, or, as in Louisiana, by natural disasters. As a predominantly rural state, North Carolina found that the PCCM model worked well because it was adaptable to either a rural or an urban setting and experienced high participation rates by PCPs. In contrast, the state's comprehensive managed care programs experienced trouble penetrating the rural market and eventually withdrew from the Medicaid program. Vermont, on the other hand, developed its Blueprint for Health program amidst sweeping health care reforms that reorganize health care delivery and financing. Louisiana was on the path to redesigning its system when devastating hurricanes prompted further reforms and greater collaboration between state agencies.

Such contextual factors must be understood in determining whether any of these models could improve care for behavioral health populations in other states. This report summarizes the key features of programs in these states in an effort to provide information that other states and policymakers can use when considering options for financing and delivering more integrated and coordinated care for individuals with behavioral health conditions.

 

References

Boyd, C., B. Leff, C. Weiss, J. Wolff, A. Hamblin, and L. Martin. "Clarifying Multimorbidity Patterns to Improve Targeting and Delivery of Clinical Services for Medicaid Populations." Center for Health Care Strategies, Inc. December 2010. Available at http://www.chcs.org/publications3960/publications_show.htm?doc_id=1261201. Accessed June 20, 2013.

Brown, J., A. Barrett, H. Ireys, E. Caffery, and K. Hourihan. "Evidence-Based Practices for Medicaid Beneficiaries with Schizophrenia and Bipolar Disorder." Report prepared for the HHS, ASPE. April 2012. Available at http://aspe.hhs.gov/daltcp/reports/2012/ebpsbd.shtml. Accessed June 11, 2013.

Bureau of TennCare. "Annual HEDIS/CAHPS 2012 Report: Comparative Analysis of Audited Results from TennCare MCOs." August 2012a. Available at http://www.tn.gov/tenncare/forms/hedis12.pdf. Accessed June 11, 2013.

Bureau of TennCare. "Extension Request: TennCare II (11-W-00151/4)." May 7, 2012b. Available at http://www.tn.gov/tenncare/forms/extensionnotice.pdf. Accessed June 11, 2013.

Bureau of TennCare. "2012 Quality Assessment and Performance Improvement Strategy and Quality Strategy: Annual Update Report." 2012c. Available at http://www.tn.gov/tenncare/forms/qualitystrategy2012.pdf. Accessed June 11, 2013.

Bureau of TennCare. "TennCare 101: Explaining the Basics of Medicaid in Tennessee: September 2012." Slide presentation. September 2012d. Available at http://www.tn.gov/tenncare/forms/tenncareoverview.pdf. Accessed June 12, 2013.

Bureau of TennCare "TennCare Plus: A Draft Proposal for an Integrated Medicaid/Medicare Model." April 13, 2012e. Available at http://www.tn.gov/tenncare/forms/dualsdemo.pdf. Accessed July 10, 2013.

Bureau of TennCare. "TennCare Operational Protocol." Nashville, TN: Bureau of TennCare, January 2013a. Available at http://www.tn.gov/tenncare/forms/operationalprotocol.pdf. Accessed June 12, 2013.

Bureau of TennCare. "TennCare Quick Guide." May 2013b. Available at http://www.tn.gov/tenncare/forms/quickguide.pdf. Accessed June 12, 2013.

Cathedral Square Corporation. "Support and Services at Home: A Caring Partnership." Available at http://cathedralsquare.org/future-sash.php. Accessed June 12, 2013.

Community Care of North Carolina. "The Community Care Story: Why a Community-Based System Really Works." Available at http://www.communitycarenc.com/about-us/. Accessed June 12, 2013a.

Community Care of North Carolina. "A History of CCNC: The Evolution of Community Care of North Carolina." Available at http://www.communitycarenc.com/about-us/history-ccnc-rev/. Accessed June 12, 2013b.

Community Care of North Carolina. "Medicare Quality Demonstration (646 Waiver)." Available at https://www.communitycarenc.org/emerging-initiatives/medicare-quality-demonstration/. Accessed June 12, 2013c.

Community Care of North Carolina. "Our Results: Making Headway on Cost and Quality." Available at https://www.communitycarenc.org/our-results/. Accessed June 12, 2013d.

Community Care of North Carolina. "Behavioral Health Integration." Available at https://www.communitycarenc.org/population-management/behavioral-health-.... Accessed June 13, 2013e.

Community Care of North Carolina. "Module 7: Creating an Informatics Center and Accountability/Feedback." Available at http://commonwealth.communitycarenc.org/toolkit/7/default.aspx. Accessed June 25, 2013f.

De Hert, M., C. Correll, J. Bobes, M. Cetkovich-Bakmas, D. Cohen, I. Asai, J. Detraux, S. Gautam, H. Moller, D.M. Ndetei, J.W. Newcomer, R. Uwakwe, and S. Leucht. "Physical Illness in Patients with Severe Mental Disorders: Prevalence, Impact of Medications and Disparities in Health Care." World Psychiatry, February 2011, vol. 10, no. 1, pp. 52-77.

Department of Vermont Health Access. "Blueprint for Health Executive Committee." June 10, 2013a. Available at http://dvha.vermont.gov/advisory-boards/blueprint-for-health-executive-committee. Accessed June 12, 2013.

Department of Vermont Health Access. "Mental Health and Substance Abuse Advisory Committee." June 20, 2012. Available at http://dvha.vermont.gov/advisory-boards/mh-sa-stakeholder-membership-9-19-12.pdf. Accessed June 12, 2013.

Department of Vermont Health Access. "Vermont Blueprint for Health 2009 Annual Report." Williston, VT: DVHA, January 2010. Available at http://hcr.vermont.gov/sites/hcr/files/Blueprint/Blueprint%20for%20Health%202012%20Annual%20Report%20%2002_14_13_FINAL.pdf. Accessed June 12, 2013.

Department of Vermont Health Access. "Vermont Blueprint for Health 2012 Annual Report." Burlington, VT: DVHA, February 15, 2013b. Available at http://hcr.vermont.gov/sites/hcr/files/Blueprint/Blueprint%20for%20Health%202012%20Annual%20Report%20%2002_14_13_FINAL.pdf. Accessed June 12, 2013.

Department of Vermont Health Access. "Vermont Blueprint for Health PPPM Payment Methodology." January 1, 2013c. Available at http://dvha.vermont.gov/advisory-boards/pppm-payment-methodology-by-payer-1-31-13.pdf. Accessed June 12, 2013.

Department of Vermont Health Access. "Vermont Chronic Care Initiative." May 29, 2013d. Available at http://dvha.vermont.gov/for-consumers/vermont-chronic-care-initiative-vcci. Accessed June 12, 2013.

Durden, E., E. Bagalman, E. Muser, J.C. Choi, C. Crivera, R. Dirani, W. Macfadden, and J.T. Haskins. "Characteristics, Healthcare Utilization and Costs of Bipolar Disorder Type I Patients With and Without Frequent Psychiatric Intervention in a Medicaid Population." Journal of Medical Economics, 2010, vol. 13, no. 4, pp. 698-704.

Finison, K. "Vermont Blueprint Evaluation: A Four-Year Overview Based on Two-Year Cohorts with Matched Controls (VHCURES Population, Age 18-64)." Manchester, ME: Onpoint Health Data, January 2012.

Gordon, D., W. Long, and C. Dungan. "Health Care Finance and Administration FY 2014 Budget Presentation." Avaliable at http://www.tn.gov/tenncare/forms/HCFAbudgetFY14.pdf. Accessed June 12, 2013.

Greenberg, L. "Best Practices Compendium for Serious Mental Illness." Report for Medicaid Health Plans of American Center for Best Practices, 2012. Available at http://www.mhpa.org/_upload/SMICompendiumFINALweb_744522.pdf. Accessed June 12, 2013.

Hamblen, W., and W.F. Fox. "The Impact of TennCare: A Survey of Recipients, 2011." Knoxville, TN: University of Tennessee Center for Business and Economic Research, November 2011. Available at http://cber.bus.utk.edu/tncare/tncare11.pdf. Accessed June 12, 2013.

Hamblin, A., J. Verdier, and M. Au. "State Options for Integrating Physical and Behavioral Health Care." ICRC Technical Assistance Brief. October 2011. Available at http://www.integratedcareresourcecenter.com/pdfs/ICrC_bh_briefing_document_1006.pdf. Accessed June 12, 2013.

James, J. "Managed Care Organization as an Integrated Care Entity." Presentation for CHCS Webinar "Integrating Physical and Behavioral Health: An Exploration of State Options." November 15, 2011. Available at http://www.chcs.org/publications3960/publications_show.htm?doc_id=1261314. Accessed June 12, 2013.

Kaiser Commission for Medicaid and the Uninsured. "Community Care of North Carolina: Putting Health Reform Ideas into Practice in Medicaid." Washington, DC: Kaiser Family Foundation, May 2009. Available at http://kff.org/health-reform/issue-brief/community-care-of-north-carolina-putting-health/. Accessed June 12, 2013.

Kim, J.Y., T.C. Higgins, D. Esposito, A.M. Gerolamo, and M. Flick. "SMI Innovations Project in Pennsylvania: Final Evaluation Report." Princeton, NJ: Mathematica Policy Research, October 2012.

Kronick, R.G., M. Bella, and T.P. Gilmer. "The Faces of Medicaid III: Refining the Portrait of People with Multiple Chronic Conditions." Hamilton, NJ: Center for Health Care Strategies, Inc., October 2009.

Louisiana Department of Health and Hospitals, Office of Behavioral Health. "Louisiana Behavioral Health Partnership Services Definition Manual Version 8." December 26, 2012a. Available at http://new.dhh.louisiana.gov/assets/docs/BehavioralHealth/LBHP/ServicesManual-Current.pdf. Accessed June 11, 2013.

Louisiana Department of Health and Hospitals, Office of Behavioral Health. "Quality Strategy for the Louisiana Behavioral Health Care Prepaid Inpatient Healthcare Plan Waiver." October 24, 2012b. Available athttp://new.dhh.louisiana.gov/assets/docs/BehavioralHealth/publications/CSOC/Doucments/LouisianaBHQualityStrategy.pdf. Accessed June 12, 2013.

Louisiana Department of Health and Hospitals, Office of Behavioral Health. "Requests for Proposals: Statewide Management Organization for the Louisiana Behavioral Health Partnership." July 8, 2011. Available at http://new.dhh.louisiana.gov/assets/docs/contracts/305PUR-DHHRFP-SMO-OBH_STATEWIDE_MANAGEMENT_ORGANIZATION.pdf. Accessed June 11, 2013.

Magellan Health Services "Provider Handbook Supplement for the Louisiana Behavioral Health Partnership." 2012. Available at http://www.magellanoflouisiana.com/media/133862/louisiana_provider_handb.... Accessed June 12, 2013.

McCarthy, D., and K. Mueller. "Community Care of North Carolina: Building Community Systems of Care Through State and Local Partnerships." Commonwealth Fund. June 2009. Available at http://www.commonwealthfund.org/~/media/Files/Publications/Case%20Study/2009/Jun/1219_McCarthy_CCNC_case_study_624_update.pdf. Accessed June 12, 2013.

National Committee for Quality Assurance. "NCQA PCMH 2011." 2011. Available at http://www.ncqa.org/portals/0/programs/recognition/PCMH_2011_Scoring_Summary.pdf. Accessed June 11, 2013.

North Carolina Foundation for Advanced Health Programs. "History of ICARE." Available at http://www.ncfahp.org/history.aspx . Accessed June 12, 2013.

Ricketts, T.C., S. Greene, P. Silberman, H.A. Howard, and S. Poley. "Evaluation of Community Care of North Carolina Asthma and Diabetes Management Initiatives: January 2000-December 2002." Chapel Hill, NC: Cecil B. Shepps Center for Health Services Research, University of North Carolina, April 2004.

Shipman, P. "Managed Care: Right or Wrong for North Carolina?" North Carolina Medical Journal, 2012, vol. 73, no. 3. Available at http://www.ncmedicaljournal.com/wp-content/uploads/2012/05/NCMJ_73306_FINAL1.pdf. Accessed June 12, 2013.

Steiner, B.D., A.C. Denbam, E. Ashkin, W.P. Newton, T. Wroth, and L.A. Dobson Jr. "Community Care of North Carolina; Improving Care through Community Networks." Annals of Family Medicine, 2008, vol. 6, pp. 361-67.

Treo Solutions. "Treo Solutions Report." 2011. Available at https://www.communitycarenc.org/elements/media/related-downloads/treo-analysis-of-ccnc-performance.pdf. Accessed June 12, 2013.

U.S. Department of Health and Human Services, Agency for Healthcare Research and Quality. "Statewide Program Supports Medical Homes Through Multidisciplinary Teams, Easy Access to Information, and Incentives, Leading to Lower Costs and Better Care." Washington, DC: HHS, AHRQ, 2012. Available at http://www.innovations.ahrq.gov/content.aspx?id=3640 . Accessed June 24, 2013.

U.S. Department of Health and Human Services, Centers for Medicare and Medicaid Services. "North Carolina Title XIX State Plan Amendment, Transmittal #NC 11-050." May 25, 2012. Available at http://www.ncdhhs.gov/dma/plan/HealthHomeApprovedSPA-Effective10012011.pdf. Accessed July 9, 2013.

Vermont Agency of Human Services. "Global Commitment to Health 11-W-00194/1, Section 1115 Demonstration Year: 6: 6 (10/1/2010-9/30/2011), Quarterly report for the period October 1, 2010-December 31, 2010." February 18, 2011.

Watkins, L.D. "Bringing it All Home: Building on the PCMH Infrastructure." Presented at the National Center for Quality Assurance's Policy GPS: Directions to Delivery System Reform, Washington, DC, December 5, 2012. Available at http://www.ncqa.org/Portals/0/Policy%20Conference/2012/Panel%204_%20Lisa%20Watkins_%20PDF.pdf. Accessed June 24, 2013.

Appendix A. State Profile Tables