There is a saying in psychiatry that "if you've seen one managed care program, you've seen one managed care program." Particularly in the public sector, each appears unique in that it is tailored to local structures, history, geography, and politics (1). With so much variability, our understanding of this new force in psychiatry is often limited to case examples that are, by nature, ever changing. Thus it is difficult for the observer to distill the essence of these attempts at implementing public-sector managed care (2).
In a previous work, Hoge and his colleagues (3) examined the historical efforts and innovations in the public sector that were designed to manage care. The intended objectives of these innovations were identified and integrated into a comprehensive definition of managed care as it should ideally be developed in the public sector. Since that publication, numerous managed care programs have been implemented for public-sector populations, and the literature discussing the design and evaluation of public-sector managed care has grown (4,5,6).
The purpose of this paper is to extend the earlier conceptual work (3) and to improve understanding of current public-sector managed care initiatives. By examining existing initiatives in the public sector, we have identified ten key dimensions on which each managed care initiative should be assessed in order to comprehend the objectives, structures, and processes that are being introduced and their likely impact on existing systems of care.
This assessment approach has at least three potential uses. First, it offers providers, consumers, and public payers a unique and concise tool to aid in understanding managed care initiatives. Second, it has value to managers as a tool for assessing the potential impact of a new managed care initiative on their agency and the patients it serves. Third, as demonstrated in the discussion section, this approach can be used as a framework for examining the current characteristics and influence of managed care as it is unfolding in the field of public mental health care.
This section describes the ten dimensions on which a managed care program should be examined and provides examples of the common ways in which programs vary on these dimensions.
Identifying the primary objectives is a starting point for understanding any managed care program. Such programs have manifest objectives or functions that are intended and anticipated, as well as latent objectives that are unintended, unanticipated, or unarticulated (7). Differentiating the principal objectives from the official hyperbole requires an examination of project documents, the public and policy statements of those initiating the project, and the political context in which the project occurs.
At least three major objectives might appear in a public-sector managed care program. Cost containment is by far the most common as states attempt to reduce their outlay for health care and shift the financial risk for escalating costs to managed care organizations and providers. Expanding eligibility to the uninsured is a second objective that has been pursued in some states as cost savings have been redirected to enroll those who were previously uncovered and ineligible for publicly funded health care. The third potential objective is to enhance the quality of care by improving access to services and the appropriateness and effectiveness of care.
Managed care programs vary tremendously in their scope. With respect to populations, the scope can be quite narrow, such as in the Connecticut ACCESS initiative (8), which focused only on Medicaid recipients qualifying through the category for needy families with dependent children. Alternatively, programs such as TennCare in Tennessee (9) have included numerous populations. The potential groups include indigent persons, elderly persons, children in the child welfare system, and disabled persons, such as those with severe mental illness, developmental disabilities, and AIDS.
With respect to population size and geographic coverage, managed care programs in the public sector have ranged from single-city projects with under 200 patients, such as the Baltimore Capitation Demonstration Project (10), to statewide programs, such as the Medicaid demonstration in Massachusetts, which has covered more than 350,000 individuals (11).
Organizational structures and authority
The third assessment dimension relates to the organizations involved in a program, the nature of their role and authority, and their interrelationships. Typically, three levels exist in the organizational structure of public-sector managed care projects, each associated with a specific type of authority.
At the top level are governmental agencies, such as the Health Care Financing Administration and state and county departments, which act as payers and regulators, setting policies, determining procedures, and providing oversight. The second level pertains to management of the program, a task that is often delegated by contract from a state or county government to a managed care organization or to a joint venture between a managed care organization and local providers. A fundamental choice is made by government payers about whether the contract for the management and delivery of mental health and substance abuse services will be carved out —given to an organization specializing in these services—or integrated (carved in) with the contract for all medical care.
The final organizational level is that of the service providers where responsibility for service provision is high but authority over any aspect of the project's policies and procedures is generally low.
The enrollment process involves placing individuals who are eligible to participate in a managed care program on the rolls of the program so that they can gain access to services when needed. Three key factors should be considered regarding enrollment. First, it is either mandatory or voluntary depending on the design of the program, the requirements of state and federal regulations, and the receipt of waivers from such regulations. Second, individuals may have numerous enrollment options. They may include a choice of fundamentally different managed care programs, as in the Massachusetts demonstration in which Medicaid recipients either selected a primary care physician and mandatory participation in a managed behavioral health carve-out or selected one of several health maintenance organizations (HMOs) that provided and managed all medical and mental health care.
Finally, the enrollment process itself must be assessed. Considerable variability exists about which organizations are empowered to recruit individuals into a managed care program and how individuals who do not actively enroll are assigned to a default program or HMO. Most controversial have been processes in which the HMOs hired recruiters to entice individuals to enroll, offering commissions to the recruiters and inducements to eligible individuals. Less controversial has been the hiring of neutral enrollment brokers by states or counties.
Each managed care program has a benefit package specifying the covered services that enrollees are eligible to receive and the excluded services that are not reimbursable. In addition, annual or lifetime limits on benefits may place a ceiling on the dollar value of services reimbursed for each enrollee in the program. The content of a benefit package is particularly important to individuals with severe mental illness enrolled in managed care programs. State-of-the-art care for this population combines clinical services with residential, social, and rehabilitative interventions. This broad conceptualization of treatment and its desired outcomes, which Anthony (12) has referred to as "places to be and symptom free," differs from the narrow, acute care view of basic benefits typically used by managed behavioral health organizations in the private sector.
Strategies for managing utilization
All managed care programs have procedures for managing utilization (5). This process begins with network development involving the selection and credentialing of providers to deliver the covered services to enrolled members. The managed care organizations most aggressively controlling utilization and costs tend to be the most restrictive about both provider selection criteria and the number of providers admitted to the network.
Controlling access to the network is a second utilization management strategy. Critical differences exist between managed care programs in whether an enrolled member can gain access to a provider directly and receive some services without obtaining prior approval or whether access to all services must be preapproved by the individual's primary care physician or by staff of the managed care organization.
Finally, enrollees in managed care programs are entitled to most covered services only if the services are deemed medically necessary. Thus an examination of the procedures, the medical necessity definitions, and the levels-of-care criteria used to review and authorize services reveals the steps an individual must take to gain access to services that are supposedly available if needed (13).
Intertwined with network development and care management are the interventions or strategies identified as optimal in a managed care program. Referred to as best practices, such interventions are defined as those that provide the best care at the lowest cost in real-world, nonlaboratory or nonexperimental settings (14). In light of the paucity of research on the treatment effectiveness of many clinical interventions in naturalistic settings, best clinical practices are currently identified through a combination of sources, including controlled research trials, treatment guidelines developed by professional organizations, peer review and utilization review studies, and provider profiling (14).
Considerable variability exists in the extent to which innovative clinical and administrative practices are employed. In public-sector managed care, best practices are most often focused on severely ill enrollees who are high utilizers of services, who are frequently admitted to inpatient units, or who tend to relapse after hospitalization. However, some managed care programs do not adequately attend to the special needs of this target population (6).
Financial arrangements are probably the most complex and variable aspect of managed care programs. Assessing this dimension involves first identifying the financial resources that are pooled. Some initiatives use a sole source of funding, such as Medicaid dollars for a single category of eligibility (for example, needy families with dependent children). Some use a combination of sources, such as Medicaid funds for multiple categories of eligibility, state general funds for mental health and child welfare, and county funds previously dedicated to behavioral health services. Pooling of previously separate funds at least theoretically increases the possibility of enhancing the coordination of purchased services (15) and achieving cost savings. Savings are accomplished by reducing duplicate administrative structures and negotiating discounted rates from providers in exchange for the promise of continued or increased referrals.
Understanding the flow of funds among the involved parties and the financing method and risk arrangements of each transaction is also essential. For example, funds may flow from a state or county to a managed care organization on a capitated basis, leaving the managed care organization with a fixed pool of resources and substantial risk for the cost of services if the cost exceeds the funds available. Alternatively, states or counties may retain the risk and hire a managed care company as an administrative services organization that assembles a network, manages care, and pays claims but bears little or no risk for the cost of the services that are rendered (16).
As the funds flow from managed care organizations to providers, an even broader range of potential financing methods and risk arrangements exist, including case rates, in which the provider receives a fixed sum to provide a course of treatment for an individual in a specific diagnostic category regardless of the length or cost of that care, and withholds, in which a managed care organization holds back a portion of payments to a provider network for a specified time and then distributes these funds based on the efficiency or financial performance of the providers or the plan. Examining the flow of funds and risk arrangements clarifies which parties in the program have incentives to either increase or decrease utilization and costs of care.
Finally, it is important to examine the amount of funds dedicated to the provision of behavioral health services in a managed care program and the allocation of those funds to cover direct services, administrative costs, and profit. Levels of funds, usually summarized in costs per member per month, vary widely and can be compared with other programs or plans that cover similar populations.
Quality management and outcomes measurement
Managed care programs also vary considerably in the extent to which they implement systems for quality management and outcome measurement. Medicaid demonstration projects that arise through a 1915 (b) federal waiver require a mandated evaluation of the demonstration's impact. However, the scope and rigor of such evaluations have varied significantly (17,18).
Beyond federally mandated evaluations, states or counties may dictate quality management and outcome monitoring requirements. In addition, the National Committee for Quality Assurance, which is an oversight group sponsored by employers and the managed care industry, has issued standards for accreditation of managed behavioral health care organizations that require them to implement a comprehensive quality management and improvement program to achieve accreditation (19). Finally, another recent strategy that has evolved in some programs to enhance quality through outcome monitoring involves performance contracting, in which financial penalties and bonuses to managed care organizations or providers are based on success in achieving specified targets on selected process and outcome measures (20).
Impact on the public system
Assessing the nine dimensions outlined above produces an overview of the structures and processes that constitute a public-sector managed care program. A more in-depth understanding often can be obtained by carefully assessing the changes that occur under managed care and analyzing the specific impact of a managed care program on the existing elements of the public mental health system, the substance abuse treatment system, and related social service systems. Especially critical is identifying how accountability for funding, managing, and delivering care is altered and whether cost shifting may occur from one element of the system to another.
From a developmental perspective, managed behavioral health care is in its early adolescence. It has evolved rapidly over the brief decade and a half of its existence and continues to change with great speed as it is now being introduced in the public sector. Just as understanding an adolescent requires a comprehensive examination of numerous variables such as symptoms, functioning, social adjustment, and family factors, understanding public-sector managed care requires a multidimensional assessment. A systematic review of public-sector initiatives reveals defining characteristics that are often not evident from a cursory examination. To illustrate the type of conclusions that this assessment approach can yield about a single project, we highlight below recurrent findings that we have derived from the application of the approach to a broad range of public-sector managed care initiatives.
Currently, most public-sector managed care initiatives focus on one major objective, giving less than optimal attention to other important dimensions. The majority of initiatives, such as the managed Medicaid demonstrations in Massachusetts (11) and Iowa (20), focus principally on cost control, while a smaller subset are driven principally by a mission to improve care to the severely mentally ill (10,21,22). Balanced attention to multiple objectives is strikingly absent.
Even when a public-sector managed care project is geographically broad, its scope in terms of the populations served and the services delivered tends to be quite narrow. Currently, most initiatives include only a small subset of the individuals covered through publicly funded insurance, and the benefit plans involve only one category of service such as mental health care. Thus the "system" of publicly funded services remains complex and highly fragmented (15). Individuals must still confront numerous bureaucracies to meet their multiple needs.
Fragmentation is similarly a problem with respect to the financing of initiatives. Typically, only one funding stream is being managed in public-sector managed care projects. With few exceptions (9), the possibility of pooling different streams to improve the coordination of care and to create economic efficiencies has not been achieved.
With respect to the organization of public-sector managed care initiatives, the integrated models in which states contract with HMOs for both general behavioral health and general medical services are, in fact, often carve-outs. They become carve-outs because the HMOs that contract with the state frequently subcontract with managed behavioral health organizations to manage mental health and substance abuse services (23,24).
Having enrollment options gives covered individuals some level of choice but too often results in differences among options in the ease of access, appropriateness, and quality of behavioral health services. These differences can result in the phenomenon of adverse selection, in which a disproportionate number of individuals with mental illness or an addiction become enrolled in one plan, burdening it financially (16).
A cursory review suggests that the utilization management policies and procedures of most managed care organizations are very similar. However, in practice, managed care programs and organizations vary substantially on this dimension. Consumers and providers can offer considerable insights about the level of complexity and restrictiveness in the day-to-day implementation of the utilization management process for a given managed care organization (6).
The clinical best practices that are increasingly being touted by managed behavioral health care organizations, such as assertive community treatment, respite care, and relapse prevention, are not new developments or advances, but rather are interventions previously developed and proven effective in public-sector systems (25).
In most projects, the efforts to assess quality of care are fairly meager, and the resulting data are quite sparse. A mixture of clinical anecdotes, process measures, and cost information gives us but a limited understanding of only the short-term impact of introducing managed care in the public sector (26). Quality-monitoring systems that are too weak minimize the potential effectiveness of the federal, state, or county government's oversight role and can undermine the managed care initiative as a whole (9).
Despite our limited knowledge about clinical outcomes of managed care, it is having a dramatic impact on the pre-existing public-sector system. Substantial shifts, which vary in each initiative, are occurring among government authorities, managed care organizations, providers, and consumer and family groups in the areas of oversight and accountability for service provision, clinical decision making, and financial risk (27).
Managed care will no doubt continue to change the face of public-sector systems. The peril lies in viewing managed care narrowly, only from a financing perspective and only as a corrective action for past economic inefficiencies. Rather, we have the opportunity to view it broadly from multiple dimensions. We can examine the various ways in which it does, or perhaps does not, influence the delivery of behavioral health care. Although infinite variations will occur as managed care is implemented across the country, a finite and select group of variables, such as those articulated in this assessment approach, should be the focus of continued planning, implementation, and evaluation efforts as we strive to achieve an optimally functioning system of care for public-sector populations.
Dr. Hoge is associate professor of psychology in psychiatry, Dr. Jacobs is professor of psychiatry, and Dr. Griffith is professor of psychiatry and African-American studies and deputy chairman for clinical affairs in the department of psychiatry at Yale University School of Medicine. Mr. Thakur is a postdoctoral fellow at the Cecil G. Sheps Center for Health Services research at the University of North Carolina in Chapel Hill. Dr. Hoge is also director of managed behavioral health care at the Yale University School of Medicine. Dr. Jacobs is also director of the Connecticut Mental Health Center in New Haven. Send correspondence to Dr. Hoge at Yale University School of Medicine, 25 Park Street, Room 622, New Haven, Connecticut 06519 (e-mail, firstname.lastname@example.org). An earlier version of this paper was presented at the American Psychiatric Association's Institute on Psychiatric Services held October 24-28, 1997, in Washington, D.C.